Split (1)

train · ~9.11M rows (showing the first 93.5k)

identifier stringlengths 11 32	pdf_url stringlengths 17 4.62k ⌀	lang stringclasses 120 values	error stringclasses 1 value	title stringlengths 2 500 ⌀	source_name stringlengths 1 435 ⌀	publication_year float64 1.9k 2.02k	license stringclasses 3 values	word_count int64 0 1.64M	text stringlengths 1 9.75M
https://openalex.org/W2964098765	https://curis.ku.dk/ws/files/226871460/e201900381.full.pdf	English	null	The miRNA<i>bantam</i>regulates growth and tumorigenesis by repressing the cell cycle regulator<i>tribbles</i>	Life science alliance	2,019	cc-by	11,733	Citation for published version (APA): Gerlach, S. U., Sander, M., Song, S., & Herranz, H. (2019). The miRNA bantam regulates growth and tumorigenesis by repressing the cell cycle regulator tribbles. Life Science Alliance, 2(4), [e201900381]. https://doi.org/10.26508/lsa.201900381 Download date: 24. Oct. 2024 university of copenhagen university of copenhagen The miRNA bantam regulates growth and tumorigenesis by repressing the cell cycle regulator tribbles Gerlach, Stephan U.; Sander, Moritz; Song, Shilin; Herranz, Héctor university of copenhagen The miRNA bantam regulates growth and tumorigenesis by repressing the cell cycle regulator tribbles Stephan U Gerlach, Moritz Sander, Shilin Song, H´ector Herranz One of the fundamental issues in biology is understanding how organ size is controlled. Tissue growth has to be carefully reg- ulated to generate well-functioning organs, and defects in growth control can result in tumor formation. The Hippo signaling pathway is a universal growth regulator and has been implicated in cancer. In Drosophila, the Hippo pathway acts through the miRNA bantam to regulate cell proliferation and apoptosis. Even though the bantam targets regulating apoptosis have been de- termined, the target genes controlling proliferation have not been identiﬁed thus far. In this study, we identify the gene tribbles as a direct bantam target gene. Tribbles limits cell proliferation by suppressing G2/M transition. We show that tribbles regulation by bantam is central in controlling tissue growth and tumorigenesis. We expand our study to other cell cycle regulators and show that deregulated G2/M transition can collaborate with oncogene activation driving tumor formation. disease progression (Esquela-Kerscher & Slack, 2006). Understanding the roles of miRNAs in development and cancer requires the identiﬁcation of the target genes responsible for their functions. disease progression (Esquela-Kerscher & Slack, 2006). Understanding the roles of miRNAs in development and cancer requires the identiﬁcation of the target genes responsible for their functions. The Hippo tumor suppressor pathway is a universal growth regu- lator and has been implicated in cancer (Pan, 2010). In Drosophila, the Hippo pathway consists of a cascade of kinases, which function se- quentially to restrict the nuclear localization of the growth-promoting transcriptional coactivator, Yorkie (Yki) (Huang et al, 2005). Yki regulates the expression of the miRNA bantam and bantam mediates its growth- promoting role (Nolo et al, 2006; Thompson & Cohen, 2006). bantam was the ﬁrst miRNA discovered in Drosophila. It was identiﬁed as an element that, when overexpressed in the wing, promotes tissue growth (Hipfner et al, 2002; Brennecke et al, 2003). bantam stimulates growth by inducing proliferation and inhibiting apoptosis (Brennecke et al, 2003). bantam can also cooperate with the oncogene epithelial growth factor receptor (EGFR) in the formation of neoplastic tumors (Herranz et al, 2012b). Years of extensive research have led to the identiﬁcation of multiple bantam targets, including suppressor of cytokine signaling at 36E (Socs36E), head involution defective (hid), capicua (cic), meiotic P26 (mei-P26), and enabled (ena) (Brennecke et al, 2003; Herranz et al, 2010, 2012a, 2012b; Becam et al, 2011). The miRNA bantam regulates growth and tumorigenesis by repressing the cell cycle regulator tribbles Gerlach, Stephan U.; Sander, Moritz; Song, Shilin; Herranz, Héctor Gerlach, Stephan U.; Sander, Moritz; Song, Shilin; Herranz, Héctor Published in: Life Science Alliance DOI: 10.26508/lsa.201900381 Publication date: 2019 Document version Publisher's PDF, also known as Version of record Document version Publisher's PDF, also known as Version of record Document license: CC BY Citation for published version (APA): Gerlach, S. U., Sander, M., Song, S., & Herranz, H. (2019). The miRNA bantam regulates growth and tumorigenesis by repressing the cell cycle regulator tribbles. Life Science Alliance, 2(4), [e201900381]. https://doi.org/10.26508/lsa.201900381 Citation for published version (APA): Gerlach, S. U., Sander, M., Song, S., & Herranz, H. (2019). The miRNA bantam regulates growth and tumorigenesis by repressing the cell cycle regulator tribbles. Life Science Alliance, 2(4), [e201900381]. https://doi.org/10.26508/lsa.201900381 Download date: 24. Oct. 2024 on 4 September, 2019 life-science-alliance.org Downloaded from http://doi.org/10.26508/lsa.201900381 Published Online: 22 July, 2019 \| Supp Info: on 4 September, 2019 life-science-alliance.org Downloaded from http://doi.org/10.26508/lsa.201900381 Published Online: 22 July, 2019 \| Supp Info: The miRNA bantam regulates growth and tumorigenesis by repressing the cell cycle regulator tribbles However, the current knowledge of the bantam target genes remains incomplete because individual depletion of any of them fails to mimic the effect of bantam driving tissue growth. DOI 10.26508/lsa.201900381 \| Received 13 March 2019 \| Revised 15 July 2019 \| Accepted 15 July 2019 \| Published online 22 July 2019 © 2019 Gerlach et al. Department of Cellular and Molecular Medicine, University of Copenhagen, Copenhagen, Denmark Correspondence: hherranz@sund.ku.dk Introduction Tissue growth has to be precisely regulated to generate organs with correct size, shape, and function. However, how growth is controlled remains a fundamental question in biology. In most animals, growth is the result of the increase in the number of cells, which is de- termined by the rate of cell division and cell death (Conlon & Raff, 1999). Deregulation in those processes can result in the formation of tumors (Hanahan & Weinberg, 2011). The wing imaginal disc of Drosophila has proven to be a useful system to study tissue growth control and to model different aspects related to tumor formation and metastasis (Gonzalez, 2013; Hariharan, 2015; Herranz et al, 2016). miRNAs are small noncoding RNAs that have emerged as central regulators in the expression and function of animal genomes. miRNAs regulate multiple processes in development, including tissue growth, stem cell development, hormone action, and the organiza- tion and function of the central nervous system (Carthew et al, 2017). In addition, miRNAs control the activity of oncogenes and tumor suppressors and, thus, play crucial roles in cancer initiation and Here, we identify the gene tribbles (trbl) as a direct bantam target gene that mediates its growth regulatory role. We ﬁnd that trbl down- regulation in the wing disc accelerates G2/M progression in the cell cycle. We show that simultaneous depletion of the bantam targets trbl and the proapoptotic gene hid reproduces the effect of bantam, promoting growth. The results presented here determine the minimal combination of bantam target genes sufﬁcient to mimic the effect of bantam in tissue growth. We provide evidence that Trbl not only mediates the bantam growth-promoting role but also limits the magnitude of tissue overgrowth induced by the proto-oncogene Yki. Finally, we ﬁnd that depletion of trbl can drive tumorigenesis in dif- ferent contexts. Taken together, these results demonstrate that, in some instances, Trbl can limit tumorigenesis in the imaginal epithe- lium. In summary, our study demonstrates that trbl is a crucial target mediating the growth-promoting role and oncogenic function of bantam. Department of Cellular and Molecular Medicine, University of Copenhagen, Copenhagen, Denmark https://doi.org/10.26508/lsa.201900381 vol 2 \| no 4 \| e201900381 1 of 12 1 of 12 to repress mRNA translation. Understanding the oncogenic and growth regulatory roles of bantam requires the identiﬁcation of its target genes. Bioinformatic analysis predicted two bantam-binding sites in the 39UTR of the gene trbl (Fig 1A). Introduction trbl encodes a kinase-like protein that controls cell proliferation by regulating the stability of the cell cycle regulator, cdc25-stg (Mata et al, 2000). bantam represses trbl miRNAs are negative regulators of gene activity. They bind to complementary sites on target mRNAs to induce mRNA cleavage or miRNAs are negative regulators of gene activity. They bind to complementary sites on target mRNAs to induce mRNA cleavage or protein that controls cell proliferation by regulating the stability of the cell cycle regulator, cdc25-stg (Mata et al, 2000). Figure 1. bantam represses trbl directly. (A) Illustration of trbl-sensor showing predicted bantam-binding sites. The transgene consists of a tubulin promoter expressing the mRNA for GFP fused to the trbl 39UTR. The mRNA is sensitive to regulation through regulatory elements in the trbl 39UTR. The trbl 39UTR includes two predicted binding sites for bantam (indicated as site 1 and 2). The predicted binding between the nucleotides of the trbl 39UTR and bantam is shown, and the predicted seed sequences are highlighted (connecting lines between the nucleotides and grey background). The trbl-mut-sensor has two nucleotides exchanged in each of the predicted seed sequences (highlighted in red). (B, C) Confocal micrographs of third-instar wing imaginal discs showing trbl-GFP expression (green and grey) in the genotypes indicated. DAPI was used to label the nuclei (blue). Larvae were kept constantly at 25°C. Scale bar: 50 μm. (D–G) Confocal micrographs of third-instar wing imaginal discs showing the expression of trbl-sensor (D, E) and trbl-mut-sensor (F, G) in the genotypes indicated. The sensors are shown in green and grey. DAPI was used to label the nuclei (blue). Larvae were kept constantly at 25°C. Scale bar: 50 μm. (H, I) bantamΔ1-mutant clones marked by the absence of β-Gal (red and grey), showing the expression of the trbl-sensor in (H) (green and grey), and showing the expression of the trbl-mut-sensor in (I) (green and grey). DAPI was used to label the nuclei (blue). To facilitate the analysis, the clones were induced in a Minute/+ background to allow the recovery of bigger clones. Scale bar: 50 μm. (J, K) wtsx1-mutant clones marked by the absence of RFP (red and grey), showing the expression of the trbl-sensor (green and grey). DAPI was used to label the nuclei (blue). (K) Shows a magniﬁcation of the wing disc shown in (J). Scale bar: 100 μm in (J) and 50 μm in (K). Figure 1. bantam represses trbl directly. Figure 1. bantam represses trbl directly. (A) Illustration of trbl-sensor showing predicted bantam-binding sites. bantam represses trbl The transgene consists of a tubulin promoter expressing the mRNA for GFP fused to the trbl 39UTR. The mRNA is sensitive to regulation through regulatory elements in the trbl 39UTR. The trbl 39UTR includes two predicted binding sites for bantam (indicated as site 1 and 2). The predicted binding between the nucleotides of the trbl 39UTR and bantam is shown, and the predicted seed sequences are highlighted (connecting lines between the nucleotides and grey background). The trbl-mut-sensor has two nucleotides exchanged in each of the predicted seed sequences (highlighted in red). (B, C) Confocal micrographs of third-instar wing imaginal discs showing trbl-GFP expression (green and grey) in the genotypes indicated. DAPI was used to label the nuclei (blue). Larvae were kept constantly at 25°C. Scale bar: 50 μm. (D–G) Confocal micrographs of third-instar wing imaginal discs showing the expression of trbl-sensor (D, E) and trbl-mut-sensor (F, G) in the genotypes indicated. The sensors are shown in green and grey. DAPI was used to label the nuclei (blue). Larvae were kept constantly at 25°C. Scale bar: 50 μm. (H, I) bantamΔ1-mutant clones marked by the absence of β-Gal (red and grey), showing the expression of the trbl-sensor in (H) (green and grey), and showing the expression of the trbl-mut-sensor in (I) (green and grey). DAPI was used to label the nuclei (blue). To facilitate the analysis, the clones were induced in a Minute/+ background to allow the recovery of bigger clones. Scale bar: 50 μm. (J, K) wtsx1-mutant clones marked by the absence of RFP (red and grey), showing the expression of the trbl-sensor (green and grey). DAPI was used to label the nuclei (blue). (K) Shows a magniﬁcation of the wing disc shown in (J). Scale bar: 100 μm in (J) and 50 μm in (K). Figure 1. bantam represses trbl directly. (A) Illustration of trbl-sensor showing predicted bantam-binding sites. The transgene consists of a tubulin promoter expressing the mRNA for GFP fused to the trbl 39UTR. The mRNA is sensitive to regulation through regulatory elements in the trbl 39UTR. The trbl 39UTR includes two predicted binding sites for bantam (indicated as site 1 and 2). The predicted binding between the nucleotides of the trbl 39UTR and bantam is shown, and the predicted seed sequences are highlighted (connecting lines between the nucleotides and grey background). https://doi.org/10.26508/lsa.201900381 vol 2 \| no 4 \| e201900381 2 of 12 1975). The bantam mutant clones were marked by the absence of lacZ expression. bantam mutant cells up-regulated the trbl-sensor (Fig 1H). In contrast to that, the expression of the trbl-mut-sensor was not affected in bantam clones (Fig 1I), providing additional evidence that trbl regulation by bantam is direct. The trbl-sensor was not affected in +/+ control clones induced in a Minute/+ background (Fig S2). In summary, these observations indicate that trbl is a direct bantam target gene and show that the trbl 39UTR is sensitive, not only to bantam overexpression but also to endog- enous bantam. To validate a miRNA-target prediction, it is necessary to correlate the expression of miRNAs with those of their potential targets. miRNAs are posttranscriptional regulators and, to be functional, need to be co-expressed with their targets. To detect Trbl protein, we used a GFP-insertion in the trbl locus, resulting in a Trbl-GFP fusion protein (Mi(Mic)trbl-MI01025) (Venken et al, 2011). We monitored bantam expression by using a P element insertion (bantam-lacZ) that reports expression from the bantam locus (Herranz et al, 2012a). Interestingly, Trbl and bantam were both observed in the wing disc, indicating that the bioinformatical prediction might reveal a functional interaction (Fig S1). The Hippo pathway regulates the expression of bantam, which is central in mediating the Yki growth regulatory role (Nolo et al, 2006; Thompson & Cohen, 2006). Hippo activation results in Warts (Wts) phosphorylation, which acts with its cofactor Mats to phosphorylate and inactivate Yki. wts-mutant tissue up-regulates Yki and overgrows (reviewed in (Pan, 2010)). As observed in wing discs up-regulating bantam, wts mutant cells down-regulated the trbl-sensor (Fig 1J and K). The trbl 39UTR is, thus, sensitive to Hippo pathway activity. p g g Next, we tested whether bantam up-regulation affected Trbl levels. Trbl showed an accumulation in two rows of cells abutting the dorsal–ventral boundary in the anterior compartment of the wing primordia (Fig 1B). Interestingly, Trbl limits G2/M progression (Mata et al, 2000) and the cells showing the highest Trbl levels correspond to the precursors of the sensory organs in the adult wing margin, which are arrested in G2 (Johnston & Edgar, 1998). We made use of the binary Gal4:UAS system to manipulate bantam levels in the wing disc (Brand & Perrimon, 1993). patched-Gal4 (ptc- Gal4) is expressed in a stripe of cells adjacent to the anterior– posterior and perpendicular to the dorsal–ventral border (Fig S1). trbl antagonizes G2/M transition in the wing disc The conserved gene trbl was ﬁrst identiﬁed in Drosophila as a gene coordinating embryonic cell division and morphogenesis (Mata et al, 2000). In the wing disc, cells overexpressing trbl show longer G2/M phases and accumulate Cyclin E (Mata et al, 2000; Reis & Edgar, 2004). However, the consequences of down- regulating trbl in the wing disc and its effects in growth control and tumorigenesis are not yet determined. The ﬂuorescence ubiquitin cell cycle indicator method (FUCCI) is a two-color sensor that provides a readout of the cell cycle phase of each cell in a population (Sakaue-Sawano et al, 2008). Drosophila FUCCI (Fly-FUCCI) marks cells in the G1-phase in green, cells in the S-phase (DNA synthesis) in red, and cells in the G2-phase in yellow (Fig 2A) (Zielke et al, 2014). We combined the use of Fly-FUCCI with ﬂow cytometry–based cell cycle analysis to study the cell cycle dynamics in discs modulating trbl. Expression of UAS-Fly-FUCCI in normal discs showed cells labeled in red, green, and yellow, in- dicating the presence of proliferating cells (Fig 2B). trbl over- expression led to an increase in cells labeled in yellow, representing an accumulation of cells in G2 (Fig 2C). We used anti-phospho- Histone H3 (PH3) as a speciﬁc marker of cells in mitosis to study whether trbl overexpression affected the rate of cell proliferation of the wing disc. We detected a reduction in the number of PH3-positive cells in discs expressing UAS-trbl when compared with control discs (Fig 2E–G). These observations are consistent with previous reports (Mata et al, 2000; Reis & Edgar, 2004) and show that trbl overexpression limits cell proliferation by dampening G2/M progression. Overexpression of UAS-bantam in the ptc domain reduced Trbl levels, as compared with the surrounding normal cells (Fig 1C). We obtained comparable results using the EP3622 element (Fig S1), an EP line inserted in the bantam locus that, when combined with a Gal4 driver, directs the expression of bantam (Hipfner et al, 2002). Consistent with those observations, Trbl-GFP levels were increased in bantam mutants (Fig S1). These results, coupled with the identiﬁcation of bantam-binding sites in the trbl 39UTR and the observations that trbl and bantam are co-expressed, suggest that trbl is a potential bantam target gene. bantam represses trbl The trbl-mut-sensor has two nucleotides exchanged in each of the predicted seed sequences (highlighted in red). (B, C) Confocal micrographs of third-instar wing imaginal discs showing trbl-GFP expression (green and grey) in the genotypes indicated. DAPI was used to label the nuclei (blue). Larvae were kept constantly at 25°C. Scale bar: 50 μm. (D–G) Confocal micrographs of third-instar wing imaginal discs showing the expression of trbl-sensor (D, E) and trbl-mut-sensor (F, G) in the genotypes indicated. The sensors are shown in green and grey. DAPI was used to label the nuclei (blue). Larvae were kept constantly at 25°C. Scale bar: 50 μm. (H, I) bantamΔ1-mutant clones marked by the absence of β-Gal (red and grey), showing the expression of the trbl-sensor in (H) (green and grey), and showing the expression of the trbl-mut-sensor in (I) (green and grey). DAPI was used to label the nuclei (blue). To facilitate the analysis, the clones were induced in a Minute/+ background to allow the recovery of bigger clones. Scale bar: 50 μm. (J, K) wtsx1-mutant clones marked by the absence of RFP (red and grey), showing the expression of the trbl-sensor (green and grey). DAPI was used to label the nuclei (blue). (K) Shows a magniﬁcation of the wing disc shown in (J). Scale bar: 100 μm in (J) and 50 μm in (K). https://doi.org/10.26508/lsa.201900381 vol 2 \| no 4 \| e201900381 2 of 12 https://doi.org/10.26508/lsa.201900381 vol 2 \| no 4 \| e201900381 2 of 12 bantam, growth control and tumorigenesis Gerlach et al. 2 of 12 bantam, growth control and tumorigenesis Gerlach et al. https://doi.org/10.26508/lsa.201900381 vol 2 \| no 4 \| e201900381 3 of 12 trbl is a direct bantam target To assess whether bantam regulates trbl directly, we cloned the trbl 39UTR downstream of a tubulin promoter-EGFP reporter plasmid (trbl-sensor, Fig 1A). The trbl-sensor was expressed ubiquitously in the wing imaginal disc (Fig 1D). Consistent with our previous ob- servations, bantam overexpression in the ptc domain led to a robust reduction of the trbl-sensor (Fig 1E). To test whether trbl regulation by bantam occurred through the predicted binding sites, we generated a mutant version of the trbl-sensor (trbl-mut-sensor), carrying modiﬁcations in the both binding sites (illustrated in Fig 1A). Remarkably, the trbl-mut-sensor was not sensitive to bantam overexpression (Fig 1F and G). Overexpression of bantam using EP- bantam (EP3622) yielded similar results, yet the regulation over the trbl-sensor was lower in magnitude than the one observed when expressing UAS-bantam (Fig S1). Next, we evaluated the role of endogenous trbl in the wing disc. Depletion of trbl by expression of a UAS-trbl-RNAi transgene caused a reduction in the number of cells marked in yellow, which indicates that those cells show faster G2/M transition (Fig 2D). The efﬁciency of the UAS-trbl-RNAi transgene is shown in Fig S3. In sum, these results conﬁrm previous observation and reveal that endogenous trbl limits G2/M progression in the wing disc. Next, we studied whether the trbl 39UTR was sensitive to en- dogenous bantam levels. To perform that analysis, we generated bantam null mutant clones (FRT bantam-Δ1) in the wing disc. bantam mutant clones grow poorly and are eliminated from the wing epithelium presumably by cell competition (Thompson & Cohen, 2006). To bypass this issue, we provided the bantam mu- tant clones with a competitive advantage over neighboring cells by inducing the clones in a Minute/+ background (Morata & Ripoll, We used the adult wing to analyze how trbl affected cell number and tissue size. In the adult wing, each hair-like structure corresponds https://doi.org/10.26508/lsa.201900381 vol 2 \| no 4 \| e201900381 3 of 12 3 of 12 Figure 2. trbl phenotype in the wing disc. (A) Illustration of ﬂy-FUCCI system. Cells are labeled in green in late mitosis and the G1 phase, in red during the S-phase (DNA synthesis), and in yellow in the G2 phase and early mitosis. (B–D) Third-instar wing imaginal discs expressing the ﬂy-FUCCI system together with UAS-trbl (C) and UAS-trbl-RNAi (D). The indicated genes were expressed under the control of the ap-Gal4 driver. ap-Gal4:UAS-FUCCI was used as control (B). Scale bar: 10 μm. trbl is a direct bantam target Bottom panels show ﬂow cytometry cell cycle analysis by quantiﬁcation of DNA content in the genetic backgrounds shown in (B–D). Flow cytometry and ﬂy-FUCCI were combined and the corresponding proﬁles are shown in the bottom of each panel. In these proﬁles, 2C and 4C DNA contents are indicated, and the cells in G1 (green) and G2 (yellow) are highlighted. (E, F) Confocal micrographs of third-instar wing imaginal discs showing mitotic cells labeled by anti-PH3 (red and grey) in the following genotypes: ap-Gal4:UAS-GFP (E) and ap-Gal4:UAS-GFP; UAS-trbl (F). DAPI was used to label the nuclei (blue). Scale bars: 50 μm. (G) Quantiﬁcation of PH3-positive cells in third-instar wing imaginal discs. Examples of quantiﬁed area in (E, F). Statistical signiﬁcance was determined by unpaired t test (n = 10; P < 0.05). (H) Cuticle preparations of adult wings of the genotypes indicated. (I) Quantiﬁcation of adult wing size. Wing area is normalized to the mean of the control. Statistical signiﬁcance was determined by unpaired t test (n = 24 [MS1096>GFP], n = 20 [MS1096>trbl- RNAi], n = 24 [MS1096>trbl]; P < 0.05; *P < 0.001). (J) High-magniﬁcation image showing cuticle preparations of adult wings of the genotypes indicated. Note the reduction in cell density reﬂecting the presence of bigger cells in the discs overexpressing trbl (MS1096>trbl). (K) Quantiﬁcation of cell number in adult wings. Examples of quantiﬁed area in (J). Cell number is normalized to the mean of the control. Statistical signiﬁcance was determined by unpaired t test (n = 13 [MS1096>GFP], n = 12 [MS1096>trbl-RNAi], n = 15 [MS1096>trbl]; ns, not signiﬁcant; *P < 0.0001). Figure 2. trbl phenotype in the wing disc. (A) ll i f ﬂ CC C ll l g p yp g (A) Illustration of ﬂy-FUCCI system. Cells are labeled in green in late mitosis and the G1 phase, in red during the S-phase (DNA synthesis), and in yellow in the G2 phase and early mitosis. (B–D) Third-instar wing imaginal discs expressing the ﬂy-FUCCI system together with UAS-trbl (C) and UAS-trbl-RNAi (D). The indicated genes were expressed under the control of the ap-Gal4 driver. ap-Gal4:UAS-FUCCI was used as control (B). Scale bar: 10 μm. Bottom panels show ﬂow cytometry cell cycle analysis by quantiﬁcation of DNA content in the genetic backgrounds shown in (B–D). Flow cytometry and ﬂy-FUCCI were combined and the corresponding proﬁles are shown in the bottom of each panel. trbl is a direct bantam target Scale bars: 50 μm. (G) Quantiﬁcation of PH3-positive cells in third-instar wing imaginal discs. Examples of quantiﬁed area in (E, F). Statistical signiﬁcance was determined by unpaired t test (n = 10; P < 0.05). (H) Cuticle preparations of adult wings of the genotypes indicated. (I) Quantiﬁcation of adult wing size. Wing area is normalized to the mean of the control. Statistical signiﬁcance was determined by unpaired t test (n = 24 [MS1096>GFP], n = 20 [MS1096>trbl- RNAi], n = 24 [MS1096>trbl]; P < 0.05; P < 0.001). (J) High-magniﬁcation image showing cuticle preparations of adult wings of the genotypes indicated. Note the reduction in cell density reﬂecting the presence of bigger cells in the discs overexpressing trbl (MS1096>trbl). (K) Quantiﬁcation of cell number in adult wings. Examples of quantiﬁed area in (J). Cell number is normalized to the mean of the control. Statistical signiﬁcance was determined by unpaired t test (n = 13 [MS1096>GFP], n = 12 [MS1096>trbl-RNAi], n = 15 [MS1096>trbl]; ns, not signiﬁcant; *P < 0.0001). contributes to the growth-promoting role of this miRNA, we rea- soned that restoring trbl expression in discs expressing bantam should reduce the magnitude of tissue overgrowth. Although trbl overexpression did not overtly affect the size of the wing disc, bantam-induced overgrowth was reduced by expression of UAS- trbl in place of the UAS-lacZ control (Fig 3A–D and G). Taken to- gether with the ﬁnding that bantam regulates trbl, these results suggest that bantam acts, at least partially, through trbl to stimulate tissue growth. to a trichome originating from a single epithelial cell. Hence, trichome density can be used as a proxy of cell size and number. trbl depletion did not affect cell or wing size. In contrast, trbl overexpression caused a mild reduction in wing size. These wings showed an increase in cell size revealed by reduced trichome density, indicating that trbl-overexpressing wings had fewer cells (Figs 2H–K and S4). The increased cell size observed upon trbl up- regulation is characteristic of cells with delayed cell cycle pro- gression (Neufeld et al, 1998) and is consistent with the PH3 staining and the FUCCI-FACS analysis. Given that bantam is an essential Yki target mediating its growth regulatory role (Nolo et al, 2006; Thompson & Cohen, 2006), we reasoned that Trbl should also restrict the growth-promoting effect of Yki. trbl is a direct bantam target Indeed, co-expression of trbl with yki was sufﬁcient to limit the magnitude of tissue overgrowth induced by yki on its own (Fig 3E–G). It is worth noting that trbl expression in an otherwise normal background did not have a strong impact on the size of the wing trbl is a direct bantam target In these proﬁles, 2C and 4C DNA contents are indicated, and the cells in G1 (green) and G2 (yellow) are highlighted. (E, F) Confocal micrographs of third-instar wing imaginal discs showing mitotic cells labeled by anti-PH3 (red and grey) in the following genotypes: ap-Gal4:UAS-GFP (E) and ap-Gal4:UAS-GFP; UAS-trbl (F). DAPI was used to label the nuclei (blue). Scale bars: 50 μm. (G) Quantiﬁcation of PH3-positive cells in third-instar wing imaginal discs. Examples of quantiﬁed area in (E, F). Statistical signiﬁcance was determined by unpaired t test (n = 10; P < 0.05). (H) Cuticle preparations of adult wings of the genotypes indicated. (I) Quantiﬁcation of adult wing size. Wing area is normalized to the mean of the control. Statistical signiﬁcance was determined by unpaired t test (n = 24 [MS1096>GFP], n = 20 [MS1096>trbl- RNAi], n = 24 [MS1096>trbl]; P < 0.05; P < 0.001). (J) High-magniﬁcation image showing cuticle preparations of adult wings of the genotypes indicated. Note the reduction in cell density reﬂecting the presence of bigger cells in the discs overexpressing trbl (MS1096>trbl). (K) Quantiﬁcation of cell number in adult wings. Examples of quantiﬁed area in (J). Cell number is normalized to the mean of the control. Statistical signiﬁcance was determined by unpaired t test (n = 13 [MS1096>GFP], n = 12 [MS1096>trbl-RNAi], n = 15 [MS1096>trbl]; ns, not signiﬁcant; P < 0.0001). g p yp g (A) Illustration of ﬂy-FUCCI system. Cells are labeled in green in late mitosis and the G1 phase, in red during the S-phase (DNA synthesis), and in yellow in the G2 phase and early mitosis. (B–D) Third-instar wing imaginal discs expressing the ﬂy-FUCCI system together with UAS-trbl (C) and UAS-trbl-RNAi (D). The indicated genes were expressed under the control of the ap-Gal4 driver. ap-Gal4:UAS-FUCCI was used as control (B). Scale bar: 10 μm. Bottom panels show ﬂow cytometry cell cycle analysis by quantiﬁcation of DNA content in the genetic backgrounds shown in (B–D). Flow cytometry and ﬂy-FUCCI were combined and the corresponding proﬁles are shown in the bottom of each panel. In these proﬁles, 2C and 4C DNA contents are indicated, and the cells in G1 (green) and G2 (yellow) are highlighted. (E, F) Confocal micrographs of third-instar wing imaginal discs showing mitotic cells labeled by anti-PH3 (red and grey) in the following genotypes: ap-Gal4:UAS-GFP (E) and ap-Gal4:UAS-GFP; UAS-trbl (F). DAPI was used to label the nuclei (blue). https://doi.org/10.26508/lsa.201900381 vol 2 \| no 4 \| e201900381 4 of 12 bantam promotes tissue growth through trbl and hid We have shown that trbl and hid are important targets con- tributing to the growth-promoting role of bantam. Tumor formation appears to require both an increase in cell division and a sup- pression of apoptosis (Green & Evan, 2002). The apoptotic response observed in wing discs depleting trbl might serve as a homeostatic process to regulate organ size and prevent the formation of tissue hyperplasia. To assess that, we expressed the baculovirus protein p35 (Hay et al, 1994) or a miRNA that inhibits the proapoptotic genes reaper, hid, and grim (miR-RHG) (Siegrist et al, 2010), as alternative means to repress apoptosis. Remarkably, suppression of apoptosis in cells depleting trbl also led to tissue overgrowth (Fig S5). The use of an independent UAS-trbl-RNAi line produced comparable results (Fig S5). Similar results were obtained when we overexpressed cdc25-stg as an alternative method of forcing G2/M transition (Fig S6). In sum, these results suggest that apoptosis in response to trbl depletion can be used as a homeostatic mechanism to offset tissue overgrowth. The results shown previously suggest that trbl is a central bantam target in growth control. Although trbl knock down in the wing disc accelerated G2/M progression, it did not result in obvious tissue overgrowth (Fig 4A, B, and E). This suggests that trbl is not the only gene through which bantam induces tissue overgrowth. Changes in cell proliferation in the imaginal discs are typically associated with an increase in cell death (Karim & Rubin, 1998; Neufeld et al, 1998). In good agreement, the use of an antibody recognizing the activated form of caspase 3 revealed the presence of apoptotic cells upon trbl depletion (Fig 4F–H). bantam inhibits cell death by repressing the proapoptotic gene hid (Brennecke et al, 2003). As in the case of trbl, hid down-regulation did not affect the size of the wing disc (Fig 4C and E). We have shown that independent depletion of the bantam targets trbl or hid did not mimic the bantam gain of function phenotype. A cooperative interaction between both genes might mediate the growth-promoting role of bantam. We hypothesized that, to promote overall tissue growth, bantam could use a dual mechanism: on the one hand, it would accelerate G2/M transition by repressing trbl and, on the other hand, it would inhibit apo- ptosis by repressing the proapoptotic gene hid. trbl is involved in the growth regulatory role of bantam The fact that Trbl regulates the cell cycle in the wing disc indicates that this gene is a promising candidate mediating the role of bantam, inducing cell proliferation. If trbl repression by bantam https://doi.org/10.26508/lsa.201900381 vol 2 \| no 4 \| e201900381 4 of 12 bantam, growth control and tumorigenesis Gerlach et al. 4 of 12 Figure 3. trbl dampens bantam- and Yki-induced tissue growth. (A–F) Confocal micrographs of third-instar wing imaginal discs. The indicated genes were expressed under the control of the ap-Gal4 driver (dorsal compartment). The discs also expressed UAS-GFP to label the region where the corresponding transgenes were induced (green). DAPI was used to label the nuclei (blue). ap-Gal4:UAS-GFP was used as control (A). Scale bar: 100 μm. (G) Quantiﬁcation of GFP-positive area in third-instar wing imaginal discs. GFP-positive area is normalized to the mean of the control. Statistical signiﬁcance was determined by unpaired t test (n = 15; P < 0.01; *P < 0.001). imaginal tissue (Fig 3B and G), indicating that the growth-repressing function of trbl is speciﬁc to the Yki and bantam axis, and not a general consequence of trbl overexpression on normal tissue growth. miRNAs under Gal4 UAS-bantam-spong (Herranz et al, 201 UAS-bantam-spong d i i (Fi Figure 3. trbl dampens bantam- and Yki-induced tissue growth. (A–F) Confocal micrographs of third-instar wing imaginal discs. The indicated genes were expressed under the control of the ap-Gal4 driver (dorsal compartment). The discs also expressed UAS-GFP to label the region where the corresponding transgenes were induced (green). DAPI was used to label the nuclei (blue). ap-Gal4:UAS-GFP was used as control (A). Scale bar: 100 μm. (G) Quantiﬁcation of GFP-positive area in third-instar wing imaginal discs. GFP-positive area is normalized to the mean of the control. Statistical signiﬁcance was determined by unpaired t test (n = 15; P < 0.01; *P < 0.001). imaginal tissue (Fig 3B and G), indicating that the growth-repressing function of trbl is speciﬁc to the Yki and bantam axis, and not a general consequence of trbl overexpression on normal tissue growth. imaginal tissue (Fig 3B and G), indicating that the growth-repressing function of trbl is speciﬁc to the Yki and bantam axis, and not a general consequence of trbl overexpression on normal tissue growth. miRNAs under Gal4 control (Loya et al, 2009). The expression of a UAS-bantam-sponge transgene causes a reduction in tissue size (Herranz et al, 2012a). trbl is involved in the growth regulatory role of bantam Notably, discs expressing simultaneously UAS-bantam-sponge, UAS-trbl-RNAi, and UAS-hid-RNAi were in- creased in size (Fig 4I–K). This suggests that co-depletion of trbl and hid is sufﬁcient to drive tissue overgrowth, even when bantam levels are reduced. bantam promotes tissue growth through trbl and hid If this is true, replacing bantam expression by co-depletion of both targets should lead to a similar outcome. To test this hypothesis, we expressed simultaneously UAS-RNAi transgenes directed against both bantam targets: UAS-trbl-RNAi and UAS-hid-RNAi. Re- markably, concurrent depletion of trbl and hid led to the for- mation of overgrown imaginal discs (Fig 4D and E), resembling bantam overexpression. hid depletion did not completely rescue the induction of apoptosis observed in wing discs expressing UAS- trbl-RNAi (Fig S5), suggesting that other proapoptotic genes, in addition to hid, might contribute to the apoptotic response to trbl knock down. Next, we studied cell proliferation in discs co-expressing p35 and trbl-RNAi. We used proliferation markers such as 5-ethynyl- 20-deoxyuridine (EdU), which marks cells in the S-phase and PH3 to label cells in mitosis. This revealed an increase in the cell division rate in cells depleting trbl and expressing p35, as com- pared with the normal cells in the ventral compartment (GFP- negative) that served as internal control (Fig S5). In conclusion, these results reﬂect that simultaneous induction of G2/M and suppression of apoptosis causes tissue hyperproliferation and hyperplasia. bantam, growth control and tumorigenesis Gerlach et al. Trbl suppresses the oncogenic function of bantam EGFR is a well-recognized cancer driver. Overexpression of wild- type EGFR in Drosophila causes the activation of the MAPK pathway and tissue hyperplasia (Herranz et al, 2012a). Cooperating factors are required for the development of neoplastic tumors. bantam has been identiﬁed as one of such factors, whereby bantam inhibits the Next, we studied the consequences of down-regulating trbl and hid in discs with reduced bantam. miRNA sponges controlled by UAS sequences allow spatially controlled down-regulation of https://doi.org/10.26508/lsa.201900381 vol 2 \| no 4 \| e201900381 5 of 12 5 of 12 Figure 4. trbl and hid are key bantam targets in growth control. ( ) suppressor of cytokine signaling Socs36E, which limits JAK/STAT signaling and tumor development (Herranz et al, 2012b). We studied whether trbl also contributed to the bantam on- cogenic function. Indeed, trbl expression in EGFR + bantam tumors impaired tumor growth leading to the formation of discs compa- rable in size with the one observed in discs expressing EGFR on its own (Fig 5A–C and G). This suggests that trbl not only limits the bantam growth-promoting role, but also reduces its oncogenic effect. Next, we analyzed whether trbl down-regulation is sufﬁcient to cooperate with the oncogene EGFR in tumorigenesis. In- terestingly, although trbl depletion did not have a big impact in normal growth, trbl knock down in a context of EGFR overexpression fueled tumor growth (Fig 5D and G). These observations indicate that trbl behaves as a tumor suppressor in a context of EGFR expression. Next, we studied the consequences of down-regulating other G2/ M repressors in a context of EGFR activation. Activation of the Cdk1–Cyclin B complex induces cells to enter mitosis from G2. Myt1 is the main Cdk1 inhibitory kinase in the wing imaginal disc (Jin et al, 2008). Consistent with our observations in cells knocking down trbl, the expression of UAS-Myt1-RNAi accelerated G2/M transition and did not have an overt effect in disc size (Fig S7). Importantly, Myt1 depletion cooperated with EGFR in tumor formation (Fig 5E and G). Similar results were obtained when we drove G2/M transition by expressing cdc25-stg in discs up-regulating EGFR (Fig 5F and G). Together, this reveals that although deregulation in G2/M transition does not have a major impact in normal tissue growth, it can be sufﬁcient to trigger tumorigenesis in combination with the onco- gene EGFR. Gal4 to mark the dorsal compartment. The RFP area was used to determine the size of the dorsal compartment for “bantam-sponge” and “bantam-sponge+trbl- RNAi+hid-RNAi.” Sizes of the dorsal compartment are normalized to the mean of the control. Statistical signiﬁcance was determined by unpaired t test (n = 15; P < 0.0001). Larvae were kept constantly at 25°C. Trbl suppresses the oncogenic function of bantam Trbl, in addition to regulate cell cycle progression, limits growth by antagonizing the insulin receptor/phosphatidylinositol 3-kinase (PI3K)/AKT protein kinase pathway (Das et al, 2014; Hong et al, 2016). This signaling cascade is hyperactivated in a wide range of human cancers (Vivanco & Sawyers, 2002). Even though inducing G2/M appears to be sufﬁcient to cooperate with EGFR in tumorigenesis (Fig 5D–G), we cannot rule out that PI3K/AKT regulation by trbl might also contribute to its tumor suppressor function. Figure 4. trbl and hid are key bantam targets in growth control. ( ) Figure 4. trbl and hid are key bantam targets in growth control. (A–D) Confocal micrographs of third-instar wing imaginal discs. The indicated genes were expressed under the control of the ap-Gal4 driver (dorsal compartment). The discs also expressed UAS-GFP to label the region where the corresponding transgenes were induced (green). DAPI was used to label the nuclei (blue). ap-Gal4:UAS-GFP was used as control (A). Scale bar: 100 μm. (E) Quantiﬁcation of GFP-positive area in third-instar wing imaginal discs. GFP- positive area is normalized to the mean of the control. Statistical signiﬁcance was determined by unpaired t test (n = 15; *P < 0.0001). (F, G) Third-instar wing imaginal discs showing staining against Cas3-activated signals (red and grey). DAPI was used to label the nuclei (blue). The indicated genes were expressed under the control of the ap-Gal4 driver (dorsal compartment). The discs also expressed UAS-GFP to label the region where the corresponding transgenes were induced (green). ap-Gal4:UAS-GFP was used as control (F). Larvae were kept constantly at 25°C. Scale bar: 50 μm. (H) Quantiﬁcation of Cas3-activated signal in the wing pouch of third-instar wing imaginal discs. The signal is represented by the mean grey value. Statistical signiﬁcance was determined by unpaired t test (n = 10; P < 0.05). Larvae were kept constantly at 25°C. (I, J) Third-instar wing imaginal discs expressing UAS-bantam-sponge-RFP (I) and UAS-bantam-sponge- RFP, UAS-trbl-RNAi, and UAS-hid-RNAi (J), under the control of the ap-Gal4 driver (dorsal compartment). Cells expressing bantam-sponge-RFP are labeled in red. DAPI was used to label the nuclei (blue). Larvae were kept constantly at 25°C. Scale bar: 100 μm. (K) Quantiﬁcation of the dorsal compartment size in third- instar wing imaginal discs. The control expressed GFP under the control of ap- bantam, growth control and tumorigenesis Gerlach et al. Trbl tumor suppressive role in other tumor contexts DAPI was used to label the nuclei (blue). Scale bar: 100 μm. (N) Quantiﬁcation of GFP-positive area in third-instar wing imaginal discs. GFP-positive area is normalized to the mean of the control. discs with reduced pnut (Gerlach et al, 2018). The oncogenic role of Yki in this context appears to be conserved and expression of YAP, the human ortholog of yki, also triggers tumorigenesis in human cells with cytokinesis failure (Ganem et al, 2014). Although trbl knock down accelerates G2/M progression, this does not produce obvious changes in tissue size. Two mechanisms might ex- plain this apparent discrepancy. (1) Cells in the wing disc compensate for perturbations in the length of speciﬁc phases of the cell cycle by altering the duration of the other phases. This serves as a mechanism to keep normal division rates of cell proliferation, even when certain phases of the cycle are affected (Reis & Edgar, 2004). Faster G2/M progression, as a consequence of trbl depletion, would slow down G1/ S to maintain a normal rate of cell division. (2) Perturbations in the cell cycle are normally accompanied by an increase in cell death (Karim & Rubin, 1998; Neufeld et al, 1998). Consistently, apoptosis can be de- tected in discs with reduced trbl, which might serve as a mechanism to prevent tissue overgrowth and tumor formation. Furthermore, knock down of the proapoptotic bantam target hid, combined with trbl depletion, induces tissue hyperplasia. Our results establish the min- imal combination of bantam targets that, when depleted, drives tissue overgrowth. Tumors commonly combine increased rates of cell di- vision with suppression of apoptosis (Green & Evan, 2002). We provide a new example, whereby defects in the cell cycle and apoptosis lead to tissue hyperplasia. In addition to trbl and hid, bantam represses other negative growth regulators (Brennecke et al, 2003; Herranz & Cohen, 2010; Herranz et al., 2012a, 2012b). Although individual depletion of those elements does not result in tissue hyperplasia, we cannot ex- clude them as important bantam targets in growth control that, to- gether with trbl and hid, might contribute to the growth-promoting role of bantam. The fact that bantam is a central Yki target led us to analyze whether trbl regulation contributes to tumorigenesis in cells knocking down pnut. Wing discs expressing UAS-pnut-RNAi activate the c-Jun N-terminal kinase (JNK) pathway, which mediates their elimination by apoptosis. Trbl tumor suppressive role in other tumor contexts Cytokinesis failure leads to the formation of tetraploid cells. This generates an unstable situation that can lead to aneuploidy and tumor formation after cell division (Ganem et al, 2007). We have recently developed a genetic model in the wing imaginal disc of Drosophila to study the oncogenic potential of cytokinesis failure. The Drosophila gene peanut (pnut) belongs to the Septin family of proteins and is required for cytokinesis (Neufeld & Rubin, 1994). RNAi-mediated depletion of pnut in the wing disc causes cytoki- nesis failure that leads to apoptosis and growth defects. We have reported that the oncogene yki is able to induce tumorigenesis in https://doi.org/10.26508/lsa.201900381 vol 2 \| no 4 \| e201900381 6 of bantam, growth control and tumorigenesis Gerlach et al. 6 of 12 Figure 5. Tumor suppression by trbl. (A–F) Confocal micrographs of third-instar wing imaginal discs. The indicated genes were expressed under the control of the ap-Gal4 driver (dorsal compartment). The discs also expressed UAS-GFP to label the region where the corresponding transgenes were induced (green). DAPI was used to label the nuclei (blue). ap-Gal4:UAS-GFP was used as control (A). Scale bar: 100 μm. (G) Quantiﬁcation of GFP-positive area in third-instar wing imaginal discs. GFP-positive area is normalized to the mean of the control. (H–M) Confocal micrographs of third-instar wing imaginal discs. The indicated genes were expressed under the control of the ap-Gal4 driver (dorsal compartment). The discs also expressed UAS-GFP to label the region where the corresponding transgenes were induced (green). DAPI was used to label the nuclei (blue). Scale bar: 100 μm. (N) Quantiﬁcation of GFP-positive area in third-instar wing imaginal discs. GFP-positive area is normalized to the mean of the control. Figure 5. Tumor suppression by trbl. (A–F) Confocal micrographs of third-instar wing imaginal discs. The indicated genes were expressed under the control of the ap-Gal4 driver (dorsal compartment). The discs also expressed UAS-GFP to label the region where the corresponding transgenes were induced (green). DAPI was used to label the nuclei (blue). ap-Gal4:UAS-GFP was used as control (A). Scale bar: 100 μm. (G) Quantiﬁcation of GFP-positive area in third-instar wing imaginal discs. GFP-positive area is normalized to the mean of the control. (H–M) Confocal micrographs of third-instar wing imaginal discs. The indicated genes were expressed under the control of the ap-Gal4 driver (dorsal compartment). The discs also expressed UAS-GFP to label the region where the corresponding transgenes were induced (green). Trbl tumor suppressive role in other tumor contexts However, those cells accumulate in G2 and suppression of apoptosis is not sufﬁcient to induce tumorigenesis (Gerlach et al, 2018). We tested whether trbl depletion could trigger tumorigenesis in cells depleting pnut and protected against apoptosis. Remarkably, simultaneous expression of pnut-RNAi, p35, and trbl-RNAi induced the formation of disorganized organs resembling tumors (Fig 5H–N). Comparable results were obtained when we expressed cdc25-stg as an alternative means of driving G2/M transition (Fig 5M and N). The expression of p35 on its own did not affect tissue growth (Fig 5K and N). In conclusion, weprovide evidence showingthattrbldown-regulation fuels tumorigenesis in two independent scenarios: EGFR-driven tumors, and discs depleting pnut and protected against apoptosis. Figure 6. Model illustrating the growth-promoting role of Yki and bantam in the wing epithelium. Yki promotes growth by inducing cell cycle progression and repressing apoptosis. Yki regulates G2/M progression directly by inducing the expression of its target gene cdc25-stg. Likewise, it limits apoptosis by inducing the expression of the repressor of apoptosis DIAP1. In parallel, Yki induces the expression of the miRNA bantam, which, in turn, represses trbl and hid. trbl acts as an inhibitor of cdc25-stg, and hid represses DIAP1. Yki triggers an FFL controlling cell cycle progression and apoptosis directly, and indirectly through bantam. The combination of these factors leads to tissue growth. The repressor of cytokine signaling, Socs36E, is one of the bantam target genes identiﬁed so far. bantam cooperates with EGFR in the formation of metastatic tumors, and Socs36E plays a central role in that process. Socs36E down-regulation causes an increase in JAK/STAT pathway activity that, in combination with EGFR, drives malignancy (Herranz et al, 2012b; Hombria & Serras, 2013). SOCS5 is the Socs36E human ortholog and, as observed in ﬂies, SOCS5 acts as a tumor suppressor in EGFR/RAS–dependent cell transformation assays (Herranz et al, 2012b). Subsequent an- alyses found that the transforming activity of oncogenic Ras-V12 relies on its ability to down-regulate SOCS5/6 in human cell lines (Hong et al, 2014). The results presented here show that trbl limits the magnitude of growth of tumors co-expressing EGFR and ban- tam, and trbl down-regulation potentiates EGFR-driven tumor formation. This effect does not appear to be unique for EGFR-driven tumors and trbl is also required to limit tumorigenesis in cells with defective cytokinesis. Cytokinesis failure can be tumorigenic, and it is proposed that around 40% of all human cancers have undergone a genome duplication event during their evolution (Zack et al, 2013). In Drosophila, cells with cytokinesis failure are eliminated by apo- ptosis as a consequence of JNK activation. Suppression of apoptosis is not sufﬁcient to promote tumorigenesis as those cells accu- mulate in G2 and proliferate poorly (Gerlach et al, 2018). Additional signals are required to induce proliferation and tumorigenesis. Here, we show that trbl is one of those elements, and trbl depletion sufﬁced to trigger tumorigenesis in cells with cytokinesis failure protected against apoptosis. trbl down-regulation is sufﬁcient to drive tumorigenesis in wing discs expressing the oncogene EGFR. However, discs with cleavage defects also require suppression of apoptosis for tumor formation. Interestingly, the EGFR pathway promotes cell survival through hid down-regulation by Ras (Bergmann et al, 1998; Kurada & White, 1998). The role of EGFR in cell survival might explain why EGFR- driven tumors are not dependent on additional blockage of apoptosis, whereas discs down-regulating pnut and trbl require suppression of apoptosis for tumor formation. Studying whether Trbl plays similar roles in human tumor models might merit further investigation. Figure 6. Model illustrating the growth-promoting role of Yki and bantam in the wing epithelium. Previous studies have shown that different growth regulators such as Ras, dMyc, and dPI3K control tissue growth by regulating G1/S progression (Johnston et al, 1999; Weinkove et al, 1999; Prober & Edgar, 2000, 2002). Yki also controls G1/S progression by regu- lating the expression of genes, including Cyclin E, E2F, and dMyc (Wu et al, 2003; Goulev et al, 2008; Neto-Silva et al, 2010). In addition to that, Yki drives G2/M progression by regulating the transcription of the phosphatase cdc25-stg (Gerlach et al, 2018). miRNAs are typi- cally embedded in regulatory loops to reduce noise and provide robustness to biological systems (Herranz & Cohen, 2010). Our results suggest the presence of a feed forward loop (FFL) down- stream of Yki that includes the miRNA bantam and modulates G2/M transition. FFLs involve at least three elements: a regulator, X (Yki), which regulates Y (bantam), and gene Z (cdc25-stg), which is regulated by both X (Yki) and Y (bantam) (Fig 6). Both branches act in the same direction on Z (cdc25-stg), hence representing a coherent FFL (Alon, 2007). Yki drives tissue growth by induc- ing proliferation and repressing apoptosis. Interestingly, a Yki promotes growth by inducing cell cycle progression and repressing apoptosis. Yki regulates G2/M progression directly by inducing the expression of its target gene cdc25-stg. Likewise, it limits apoptosis by inducing the expression of the repressor of apoptosis DIAP1. In parallel, Yki induces the expression of the miRNA bantam, which, in turn, represses trbl and hid. trbl acts as an inhibitor of cdc25-stg, and hid represses DIAP1. Yki triggers an FFL controlling cell cycle progression and apoptosis directly, and indirectly through bantam. The combination of these factors leads to tissue growth. Discussion Here, we identify the cell cycle regulator trbl as a direct bantam target gene involved in the growth regulatory role of this miRNA. https://doi.org/10.26508/lsa.201900381 vol 2 \| no 4 \| e201900381 7 of 12 bantam, growth control and tumorigenesis Gerlach et al. bantam, growth control and tumorigenesis Gerlach et al. https://doi.org/10.26508/lsa.201900381 vol 2 \| no 4 \| e201900381 8 of 12 Transgene expression in the wing imaginal disc The apterous and patched enhancers were used to express Gal4 in the wing imaginal disc. To prevent early lethality due to transgene expression, the tubulin-Gal80TS construct was used for gene manipulation experiments. It allows the control of the timing of UAS transgene expression in a temperature dependent manner (Zeidler et al, 2004). Drosophila crosses were kept for 2 days at 18°C to lay eggs and larvae were kept for ﬁve additional days at 18°C. After that, the larvae were transferred to 29°C to induce transgene expression. Only third-instar wandering larvae were dissected to ensure that the animals analyzed were in the same developmental stage. Figure legends indicate when the timing and temperature differed from this procedure. Genetic mosaics The FLP/FRT technique (Xu & Rubin, 1993) was used to generate bantam and wts loss of function clones. Second-instar larvae were heat-shocked for 1 h at 37°C. Next, the larvae were kept at 25°C. Clones were visualized by either loss of RFP or β-Gal expression. Drosophila strains The following Drosophila strains are described in the cited refer- ences: ap-Gal4 and nub-Gal4 (Calleja et al, 1996); ptc-Gal4 (Wilder & Perrimon, 1995); MS1096-Gal4 (Capdevila & Guerrero, 1994); bantam- Δ1, UAS-bantam-A, and UAS-bantam-D (Brennecke et al, 2003); EP3622 (Hipfner et al, 2002); UAS-miR-RHG (Siegrist et al, 2010); UAS- ﬂy-FUCCI (Zielke et al, 2014); UAS-yki (Huang et al, 2005); UAS-EGFR (Buff et al, 1998); UAS-bantam-sponge and bantam-lacZ (P{lacW} banL1170a) (Herranz et al, 2012a); FRT-wtsX1 (Xu et al, 1995). Other stocks are described in FlyBase. Immunohistochemistry The following primary antibodies were used: anti-ß-Gal (De- velopmental Studies Hybridoma Bank; 40-1a; mouse; 1:20); anti- Cas3-act (no. 9661; rabbit; 1:100; Cell Signaling Technology); anti-PH3 (no. 9701; rabbit; 1:100; Cell Signaling Technology). The following secondary antibodies were used: antimouse Alexa Fluor 555 (A- 21425; goat; 1:400; Invitrogen); antimouse Alexa Fluor 647 (A-21237; goat; 1:400; Invitrogen); anti-rabbit Alexa Fluor 555 (A-21430; goat; 1: 400; Invitrogen); and antirabbit Alexa Fluor 647 (A-21246; goat; 1:400; Invitrogen). The following stocks were provided by the Bloomington Dro- sophila Stock Center (BDSC): UAS-p35 (#6298); UAS-stg (#56562); UAS-pnut-RNAi (#11791); UAS-Myt1-RNAi (#62892); and trbl-GFP (#61654). The following stocks were provided by the Vienna Drosophila RNAi Center (VDRC): UAS-trbl-RNAi (#22114); UAS-hid-RNAi (#8269); and UAS-trbl-RNAi-KK (#106774). Third-instar larvae were dissected in PBS. Fixation of the sample was performed with 4% formaldehyde solution for 20 min at room temperature. After that, the samples were washed three times for 10 min in phosphate-buffered saline-tween (PBT) and blocked for 20 min in PBT-BSA (BBT). The ﬁrst antibody diluted in BBT was in- cubated with the samples at room temperature overnight followed by three 15-min washes in BBT. The secondary antibody together with DAPI in BBT was incubated with the sample for 2 h at room temperature followed by four 15-min washes in PBT. The samples were mounted in 90% glycerol with PBS containing 0.05% propyl gallate. Images of wing imaginal discs were acquired with a Leica SP8 confocal laser-scanning microscope and processed with ImageJ and Adobe Photoshop CC. The complete list of genotypes can be found as Supplemental Data 1. trbl 39UTR sensor The trbl 39UTR was ampliﬁed from cDNA (DGRC clone RH69304) with the primers wt-fwd and wt-rev and cloned into the tub-EGFP pCaSpeR4 sensor vector as an XhoI–NotI fragment. We generated a mutation in the bantam 39 site, via PCR using the wt-fwd and 39- mt-rev primers, exchanging TC with AG. To mutate the bantam 59 site (located more upstream in the trbl 39UTR), we performed two PCRs with (1) the wt-fwd and 59mt-rev primers, exchanging AT with GC and (2) the 59mt-fwd and wt-rev primers, again exchanging AT with GC. We generated the full trbl 39UTR with the bantam 59 site mutated in a subsequent PCR using these overlapping fragments as templates together with the wt-fwd and wt-rev primers. Using the same strategy but combining the wt-fwd- with the 39-mt-rev primers, we generated the trbl 39UTR with both bantam sites mutated. The following primers were used: activation of both arms in the FFL, resulting in an excess of pro- liferation and reduced apoptosis and consequently tumor for- mation. In fact, these kinds of loops have also been reported operating in human cancer (Iliopoulos et al, 2009). wt-fwd: 59-GACTgcggccgcTTGGAGCTCGTGGAGTCACCC wt-rev: 59-TCAGctcgagCTTGGTGAGATCAAATTCCAATGTATTATGTTC TTT 39-mt-rev: 59-TCAGctcgagCTTGGTGActTCAAATTCCAATGTATTATGT TCTTTTTTGCAATTTTCACTT 59mt-rev: 5'-CGACTGTCCTTTGATAAGgcCAACAGTTTGCAAAAAAATTA AAATTAAAATGTTAAATGT 59mt-fwd: 5'-TGCAAACTGTTGgcCTTATCAAAGGACAGTCGCCC. similar coherent FFL involving bantam is present in the “anti- apoptotic branch.” Yki represses apoptosis directly by inducing the expression of DIAP1, and indirectly by inducing the expression of bantam that in turn represses the proapoptotic gene hid (Fig 6). These types of circuits can be used to reduce noise and ensure developmental decisions leading to an “all-or-none” outcome (Herranz & Cohen, 2010). This type of loop can serve to drive tu- morigenesis, whereby sustained Yki activity would lead to the https://doi.org/10.26508/lsa.201900381 vol 2 \| no 4 \| e201900381 8 of 12 bantam, growth control and tumorigenesis Gerlach et al. 8 of 12 bantam, growth control and tumorigenesis Gerlach et al. Quantiﬁcation of cells in mitosis Third-instar wing imaginal discs were stained against PH3, which speciﬁcally labels mitotic cells. Images were acquired of the apical part of the wing disc with a Leica SP8 confocal laser-scanning microscope. An area of 250 × 250 pixels was used to count PH3- positive nuclei. Unpaired t test was used to determine statistical signiﬁcance, and graphs were generated with GraphPad Prism 7. Adult wing size and cell number quantiﬁcation Eppendorf Centrifuge 5424, Rotor FA-45-24-11), and the supernatant used further. The BCA protein assay (Thermo Fisher Scientiﬁc) was used to determine the protein concentration. 10 μg of protein was loaded per sample for the Western blots. Antibody against GFP (ab13970, chicken; Abcam) was used to detect Trbl–GFP fusion protein. Antibody against Actin (A5441, mouse; Sigma-Aldrich) was used as a control. Secondary HRP goat antichicken (A16054; Invi- trogen) and goat antimouse (P0447; Dako) antibodies were used together with the Pierce ECL Western Blotting Substrate (Thermo Fisher Scientiﬁc) to visualize protein levels. The MS1096 and nubbin enhancers were used to express Gal4 in the wing. Adult ﬂies were ﬁxed in glycerol-ethanol (25% glycerol, 75% ethanol) for three days and the wings were mounted in glycerol for imaging. Wing images of adult ﬂies were taken with a Leica DM 5500B bright-ﬁeld microscope. Tissue area measurement was performed with ImageJ. Bristle counting to determine cell density was obtained from high-magniﬁcation pictures of the region im- mediately anterior to the cross vein between veins four and ﬁve. The number of bristles in a 480 × 170 pixels area was quantiﬁed, and unpaired t test was used to determine statistical signiﬁcance. Graphs were generated with GraphPad Prism 7. Flow cytometry Dissection of wing imaginal discs of third-instar larvae was per- formed in PBS and trypsin–EDTA solution was used to dissociate the sample for 45 min. Cells were ﬁxed with 4% formaldehyde solution for 20 min and stored in 70% ethanol for 2 h. The samples were incubated with DAPI in PBT overnight at 4°C. Flow cytometry was used to measure ﬂuorescence of DAPI, GFP, and RFP in cells with a BD FACSAria Fusion. Doublet cells were excluded with an integral/ peak dot plot of DAPI ﬂuorescence. https://doi.org/10.26508/lsa.201900381 vol 2 \| no 4 \| e201900381 9 of 12 bantam, growth control and tumorigenesis Gerlach et al. 9 of 12 Conﬂict of Interest Statement The authors declare that they have no conﬂict of interest. Wing imaginal disc size quantiﬁcation Eggs were collected on apple agar plates with yeast paste at 25°C for 24 h and larvae were kept on the plates for another 24 h. A maximum of 30 larvae were transferred to each vial with standard Drosophila food and stored for 4 days on 18°C. Then, the vials were switched to 29°C to activate transgene expression. Larvae were dissected as third-instar wandering larvae to ensure same de- velopmental timing. Images were acquired with a Leica SP8 con- focal laser-scanning microscope. The option “Threshold” was used to mark the whole GFP-positive area in ImageJ. After this, the option “Analyze Particles” was used to determine the area size. Unpaired t test was used to determine statistical signiﬁcance and graphs were generated with GraphPad Prism 7. Figure legends indicate when the timing, temperature, or the ﬂuorophore used for the area quan- tiﬁcation differed from this procedure. Author Contributions SU Gerlach: conceptualization, data curation, formal analysis, validation, investigation, visualization, methodology, and wri- ting—original draft, review, and editing. Acknowledgements We thank Carlos Estella for reagents; Melissa Alexandra Visser for technical support; Jan Michael Kugler for technical guidance; Olive Ben´et Kersey for comments on the manuscript; and the DSHB, VDRC, and BDSC for antibodies and ﬂy strains. This work was supported by Novo Nordisk Foundation (grant number NNF0052223), a grant by the Neye Foundation for genetic models for cancer gene discovery, and a grant by Læge Sofus Carl Emil Friis og Hustru Olga Doris Friis’ Legat. EdU incorporation Third-instar larvae were dissected in PBS and incubated in 300 mM EdU in PBS for 30 min. Fixation of the sample was performed with 4% formaldehyde solution for 20 min at room temperature. Fol- lowing, the samples were washed three times for 10 min in PBT and blocked for 20 min in BBT. The Click-iT EdU Alexa Fluor 488 Imaging Kit (Invitrogen) was used, and the samples were incubated for 1 h with the reaction mix provided by the Imaging Kit followed by one BBT wash for 10 min. The samples were incubated with DAPI in BBT for 2 h at room temperature followed by four 15-min washes with PBT. The samples were mounted in 90% glycerol with PBS con- taining 0.05% propyl gallate. Quantiﬁcation of apoptosis M Sander: data curation, formal analysis, investigation, method- ology, and writing—review and editing. Third-instar wing imaginal discs were stained against Cas3-act, which marks cells undergoing apoptosis. Images were acquired of the basal part of the wing disc with a Leica SP8 confocal laser- scanning microscope. An area of 250 × 250 pixels in the wing pouch was used to quantify the signal. The overall intensity of Cas3-act per genotype was quantiﬁed by using the measurement “Mean Grey Value” in ImageJ. Unpaired t test was used to determine statistical signiﬁcance, and graphs were generated with GraphPad Prism 7. S Song: conceptualization, formal analysis, investigation, visuali- zation, and methodology. H Herranz: conceptualization, resources, supervision, funding ac- quisition, investigation, methodology, project administration, and writing—original draft, review, and editing. Becam I, Rafel N, Hong X, Cohen SM, Milan M (2011) Notch-mediated repression of bantam miRNA contributes to boundary formation in the Drosophila wing. Development 138: 3781–3789. doi:10.1242/ dev.064774 Herranz H, Eichenlaub T, Cohen SM (2016) Cancer in Drosophila: Imaginal discs as a model for epithelial tumor formation. Curr Top Dev Biol 116: 181–199. doi:10.1016/bs.ctdb.2015.11.037 Herranz H, Hong X, Cohen SM (2012a) Mutual repression by bantam miRNA and Capicua links the EGFR/MAPK and Hippo pathways in growth control. Curr Biol 22: 651–657. doi:10.1016/j.cub.2012.02.050 Bergmann A, Agapite J, McCall K, Steller H (1998) The Drosophila gene hid is a direct molecular target of Ras-dependent survival signaling. Cell 95: 331–341. doi:10.1016/s0092-8674(00)81765-1 Herranz H, Hong X, Hung NT, Voorhoeve PM, Cohen SM (2012b) Oncogenic cooperation between SOCS family proteins and EGFR identiﬁed using a Drosophila epithelial transformation model. Genes Dev 26: 1602–1611. doi:10.1101/gad.192021.112 Brand AH, Perrimon N (1993) Targeted gene expression as a means of altering cell fates and generating dominant phenotypes. Development 118: 401–415. Brennecke J, Hipfner DR, Stark A, Russell RB, Cohen SM (2003) bantam encodes a developmentally regulated microRNA that controls cell proliferation and regulates the proapoptotic gene hid in Drosophila. Cell 113: 25–36. doi:10.1016/s0092-8674(03)00231-9 Herranz H, Hong X, Perez L, Ferreira A, Olivieri D, Cohen SM, Milan M (2010) The miRNA machinery targets Mei-P26 and regulates Myc protein levels in the Drosophila wing. EMBO J 29: 1688–1698. doi:10.1038/ emboj.2010.69 Buff E, Carmena A, Gisselbrecht S, Jimenez F, Michelson AM (1998) Signalling by the Drosophila epidermal growth factor receptor is required for the speciﬁcation and diversiﬁcation of embryonic muscle progenitors. Development 125: 2075–2086. Hipfner DR, Weigmann K, Cohen SM (2002) The bantam gene regulates Drosophila growth. Genetics 161: 1527–1537. Hombria JC, Serras F (2013) Why should we care about ﬂy tumors?: The case of JAK-STAT and EGFR cooperation in oncogenesis. JAKSTAT 2: e23203. doi:10.4161/jkst.23203 Calleja M, Moreno E, Pelaz S, Morata G (1996) Visualization of gene expression in living adult Drosophila. Science 274: 252–255. doi:10.1126/ science.274.5285.252 Hong SH, Kang M, Lee KS, Yu K (2016) High fat diet-induced TGF-beta/Gbb signaling provokes insulin resistance through the tribbles expression. Sci Rep 6: 30265. doi:10.1038/srep30265 Capdevila J, Guerrero I (1994) Targeted expression of the signaling molecule decapentaplegic induces pattern duplications and growth alterations in Drosophila wings. EMBO J 13: 4459–4468. doi:10.1002/j.1460- 2075.1994.tb06768.x Hong X, Nguyen HT, Chen Q, Zhang R, Hagman Z, Voorhoeve PM, Cohen SM (2014) Opposing activities of the Ras and Hippo pathways converge on regulation of YAP protein turnover. EMBO J 33: 2447–2457. doi:10.15252/ embj.201489385 Carthew RW, Agbu P, Giri R (2017) MicroRNA function in Drosophila melanogaster. Semin Cell Dev Biol 65: 29–37. doi:10.1016/ j.semcdb.2016.03.015 Huang J, Wu S, Barrera J, Matthews K, Pan D (2005) The Hippo signaling pathway coordinately regulates cell proliferation and apoptosis by inactivating Yorkie, the Drosophila Homolog of YAP. Cell 122: 421–434. doi:10.1016/j.cell.2005.06.007 Conlon I, Raff M (1999) Size control in animal development. Cell 96: 235–244. doi:10.1016/s0092-8674(00)80563-2 Das R, Sebo Z, Pence L, Dobens LL (2014) Drosophila tribbles antagonizes insulin signaling-mediated growth and metabolism via interactions with Akt kinase. PLoS One 9: e109530. doi:10.1371/journal.pone.0109530 Iliopoulos D, Hirsch HA, Struhl K (2009) An epigenetic switch involving NF- kappaB, Lin28, Let-7 MicroRNA, and IL6 links inﬂammation to cell transformation. Cell 139: 693–706. doi:10.1016/j.cell.2009.10.014 Esquela-Kerscher A, Slack FJ (2006) Oncomirs: microRNAs with a role in cancer. Nat Rev Cancer 6: 259–269. doi:10.1038/nrc1840 Jin Z, Homola E, Tiong S, Campbell SD (2008) Drosophila myt1 is the major cdk1 inhibitory kinase for wing imaginal disc development. Genetics 180: 2123–2133. doi:10.1534/genetics.108.093195 Ganem NJ, Cornils H, Chiu SY, O’Rourke KP, Arnaud J, Yimlamai D, Thery M, Camargo FD, Pellman D (2014) Cytokinesis failure triggers hippo tumor suppressor pathway activation. Cell 158: 833–848. doi:10.1016/ j.cell.2014.06.029 Johnston LA, Edgar BA (1998) Wingless and Notch regulate cell-cycle arrest in the developing Drosophila wing. Nature 394: 82–84. doi:10.1038/ 27925 Ganem NJ, Storchova Z, Pellman D (2007) Tetraploidy, aneuploidy and cancer. Curr Opin Genet Dev 17: 157–162. doi:10.1016/j.gde.2007.02.011 Johnston LA, Prober DA, Edgar BA, Eisenman RN, Gallant P (1999) Drosophila myc regulates cellular growth during development. Cell 98: 779–790. doi:10.1016/s0092-8674(00)81512-3 Gerlach SU, Eichenlaub T, Herranz H (2018) Yorkie and JNK control tumorigenesis in Drosophila cells with cytokinesis failure. Cell Rep 23: 1491–1503. doi:10.1016/j.celrep.2018.04.006 Karim FD, Rubin GM (1998) Ectopic expression of activated Ras1 induces hyperplastic growth and increased cell death in Drosophila imaginal tissues. Development 125: 1–9. Gonzalez C (2013) Drosophila melanogaster: A model and a tool to investigate malignancy and identify new therapeutics. Nat Rev Cancer 13: 172–183. doi:10.1038/nrc3461 Kurada P, White K (1998) Ras promotes cell survival in Drosophila by downregulating hid expression. Cell 95: 319–329. References 15 third-instar larvae were transferred into protein extraction and immunoprecipitation buffer (RIPA; Sigma-Aldrich), cOmplete Pro- tease Inhibitor Cocktail (Roche) added, and the larvae homoge- nized. The homogenate was centrifuged (15 min, 13,000 rpm, Alon U (2007) Network motifs: Theory and experimental approaches. Nat Rev Genet 8: 450–461. doi:10.1038/nrg2102 https://doi.org/10.26508/lsa.201900381 vol 2 \| no 4 \| e201900381 10 of 12 https://doi.org/10.26508/lsa.201900381 vol 2 \| no 4 \| e201900381 10 of 12 bantam, growth control and tumorigenesis Gerlach et al. Neufeld TP, de la Cruz AF, Johnston LA, Edgar BA (1998) Coordination of growth and cell division in the Drosophila wing. Cell 93: 1183–1193. doi:10.1016/s0092-8674(00)81462-2 Vivanco I, Sawyers CL (2002) The phosphatidylinositol 3-Kinase AKT pathway in human cancer. Nat Rev Cancer 2: 489–501. doi:10.1038/nrc839 Weinkove D, Neufeld TP, Twardzik T, Waterﬁeld MD, Leevers SJ (1999) Regulation of imaginal disc cell size, cell number and organ size by Drosophila class I(A) phosphoinositide 3-kinase and its adaptor. Curr Biol 9: 1019–1029. doi:10.1016/s0960-9822(99)80450-3 Neufeld TP, Rubin GM (1994) The Drosophila peanut gene is required for cytokinesis and encodes a protein similar to yeast putative bud neck ﬁlament proteins. Cell 77: 371–379. doi:10.1016/0092-8674(94)90152-x Neufeld TP, Rubin GM (1994) The Drosophila peanut gene is required for cytokinesis and encodes a protein similar to yeast putative bud neck ﬁlament proteins. Cell 77: 371–379. doi:10.1016/0092-8674(94)90152-x Wilder EL, Perrimon N (1995) Dual functions of wingless in the Drosophila leg imaginal disc. Development 121: 477–488. Nolo R, Morrison CM, Tao C, Zhang X, Halder G (2006) The bantam microRNA is a target of the hippo tumor-suppressor pathway. Curr Biol 16: 1895–1904. doi:10.1016/j.cub.2006.08.057 Wu S, Huang J, Dong J, Pan D (2003) hippo encodes a Ste-20 family protein kinase that restricts cell proliferation and promotes apoptosis in conjunction with salvador and warts. Cell 114: 445–456. doi:10.1016/ s0092-8674(03)00549-x Pan D (2010) The hippo signaling pathway in development and cancer. Dev Cell 19: 491–505. doi:10.1016/j.devcel.2010.09.011 Prober DA, Edgar BA (2000) Ras1 promotes cellular growth in the Drosophila wing. Cell 100: 435–446. doi:10.1016/s0092-8674(00)80679-0 Xu T, Rubin GM (1993) Analysis of genetic mosaics in developing and adult Drosophila tissues. Development 117: 1223–1237. Prober DA, Edgar BA (2002) Interactions between Ras1, dMyc, and dPI3K signaling in the developing Drosophila wing. Genes Dev 16: 2286–2299. doi:10.1101/gad.991102 Xu T, Wang W, Zhang S, Stewart RA, Yu W (1995) Identifying tumor suppressors in genetic mosaics: The Drosophila lats gene encodes a putative protein kinase. Development 121: 1053–1063. Reis T, Edgar BA (2004) Negative regulation of dE2F1 by cyclin-dependent kinases controls cell cycle timing. Cell 117: 253–264. doi:10.1016/s0092- 8674(04)00247-8 Zack TI, Schumacher SE, Carter SL, Cherniack AD, Saksena G, Tabak B, Lawrence MS, Zhsng CZ, Wala J, Mermel CH, et al (2013) Pan-cancer patterns of somatic copy number alteration. Nat Genet 45: 1134–1140. doi:10.1016/s0092- 8674(00)81764-x Goulev Y, Fauny JD, Gonzalez-Marti B, Flagiello D, Silber J, Zider A (2008) SCALLOPED interacts with YORKIE, the nuclear effector of the hippo tumor-suppressor pathway in Drosophila. Curr Biol 18: 435–441. doi:10.1016/j.cub.2008.02.034 Loya CM, Lu CS, Van Vactor D, Fulga TA (2009) Transgenic microRNA inhibition with spatiotemporal speciﬁcity in intact organisms. Nat Methods 6: 897–903. doi:10.1038/nmeth.1402 Green DR, Evan GI (2002) A matter of life and death. Cancer Cell 1: 19–30. doi:10.1016/s1535-6108(02)00024-7 Mata J, Curado S, Ephrussi A, Rorth P (2000) Tribbles coordinates mitosis and morphogenesis in Drosophila by regulating string/CDC25 proteolysis. Cell 101: 511–522. doi:10.1016/s0092-8674(00)80861-2 Hanahan D, Weinberg RA (2011) Hallmarks of cancer: The next generation. Cell 144: 646–674. doi:10.1016/j.cell.2011.02.013 Hariharan IK (2015) Organ size control: Lessons from Drosophila. Dev Cell 34: 255–265. doi:10.1016/j.devcel.2015.07.012 Morata G, Ripoll P (1975) Minutes: Mutants of drosophila autonomously affecting cell division rate. Dev Biol 42: 211–221. doi:10.1016/0012- 1606(75)90330-9 Hay BA, Wolff T, Rubin GM (1994) Expression of baculovirus P35 prevents cell death in Drosophila. Development 120: 2121–2129. Neto-Silva RM, de Beco S, Johnston LA (2010) Evidence for a growth- stabilizing regulatory feedback mechanism between Myc and Yorkie, the Drosophila homolog of Yap. Dev Cell 19: 507–520. doi:10.1016/ j.devcel.2010.09.009 Herranz H, Cohen SM (2010) MicroRNAs and gene regulatory networks: Managing the impact of noise in biological systems. Genes Dev 24: 1339–1344. doi:10.1101/gad.1937010 https://doi.org/10.26508/lsa.201900381 vol 2 \| no 4 \| e201900381 11 of 12 https://doi.org/10.26508/lsa.201900381 vol 2 \| no 4 \| e201900381 11 of 12 bantam, growth control and tumorigenesis Gerlach et al. 11 of 12 doi:10.1038/ng.2760 Sakaue-Sawano A, Kurokawa H, Morimura T, Hanyu A, Hama H, Osawa H, Kashiwagi S, Fukami K, Miyata T, Miyoshi H, et al (2008) Visualizing spatiotemporal dynamics of multicellular cell-cycle progression. Cell 132: 487–498. doi:10.1016/j.cell.2007.12.033 Zeidler MP, Tan C, Bellaiche Y, Cherry S, Hader S, Gayko U, Perrimon N (2004) Temperature-sensitive control of protein activity by conditionally splicing inteins. Nat Biotechnol 22: 871–876. doi:10.1038/nbt979 Siegrist SE, Haque NS, Chen CH, Hay BA, Hariharan IK (2010) Inactivation of both Foxo and reaper promotes long-term adult neurogenesis in Drosophila. Curr Biol 20: 643–648. doi:10.1016/j.cub.2010.01.060 Zielke N, Korzelius J, van Straaten M, Bender K, Schuhknecht GF, Dutta D, Xiang J, Edgar BA (2014) Fly-FUCCI: A versatile tool for studying cell proliferation in complex tissues. Cell Rep 7: 588–598. doi:10.1016/ j.celrep.2014.03.020 Thompson BJ, Cohen SM (2006) The Hippo pathway regulates the bantam microRNA to control cell proliferation and apoptosis in Drosophila. Cell 126: 767–774. doi:10.1016/j.cell.2006.07.013 Venken KJ, Schulze KL, Haelterman NA, Pan H, He Y, Evans-Holm M, Carlson JW, Levis RW, Spradling AC, Hoskins RA, et al (2011) MiMIC: A highly versatile transposon insertion resource for engineering Drosophila melanogaster genes. Nat Methods 8: 737–743. doi:10.1038/nmeth.1662 License: This article is available under a Creative Commons License (Attribution 4.0 International, as described at https://creativecommons.org/ licenses/by/4.0/). https://doi.org/10.26508/lsa.201900381 vol 2 \| no 4 \| e201900381 https://doi.org/10.26508/lsa.201900381 vol 2 \| no 4 \| e201900381 12 of 12 bantam, growth control and tumorigenesis Gerlach et al.
https://openalex.org/W3111823557	https://trialsjournal.biomedcentral.com/counter/pdf/10.1186/s13063-020-04949-0	English	null	Efficacy of a web application for stress management among Iranian college students during COVID-19 outbreak: a study protocol for randomized controlled trials	Trials	2,020	cc-by	5,231	STUDY PROTOCOL Open Access Efficacy of a web application for stress management among Iranian college students during COVID-19 outbreak: a study protocol for randomized controlled trials Fatemeh Khademian1, Azam Aslani2,3, Ramin Ravangard4* , Peivand Bastani3, Mohammad Nami5,6,7 and Peyman Jafari8 © The Author(s). 2020 Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated in a credit line to the data. Abstract Background: The prevalence of mental health disorders is increasing globally, and the prevalence of COVID-19 has made it worse. Evidence has indicated a major mental health burden and elevated anxiety associated with the new coronavirus outbreak in the general population. This study aims to evaluate an evidence-based web application (Naranj) for stress management among Iranian college students. Methods and design: This study aims to present a protocol related to a randomized controlled trial among Iranian college students. The study will be conducted on 100 students from two colleges of Shiraz University of Medical Sciences in Iran. The participants will be randomly assigned to the intervention and control groups. The intervention group participants will be provided with a web application, whereas the control group ones will be provided with an app unrelated to stress management. The primary outcome for this study will be the Perceived Stress Scale, and the two groups will be compared with respect to stress level and sleep quality. Discussion: A web application will be developed according to psychological theories and will be scientifically approved for managing college students’ stress and improving their sleep quality during the COVID-19 outbreak. Trial registration: Iranian Registry of Clinical Trials IRCT20160427027647N2. Registered on 14 May 2020 Keywords: Randomized controlled trial, Telemedicine, Stress, COVID-19, Mobile health, College student * Correspondence: ra_ravangard@yahoo.com * Correspondence: ra_ravangard@yahoo.com 4Health Human Resources Research Center, School of Management and Medical Information Sciences, Shiraz University of Medical Sciences, Almas Building, Alley 29, Qasrodasht Ave., Shiraz, Iran Full list of author information is available at the end of the article Khademian et al. Trials (2020) 21:1023 https://doi.org/10.1186/s13063-020-04949-0 Khademian et al. Trials (2020) 21:1023 https://doi.org/10.1186/s13063-020-04949-0 * Correspondence: ra_ravangard@yahoo.com 4Health Human Resources Research Center, School of Management and Medical Information Sciences, Shiraz University of Medical Sciences, Almas Building, Alley 29, Qasrodasht Ave., Shiraz, Iran Full list of author information is available at the end of the article Background Stress can lead to serious health problems and can dam- age almost all body systems [28]. Studies have found that stress was associated with an increased risk of rheumatoid arthritis [29], cardiovascular disease [30], cancer [31], and autoimmune diseases such as systemic lupus erythemato- sus, type 1 diabetes, and multiple sclerosis [32]. Despite the importance of mental health problems, WHO reported that nearly two thirds of people with a known mental health problem never sought professional help. It was also stated that stigma, discrimination, and ignorance prevented people from receiving treatment and care [10]. Therefore, societies should provide people with an easy, inexpensive, affordable, and reliable way to access medical services remotely. Digital technology ini- tiatives, such as m-Health, could be helpful in this re- gard. M-Health refers to the utilization of mobile phones and other wireless technologies in medical care. Some studies have shown evidence supporting the effect of m-Health on different fields of health [33–35]. The good news is that from more than 15,000 mobile apps, at least 29% have been designed in the field of mental health [36]. The bad news is that a small number of apps are evidence-based [37]. Besides, a study that appraised mental health apps showed that only few of them could be personalized [38]. Moreover, the results of a study demonstrated that more than 50% of students reported low levels of eHealth literacy [39]. Therefore, users have to be empowered to use digital technologies effectively. In general, the prevalence of mental disorders is in- creasing globally [9]. Statistics have shown that about 970 million people worldwide suffered from some sort of mental illnesses in 2017 [10], 75% of whom lived in low- and middle-income countries [11]. Mental health problems are one of the main causes of the global disease burden, such a way that they account for 13% of the total global burden of disease [12]. Mental illnesses also account for 25.3% and 33.5% of all years lived with a disability in low- and middle- income countries, respectively [12]. According to the latest Ministry of Health nationwide study regarding mental health disorders in Iran in 2015, about 23.4% of the population were suspected to suffer from a kind of mental disorder. According to this report, about one out of every four people was suspected to suffer from mental disorders [13]. These results were in line with the global average [14]. Background cause discomfort, which can lead to depression and health problems [18]. In Iran, limited research has been conducted in this field among the general population. The results of a study conducted among the people aged 18 years in Tehran (2017) showed that more than 80% of the people had experienced at least one severe episode of stress during the past year [19]. The results of another study conducted on 10,000 adult residents of Yazd Prov- ince revealed that 34.8% of the people had experienced stress [20]. Furthermore, previous research showed that college students had experienced high levels of stress. Studies have also reported some levels of stress and anx- iety among students in China (74.3%), Saudi Arabia (63.5%), Bangladesh (61%), and Egypt (59.9%) [21–24]. A case report performed in Iran also indicated that 83% of the students had experienced stress [25]. The most com- mon stressors among the students were parental expec- tations, university/life balance, financial/work balance, living away from home, sleeping difficulties, and aca- demic issues [26, 27]. g A new infectious respiratory disease, named COVID- 19, appeared in Wuhan, Hubei province, China, and it has affected almost all other countries throughout the world. According to the World Health Organization (WHO, 6 July 2020), COVID-19 has affected over 11, 327,790 people and has killed 532,340 people in more than 200 countries worldwide [1]. A study found that similar to the SARS-induced psychological burden [2], anxiety symptom levels increased when COVID-19 occurred in China [3]. Moreover, initial studies indi- cated a major mental health burden among the public, increased depressive symptoms, decreased sleep qual- ity, and elevated anxiety during the COVID-19 out- break [3–5]. In this regard, a study in China revealed that stress, anxiety, and depression were more preva- lent among students and healthcare professionals in comparison with other professions [6]. Moreover, another study showed that Chinese young people and those who spent more time thinking about COVID-19 were more at risk [3]. The probable cause of mental health problems during the coronavirus pandemic could be related to the “hypochondria concerns” the- ory that means being worried about being infected [7]. According to a Spanish study, it could also be related to important life changes in a very short period of time, including changes in financial or occupational status, restrictions on movement, and cancelation of important activities [8]. © The Author(s). 2020 Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated in a credit line to the data. Khademian et al. Trials (2020) 21:1023 Page 2 of 7 Page 2 of 7 Sample size The sample size is calculated based on a similar previous study [41] where the mean (SD) of perceived stress in groups 1 and 2 were 7.43 (2.93) and 9.49 (3.06), respect- ively. This study was a research aimed to evaluate the efficacy of a stress management app for college students. We chose this randomized controlled trial study because it was very similar to our study in terms of the interven- tion, outcome, and participants. After that, the sample size is calculated using the NCSS software. The mini- mum sample size based on power = 80%, and α = 0.05 is estimated at 35 individuals in each group. In order to compensate for the loss to follow-up, we will increase the sample size to 50 individuals in each group. Aims and objectives This study aims to assess the efficacy of a web applica- tion (Naranj) for stress management among Iranian college students during the COVID-19 outbreak. Design and setting This is a non-blind, parallel-group (ratio 1:1), random- ized controlled trial (superiority framework). The partici- pants will be recruited from two colleges of Shiraz University of Medical Sciences in Iran: (1) School of Management and Medical Information (Health Informa- tion Management, Medical Informatics, and Health Services Management) and (2) School of Health. Randomization h The participants will be selected by the systematic random sampling method based on two separate lists of all eligible BSc and MSc students in both colleges. We will select 50 students at each college. Then, the selected students will be provided with an explanation of the aims and scope of the project and will be orally invited to participate in the research. If they accept the invita- tion, they will be asked to sign informed consent forms. However, if they do not give informed consent, the next student in the systematic random sampling list will be substituted. Then, we will allocate the partici- pants of each college into 2 groups of intervention and control based on two separate 1:1 permuted block randomization lists, so that five blocks of size 10 in each college will be formed. The randomization sequence will be generated by the biostatistician using the Random Allocation Software with a 1:1 allocation. Assigning participants to their groups will be carried out via random allocation. Blinding Due to the nature of the study, the existence of the two arms in the experiment cannot be concealed (this will be a study limitation), but some levels of blinding will be considered. For example, the participants on each arm will not be informed about what the other arm will receive. Participants I l i it i The Naranj app includes a user profile, stress manage- ment strategies, consultation requests, and COVID-19 sections. The user profile includes account data, demo- graphics, and health history (nutrition, physical activity, smoking, diseases, drug and alcohol consumption, etc.). Stress management strategies are based on the Accept- ance and Commitment Therapy (ACT), Stress Reduction Theory (SRT), and some other evidence-based strategies. This section consists of various parts, including Procedures All the participants will fill out the demographic data and other study questionnaires electronically. Then, the researcher will send a link to the intervention group to have access to the stress management app (Naranj) and another link to the control group to download an app unrelated to stress management. Background It has been claimed that the prevalence of anxiety and depression has increased by about 40% over the past 30 years [15]. It has also been estimated that around 4% of the population had an anxiety disorder [16]. In addition, stress has been classified as the health epidemic of the twenty-first century by the WHO [17]. Stress has been defined as the body’s response to painful situations that The results of a survey revealed that college students had higher anxiety levels compared to the general popu- lation after the outbreak of COVID-19. This implied that COVID-19 had negative psychological impacts on Chinese college students’ anxiety levels [40]. Considering the spread of mental health problems and the prevalence Khademian et al. Trials (2020) 21:1023 Page 3 of 7 Khademian et al. Trials (2020) 21:1023 hypothyroidism, hyperthyroidism, asthma); (8) not having physical disabilities (including blindness and deafness); (9) being willing to participate in the study; and (10) signed in- formed consent forms. of COVID-19 that has made it worse, it is necessary to take appropriate and timely measures. In this context, mental health support should be provided and continued for all people, especially the vulnerable groups. Such support should include accurate information as well as training of suitable coping strategies to manage the situ- ation. Therefore, the present study aims to evaluate an evidence-based web application (Naranj) for stress man- agement among Iranian college students during the COVID-19 outbreak. It has been hypothesized that the application is effective in decreasing college students’ psychological stress. The secondary expected outcome is the improvement of sleep quality. Exclusion criteria The students will be excluded if they are not willing to take part in the research. Inclusion criteria Students will have to meet the following eligibility criteria to be included in the study: (1) aging 20–35 years; (2) being a BSc (second and higher semesters) or an MSc student; (3) having a smartphone, computer, laptop, or tablet; (4) not suffering from known psychological problems; (5) not con- suming psychiatric drugs; (6) not being addicted according to the students’ self-expression; (7) not having chronic physical diseases (including diabetes, hypertension, Khademian et al. Trials (2020) 21:1023 Page 4 of 7 Fig. 1 The participants’ timeline Fig. 1 The participants’ timeline Fig. 1 The participants’ timeline metaphors based on ACT, 4-7-8 breathing technique, Emotional Freedom Technique (EFT), pictures of nature, relaxing music, relaxing videos, educational quotes, my diary, forum, and meditation. In “my diary,” students have a space to enter their thoughts, emotions, and experiences of using the strategies. The “consultation request” section provides users with the opportunity to use emergency (receiving service during 24 h) and usual (receiving service during 48 h) consultations with psy- chological consultants. The “COVID-19” section pro- vides educational content on COVID-19. In addition, this section is hyperlinked to the “stress management strategies” section. The Naranj app includes “privacy policy and terms of use,” app tutorial, contact us, and frequently asked questions, as well. content. The rationale for using this app is to determine whether there will be an improvement in the outcome measures due to the stress management content in the Naranj app. Overall, both groups are provided with the same conditions. In this way, the effect of the con- founding factors, if any, can be controlled and the Hawthorne effect will be minimized. According to the Hawthorne effect, individuals will alter their behaviors in response to being studied and usually refer to posi- tive changes [42]. Both groups will have access to their specified apps for 3 months. All participants in the intervention and con- trol groups will be requested to complete the question- naires again after completion of the intervention. An app regarding calorie uptake management will be offered to the control group. This app is comparable to the Naranj app in terms of the need for daily use by the user. However, it is different from Naranj in terms of Secondary outcome measure The secondary outcome of the study will be sleep quality, which will be evaluated using the Pittsburgh Sleep Quality Index (PSQI). PSQI is a valid and reli- able questionnaire used to measure stress among Iranian college students [44]. PSQI contains 18 items on a 4-point Likert scale ranging from 0 (not at all) to 3 (three or more times a week). These items index in seven subscales: subjective sleep efficiency, sleep latency, sleep duration, sleep quality, sleep disturb- ance, sleep medication use, and daytime dysfunction due to sleepiness. The total score ranges from 0 to 21 where higher scores indicate worse sleep quality. A PSQI score > 5 indicates poor sleep quality. Ethical aspects of the study The present study was approved by the Ethics Committee of Shiraz University of Medical Sciences, Shiraz, Iran (IR.SUMS.REC.1398.664). Informed consent will be obtained from all participants, and confidentiality of the information will be assured. Discussion This study aims to develop a web application and to determine its impact on stress management among college students. Development of the suggested web application and evaluation of its impact on the students will have the potential to develop useful media for man- aging stress, improving sleep quality, and providing in- formation about the effectiveness of stress management strategies among college students who have been shown to experience high-stress levels. It has been hypothesized that the web application can reduce the stress level and improve the sleep quality among the target college stu- dents. It will also promote students’ healthy behaviors. Considering the fact that mental health services are lim- ited in developing countries, many people cannot access fair resources, and only few applications developed in this field are evidence-based. Therefore, this theory- based web application is expected to be a promising tool to help students in Iran. To avoid the risk of contamin- ation between the intervention and control groups, the web application will be available only to the intervention group by putting all intervention group participants in a white list and registering them in the web application. Then, web application registration will be disabled tem- porarily during the intervention, so that it cannot be used by anyone else. Overall, the study results will indi- cate an important step towards developing a technology platform for delivering a range of psychosocial strategies for managing stress and sleep problems. Data collection The participants will be asked to fill out three electronic questionnaires. At baseline, they will be asked to complete the demographic information form, Perceived Khademian et al. Trials (2020) 21:1023 Page 5 of 7 Page 5 of 7 Stress Scale (PSS), and Pittsburgh Sleep Quality Index (PSQI). After the completion of the intervention, they will be requested to complete the PSS and PSQI again (Fig. 1). Since the Iranian government has advised the public to minimize face-to-face interactions and iso- late themselves at home during the COVID-19 pan- demic, the electronic versions of the questionnaires will be used. Information on the app use, such as daily stress level and daily quality of sleep measured daily using the visual analog scale (VAS) in the web app, as well as the frequency of using different app sections will be col- lected, as well. Data on the app use and the possible use of other apps, including apps related to stress manage- ment, will also be collected. test will be applied. In addition, the Mann-Whitney U and chi-square tests will be used to compare the two groups regarding non-parametric and categorical vari- ables, respectively. In case of confounders, analysis of co- variance or linear regression model will be used. Moreover, the paired samples t test will be utilized in order to compare the mean scores of stress and sleep quality in each group. Intention-to-treat analysis will be employed, as well. All analyses will be conducted using the SPSS 16 software, and p ≤0.05 will be considered statistically significant. Statistical analysis The Kolmogorov-Smirnov test will be used to test the normal distribution of the data. The comparability of the intervention and control groups will be assessed at base- line. The quantitative variables will be described using mean and standard deviation if they follow a normal dis- tribution and using median and range in case of non- normal distribution. The qualitative variables will be pre- sented as percentages and confidence intervals. In order to compare the study groups regarding the mean scores of stress and sleep quality, the independent samples t Primary outcome measure The primary outcome of the study will be perceived stress, which will be assessed using the Perceived Stress Scale (PSS). PSS is a valid and reliable questionnaire used to measure stress among Iranian college students [43]. PSS contains 14 items scored based on a 5-point Likert scale ranging from 0 (never) to 4 (very often). Items 4, 7, 9, 10, and 13 are scored reversely. The total score of PSS will be obtained by adding all the items. The scores can range from 0 to 56, with higher scores representing the higher perceived stress. Trial status This trial was registered in the Iranian Registry of Clinical Trials (IRCT20160427027647N2) on 14 May 2020. The recruitment began on 14 June 2020 and ended on 05 September 2020. Page 6 of 7 Page 6 of 7 Khademian et al. Trials (2020) 21:1023 Khademian et al. Trials (2020) 21:1023 Abbreviations 7. Furer P, Walker JR, Chartier MJ, Stein MB. Hypochondriacal concerns and somatization in panic disorder. Depress Anxiety. 1997;6(2):78–85. ACT: Acceptance and commitment therapy; SRT: Stress Reduction Theory; PSS: Perceived Stress Scale; PSQI: Pittsburgh Sleep Quality Index ACT: Acceptance and commitment therapy; SRT: Stress Reduction Theory; PSS: Perceived Stress Scale; PSQI: Pittsburgh Sleep Quality Index 8. Rodríguez-Rey R, Garrido-Hernansaiz H, Collado S. Psychological impact and associated factors during the initial stage of the coronavirus (COVID-19) pandemic among the general population in Spain. Front Psychol. 2020;11: 1540. The authors declare no competing interests. The authors declare no competing interests. 18. Kulkarni SH, Vaidya RR, Jadhav SA. Stress, its consequences and ways to combat it in modern era. Stud Indian Place Names. 2020;40(42):185–91. Authors’ contributions All authors made substantial contributions to the conception and design and finally approved the manuscript. FKh drafted the article and coordinated the revision phases. RR, AA, PB, and MN participated in the design of the methodology. PJ organized the sample size and the statistical plan. All authors contributed to the drafting and preparation of the manuscript and the study protocol and provided the approval of the final version of the manuscript. 12. World Health Assembly. Global burden of mental disorders and the need for a comprehensive, coordinated response from health and social sectors at the country level: report by the Secretariat. World Health Organization. 2012. Available from: https://apps.who.int/iris/handle/10665/78898. 13. Noorbala AA, Faghihzadeh S, Kamali K, Yazdi SAB, Hajebi A, Mousavi MT, et al. Mental health survey of the Iranian adult population in 2015. Arch Iran Med. 2017;20(3):0. Acknowledgements This article was extracted from the Ph.D. dissertation written by Fatemeh Khademian and was financially supported by Shiraz University of Medical Sciences (grant No 98-01-07-19451). The authors would like to thank Ms. A. Keivanshekouh at the Research Improvement Center of Shiraz University of Medical Sciences for improving the use of English in the manuscript. 9. Steel Z, Marnane C, Iranpour C, Chey T, Jackson JW, Patel V, et al. The global prevalence of common mental disorders: a systematic review and meta- analysis 1980-2013. Int J Epidemiol. 2014;43(2):476–93. 10. World Health Organization. The World Health Report 2001: Mental health: new understanding, new hope. World Health Organization; 2001. 11. Organization WH. mhGAP: Mental Health Gap Action Programme: scaling up care for mental, neurological and substance use disorders. Geneva: World Health Organization; 2008. Received: 31 July 2020 Accepted: 7 December 2020 25. Borjalilu S, Mohammadi A, Mojtahedzadeh R. Sources and severity of perceived stress among Iranian medical students. Iran Red Crescent Med J. 2015;17(10):e17767. Consent for publication Not applicable. 17. Fink G. Stress: The health epidemic of the 21st century. Elsevier SciTech Connect; 2016. Funding Funding This trial will be conducted with no internal and external funding. 14. World Health Organization. The World Health Report: Mental disorders affect one in four people; 2001. g This trial will be conducted with no internal and external funding. Availability of data and materials Not applicable. Availability of data and materials 15. Metrics IfH, Evaluation. Global Burden of Disease Collaborative Network. Global Burden of Disease Study 2016 (GBD 2016) results. Institute for Health Metrics and Evaluation Seattle; 2017. 16. Roser M, Ritchie H, Ortiz-Ospina E. Our World in Data. 2020. Burden of disease URL: https://ourworldindata.org/burden-of-disease. [Accessed 2020- 02-09] [WebCite Cache]. 2020. Author details 1 d 19. Noorbala AA, Rafiey H, Alipour F, Moghanibashi-Mansourieh A. Psychosocial stresses and concerns of people living in Tehran: a survey on 6000 adult participants. Iran J Psychiatry. 2018;13(2):94. 1Student Research Committee, Department of Health Information Management, School of Management and Medical Information Sciences, Shiraz University of Medical Sciences, Shiraz, Iran. 2National Agency of Strategic Research in Medical Education, Tehran, Iran. 3Health Human Resources Research Center, School of Management & Information Sciences, Shiraz University of Medical Sciences, Shiraz, Iran. 4Health Human Resources Research Center, School of Management and Medical Information Sciences, Shiraz University of Medical Sciences, Almas Building, Alley 29, Qasrodasht Ave., Shiraz, Iran. 5Department of Neuroscience, School of Advanced Medical Sciences and Technologies, Shiraz University of Medical Sciences, Shiraz, Iran. 6Neuroscience Center, Instituto de Investigaciones Científicas y Servicios de Alta Tecnología (INDICASAT AIP), City of Knowledge 084301103, Panama City, Panama. 7Department of Cognitive Neuroscience, Institute for Cognitive Science Studies, Pardis, Tehran, Iran. 8Department of Biostatistics, Faculty of Medicine, Shiraz University of Medical Sciences, Shiraz, Iran. 20. Mirzaei M, Ardekani SMY, Mirzaei M, Dehghani A. Prevalence of depression, anxiety and stress among adult population: results of Yazd Health Study. Iran J Psychiatry. 2019;14(2):137. 21. Lian Z, Wallace BC. Prevalence of past-year mental disorders and its correlates among Chinese international students in US higher education. J Am Coll Heal. 2020;68(2):176–84. 22. Alotaibi AD, Alosaimi FM, Alajlan AA, Abdulrahman KAB. The relationship between sleep quality, stress, and academic performance among medical students. J Fam Community Med. 2020;27(1):23. 23. Islam S, Akter R, Sikder T, Griffiths MD. Prevalence and factors associated with depression and anxiety among first-year university students in Bangladesh: a cross-sectional study. Int J Ment Heal Addict. 2020:1–14. 24. Fawzy M, Hamed SA. Prevalence of psychological stress, depression and anxiety among medical students in Egypt. Psychiatry Res. 2017;255:186–94. Received: 31 July 2020 Accepted: 7 December 2020 Received: 31 July 2020 Accepted: 7 December 2020 Khademian et al. Trials (2020) 21:1023 Khademian et al. Trials (2020) 21:1023 progression of diabetes and cancer. Int J Endocrinol Metab. 2019;17(2): e67400. progression of diabetes and cancer. Int J Endocrinol Metab. 2019;17(2): e67400. 32. Ramani S, Pathak A, Dalal V, Paul A, Biswas S. Oxidative stress in autoimmune diseases: an under dealt malice. Curr Protein Pept Sci. 2020 21(6):611–21. 33. Khademian F, Rezaee R, Pournik O. Randomized controlled trial: the effects of short message service on mothers’ oral health knowledge and practice. Community Dent Health. 2020;37:1–7. 34. Masaki K, Tateno H, Nomura A, Muto T, Suzuki S, Satake K, et al. A randomized controlled trial of a smoking cessation smartphone application with a carbon monoxide checker. NPJ Digital Med. 2020;3(1):1–7. 35. Huberty J, Puzia M, Eckert R, Larkey L. Cancer patients’ and survivors’ perceptions of the calm app: cross-sectional descriptive study. JMIR Cancer. 2020;6(1):e16926. 36. Anthes E. Pocket psychiatry: mobile mental-health apps have exploded onto the market, but few have been thoroughly tested. Nature. 2016; 532(7597):20–4. 37. Grist R, Porter J, Stallard P. Mental health mobile apps for preadolescents and adolescents: a systematic review. J Med Internet Res. 2017;19(5):e176. 38. Salehi F, Kermani ZA, Khademian F, Aslani A. Critical appraisal of mental health applications. Stud Health Technol Inform. 2019;261:303–8. 38. Salehi F, Kermani ZA, Khademian F, Aslani A. Critical appraisal of mental health applications. Stud Health Technol Inform. 2019;261:303–8. 39. KHademian F, Montazer MRA, Aslani A. Web-based health information seeking and eHealth literacy among college students: a self-report study. Investigación y educación en enfermería. 2020;38(1):3. 40. Wang C, Zhao H. The impact of COVID-19 on anxiety in Chinese university students. Front Psychol. 2020;11:1168–75. 40. Wang C, Zhao H. The impact of COVID-19 on anxiety in Chinese university students. Front Psychol. 2020;11:1168–75. 41. Harrer M, Adam SH, Fleischmann RJ, Baumeister H, Auerbach R, Bruffaerts R, et al. Effectiveness of an internet-and app-based intervention for college students with elevated stress: randomized controlled trial. J Med Internet Res. 2018;20(4):e136. 42. Adair JG. The Hawthorne effect: a reconsideration of the methodological artifact. J Appl Psychol. 1984;69(2):334. 43. Sepahvand T, Gilani B, Zamani R. The relationship between explanatory style. J Psychol Educ. 2009;38(4):27–43. 44. Afkham Ebrahimi A, Bandi G, Salehi M, Tafti K, Vakili Y, Farsi A. Sleep parameters and the factors affecting the quality of sleep in patients attending selected clinics of Rasoul-e-Akram Hospital. Razi J Med Sci. 2008; 15:31–8. References 26. Tariq S, Tariq S, Tariq S, Jawed S. Perceived stress, severity and sources of stress among female medical students in a private medical college in Pakistan. J Pak Med Assoc. 2020;70(1):162–7. 1. World Health Organization. Coronavirus Disease (COVID-19), Situation report–168 2020. Available from: https://www.who.int/docs/default-source/ coronaviruse/situation-reports/20200706-covid-19-sitrep-168.pdf?sfvrsn= 7fed5c0b_2. Accessed 20 Aug 2020. 1. World Health Organization. Coronavirus Disease (COVID-19), Situation report–168 2020. Available from: https://www.who.int/docs/default-source/ coronaviruse/situation-reports/20200706-covid-19-sitrep-168.pdf?sfvrsn= 7fed5c0b_2. Accessed 20 Aug 2020. 1. World Health Organization. Coronavirus Disease (COVID-19), Situation report–168 2020. Available from: https://www.who.int/docs/default-source/ coronaviruse/situation-reports/20200706-covid-19-sitrep-168.pdf?sfvrsn= 7fed5c0b_2. Accessed 20 Aug 2020. 27. Pitt A, Oprescu F, Tapia G, Gray M. An exploratory study of students’ weekly stress levels and sources of stress during the semester. Act Learn High Educ. 2018;19(1):61–75. 2. Su T-P, Lien T-C, Yang C-Y, Su YL, Wang J-H, Tsai S-L, et al. Prevalence of psychiatric morbidity and psychological adaptation of the nurses in a structured SARS caring unit during outbreak: a prospective and periodic assessment study in Taiwan. J Psychiatr Res. 2007;41(1–2):119–30. 28. Segal J, Smith M, Melinda S, MA R, Robinson L. Stress Symptoms, Signs and Causes: Improving your Ability to Handle Stress. Trusted guide to mental, emotional & social health HELPGUIDE ORG. 2018. 3. Huang Y, Zhao N. Generalized anxiety disorder, depressive symptoms and sleep quality during COVID-19 outbreak in China: a web-based cross- sectional survey. Psychiatry Res. 2020;288:112954. 29. Polinski KJ, Bemis EA, Feser M, Seifert J, Demoruelle MK, Striebich CC, et al. Perceived Stress and Inflammatory Arthritis: A Prospective Investigation in the Studies of the Etiologies of Rheumatoid Arthritis Cohort. Arthritis Care Res. 2020;72(12):1766–71. 30. Silverman AL, Herzog AA, Silverman DI. Hearts and minds: stress, anxiety, and depression: unsung risk factors for cardiovascular disease. Cardiol Rev. 2019;27(4):202–7. 6. Rehman U, Shahnawaz MG, Khan NH, Kharshiing KD, Khursheed M, Gupta K, et al. Depression, anxiety and stress among Indians in times of Covid-19 lockdown. Community Ment Health J. 2020:1–7. 31. Afrisham R, Paknejad M, Soliemanifar O, Sadegh-Nejadi S, Meshkani R, Ashtary-Larky D. The influence of psychological stress on the initiation and 31. Afrisham R, Paknejad M, Soliemanifar O, Sadegh-Nejadi S, Meshkani R, Ashtary-Larky D. The influence of psychological stress on the initiation and Page 7 of 7 Khademian et al. Trials (2020) 21:1023 Publisher’s Note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
https://openalex.org/W4221052431	https://tropmedhealth.biomedcentral.com/track/pdf/10.1186/s41182-022-00415-6	English	null	Parental COVID-19 vaccine hesitancy for children with neurodevelopmental disorders: a cross-sectional survey	Tropical medicine and health	2,022	cc-by	7,254	Tropical Medicine and Health Ali et al. Tropical Medicine and Health (2022) 50:24 https://doi.org/10.1186/s41182-022-00415-6 Abstract Higher odds of vaccine hesitancy were found among parents who lived in the northern zone (AOR = 17.15, 95% CI = 5.86–50.09; p < 0.001), those who thought vaccines would not be safe and effective for Bang‑ ladeshi children (AOR = 3.22, 95% CI = 1.68–15.19; p < 0.001), those who were either not vaccinated or did not receive the COVID-19 vaccine themselves (AOR = 12.14, 95% CI = 8.48–17.36; p < 0.001), those who said that they or their family members had not tested positive for COVID-19 (AOR = 2.13, 95% CI = 1.07–4.25), and those who did not lose a family member to COVID-19 (AOR = 2.12, 95% CI = 1.03–4.61; p = 0.040). Furthermore, parents who were not likely to believe that their children or a family member could be infected with COVID-19 the following year (AOR = 4.99, 95% CI = 1.81–13.77; p < 0.001) and who were not concerned at all about their children or a family member being infected the following year (AOR = 2.34, 95% CI = 1.65–8.37; p = 0.043) had significantly higher odds of COVID-19 vaccine hesitancy. Conclusions: Given the high prevalence of vaccine hesitancy, policymakers, public health practitioners, and pediatri‑ cians can implement and support strategies to ensure that children with NDD and their caregivers and family mem‑ bers receive the COVID-19 vaccine to fight pandemic induced hazards. rds: Bangladesh, COVID-19, Pediatrics, Neurodevelopmental disorders, Parental vaccine hesitancy © The Author(s) 2022. Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/. Parental COVID‑19 vaccine hesitancy for children with neurodevelopmental disorders: a cross‑sectional survey Mohammad Ali1,2* , Tasnuva Shamarukh Proma3, Zarin Tasnim4, Md. Ariful Islam5, Tania Akter Urmi6, Sohel Ahmed7, Abu‑sufian Sarkar8, Atia Sharmin Bonna9 and Umme Salma Khan10 Abstract Background: Little is known about parental coronavirus disease 2019 (COVID-19) vaccine hesitancy in children with neurodevelopmental disorders (NDD). This survey estimated the prevalence and predictive factors of vaccine hesi‑ tancy among parents of children with NDD. Methods: A nationally representative cross-sectional survey was conducted from October 10 to 31, 2021. A struc‑ tured vaccine hesitancy questionnaire was used to collect data from parents aged ≥ 18 years with children with NDD. In addition, individual face-to-face interviews were conducted at randomly selected places throughout Bangladesh. Multiple logistic regression analysis was conducted to identify the predictors of vaccine hesitancy. Results: A total of 396 parents participated in the study. Of these, 169 (42.7%) parents were hesitant to vacci‑ nate their children. Higher odds of vaccine hesitancy were found among parents who lived in the northern zone (AOR = 17.15, 95% CI = 5.86–50.09; p < 0.001), those who thought vaccines would not be safe and effective for Bang‑ ladeshi children (AOR = 3.22, 95% CI = 1.68–15.19; p < 0.001), those who were either not vaccinated or did not receive the COVID-19 vaccine themselves (AOR = 12.14, 95% CI = 8.48–17.36; p < 0.001), those who said that they or their family members had not tested positive for COVID-19 (AOR = 2.13, 95% CI = 1.07–4.25), and those who did not lose a family member to COVID-19 (AOR = 2.12, 95% CI = 1.03–4.61; p = 0.040). Furthermore, parents who were not likely to believe that their children or a family member could be infected with COVID-19 the following year (AOR = 4.99, 95% CI = 1.81–13.77; p < 0.001) and who were not concerned at all about their children or a family member being infected the following year (AOR = 2.34, 95% CI = 1.65–8.37; p = 0.043) had significantly higher odds of COVID-19 vaccine hesitancy. Results: A total of 396 parents participated in the study. Of these, 169 (42.7%) parents were hesitant to vacci‑ nate their children. Background Vaccine hesitancy has been declared as one of the top 10 global health threats by the World Health Organization in 2019 [1]. Vaccine hesitancy can be defined as a delay in refusal or acceptance of a vaccine during the availability of a vaccination service [2]. For children, parental vaccine *Correspondence: m180002@student.bup.edu.bd 1 Department of Physiotherapy and Rehabilitation, Uttara Adhunik Medical College and Hospital, Uttara Model Town, Dhaka 1230, Bangladesh angladesh ull list of author information is available at the end of the article Tropical Medicine and Health Tropical Medicine and Health Tropical Medicine and Health Study design and participants Anonymous data were collected from October 10 to 31, 2021, for this nationally representative cross-sectional survey. Individuals aged ≥ 18 years who were parents of at least one child aged under 18 years with neurodevel- opmental disorders were eligible for this study. Individu- alized interviews using a structured questionnaire were conducted in homes and health care centers. Parents aged < 18 years were excluded from the study. A margin of 5% error, a confidence level of 95%, and a response dis- tribution of 50% were used to calculate the sample size to target fathers/mothers of approximately 8 million chil- dren and secure a minimum sample size of 384 partici- pants [16, 17]. Data were collected from 408 individuals. Furthermore, NDD patients affected by COVID-19 can experience difficulties accessing required healthcare and possess other characteristics that increase their dif- ficulties faced due to COVID-19. These include mobility limitation, direct supervision requirements, challenges practicing prophylactic measures, and communicating disease symptoms [9]. Furthermore, evidence suggests that COVID-19 incidence, hospitalization, ICU admis- sion, and death are much higher among COVID-19 patients with NDD [10]. Methods Many parents of a child with NDD believe that vaccines can negatively impact their child’s disability conditions [6]. Research has further demonstrated that parents of a child with NDD are also vaccine-hesitant for their other healthy children [7]. Conversely, children with NDD and their families are significant victims of the current coro- navirus disease 2019 (COVID-19) pandemic [8]. Despite the limitations of public health data, some evidence sug- gests that children with NDD might be disproportion- ately affected by the COVID-19 itself or the pandemic’s impact on receiving services. In addition, children with NDD often have medical conditions that contribute to a higher risk of severe COVID-19 disease [9]. © The Author(s) 2022. Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/. Ali et al. Tropical Medicine and Health (2022) 50:24 Page 2 of 9 (1) estimate the prevalence of vaccine hesitancy within a nationally representative sample of parents of children with NDD and (2) identify the predictors of vaccine hesi- tancy considering socioeconomic and behavioral factors and COVID-19 threat perception. hesitancy is a significant barrier to a smooth vaccination program for fighting vaccine-preventable diseases [3]. In addition, previous studies have suggested that parents of children with a neurodevelopmental disorder (NDD) have a significantly higher vaccine hesitancy than parents with neurotypical children [4, 5]. Sampling technique Data were collected from all eight divisions in Bang- ladesh. We identified approximately 150 government and non-government centers dealing with children with NDD. In addition, we visited 38 randomly selected cent- ers to collect data of the parents. Approximately 450 par- ents were conveniently approached for the interviews. Previous studies have found a high prevalence of COVID-19 vaccine hesitancy among parents for their neurotypical children in several settings; this was a mat- ter of great concern when discussing immunization of an optimum population percentage to gain community herd immunity for fighting the pandemic [11, 12]. In addition, existing literature suggested that the higher COVID-19 vaccine hesitancy among parents of a child with NDD would raise further concern worldwide [13]. Data analysis Descriptive statistics were employed to elaborate on the demographic characteristics of the study participants. Χ2 test or Fisher’s exact test was used to calculate vaccine hesitancy proportions and draw comparisons between subgroups. Responses were compared for various soci- odemographic characteristics by dichotomizing the vari- able as positive (“very likely” and “somewhat likely”) or a negative (“not likely” and “definitely not”) attitude toward the COVID-19 vaccine, indicating the prevalence of vac- cine hesitancy. Multiple logistic regression analyses were performed with vaccine hesitancy as a dependent vari- able and sociodemographic characteristics and perceived COVID-19 threat as predictor variables for vaccine hesi- tancy to calculate adjusted odds ratios (AOR) with a 95% confidence interval (CI). In this model, we included fac- tors significantly associated with vaccine hesitancy in the descriptive analysis. The Hosmer–Lemeshow good- ness-of-fit test was employed to ensure that the models adequately fit the data. Statistical significance was set at p < 0.05. SPSS version 0.22.0 (IBM Corp., USA) was used for all data analyses. Overall, 42.7% (CI = 37.82–47.57) of parents were hesi- tant to vaccinate their children with NDD against COVID- 19. Table 1 displays the result of descriptive analysis. The highest prevalence of vaccine hesitancy was observed among parents of children aged 0–4 years (57.9%; p < 0.001), parents aged 31–35 years (52.3%; p < 0.001), those living in the northern zone (59.5%; p < 0.001), those who did not believe the vaccine will be safe and effective for Bangladeshi children (86.4%; p < 0.001), those who were either not vaccinated or chose not receive the COVID-19 vaccination themselves (79%, p < 0.001). Furthermore, vac- cine hesitancy was highest among parents who said that they or their family members had not tested positive for COVID-19 (50.6%; p < 0.001) and those who did not lose any family member due to COVID-19 (45%; p = 002). Moreover, participants who were not likely to believe that their children or a family member could be infected with COVID-19 the following year (66.7%; p < 0.001) and those who were not concerned at all about their children or a family member getting infected in the following year (62.9%; p < 0.001) showed high levels of COVID-19 vaccine hesitancy. Figure 1 shows the breakdown of vaccine accept- ance or refusal. Table 2 displays the results of multiple logistic regression analyses. The questionnaire P l l Previously employed vaccine hesitancy and COVID-19 threat perception questionnaire modified for parents were used in this study [14]. In the first part, parents were asked about the likelihood of vaccinating their chil- dren with NDD. Vaccine hesitancy was measured using the question, “If a vaccine that would be effective against coronavirus among children was available, how likely are you to get your children with NDD vaccinated?” The response options for this question were “very likely,” “somewhat likely,” “not likely,” and “definitely not.” Partici- pants were also asked two questions regarding the per- ceived COVID-19 threat: (1) “How likely is it that your children or a family member could get infected with coronavirus in the next one year?” The response options for this were “very likely,” “somewhat likely,” “not likely,” and “definitely not.” (2) “How concerned are you that your children or a family member could get infected with COVID-19 in the next year?” Here, response options were “very concerned,” “concerned,” “slightly concerned,” and “not concerned at all.” The second part of the ques- tionnaire included a wide range of sociodemographic, In Bangladesh, the COVID-19 vaccine roll-out was initiated on January 27, 2021, for adults aged 18 years and above [14]. Vaccination among students aged 12–17 started on November 1, 2021 [15]. However, very little is known about the COVID-19 vaccine hesitancy among the parents of children aged under 18. Furthermore, it is crucial to determine whether the higher prevalence of vaccine hesitancy among parents of a child with NDD is consistent with the COVID-19 vaccination program, considering the public health importance of widespread COVID-19 vaccination in Bangladesh and other parts of the world. In addition, it is vital to identify other factors associated with vaccine hesitancy in this population that could inform the design of targeted, preemptive vaccine interventions. With this in mind, our objectives were to Ali et al. Tropical Medicine and Health (2022) 50:24 Page 3 of 9 Ali et al. Tropical Medicine and Health behavioral, and COVID-19-related questions regarding both the child and parent. behavioral, and COVID-19-related questions regarding both the child and parent. tested positive for COVID-19 earlier, and 8.6% had lost a loved one due to COVID-19. The questionnaire P l l Details of the responses to the questions regarding the likelihood of children or fam- ily members’ COVID-19 infection and the level of con- cern about children or family members being infected in the next year are shown in Table 1. Results With a 90.7% response rate, 408 parents provided informed consent and data for this study. Seven par- ents refused to answer the entire question and were thus excluded. We also excluded five more data points for inconsistent answers to the questions. Finally, a total of 396 parents (60.4% men) aged 34.54 ± 6.56 years (mean ± standard deviation) included in the analysis. Of the children, 60.9% were male, and 40.2% were in the 0–4-year-old group. The highest number of parents (27.5%) were in the 31–35-year-old group. Overall, 87.4% of parents were Muslim, 70.7% were belonged to nuclear family members, 44.2% had two children, 29.8% had a low education level, 24.2% were service holders, and 33.1% had a high household income. Among all partici- pants, 49.2% were from the village, 62.6% lived in the cen- tral zone, including Dhaka, 76.3% were not tobacco users, 76.8% were regular religious practitioners, and 37.9% were politically neutral respondents. However, 3.3% of parents did not adhere to the standard government vac- cination (other than COVID-19) programs, and 53.5% remained skeptical about COVID-19 vaccine safety and effectiveness for Bangladeshi children. Furthermore, 20.5% of parents were either not vaccinated or did not receive the COVID-19 vaccine themselves, but 39.6% of parents reported that they or their family members had Data analysis The subgroup of participants with the highest odds of vaccine hesitancy were those who lived in the northern zone (AOR = 17.15, 95% CI = 5.86– 50.09; p < 0.001), who thought vaccines would not be safe and effective for Bangladeshi children (AOR = 3.22, 95% CI = 1.68–15.19; p < 0.001), who were either not vacci- nated or did not receive the COVID-19 vaccine themselves (AOR = 12.14, 95% CI = 8.48–17.36; p < 0.001), who said that they or their family members had not tested positive for COVID-19 (AOR = 2.13, 95% CI = 1.07–4.25), and who did not lose a family member to COVID-19 (AOR = 2.19, 95% CI = 1.07–4.61; p = 0.040). Furthermore, parents who were not likely to believe that their children or a family member could be infected with COVID-19 the following year (AOR = 4.99, 95% CI = 1.81–13.77; p < 0.001) and not concerned at all about their children or a family member getting infected the next year (AOR = 2.34, 95% CI = 1.65– 8.37; p = 0.043) had significantly higher odds of COVID-19 vaccine hesitancy. Discussion To the best of our knowledge, this is the first study to analyze parental COVID-19 vaccine hesitancy in parents of children with NDD. This nationally Ali et al. Discussion Tropical Medicine and Health (2022) 50:24 Page 5 of 9 Table 1 (continued) Variables Total sample n (%) Likelihood of va Not likely/defin Monthly household income (৳) < ৳15 000 103 (26) 45 (43.7) ৳ 15,000–30,000 86 (21.7) 35 (40.7) ৳ 31,000–45,000 76 (19.2) 32 (42.1) > ৳ 45,000 131 (33.1) 57 (43.5) Current residence type Own 186 (47) 81 (43.5) Rented 190 (48) 79 (41.6) Others 20 (5) 9 (45) Permanent address Village 195 (49.2) 88 (45.1) Semi city 84 (21.2) 40 (47.6) City 117 (29.5) 41 (35) Current living location Central zone 248 (62.6) 107 (43.1) North zone 74 (18.7) 44 (59.5) South zone 74 (18.7) 18 (24.3) Present tobacco user No 302 (76.3) 127 (42.1) Yes 94 (23.7) 42 (44.7) Regular religious practice No 92 (23.2) 39 (42.4) Yes 304 (76.8) 130 (42.8) Political affiliation Ruling party 126 (31.8) 45 (35.7) Opposition 49 (12.4) 19 (38.8) Neutral 150 (37.9) 73 (48.7) Prefer not to say 71 (17.9) 32 (45.1) Vaccinated/plan to vaccinate children under regular govt. vaccination programs (other than COVID-19) No 13 (3.3) 6 (46.2) Yes 383 (96.7) 163 (42.6) Do you think the COVID-19 vaccine will be safe and effective for Bangladeshi children? No 22 (5.6) 19 (86.4) Yes 162 (40.9) 12 (7.4) Skeptical 212 (53.5) 138 (65.1) Have you taken or planned to take the COVID-19 vaccine? No 81 (20.5) 64 (79) Yes 315 (79.5) 105 (33.3) Were you or your family member(s) tested positive for COVID-19? No 239 (60.4) 121 (50.6) Yes 157 (39.6) 48 (30.6) Have you lost any of your family member(s) for COVID-19? Discussion No 362 (91.4) 163 (45) Yes 34 (8.6) 6 (17.6) Perceived likelihood of children or family members’ infection in the next year Very likely 56 (14.1) 9 (16.1) Somewhat likely 240 (60.6) 102 (42.5) Not likely 73 (18.4) 40 (54.8) Definitely not 27 (6.8) 18 (66.7) Table 1 (continued) Variables Total sample n (%) Likelihood of vaccinating children p value Not likely/definitely not n (%) Very likely/somewhat-likely n (%) Monthly household income (৳) < ৳15 000 103 (26) 45 (43.7) 58 (56.3) 0.973 ৳ 15,000–30,000 86 (21.7) 35 (40.7) 51 (59.3) ৳ 31,000–45,000 76 (19.2) 32 (42.1) 44 (57.9) > ৳ 45,000 131 (33.1) 57 (43.5) 74 (56.5) Current residence type Own 186 (47) 81 (43.5) 105 (56.5) 0.907 Rented 190 (48) 79 (41.6) 111 (58.4) Others 20 (5) 9 (45) 11 (55) Permanent address Village 195 (49.2) 88 (45.1) 107 (54.9) 0.128 Semi city 84 (21.2) 40 (47.6) 44 (52.4) City 117 (29.5) 41 (35) 76 (65) Current living location Central zone 248 (62.6) 107 (43.1) 141 (56.9) < 0.001 North zone 74 (18.7) 44 (59.5) 30 (40.5) South zone 74 (18.7) 18 (24.3) 56 (75.7) Present tobacco user No 302 (76.3) 127 (42.1) 175 (57.9) 0.653 Yes 94 (23.7) 42 (44.7) 52 (55.3) Regular religious practice No 92 (23.2) 39 (42.4) 53 (57.6) 0.950 Yes 304 (76.8) 130 (42.8) 174 (57.2) Political affiliation Ruling party 126 (31.8) 45 (35.7) 81 (64.3) 0.160 Opposition 49 (12.4) 19 (38.8) 30 (61.2) Neutral 150 (37.9) 73 (48.7) 77 (51.3) Prefer not to say 71 (17.9) 32 (45.1) 39 (54.9) Vaccinated/plan to vaccinate children under regular govt. vaccination programs (other than COVID-19) No 13 (3.3) 6 (46.2) 7 (53.8) 0.779 Yes 383 (96.7) 163 (42.6) 220 (57.4) Do you think the COVID-19 vaccine will be safe and effective for Bangladeshi children? No 22 (5.6) 19 (86.4) 3 (13.6) < 0.001 Yes 162 (40.9) 12 (7.4) 150 (92.6) Skeptical 212 (53.5) 138 (65.1) 74 (34.9) Have you taken or planned to take the COVID-19 vaccine? No 81 (20.5) 64 (79) 17 (21) < 0.001 Yes 315 (79.5) 105 (33.3) 210 (66.7) Were you or your family member(s) tested positive for COVID-19? No 239 (60.4) 121 (50.6) 118 (49.4) < 0.001 Yes 157 (39.6) 48 (30.6) 109 (69.4) Have you lost any of your family member(s) for COVID-19? Discussion Tropical Medicine and Health (2022) 50:24 Page 4 of 9 Table 1 Descriptive analysis: sociodemographic characteristics, COVID-19 threat, and parental vaccine hesitancy Variables Total sample n (%) Likelihood of vaccinating children p value Not likely/definitely not n (%) Very likely/somewhat-likely n (%) All participants 396 (100) 169 (42.7) 227 (57.3) N/A Children’s age group (in year) 0–4 159 (40.2) 92 (57.9) 67 (42.1) < 0.001 5–9 156(39.4) 60 (38.5) 96 (61.5) 10–14 63 (15.9) 13 (20.6) 50 (79.4) 15–17 18 (4.5) 4 (22.2) 14 (77.8) Children’s gender Male 241 (60.9) 104 (43.2) 137 (56.8) 0.811 Female 155 (39.1) 65 (41.9) 90 (58.1) Parents’ age group (in year) 18–25 27 (6.8) 13 (48.1) 14 (51.9) < 0.001 26–30 94 (23.7) 49 (52.1) 45 (47.9) 31–35 109 (27.5) 57 (52.3) 52 (47.7) 36–40 101 (25.5) 36 (35.6) 65 (64.4) 41–45 45 (11.4) 14 (31.1) 31 (68.9) 46–50 15 (3.8) 0 (0) 15 (100) ≥ 51 5 (1.3) 0 (0) 5 (100) Parents’ gender Female 157 (39.6) 74 (47.1) 83 (52.9) 0.146 Male 239 (60.4) 95 (39.7) 144 (60.3) Marital status Married 376 (94.9) 163 (43.4) 213 (56.6) 0.239 Divorced or widow 20 (5.1) 6 (30) 14 (70) Religion Muslim 346 (87.4) 149 (43.1) 197 (56.9) 0.521 Hindu 47 (11.9) 20 (42.6) 27 (57.4) Buddha 1 (.3) 0 (0) 1 (100) Cristian 2 (.5) 0 (0) 2 (100) Type of family Joint family 116 (29.3) 56 (48.3) 60 (51.7) 0.147 Nuclear family 280 (70.7) 113 (40.4) 167 (59.6) Number of children One 131 (33.1) 66 (50.4) 65 (49.6) 0.090 Two 175 (44.2) 67 (38.3) 108 (61.7) Three or more 90 (22.7) 36 (40) 54 (60) Educational qualification ≤ High school 118 (29.8) 46 (39) 72 (61) 0.176 College education 99 (25) 44 (44.4) 55 (55.6) Graduate 103 (26) 52 (50.5) 51 (49.5) Post graduate 76 (19.2) 27 (35.5) 49 (64.5) Occupation Service 96 (24.2) 41 (42.7) 55 (57.3) 0.419 Business 60 (15.2) 32 (53.3) 28 (46.7) Unemployed 21 (5.3) 10 (47.6) 11 (52.4) Student 6 (1.5) 2 (33.3) 4 (66.7) Home maker 157 (39.6) 65 (41.4) 92 (58.6) Healthcare 26 (6.6) 7 (26.9) 19 (73.1) Daily labor 30 (7.6) 12 (40) 18 (60) Table 1 Descriptive analysis: sociodemographic characteristics, COVID-19 threat, and parental vaccine hesitancy V i bl T l l (%) Lik lih d f i i hild Table 1 Descriptive analysis: sociodemographic characteristics, COVID-19 threat, and parental vaccine hesitancy Table 1 Descriptive analysis: sociodemograph Variables Total sample All participants 396 (100) Children’s age group (in year) 0–4 159 (40.2) 5–9 156(39.4) 10–14 63 (15.9) 15–17 18 (4.5) Children’s gender Male 241 (60.9) Female 155 (39.1) Parents’ age group (in year) 18–25 27 (6.8) 26–30 94 (23.7) 31–35 109 (27.5) 36–40 101 (25.5) 41–45 45 (11.4) 46–50 15 (3.8) ≥ 51 5 (1.3) Parents’ gender Female 157 (39.6) Male 239 (60.4) Marital status Married 376 (94.9) Divorced or widow 20 (5.1) Religion Muslim 346 (87.4) Hindu 47 (11.9) Buddha 1 (.3) Cristian 2 (.5) Type of family Joint family 116 (29.3) Nuclear family 280 (70.7) Number of children One 131 (33.1) Two 175 (44.2) Three or more 90 (22.7) Educational qualification ≤ High school 118 (29.8) College education 99 (25) Graduate 103 (26) Post graduate 76 (19.2) Occupation Service 96 (24.2) Business 60 (15.2) Unemployed 21 (5.3) Student 6 (1.5) Home maker 157 (39.6) Healthcare 26 (6.6) Daily labor 30 (7.6) Ali et al. Discussion No 362 (91.4) 163 (45) 199 (55) 0.002 Yes 34 (8.6) 6 (17.6) 28 (82.4) Perceived likelihood of children or family members’ infection in the next year Very likely 56 (14.1) 9 (16.1) 47 (83.9) < 0.001 Somewhat likely 240 (60.6) 102 (42.5) 138 (57.5) Not likely 73 (18.4) 40 (54.8) 33 (45.2) Definitely not 27 (6.8) 18 (66.7) 9 (33.3) Ali et al. Tropical Medicine and Health (2022) 50:24 Page 6 of 9 Table 1 (continued) Variables Total sample n (%) Likelihood of vaccinating children p value Not likely/definitely not n (%) Very likely/somewhat-likely n (%) Level of concern about children or family members’ infection in the next year Very concerned 49 (12.4) 14 (28.6) 35 (71.4) < 0.001 Concerned 128 (32.3) 38 (29.7) 90 (70.3) Slightly concerned 122 (30.8) 56 (45.9) 66 (54.1) Not concerned at all 97 (24.5) 61 (62.9) 36 (37.1) Bold faces are significant at a 5% significance level Bold faces are significant at a 5% significance level non-COVID-19 vaccine hesitancy among parents of children with NDD in another country [4]. Nonethe- less, a contemporary study conducted in Taiwan among caregivers of children with attention–deficit/hyperac- tivity disorder found that 37% of caregivers hesitated or refused to vaccinate their children against COVID-19 [18]. In addition, 42% and 52% of parents of healthy chil- dren in the USA and China, respectively, were hesitant to vaccinate their children against COVID-19 [12, 19]. representative survey found a significantly high preva- lence of parental vaccine hesitancy in Bangladesh. Logistic regression analysis suggested that location of residence, perception about vaccine safety and effec- tiveness for children, experience of family members being testing positive for COVID-19, and a family members’ death due to COVID-19 strongly predicted vaccine hesitancy. Furthermore, COVID-19 threat per- ceptions were significantly associated with parental vaccine hesitancy for their children with NDD. In line with the findings of previous studies [19, 20], we found higher vaccine hesitancy among younger parents as well as parents of younger children. The safety and effi- cacy of the COVID-19 vaccine in children aged < 12 years remains unclear, which might cause higher vaccine hesi- tancy among parents of young children [21]. However, COVID-19 vaccine hesitancy is generally high among Bangladeshi adults [14]; however, we found a higher parental vaccine hesitancy among parents for their chil- dren with NDD in this study (32.5% vs. 42.7%). Discussion This prevalence was also much higher than that of common Fig. 1 Likelihood of COVID-19 vaccine acceptance/refusal by Bangladeshi parents for children with neurodevelopmental disorders ood of COVID-19 vaccine acceptance/refusal by Bangladeshi parents for children with neurodevelopmental disorders Ali et al. Tropical Medicine and Health (2022) 50:24 Page 7 of 9 Table 2 Multiple logistic regression: predictors of parental vaccine hesitancy in study participants Bold faces are significant at a 5% significance level Variables Adjusted OR SE 95% CI p value Children’s age group (in year) 0–4 1.57 0.83 0.31 8.02 0.588 5–10 0.61 0.82 0.12 3.04 0.550 11–14 0.41 0.86 0.08 2.23 0.304 15–< 18 References Parents’ age group (in year) 18–25 Reference 26–30 1.25 0.62 0.37 4.18 0.722 31–35 1.31 0.62 0.39 4.43 0.667 36–40 0.97 0.64 0.28 3.40 0.959 41–45 0.79 0.74 0.19 3.33 0.745 ≥ 46 1.10 0.81 0.31 3.54 0.878 Current living location Central zone including Dhaka 7.25 0.45 2.98 17.67 < 0.001 North zone 17.15 0.55 5.87 50.09 < 0.001 South zone Reference Do you think the COVID-19 vaccine will be safe and effective for Bang‑ ladeshi children No 3.22 0.79 1.68 15.19 0.049 Yes 0.04 0.39 0.02 0.09 < 0.001 Skeptical Reference Have you taken or plan to take the COVID-19 vaccine No 12.14 0.18 8.48 17.36 < 0.001 Yes Reference Were you or your family member(s) tested positive for COVID-19 No 2.13 0.35 1.07 4.25 0.032 Yes Reference Have you lost any of your family member(s) for COVID-19 No 2.19 0.38 1.04 4.62 0.040 Yes Reference Perceived likelihood of children or family members’ infection in the next year Very likely Reference Somewhat likely 4.54 0.52 0.99 22.36 0.051 Not likely 4.71 0.63 1.32 15.67 0.017 Definitely not 4.99 0.79 1.81 13.78 0.002 Level of concern about children or family members’ infection in the next year Very concerned Reference Concerned 0.69 0.58 0.22 2.16 0.527 Slightly concerned 1.04 0.56 0.34 3.21 0.943 Not concerned at all 2.34 0.65 1.65 8.37 0.043 Table 2 Multiple logistic regression: predictors of parental vaccine hesitancy in study participants higher vaccine hesitancy among parents who lived in the central and northern zones of Bangladesh [14]. How- ever, it is impossible to infer the cause of higher vaccine hesitancy in these areas from our data, thus indicat- ing the need for additional studies to understand causal relationships. Discussion Our analyses found that one-fifth of the participants were either not vaccinated or chose not receive the COVID-19 vaccination themselves. Unsurprisingly, 80% of the non-vaccinated parents were hesitant to vacci- nate their children. Previous studies have also observed a significant association between parental willingness to get vaccinated and reported intentions to vaccinate children [19, 21]. However, we found a lower prevalence of parental vaccine hesitancy for their children among participants who reported that their family members tested positive for or died of COVID-19. Therefore, it can be assumed that perception of disease threats drove these parents to make a favorable decision regarding vaccination. Disease threat perception is the key to theories of many health behaviors. A systematic review and meta-analysis concluded that the likelihood of risk, susceptibility, and severity from the disease is significantly associated with vaccine hesitancy [22]. Furthermore, a recent study revealed that high COVID-19 risk perception was associ- ated with reduced vaccine hesitancy [23]. Another study suggested that reduced risk perception is associated with reduced COVID-19 vaccine uptake [24]. In agree- ment with these findings, our study found that perceived COVID-19 threat was one of the strongest predictors of parental vaccine hesitancy. In contrast, an empirical investigation revealed that, among others, the safety of the COVID-19 vaccine outweighs disease risk percep- tion when predicting vaccine hesitancy [25]. Our study also found a high prevalence of vaccine hesitancy among parents who remained skeptical or did not perceive that the vaccine would be safe and effective for Bangladeshi children. Perceived likelihood of children or family members’ infection in the next year Very likely Reference Somewhat likely 4.54 0.52 0.99 22.36 0.051 Not likely 4.71 0.63 1.32 15.67 0.017 Definitely not 4.99 0.79 1.81 13.78 0.002 This study had several limitations. First, the cross-sec- tional nature of this study only portrays the community response at the time of the study. However, studies have found that vaccine hesitancy is complex in disposition and is adherence-specific, varying over time, location, and the perceived behavioral nature of the community [26–28]. Second, the inclusion of the Bangladeshi tribal population may help increase the generalizability of the findings. Despite these limitations, this is the first study to provide baseline evidence regarding COVID-19 vac- cine hesitancy among parents of children with NDD, identifying a range of subgroups of the parents that must be considered during widespread vaccination dis- cussions in low- and middle-income countries. Acknowledgements f 5. Zerbo O, Modaressi S, Goddard K, Lewis E, Fireman BH, Daley MF, et al. Vaccination patterns in children after autism spectrum disorder diagnosis and in their younger siblings. JAMA Pediatr [Internet]. 2018;172(5):469. 5. Zerbo O, Modaressi S, Goddard K, Lewis E, Fireman BH, Daley MF, et al. Vaccination patterns in children after autism spectrum disorder diagnosis and in their younger siblings. JAMA Pediatr [Internet]. 2018;172(5):469. The authors are thankful to the participants for providing the information used to conduct the study. We also thank Professor Dr. Ahmed Hossain, PhD (Biostatistics) for his help in data analysis and result interpretations. 6. Goin-Kochel RP, Mire SS, Dempsey AG. Emergence of autism spectrum disorder in children from simplex families: relations to parental percep‑ tions of etiology. J Autism Dev Disord [Internet]. 2015;45(5):1451–63. https://doi.org/10.1007/s10803-014-2310-8. Funding This research did not receive any specific Grant from funding agencies in the public, commercial, or not-for-profit sectors. 9. Tinker SC, Cogswell ME, Peacock G, Ryerson AB. Important considerations for COVID-19 vaccination of children with developmental disabilities. Pediatrics [Internet]. 2021;148(4):e2021053190. 9. Tinker SC, Cogswell ME, Peacock G, Ryerson AB. Important considerations for COVID-19 vaccination of children with developmental disabilities. Pediatrics [Internet]. 2021;148(4):e2021053190. Abbreviations d Abbreviations NDD: Neurodevelopmental disorders; AOR: Adjusted odds ratio; CI: Confi‑ dence interval; COVID-19: Coronavirus disease of 2019. Data availability Data will be made available on request. 10. Gleason J, Ross W, Fossi A, Blonsky H, Tobias J, Stephens M. The devastat‑ ing impact of Covid-19 on individuals with intellectual disabilities in the United States. NEJM Catal Innov Care Deliv. 2021;i 10. Gleason J, Ross W, Fossi A, Blonsky H, Tobias J, Stephens M. The devastat‑ ing impact of Covid-19 on individuals with intellectual disabilities in the United States. NEJM Catal Innov Care Deliv. 2021;i Declarations 11. Alfieri NL, Kusma JD, Heard-Garris N, Davis MM, Golbeck E, Barrera L, et al. Parental COVID-19 vaccine hesitancy for children: vulnerability in an urban hotspot. BMC Public Health [Internet]. 2021;21(1):1662. https://doi. org/10.1186/s12889-021-11725-5. 11. Alfieri NL, Kusma JD, Heard-Garris N, Davis MM, Golbeck E, Barrera L, et al. Parental COVID-19 vaccine hesitancy for children: vulnerability in an urban hotspot. BMC Public Health [Internet]. 2021;21(1):1662. https://doi. org/10.1186/s12889-021-11725-5. Discussion Finally, Level of concern about children or family members’ infection in the next year Bold faces are significant at a 5% significance level Very concerned Reference Concerned 0.69 0.58 0.22 2.16 0.527 Slightly concerned 1.04 0.56 0.34 3.21 0.943 Not concerned at all 2.34 0.65 1.65 8.37 0.043 more studies are warranted to understand what drives young parents not to vaccinate their children against COVID-19. Conversely, in agreement with our previ- ous study on the adult population, we found remarkably Ali et al. Tropical Medicine and Health (2022) 50:24 Ali et al. Tropical Medicine and Health (2022) 50:24 Page 8 of 9 Page 8 of 9 face-to-face data collection from randomly selected regions throughout Bangladesh would have reduced non- response bias and provided a better representation of the population in the sample, thus increasing the generaliz- ability of the study. Hena Pain Physiotherapy and Public Health Research Center (HPRC), Uttara Model Town, Dhaka 1230, Bangladesh. 3 Advanced Physiotherapy and Rehab Solution, Women’s Children’s and General Hospital, Dhanmondi, Dhaka 1209, Bangladesh. 4 Department of Neurosurgery, Ibrahim Cardiac Hospital and Research Institute, Dhaka, Bangladesh. 5 Zaman Mordan Hospital, Sherpur Sadar, Sherpur 2100, Bangladesh. 6 Jatio Protibondhi Seba O Sahajjo Kendro, Gopalgonj Sadar, Gopalgonj 8100, Bangladesh. 7 Department of Physiotherapy, Mount Adora Hospital, Akhalia, Sylhet 3100, Bangladesh. 8 Bashundhara Kings Football Club, Bashundhara R/A, Dhaka 1229, Bangladesh. 9 Save the Chil‑ dren, Gulshan, 2, Dhaka 1212, Bangladesh. 10 University of Turku, 20014 Turun yliopisto, Finland. Competing interests 14. Ali M, Hossain A. What is the extent of COVID-19 vaccine hesitancy in Bangladesh? A cross-sectional rapid national survey. BMJ Open [Internet]. 2021;11(8):e050303. https://doi.org/10.1136/bmjopen-2021-050303. p g All authors declare no conflicts of interest. Authors’ contributions MA: conceived and designed the experiments; performed the experiments; analyzed and interpreted the data; contributed reagents, materials, analysis tools, or data; wrote the paper. SA, ASB, A-sS, MdAI, TAU, TSP, ZT, and USK: performed the experiments; contributed reagents, materials, analysis tools, or data; wrote the paper. All authors read and approved the final manuscript. 7. Kuwaik GA, Roberts W, Zwaigenbaum L, Bryson S, Smith IM, Szatmari P, et al. Immunization uptake in younger siblings of children with autism spectrum disorder. Autism [Internet]. 2014;18(2):148–55. https://doi.org/ 10.1177/1362361312459111. 7. Kuwaik GA, Roberts W, Zwaigenbaum L, Bryson S, Smith IM, Szatmari P, et al. Immunization uptake in younger siblings of children with autism spectrum disorder. Autism [Internet]. 2014;18(2):148–55. https://doi.org/ 10.1177/1362361312459111. 8. Rizzo R, Karlov L, Maugeri N, Di Silvestre S, Eapen V. Impact of the COVID- 19 pandemic on family wellbeing in the context of neurodevelopmental disorders. Neuropsychiatr Dis Treat [Internet]. 2021;17:3007–14. 8. Rizzo R, Karlov L, Maugeri N, Di Silvestre S, Eapen V. Impact of the COVID- 19 pandemic on family wellbeing in the context of neurodevelopmental disorders. Neuropsychiatr Dis Treat [Internet]. 2021;17:3007–14. Consent for publication Not applicable. Not applicable. References 1. Aranda S. Ten threats to global health in 2019 [Internet]. World Health Organisation (WHO). 2019; p. 1–18. 1. Aranda S. Ten threats to global health in 2019 [Internet]. World Health Organisation (WHO). 2019; p. 1–18. 1. Aranda S. Ten threats to global health in 2019 [Internet]. World Health Organisation (WHO). 2019; p. 1–18. 2. MacDonald NE. Vaccine hesitancy: definition, scope and determinants. Vaccine [Internet]. 2015;33(34):4161–4. 2. MacDonald NE. Vaccine hesitancy: definition, scope and determinants. Vaccine [Internet]. 2015;33(34):4161–4. 3. Olusanya OA, Bednarczyk RA, Davis RL, Shaban-Nejad A. Addressing parental vaccine hesitancy and other barriers to childhood/adolescent vaccination uptake during the coronavirus (COVID-19) pandemic. Front Immunol [Internet]. 2021;12:855. https://doi.org/10.3389/fimmu.2021. 663074/full. 3. Olusanya OA, Bednarczyk RA, Davis RL, Shaban-Nejad A. Addressing parental vaccine hesitancy and other barriers to childhood/adolescent vaccination uptake during the coronavirus (COVID-19) pandemic. Front Immunol [Internet]. 2021;12:855. https://doi.org/10.3389/fimmu.2021. 663074/full. 4. Bonsu NEM, Mire SS, Sahni LC, Berry LN, Dowell LR, Minard CG, et al. Understanding vaccine hesitancy among parents of children with autism spectrum disorder and parents of children with non-autism developmen‑ tal delays. J Child Neurol [Internet]. 2021;36(10):911–8. https://doi.org/10. 1177/08830738211000505. 4. Bonsu NEM, Mire SS, Sahni LC, Berry LN, Dowell LR, Minard CG, et al. Understanding vaccine hesitancy among parents of children with autism spectrum disorder and parents of children with non-autism developmen‑ tal delays. J Child Neurol [Internet]. 2021;36(10):911–8. https://doi.org/10. 1177/08830738211000505. Conclusions Given the increased prevalence of underlying health con- ditions, suboptimal vaccination rates, and systemic ineq- uities, children with NDD are likely to have a higher risk of being infected with COVID-19 and its outcomes. Poli- cymakers, public health practitioners, and pediatricians should address the findings of this study when discussing and implementing strategies to ensure that children with NDD, their caregivers and family members, and service providers receive the COVID-19 vaccine. In addition, highlighting the unique considerations for COVID-19 vaccination for children with NDD can support equitable vaccination access for these children and their families. Received: 30 November 2021 Accepted: 15 March 2022 Received: 30 November 2021 Accepted: 15 March 2022 Ethics approval and consent to participate Ethical clearances for this study were obtained from the institutional review board of Uttara Adhunik Medical College and Hospital. The voluntary nature of the interview was explained to the participants. All the participants have provided written informed consent. The Strengthening the Reporting of Observational Studies in Epidemiology guidelines were strictly followed throughout the study. 12. Zhang M-X, Lin X-Q, Chen Y, Tung T-H, Zhu J-S. Determinants of parental hesitancy to vaccinate their children against COVID-19 in China. Expert Rev Vaccines [Internet]. 2021. https://doi.org/10.1080/14760584.2021. 1967147. 12. Zhang M-X, Lin X-Q, Chen Y, Tung T-H, Zhu J-S. Determinants of parental hesitancy to vaccinate their children against COVID-19 in China. Expert Rev Vaccines [Internet]. 2021. https://doi.org/10.1080/14760584.2021. 1967147. 13. Aggarwal S, Madaan P, Sharma M. Vaccine hesitancy among par‑ ents of children with neurodevelopmental disabilities: a possible threat to COVID-19 vaccine coverage. J Child Neurol [Internet]. 2021;13(10):088307382110421. https://doi.org/10.1177/0883073821 1042133. cross-sectional study. F1000Research [Internet]. 2022;11:90. https://doi. org/10.12688/f1000research.76181.2. Acceptance of a pandemic influenza vaccine: a systematic review of surveys of the general public. Infect Drug Resist [Internet]. 2011;4(1):197. Author details 1 Author details 1 Department of Physiotherapy and Rehabilitation, Uttara Adhunik Medical College and Hospital, Uttara Model Town, Dhaka 1230, Bangladesh. 2 Hasna 15. Ali M, Ahmed S, Bonna AS, Sarkar A, Islam MA, Urmi TA, et al. Parental coronavirus disease vaccine hesitancy for children in Bangladesh: a 15. Ali M, Ahmed S, Bonna AS, Sarkar A, Islam MA, Urmi TA, et al. Parental coronavirus disease vaccine hesitancy for children in Bangladesh: a Page 9 of 9 Ali et al. Tropical Medicine and Health (2022) 50:24 Ali et al. Tropical Medicine and Health (2022) 50:24 cross-sectional study. F1000Research [Internet]. 2022;11:90. https://doi. org/10.12688/f1000research.76181.2. cross-sectional study. F1000Research [Internet]. 2022;11:90. https://doi. org/10.12688/f1000research.76181.2. cross-sectional study. F1000Research [Internet]. 2022;11:90. https://doi. org/10.12688/f1000research.76181.2. 16. Taherdoost H. Determining sample size; How to calculate survey sample size. Int J Econ Manag Syst [Internet]. 2017;2(2):237–9. 17. James E, Joe W, Chadwick C. Organizational research: determining appropriate sample size in survey research. Inf Technol Learn Perform J. 2001;19(1):43–50. 18. Tsai CS, Hsiao RC, Chen YM, Yen CF. Factors related to caregiver intentions to vaccinate their children with attention-deficit/hyperactivity disorder against covid-19 in Taiwan. Vaccines [Internet]. 2021;9(9):983. 19. Szilagyi PG, Shah MD, Delgado JR, Thomas K, Vizueta N, Cui Y, et al. Parents’ intentions and perceptions about COVID-19 Vaccination for their children: results from a national survey. Pediatrics [Internet]. 2021;148(4):e2021052335. 20. Yılmaz M, Sahin MK. Parents’ willingness and attitudes concerning the COVID-19 vaccine: a cross-sectional study. Int J Clin Pract [Internet]. 2021;75(9):e14364. https://doi.org/10.1111/ijcp.14364. 21. Teasdale CA, Borrell LN, Shen Y, Kimball S, Rinke ML, Fleary SA, et al. Paren‑ tal plans to vaccinate children for COVID-19 in New York city. Vaccine [Internet]. 2021;39(36):5082–6. https://doi.org/10.1016/j.vaccine.2021.07. 058. 22. Brewer NT, Chapman GB, Gibbons FX, Gerrard M, McCaul KD, Weinstein ND. Meta-analysis of the relationship between risk perception and health behavior: the example of vaccination. Heal Psychol [Internet]. 2007;26(2):136–45. https://doi.org/10.1037/0278-6133.26.2.136. 23. Caserotti M, Girardi P, Rubaltelli E, Tasso A, Lotto L, Gavaruzzi T. Associa‑ tions of COVID-19 risk perception with vaccine hesitancy over time for Italian residents. Soc Sci Med [Internet]. 2021;272:113688. 24. Kucukkarapinar M, Karadag F, Budakoglu I, Aslan S, Ucar O, Yay A, et al. COVID-19 vaccine hesitancy and its relationship with illness risk percep‑ tions, affect, worry, and public trust: an online serial cross-sectional survey from Turkey. Psychiatry Clin Psychopharmacol [Internet]. 2021;31(1):98–109. 25. Karlsson LC, Soveri A, Lewandowsky S, Karlsson L, Karlsson H, Nolvi S, et al. Fearing the disease or the vaccine: the case of COVID-19. Pers Individ Dif [Internet]. 2021;172:110590. 26. Xiao X, Wong RM. Vaccine hesitancy and perceived behavioral control: a meta-analysis. Vaccine [Internet]. 2020;38(33):5131–8. 27. Larson HJ, Jarrett C, Eckersberger E, Smith DMD, Paterson P. Understand‑ ing vaccine hesitancy around vaccines and vaccination from a global per‑ spective: a systematic review of published literature, 2007–2012. Vaccine. 2014;32:2150–9. 28. Wilson K, Nguyen HH, Brehaut H. Acceptance of a pandemic influenza vaccine: a systematic review of surveys of the general public. Infect Drug Resist [Internet]. 2011;4(1):197. 28. Wilson K, Nguyen HH, Brehaut H. • fast, convenient online submission • thorough peer review by experienced researchers in your ﬁeld • rapid publication on acceptance • support for research data, including large and complex data types • gold Open Access which fosters wider collaboration and increased citations maximum visibility for your research: over 100M website views per year • At BMC, research is always in progress. Learn more biomedcentral.com/submissions Ready to submit your research Ready to submit your research ? Choose BMC and benefit from: ? Choose BMC and benefit from: Publisher’s Note Springer Nature remains neutral with regard to jurisdictional claims in pub‑ lished maps and institutional affiliations. Springer Nature remains neutral with regard to jurisdictional claims in pub‑ lished maps and institutional affiliations. Springer Nature remains neutral with regard to jurisdictional claims in pub‑ lished maps and institutional affiliations. • fast, convenient online submission • thorough peer review by experienced researchers in your ﬁeld • rapid publication on acceptance • support for research data, including large and complex data types • gold Open Access which fosters wider collaboration and increased citations maximum visibility for your research: over 100M website views per year • At BMC, research is always in progress. Learn more biomedcentral.com/submissions Ready to submit your research Ready to submit your research ? Choose BMC and benefit from: ? Choose BMC and benefit from: • fast, convenient online submission • thorough peer review by experienced researchers in your ﬁeld • rapid publication on acceptance • support for research data, including large and complex data types • gold Open Access which fosters wider collaboration and increased citations maximum visibility for your research: over 100M website views per year • At BMC, research is always in progress. Learn more biomedcentral.com/submissions Ready to submit your research Ready to submit your research ? Choose BMC and benefit from: ? Choose BMC and benefit from:
https://openalex.org/W2740293514	https://discovery.ucl.ac.uk/10052247/1/c7tb02899g.pdf	English	null	Dual controlled delivery of squalenoyl-gemcitabine and paclitaxel using thermo-responsive polymeric micelles for pancreatic cancer	Journal of controlled release	2,017	cc-by	9,284	Introduction at very high doses via 30 minute intravenous infusion sessions at 41000 mg m2. Pancreatic cancer is the 4th most lethal cancer in Europe with disappointing prognosis and survival rates for 1, 5, and 10 years patients not exceeding 21%, 3%, and 1%, respectively1,2 – nearly the same since the early 1970s. More than 100 000 patients are diagnosed with pancreatic cancer in Europe annually which is now ranked as the 9th most common cancer in Europe and the 4th most common cause of cancer death in the Western world. Gemcitabine (2,2-difluoro 2-deoxycytidine) (GEM) is the front line drug for pancreatic cancer, with relatively moderate therapeutic potency due to poor pharmacokinetics, high dosage required for effective therapy, and the rapid development of drug resistance.3,4 GEM is a highly soluble molecule that is rapidly de-aminated (intra- and extra-cellularly) by cytidine deaminase (CDA) – which is abundant in blood plasma and the liver- and results in the non- cytotoxic metabolite 2-deoxy-2,2-difluorouridine.5 As a result, GEM has an extremely low plasma half-life (o60 minutes, in human subjects, data can be accessed here6) and hence, to compensate for the poor pharmacokinetics, GEM is administered Recently, the combination of GEM with other chemotherapeutic agents has been explored as an aggressive means to maximize the potency of therapeutic protocols, with the most notable examples the FOLFIRINOX7 protocol, and the co-administration of GEM with paclitaxel (in the form of an albumin carrying PTX vehicle, nab-PTX). In a large scale phase III clinical study, it was found that the rate of survival was significantly higher in the nab- PTX–GEM group than in the GEM group—by 59% at 1 year (35% vs. 22%) and by more than 100% at 2 years (9% vs. 4%).8 Interestingly, nab-PTX was found to improve the tumor uptake of GEM in solid tumors and increased the levels of the non- deaminated active molecule by suppressing the activity of CDA through the activation of intracellular reactive oxygen species (ROS) secretion;9 in addition, PTX was found to increase the permeability of the tumors by decreasing the stiffness and loosening of the tumor outer stroma.10 These findings imply that the co-delivery of GEM with nab-PTX could further enhance the therapeutic efficiency by exerting augmented synergism between the two drugs. a University College London, School of Pharmacy, 29-39 Brunswick Square, London WC1N 1AX, UK. E-mail: g.pasparakis@ucl.ac.uk b Institut Galien, UMR 8612, CNRS, Universite´ Paris-Sud, Universite´ Paris-Saclay, Faculte´ de Pharmacie, 5 rue Jean-Baptiste Cle´ment, Chaˆtenay-Malabry Cedex F-92296, France † Electronic supplementary information (ESI) available. See DOI: 10.1039/c7tb02899g Open Access Article. Published on 28 March 2018. Downloaded on 7/11/2018 3:28:19 PM. This article is licensed under a Creative Commons Attribution 3.0 Unported Licence. Cite this: J. Mater. Chem. B, 2018, 6, 2230 Cite this: J. Mater. Chem. B, 2018, 6, 2230 Mandana Emamzadeh,a Didier Desmae¨le,b Patrick Couvreurb and In this study we report the synthesis of a themroresponsive block copolymer by reversible addition fragmentation transfer polymerization comprising poly(2-ethylhexyl methacrylate)-b-poly[di(ethylene glycol)methyl ether methacrylate-co-oligo(ethylene glycol)methyl ether methacrylate] as hydrophobic and thermoresponsive blocks respectively. The polymer self-assembles into sub-50 micelles and can carry simultaneously two drug molecules, namely squalene-gemcitabine and paclitaxel. Both drugs can be released from the micellar compartment in a thermally controlled manner owing to the controllable disruption of the micellar corona above the lower critical solution temperature of the polymer. We demonstrate that the formulation augments synergistically the cytotoxicity of the two drugs in vitro against a model pancreatic cancer cell line. More importantly, it is shown that the polymer exerts a direct interaction with the cell membrane which further augments the cytotoxicity of the drug cargo in a thermally controlled manner. Received 7th November 2017, Accepted 9th March 2018 DOI: 10.1039/c7tb02899g rsc.li/materials-b Journal of Materials Chemistry B Journal of Materials Chemistry B This journal is ©The Royal Society of Chemistry 2018 Introduction Therefore, we hypothesized that the development of drug delivery systems that co-carry multiple GEM and PTX molecules in a single carrier could augment the synergistic activity of the two drugs and also, possibly, alleviate the pharmacokinetic limitations, such as the rapid deamination of GEM in blood plasma, and the absence of alternative formulations of nab-PTX, 2230 \| J. Mater. Chem. B, 2018, 6, 2230--2239 This journal is ©The Royal Society of Chemistry 2018 View Article Online Paper Journal of Materials Chemistry B especially given the diﬃculty to formulate the PTX molecule in more compliant forms (it is currently administered in alcohol/ castor oil mixtures).11 Materials All solvents were of analytical or HPLC grade and were pur- chased from Sigma-Aldrich. Double distilled water was used throughout all the experiments. 2,20-Azobis(2-methylpropionitrile) (AIBN), 2-butanone, 2-cyano-2-propyl benzodithioate (CTA/RAFT agent), 2-ethylhexyl methacrylate (EHMA), chloroform, diethyl ether, di(ethylene glycol)methyl ether methacrylate (DiEGMA), pyrene, fetal Bovine serum (FBS), hexane, oligo(ethylene glycol)- methyl ether methacrylate (OEGMA300), phosphate buﬀered saline (PBS), penicillin–Streptomycin, L-glutamine, Dulbecco’s modified eagle’s medium-high glucose (DMEM), dimethyl sulfoxide (DMSO), dimethylformamide (DMF) thiazolyl blue tetrazolium bromide (MTT) were purchased from Sigma-Aldrich. Paclitaxel (PTX), Taxus brevifolia was purchased from Fluorochem Ltd. Squalenoyl-gemcitabine (Sq-Gem) was synthesized by our pre- viously published synthesis route.21 Synthesis of fluorescent polymer for cellular uptake assessment A co-polymer of DiEGMA and OEGMA300 with molar ratio of 60 : 40 and degree of polymerization of 120 was synthesized. DiEGMA (6 mmol, 1.107 mL), OEGMA300 (4 mmol, 1.143 mL), AIBN (0.02 mmol, 3.3 mg) and CTA (0.08 mmol, 18 mg) were dissolved in 2-butanone (8 mL) in a 25 mL round bottom flask. The reaction mixture was sealed with rubber septum and purged with argon. The polymerization started by placing the vessel in an oil bath at 75 1C and was left under magnetic stirring. After 24 hours, the hydrophobic block with degree of polymerization of 50 was synthesized by introducing an argon- purged solution (3 mL, 2-butanone) with EHMA (3.95 mmol, 885 mL) and fluorescein O-methacrylate (FOMA) (0.05 mmol, 20 mg) to the reaction via a fine needle. The reaction was stopped after 24 hours by cooling down the reaction and exposing it to air. The polymers were washed twice in excess hexane in order to dissolve unreacted monomer and precipitate synthesized polymers. The precipitated polymers were dried under reduced pressure and collected as a pink viscous fluid (yield: 73%). In the present study, we report on a novel thermoresponsive block copolymer that self-assembles into well defined, sub-50 nm micelles that can co-carry two drug molecules, namely, PTX and a hydrophobic derivative of GEM, that is, squalene-GEM (Sq-GEM). It is demonstrated that the polymeric micelles, not only augment the synergism of the two drugs above the polymers’ LCST, but also exert an active interaction with the cell membranes, which evidently seems to accelerate the translocation of the drug molecules in the cytoplasm. To the best of our knowledge, this is one of the very first reports on drug synergism augmentation by thermoresponsive micelles accompanied by simultaneous cell membrane interaction. Open Access Article. Published on 28 March 2018. Downloaded on 7/11/2018 3:28:19 PM. This article is licensed under a Creative Commons Attribution 3.0 Unported Licence. In principle, this concept could significantly improve the delivery of drugs with a narrow therapeutic index and more significantly elevate the maximum tolerated dose. Interestingly, this approach becomes even more relevant in the case of multi-drug therapies where cumulative toxicity is a severe limitation; sur- prisingly, the number of studies on thermoresponsive micelles used for multidrug therapeutics has only very recently emerged despite the apparent formulation advantages they potentially exert.20 This journal is ©The Royal Society of Chemistry 2018 J. Mater. Chem. B, 2018, 6, 2230--2239 \| 2231 Experimental Materials Liquid state 1H NMR spectra of the polymers were recorded on a Bruker NMR spectrometer (Ultrashield 400 MHz). The polymers were dissolved in deuterated chloroform (CDCl3) for analysis. One pot synthesis of poly(DiEGMA-co-OEGMA300)-b-EHMA (P1) by RAFT polymerization DiEGMA (6 mmol, 1.107 mL), OEGMA300 (4 mmol, 1.143 mL), AIBN (0.02 mmol, 3.3 mg) and CTA (0.08 mmol, 18 mg) were dissolved in 2-butanone (8 mL) in a 25 mL round bottom flask. The reaction mixture was sealed with rubber septum and purged with argon for 10 minutes. The polymerization started by placing the vessel in an oil bath at 75 1C and was left under magnetic stirring. After 24 hours, EHMA (4 mmol, 896 mL) was dissolved in 3 mL 2-butanone purged with Ar, and introduced to the reaction via a fine needle. The reaction was stopped after 24 hours by cooling down the reaction and exposing it to air. The polymer was precipitated in excess hexane, dried under reduced pressure, and collected as a pink viscous fluid (yield: 78%). The targeted DP for hydrophilic block and hydrophobic block were 120 and 50 respectively. Of particular interest are thermoresponsive micelles com- prising the use of polymeric building blocks that exhibit a so called lower critical solution temperature (LCST).12–14 These polymers exist in well hydrated form below the LCST, and dehydrate above the LCST which in turn induces a collapsing of the micellar compartment leading to triggered release of the molecular cargo. Therefore, in a therapeutic scenario, these micelles can (1) protect the drug molecules from premature hydrolysis/deactivation (as in the case of GEM) (2) carry multiple drug molecules (3) be accumulated at the tumor sites at higher rates compared to other tissues/organs owing to the enhanced permeation and retention effect (EPR) which is observed in certain animal models (and partly in pancreatic cancer).15–19 Formulation of drug loaded micelles The drug release profile from the micelles was evaluated using a Slide-A-Lyzer dialysis cassette (MWCO 7 kDa, Spectrum laboratories). Drug loaded micelles (polymer 5 mg mL1, and PTX/Sq-Gem 500 mg mL1 each) dispersed in 5 mL of PBS pH 7.4 (0.01 M) were transferred in the dialysis cassette by using a syringe, which was then immersed in 200 mL PBS pH 7.4 (0.01 M) containing 1% Tween 20. The drug release experiments were performed under mild stirring below and above LCST, 37 1C and 40 1C respectively, in order to compare the drug release profiles at diﬀerent temperatures. At predetermined time intervals, aliquots (0.5 mL) were withdrawn from the cassette, which were replenished with fresh PBS pH 7.4 (0.01 M). The collected samples were centrifuged (Thermo Scientific Heraeus Fresco 17) at 13 300 rpm for 10 min and the drug(s) released were collected as pellets and analyzed by HPLC to measure the drug concentration released at diﬀerent time intervals. The thin-film hydration method was used to load the drug molecules in the polymeric micelles. Polymer P1 (5 mg mL1) and PTX/Sq-Gem (500 mg mL1 each) were dissolved in acetone (2 mL) and transferred in a 10 mL round bottom flask, followed by slow vacuum drying until a thin film was formed on the walls of the flask. Subsequently, hydration with 1 mL of PBS (pH 7.4) under vigorous stirring at room temperature for 5 min took place in order to form the micelles. The as prepared micelles were further sonicated for 5 min before passing them through a 0.2 mm filter in order to isolate them as a nearly monodispersed suspension. The suspension was further centrifuged (Sigma 3–16 KL) at 13300 rpm for 10 min at room temperature to confirm the complete absence of pellet formation, indicative of virtually quantitative loading efficiency under this protocol. Size exclusion chromatography Size exclusion chromatography (SEC) was conducted with DMF as the mobile phase containing 5 mM NH4BF4 additive at 70 1C at a flow rate of 1.00 mL min1. The SEC system was equipped with a solvent pump (Viscotek VE 1121), a degasser (Viscotek VE 7510), two Styragel columns (MGHHR-M E0057 and MGHHR-M E0058) and a refractive index detector (Viscotek VE 3580). DMF was used as the mobile phase at a flow rate of 1 mL min1. The apparatus was calibrated prior to analysis by using linear poly(methyl methacrylate) polymers with narrow molecular weight distributions as standards. Then, 100 mL of polymer aliquots in DMF (40 mg mL1) was injected to the instrument to determine the average molecular weight (Mn) and index of polydispersity (ÐM). J. Mater. Chem. B, 2018, 6, 2230--2239 \| 2231 This journal is ©The Royal Society of Chemistry 2018 View Article Online View Article Online Journal of Materials Chemistry B Paper Determination of the polymers’ critical micelle concentration (CMC) The morphology and size of the polymeric micelles was determined by TEM on a FEI/PHILIPS CM120 BioTwin. A carbon coated copper grid was soaked by a drop of distilled water and after a few seconds it was removed by filter paper. Then a drop of the micellar suspension was placed on the grid and left to dry in open air. Before imaging, a drop of a staining solution (1% uranyl acetate reagent) was applied to enhance the contrast of the sample and a few seconds later it was also gently absorbed by filter paper. After drying, the grid was transferred to the sample holder ready to be introduced to the electron beam. The recorded imaged were analysed by ImageJ to construct a frequency histogram and to determine the average diameter of the micelles. The critical micelle concentration of the polymers was studied using the standard pyrene protocol.22 Briefly, 20 mL aliquots of pyrene in acetone (6.5 105 M) were added to series of glass vials and the acetone was allowed to evaporate in open air. Subsequently, 1 mL of various concentrations of block copolymer samples in water (from 0.002 to 10 mg mL1) was added to each glass vial and left to equilibrate overnight under vigorous stirring. Fluorescent intensity of pyrene was measured using a spectro- fluorometer (SpectraMax multi-mode microplate reader, molecular device) where the excitation spectra were scanned from 280 to 350 nm and emission wavelength was fixed at 370 nm. Determination of the polymers’ lower critical solution temperature (LCST) Determination of the polymers’ lower critical solution temperature (LCST) Particle size measurements of the polymeric micelles were recorded using a Zetasizer Nano ZS, Malvern Instruments Ltd, UK. The scattered laser beam (632.8 nm) was measured at an angle of 1751. Samples were prepared using polymer (5 mg mL1) and PTX/Sq-Gem (500 mg mL1 each) dispersed in 1 mL of PBS pH 7.4 (0.01 M). All samples were measured in a quartz cuvette at 25 1C and the data were recorded by Malvern Zetasizer software 7.11. The polymers (5 mg) were dissolved in 1 mL of phosphate- buﬀered saline (PBS) pH 7.4 and gradually heated by immersing the solution in water bath of predetermined temperature and the LCST value was determined as the onset of optical turbidity for each sample by a Agilent Cary series UV-vis spectrophotometer at 550 nm. This journal is ©The Royal Society of Chemistry 2018 Calculation of the combination index The plating eﬃciency (PE) was calculated (eqn (2)) by dividing the number of colonies counted by the number of cells plated and then multiplying by 100: The combination index (CI) was investigated in order to measure the combinatorial therapeutic eﬀect resulting from the co-delivery of PTX and Sq-Gem. CI 4 1 implies antagonistic behavior, CI = 1 corresponds to additive behavior and CI o 1 represents synergistic behavior. The CI was calculated based on the IC50 values obtained from the MTT assay by using the formula (eqn (1)): % Plating efficiency ¼ No: of colonies counted No: of cells plated 100 (2) % Plating efficiency ¼ No: of colonies counted No: of cells plated 100 (2) (2) PE was determined to investigate the percentage of the single cells seeded in the plates that formed a colony. PE of the control was considered as 100%. By determining the PE, the survival fraction of the single cells seeded in the plates was also calculated (eqn (3)) by dividing the PE of the treated cells by the PE of the control and then multiplying by 100: CI ¼ IC50ðA þ BÞ IC50ðAÞ þ IC50ðA þ BÞ IC50ðBÞ (1) (1) where IC50 (A) and IC50 (B) are the IC50 values obtained from each drug separately. IC50 (A + B) is the IC50 value of both drugs in combination. % Survival fraction ¼ PE of the treated cells PE of the control 100 (3) (3) Polymer synthesis and design rationale Our proposed system (shown in Scheme 1) comprises the synthesis of a thermoresponsive block copolymer with a hydro- phobic block serving as the core and is made of poly(2-ethylhexyl methacrylate) (p(EHMA)). This polymer synthon was chosen by preliminary experiments which were performed with lauryl- methacrylate and n-butyl methacrylate as potential hydrophobic building blocks; p(EHMA) was shortlisted as it exhibited the lowest critical micelle concentration and optimum colloidal stability compared to the other two candidates (data not shown). The thermoresponsive corona was based on the copolymer of di(ethylene glycol)methyl ether methacrylate (DiEGMA) and oligo(ethylene glycol)methyl ether methacrylate (OEGMA300);23–25 these co-monomers were chosen as they resemble the structure of poly(ethylene glycol) which is biocompatible, and protein repellent which allows for prolonged circulation of the carriers in the bloodstream with minimum opsonisation by the mono- nuclear phagocyte system.26 In addition, this monomer system may be polymerized by controlled polymerization reactions, allowing for precise control on the final polymer structure and it is also possible to tune the LCST of the polymer by adjusting the DiEGMA : OEGMA300 monomer feed ratio. Thermo-dependent cellular uptake of fluorescent polymer The influence of the temperature on the cellular uptake of thermo-responsive nanoparticles below and above their thermal transition temperatures was studied using fluorescence micro- scopy. MiaPaCa-2 cells were seeded in 6-well plates at a concen- tration of 1 105 cells per well and incubated at 37 1C and 5% CO2 to allow the cells to attach and reach confluency. The media in each well was replaced with 2 mL of media containing 2 mg of fluorescent polymer and incubated at 37 1C (below LCST) or 40 1C (above LCST) for 30 min. The media was removed and the cells were rinsed with Dulbecco’s phosphate buﬀered saline (DPBS) once, at room temperature. The washed cells were then fixed using 4% paraformaldehyde phosphate buﬀer solution for 20 min and rinsed twice with DPBS. The cells were observed in 6-well plates using EVOSs FL Imaging microscope. Flow cytometry analysis MiaPaCa-2 cells were plated in 6-well plate at a density of 1 105 cells per well. After the cells reached confluency, each well was treated with 2 mg of poly(DiEGMA-co-OEGMA300)-b-(EHMA- co-FOMA) dissolved in 2 mL of culture medium. After 30 min, the cells were washed with cold PBS (4 1C) several times to stop the intake. 1 mL of cold PBS (4 1C) was added, the cells were scraped and transferred into Nalgenes centrifuge tubes to be analyzed by MACSQuant Analyzer 10 Flow Cytometer. these co-monomers were chosen as they resemble the structure of poly(ethylene glycol) which is biocompatible, and protein repellent which allows for prolonged circulation of the carriers in the bloodstream with minimum opsonisation by the mono- nuclear phagocyte system.26 In addition, this monomer system may be polymerized by controlled polymerization reactions, allowing for precise control on the final polymer structure and it is also possible to tune the LCST of the polymer by adjusting the DiEGMA : OEGMA300 monomer feed ratio. In vitro cytotoxicity assay An HPLC assay for the quantification of both PTX and Sq-Gem was developed using an Agilent Technologies 1200 Series HPLC system. The chromatographic separation was achieved using a Phenomenex Synergit 4 mm Polar-RP 80 Å, LC Column 250 4.6 mm. Methanol and water (85 : 15) was introduced to the column as mobile phase under isocratic conditions. The mobile phase was pumped through the column at a flow rate of 1 mL min1. The UV detector was set at 250 nm and the injection volume was 20 mL. The PTX and Gem-Sq standard solutions used for quantification were prepared by suitably diluting a 100 mg mL1 working standard of each drug in a mixture of 80 : 20 methanol : water. The data was acquired and analysed using ChemStation for LC, software by Agilent Technologies, UK. The cytotoxicity studies were performed with the use of the MTT assay. MiaPaca-2 cells were seeded in a 96-well plate at a density of 1 104 cells per well. Cells were incubated in Dulbecco’s modified eagle’s medium-high glucose and supplemented with 10% fetal bovine serum, 1% penicillin–Streptomycin and 1% L-glutamine. Incubation was performed either at 37 1C (below LCST) or 40 1C (above LCST) in humidified atmosphere with 5% CO2 for 24 h before the assay. Then, the medium was replaced with 200 mL medium containing either empty block copolymer micelles or loaded with 0.0001–10 mM of each drug. The free drugs were dissolved in DMSO and then diluted in the medium; briefly, concentrated stock solutions of Sq-Gem and PTX were prepared (1 mg/200 mL of DMSO). 3 mL of stock solution was 2232 \| J. Mater. Chem. B, 2018, 6, 2230--2239 This journal is ©The Royal Society of Chemistry 2018 View Article Online Paper Journal of Materials Chemistry B After 24 hours, the treated and untreated flasks were trypsinized, counted and subsequently 250 cells from each flask were seeded in the 6 well plate. After 14 days, the colonies were washed with PBS and then fixed using methanol and acetic acid in ratio 3:1. The colonies were stained with 0.5% crystal violet solution (diluted with methanol) for 5 min. The stained plates were rinsed in the tray full of distilled water and left in the fume hood overnight to dry. Colonies appeared as clusters of violet stained cells which could be visualized with the naked eye. J. Mater. Chem. B, 2018, 6, 2230--2239 \| 2233 This journal is ©The Royal Society of Chemistry 2018 In vitro cytotoxicity assay The number of air-dried colonies for the average of three colony counts for each plate was recorded. A cluster of 50 or more cells was counted as one colony. diluted in 1997 mL of medium to achieve 10 mM for each drug. Further dilutions were carried out to obtain less concentrated samples. After incubation for 72 h, the medium was replaced by 100 mL of fresh medium and 25 mL of MTT stock solution (5 mg mL1 in PBS) and incubated for an additional 4 h. Subsequently, the medium was removed and the formazan crystals were dissolved in 200 mL of DMSO. The plates were shaken for 2 min at room temperature before measuring the optical density (OD) at 570 nm on a SpectraMaxs M2/M2e Multi- mode Microplate Reader, with SoftMaxs Pro Software. After 24 hours, the treated and untreated flasks were trypsinized, counted and subsequently 250 cells from each flask were seeded in the 6 well plate. After 14 days, the colonies were washed with PBS and then fixed using methanol and acetic acid in ratio 3:1. The colonies were stained with 0.5% crystal violet solution (diluted with methanol) for 5 min. The stained plates were rinsed in the tray full of distilled water and left in the fume hood overnight to dry. Colonies appeared as clusters of violet stained cells which could be visualized with the naked eye. The number of air-dried colonies for the average of three colony counts for each plate was recorded. A cluster of 50 or more cells was counted as one colony. Clonogenic cell survival assay The block copolymer was synthesized by RAFT polymerization of the thermoresponsive block at DiEGMA: OEGMA300 ratio of 60 : 40 in order to achieve an optimum LCST onset of 40 1C, slightly above physiological temperature (Fig. S1, ESI†). The corona block was then used as a macro-RAFT transfer agent to 2 105 MiaPaCa-2 cells were seeded in 25 cm2 T-flask (using 5 mL of culture medium). Once the cells reached confluency, each flask was treated with diﬀerent concentrations of drugs in combinations. One flask was remained untreated as a control. J. Mater. Chem. B, 2018, 6, 2230--2239 \| 2233 This journal is ©The Royal Society of Chemistry 2018 View Article Online View Article Online Journal of Materials Chemistry B Scheme 1 The chemical structures of poly(2-ethylhexyl methacrylate)-b-poly[di(ethylene glycol)-oligo(ethylene glycol)methyl ether methacrylate] (P1), squalene-gemcitabine (Sq-GEM), paclitaxel (PTX) and key features of our proposed formulation below and above the lower critical solution of the polymer micelles. 8 March 2018. Downloaded on 7/11/2018 3:28:19 PM. under a Creative Commons Attribution 3.0 Unported Licence. Open Access Article. Published on 28 March 2018. Downloaded on 7/11/2018 3:28:19 PM. This article is licensed under a Creative Commons Attribution 3.0 Unported Licence. Scheme 1 The chemical structures of poly(2-ethylhexyl methacrylate)-b-poly[di(ethylene glycol)-oligo(ethylene glycol)methyl ether methacrylate] (P1), squalene-gemcitabine (Sq-GEM), paclitaxel (PTX) and key features of our proposed formulation below and above the lower critical solution of the polymer micelles. grow the p(EHMA) block. The final polymer structure was con- firmed by 1H NMR where all the proton peaks could be assigned successfully (Fig. 1a). From the GPC data it was possible to confirm that the obtained products had a Mn of 31 600 Da and 39 500 Da for p(DiEGMA–OEGMA300) and final P1, respectively; the degree of polymerization (DP) was close to the initial feed for both blocks (DPGPC for p(EHMA) and p(DiEGMA–OEGMA) was 43 and 116, respectively); also the final polymer had a narrow polydispersity index of 1.3 ÐM. Only a very minor tailing of the GPC trace was observed which was also persistent in the final P1 chromatogram, but this was negligible given the relatively low ÐM value. The critical micelle concentration of the polymer was determined in water by the pyrene assay and was found to be ca. 5 mg L1 (Fig. S2, ESI†), which is in accord with similar studies27 and also verify the colloidal stability of the polymer in solution. Fig. Clonogenic cell survival assay 1 (a) 1H NMR spectrum of P1 and (b) the GPC trace of the thermo- responsive block and P1. 2234 \| J. Mater. Chem. B, 2018, 6, 2230--2239 Drug loading and triggered release studies Indeed, it was possible to load PTX, Sq-GEM, or both, in the micelles at a drug loading efficiency of 10% per mass of polymer (i.e. 0.5 mg of each drug per 5 mg of polymer) without any measurable loss of drug during preparation. Remarkably, it was possible to retain an average size below 50 nm in the micelles co-loaded with both drugs. Also, from the DLS size distribution histogram, we could not trace the presence of larger particles implying the complete absence of free Sq-GEM particles (Fig. 2j–l). This means that it was possible to achieve nearly quantitative encapsulation efficiency with our drug loading protocol. We next studied the drug release profile of the two drugs separately loaded or co-loaded, below and above the polymers’ LCST (Fig. 3). In the case of Sq-GEM (Fig. 3a), below the LCST the release rate followed a zero-order rate, reaching ca. 37% and 67% within 2 and 4 hours, respectively. Above the LCST, the release rapidly reached nearly 80% which plateaued at ca. 90% in 4 hours. For PTX (Fig. 3b), it was found that 50% of the drug was released in one hour followed by a slower release rate which eventually resulted in ca. 80% release in 4 hours, below the LCST. Above LCST, the PTX release was significantly more pronounced, reaching ca. 85% in 1 hour before plateauing to 490% release within 4 hours. In the case of the sq-GEM and PTX co-loaded micelles (Fig. 3c), it was observed that PTX exhibited a more linear release profile both below and above the LCST and the release reached ca. 57% and 40% above and below the LCST, respectively. The temperature effect was more pronounced in the release pattern of sq-GEM where it was found that ca. 65% was released within 2 hours above the LCST (vs. 42% below the LCST) and reached 470% in 4 hours above the LCST compared to ca 57% below the LCST. Drug loading and triggered release studies Paclitaxel and squalene-gemcitabine were chosen as a combi- national drug ensemble as they exert synergistic activity and are currently indicated in clinical practice for patients with locally advanced pancreatic cancer (LAPC). PTX is already highly lipophilic and could be incorporated in the hydrophobic compart- ment of the polymer micelles, whereas gemcitabine in its parent form is highly water soluble and hence a synthetic more lipophilic28 derivative was chosen; squalene-gemcitabine was synthesized by the amide coupling of squalenic acid with gemcitabine at the 4-amino position that results in the formation of an amphiphilic drug with its own self-assembly properties: Sq-gem forms nanoparticles of ca. 130 nm in solution. Therefore, it was hypothesized that the squalene moiety could interact with the p(EHMA) segment of Fig. 1 (a) 1H NMR spectrum of P1 and (b) the GPC trace of the thermo- responsive block and P1. 2234 \| J. Mater. Chem. B, 2018, 6, 2230--2239 This journal is ©The Royal Society of Chemistry 2018 View Article Online Paper Journal of Materials Chemistry B the micelles so that Sq-GEM could be co-compartmentalized with PTX in the micellar core. The polymer could self-assemble in PBS pH 7.4 (0.01 M) into relatively narrow dispersed spherical micelles with an average diameter of 34.55 nm as measured by TEM (Fig. 2a and e). The size and the morphology of the drug loaded micelles was also studied by TEM (Fig. 2b–d and f–h); it was found that the loaded micelles had larger average diameter compared to the empty ones (45.93 nm, 47.71 nm, and 47.25 nm, for Sq-GEM, PTX, and Sq-GEM/PTX co-loaded micelles, respectively) which was expected due to the loading of the molecular cargo. The results were further confirmed by DLS measurements which were in almost perfect agreement with the TEM results (Fig. 2i–l). Intriguingly, it was observed that the doubly loaded micelles had only a small increase in their diameter compared to the singly loaded ones, however a careful inspection of the TEM histograms showed that the population of the micelles in the frequency bands between 40 and 45 nm slightly increased. Moreover, it is likely that Sq-GEM further enhanced the colloidal stability of the system owing to its own amphiphilic properties which could in turn enhance the drug loading capacity of the system without significant increase in the overall diameter of the formulation. Drug loading and triggered release studies Overall, it can be concluded that the effect of the micelle disruption due to the thermal collapsing of the p(DiEGMA–OEGMA300) corona had a more pronounced impact on the formulations of singly loaded drugs and to lesser extent to the PTX/Sq-GEM co-loaded micelles; it may be assumed that in the co-loaded sample, hydrophobic intermole- cular interactions between PTX, Sq-GEM and the p(EHMA) polymer segments may be stronger and hence somewhat stabilize the overall formulation against the thermal collapsing of the micellar corona. The possible interaction of PTX with Sq-GEM can be further supported by the higher drug loading efficiency Open Access Article. Published on 28 March 2018. Downloaded on 7/11/2018 3:28:19 PM. This article is licensed under a Creative Commons Attribution 3.0 Unported Licence. It was not possible to calculate a CI value for the micelles below the LCST as Sq-GEM loaded micelles were virtually non-toxic under the experimental conditions. Above LCST, the CI was 0.27 indicating a further enhancement of the synergistic activity of PTX and Sq-GEM in the formulated form. Overall, these results indicate that in all cases, the encapsulation of the drugs, individually or in combination, resulted in retention or significant enhancement of the cytotoxicity above the LCST and in a reduction of the cytotoxicity below the LCST. However, a careful interpretation of the data based merely on the drug release patterns cannot justify the large diﬀerences observed. Therefore, we hypothesized that the polymer itself may be actively interacting with cellular organelles (i.e. the cell membrane) in a temperature dependent manner which could further augment the drug uptake rates above the LCST. For this, we synthesized a block copolymer with a fluorescent tag that could be trackable with fluorescence spectrometry. The polymer was incubated with MiaPaCa-2 cells for 30 min below or above the LCST, and then the cells were visualized by fluorescence microscopy. Interestingly, it was found that the polymer only adsorbed onto the cellular membrane and did not permeate inside cells when incubated below LCST (at 37 1C) (Fig. 4a) as the fluorescence signal was persistently tracked at the periphery was not possible to extract an IC50 value (Table 1). Conversely, above the polymers’ LCST, the Sq-GEM-loaded formulation had an IC50 of 310.5 nM which is a 3.6 fold improvement compared to the non-encapsulated sample. Second, PTX was found to have an IC50 of 9.44 nM, which is in accord with previous reports.31 The micelles could reduce the cytotoxicity of the parent drug by almost 3-fold as the formulation had an IC50 of 29.22 nM which could be fully restored (and marginally improved) above the LCST (IC50 = 7.25 nM) (Table 1). Finally, the simultaneous addition of the two drugs resulted in a reduced IC50 (3.03 nM) which was significantly lower than each drug individually, indicative of the synergistic activity of the two drugs. In the micellar form, the IC50 was merely increased below the LCST (IC50 = 4.23 nM) but was significantly enhanced above the LCST (IC50 = 1.93 nM) (Table 1). Open Access Article. Published on 28 March 2018. Downloaded on 7/11/2018 3:28:19 PM. This article is licensed under a Creative Commons Attribution 3.0 Unported Licence. The synergistic activity of the two drugs was verified by the Chou-Talalay32 combination index (CI)(eqn (1)); it was found that the CI of the non-loaded drugs was 0.32 which verifies the synergism of the two drug molecules (CI o 1 is indicative of synergistic activity). It was not possible to calculate a CI value for the micelles below the LCST as Sq-GEM loaded micelles were virtually non-toxic under the experimental conditions. Above LCST, the CI was 0.27 indicating a further enhancement of the synergistic activity of PTX and Sq-GEM in the formulated form. Overall, these results indicate that in all cases, the encapsulation of the drugs, individually or in combination, resulted in retention or significant enhancement of the cytotoxicity above the LCST and in a reduction of the cytotoxicity below the LCST. However, a careful interpretation of the data based merely on the drug release patterns cannot justify the large diﬀerences observed. Therefore, we hypothesized that the polymer itself may be actively interacting with cellular organelles (i.e. the cell membrane) in a temperature dependent manner which could further augment the drug uptake rates above the LCST. For this, we synthesized a block copolymer with a fluorescent tag that could be trackable with fluorescence spectrometry. The polymer was incubated with MiaPaCa-2 cells for 30 min below or above the LCST, and then the cells were visualized by fluorescence microscopy. Interestingly, it was found that the polymer only adsorbed onto the cellular membrane and did not permeate inside cells when incubated below LCST (at 37 1C) (Fig. 4a) as the fluorescence signal was persistently tracked at the periphery Fig. 3 Drug release of (a) Sq-GEM, (b) PTX, and (c) Sq-GEM and PTX, from P1 micelles below and above the LCST. Fig. 3 Drug release of (a) Sq-GEM, (b) PTX, and (c) Sq-GEM and PTX, from P1 micelles below and above the LCST. that could be achieved (i.e. 20%) compared to the singly loaded formulations (10%); it should be noted that Sq-GEM is colloidally stable due to its own amphiphilic character and hence it could possibly contribute to the further stabilization of the formulation. that could be achieved (i.e. 20%) compared to the singly loaded formulations (10%); it should be noted that Sq-GEM is colloidally stable due to its own amphiphilic character and hence it could possibly contribute to the further stabilization of the formulation. Open Access Article. Published on 28 March 2018. Downloaded on 7/11/2018 3:28:19 PM. This article is licensed under a Creative Commons Attribution 3.0 Unported Licence. Fig. 2 TEM images of blank, Sq-GEM loaded, PTX loaded and Sq-GEM/PTX co-loaded micelles in (a–d) with their frequency histograms in (e–h), and their DLS measurements in (i–l). Fig. 2 TEM images of blank, Sq-GEM loaded, PTX loaded and Sq-GEM/PTX co-loaded micelles in (a–d) with their frequency histograms in (e–h), and their DLS measurements in (i–l). J. Mater. Chem. B, 2018, 6, 2230--2239 \| 2235 This journal is ©The Royal Society of Chemistry 2018 View Article Online View Article Online Journal of Materials Chemistry B Fig. 3 Drug release of (a) Sq-GEM, (b) PTX, and (c) Sq-GEM and PTX, from P1 micelles below and above the LCST. Table 1 IC50 values with their standard deviation from three experiments (S.D.) of Sq-GEM, PTX, and their combination in non-encapsulated (N.E.) and encapsulated formulations below and above the LCST Drug LCST IC50 (nM) S.D. Sq-Gem N.E. 1122 258.59 Below — — Above 310.4 55.44 PTX N.E. 9.44 0.17 Below 29.22 1.75 Below 7.25 0.16 Sq-Gem and PTX N.E. 3.03 0.30 Below 4.24 0.16 Above 1.93 0.10 Open Access Article. Published on 28 March 2018. Downloaded on 7/11/2018 3:28:19 PM. This article is licensed under a Creative Commons Attribution 3.0 Unported Licence. was not possible to extract an IC50 value (Table 1). Conversely, above the polymers’ LCST, the Sq-GEM-loaded formulation had an IC50 of 310.5 nM which is a 3.6 fold improvement compared to the non-encapsulated sample. Second, PTX was found to have an IC50 of 9.44 nM, which is in accord with previous reports.31 The micelles could reduce the cytotoxicity of the parent drug by almost 3-fold as the formulation had an IC50 of 29.22 nM which could be fully restored (and marginally improved) above the LCST (IC50 = 7.25 nM) (Table 1). Finally, the simultaneous addition of the two drugs resulted in a reduced IC50 (3.03 nM) which was significantly lower than each drug individually, indicative of the synergistic activity of the two drugs. In the micellar form, the IC50 was merely increased below the LCST (IC50 = 4.23 nM) but was significantly enhanced above the LCST (IC50 = 1.93 nM) (Table 1). The synergistic activity of the two drugs was verified by the Chou-Talalay32 combination index (CI)(eqn (1)); it was found that the CI of the non-loaded drugs was 0.32 which verifies the synergism of the two drug molecules (CI o 1 is indicative of synergistic activity). This journal is ©The Royal Society of Chemistry 2018 In vitro cytotoxicity studies The cytotoxicity of the formulations was tested against model pancreatic cancer cell line MiaPaCa-2, using the MTT assay. Initially, we tested the possible toxicity of temperature increase (to 40 1C), DMSO (which was used as a co-solvent to dissolve free drugs), and the polymer without drugs loaded above and below its LCST; it was found that these parameters did not induce any detectable cytotoxicity (Fig. S3, ESI†). Having established that the experimental conditions were not toxic to the cells, we determined the half maximal inhibitory concentration (IC50) of the drugs, the formulations and their combinations under various conditions. First, Sq-GEM was found to have an IC50 of 1122 nM which is significantly higher than the parent drug (Table 1).21,29 Also, the obtained IC50 value is considerably higher than the IC50 of Sq-GEM in its nano-assembled form (i.e. 40 nM on MiaPaCa-2 cells30), which signifies the marked variances in the cytotoxicity profile of the same molecule under diﬀerent for- mulations. Remarkably, Sq-GEM that was loaded in the micelles did not show any cytotoxicity under the conditions tested and it 2236 \| J. Mater. Chem. B, 2018, 6, 2230--2239 This journal is ©The Royal Society of Chemistry 2018 View Article Online Paper Paper Journal of Materials Chemistry B Fig. 4 Fluorescence microscopy images of MiaPaCa-2 cells incubated with P1 (a) below, and (b) above the LCST of a fluorescent P1 derivative with their respective flow cytometry graphs in (c). Scale bar is 400 mm. This article is licensed under a Creative Commons Attribution 3.0 Unported Licence. of the cells. On the other hand, when incubated with the micelles above the LCST (at 40 1C), the cells displayed stronger fluorescent signal inside the cell cytosol (Fig. 4b). These results indicated that increasing the temperature eﬀectively increases the cell internalization of the fluorescent polymer first by eliciting a direct interaction of the lipophilic pEHMA block of the polymer with the cell membrane followed by enhanced internalization, likely driven by the thermal collapsing of the PEGMA block. Open Access Article. Published on 28 March 2018. Downloaded on 7/11/2018 3:28:19 PM. This article is licensed under a Creative Commons Attribution 3.0 Unported Licence. The consistency of this internalization pattern was further confirmed by flow cytometry experiments. Fig. 4c shows a standard curve divided into two curves, one with lower intensity and the other one shifting towards higher intensity. In vitro cytotoxicity studies The first band was attributed to the fluorescence signal from the polymer adsorbed on the cell membrane while the second band corresponded to the signal resulting from the cell internalized polymer. The flow cyto- metry profile above the polymers’ LCST showed more cells with more fluorescent polymer adsorbed on the cell membrane when compared to the curve below LCST. Moreover, the fluorescent signal corresponding to internalized polymer increased significantly above the LCST as evidenced by the pronounced right-shifting of the fluorescent band. These results corroborate with our hypothesis that the polymer was initially adsorbed on the cell membrane via hydrophobic interactions primarily due to the pEHMA followed by enhanced cellular uptake driven by the collapsing of the thermo- responsive corona. In eﬀect, this mechanism could adequately explain the enhanced cytotoxicity profiles and point to a more complex and collective cytotoxic mechanism pattern which is justified by: (1) The enhanced drug release controlled by the temperature stimulus which eﬀectively leads to increased drug translocation across the cell membrane, (2) The direct inter- action of the polymer with the cell membrane which potentially results in an increased permeability of the cell membrane above the polymers LCST which could further augment the trans- location rates of both lipophilic drugs and 3. The enhanced cellular uptake of the micelles above the LCST via endocytosis mechanism.33–35 Finally, a clonogenic cell survival assay was conducted with MiaPaCa-2 cells to investigate the proliferative ability of survived cells post treatment (Fig. 5). The clonogenic Fig. 4 Fluorescence microscopy images of MiaPaCa-2 cells incubated with P1 (a) below, and (b) above the LCST of a fluorescent P1 derivative with their respective flow cytometry graphs in (c). Scale bar is 400 mm. Fig. 4 Fluorescence microscopy images of MiaPaCa-2 cells incubated with P1 (a) below, and (b) above the LCST of a fluorescent P1 derivative with their respective flow cytometry graphs in (c). Scale bar is 400 mm. Fig. 5 Digital photographs of the clonogenic assay of MiaPaCa-2 cells performed with the combination of free Sq-GEM and PTX, or loaded in micelles below and above the LCST for 14 consecutive days in 6-well plates (the diameter of each well is approx. 34 mm). Fig. Acknowledgements However, above the LCST, the survival fraction was still more diminished and was very close to the values for the free drugs (note that for 0.001 mM Sq-GEM/PTX it was statistically improved) (Fig. 6). This longer term effect of the polymer micelles further supports our proposed mechanism of action as previously discussed, especially given the fact that the release of the drugs is controlled only for a few hours by temperature (Fig. 3); therefore, a longer term mechanism as evidenced by the cell membrane interaction studies could indeed explain the longer term effects of temperature on the cytotoxicity profiles of the drugs. Notes and references 1 M. Malvezzi, P. Bertuccio, F. Levi, C. La Vecchia and E. Negri, Ann. Oncol., 2014, 25, 1650–1656. 2 D. Li, K. Xie, R. Wolﬀand J. L. Abbruzzese, Lancet, 2004, 363, 1049–1057. 3 H. A. Burris, 3rd, M. J. Moore, J. Andersen, M. R. Green, M. L. Rothenberg, M. R. Modiano, M. C. Cripps, R. K. Portenoy, A. M. Storniolo, P. Tarassoﬀ, R. Nelson, F. A. Dorr, C. D. Stephens and D. D. Von Hoﬀ, J. Clin. Oncol., 1997, 15, 2403–2413. 4 E. Mini, S. Nobili, B. Caciagli, I. Landini and T. Mazzei, Ann. Oncol., 2006, 17, v7–v12. 5 L. A. Shipley, T. J. Brown, J. D. Cornpropst, M. Hamilton, W. D. Daniels and H. W. Culp, Drug Metab. Dispos., 1992, 20, 849–855. 6 https://www.fda.gov/ohrms/dockets/ac/06/briefing/2006-4254b_ i bi l b l df 6 https://www.fda.gov/ohrms/dockets/ac/06/briefing/2006-4254b_ 11_04_KP%20GemcitabineFDAlabel42005.pdf. 6 https://www.fda.gov/ohrms/dockets/ac/06/briefing/2006 4254b_ 11_04_KP%20GemcitabineFDAlabel42005.pdf. 11_04_KP%20GemcitabineFDAlabel42005.pdf. 7 T. Conroy, F. Desseigne, M. Ychou, O. Bouche´, R. Guimbaud, Y. Be´couarn, A. Adenis, J. L. Raoul, S. Gourgou-Bourgade, C. de la Fouchardie`re, J. Bennouna, J.-B. Bachet, F. Khemissa-Akouz, D. Pe´re´-Verge´, C. Delbaldo, E. Assenat, B. Chauﬀert, P. Michel, C. Montoto-Grillot and M. Ducreux, N. Engl. J. Med., 2011, 364, 1817–1825. Conflicts of interest There are no conflicts to declare. Fig. 6 The survival fraction percentile of MiaPaCa-2 cells post treatment with free Sq-GEM and PTX, and Sq-GEM/PTX loaded micelles for 14 days. , , and ** denote moderate (0.01 4 p 4 0.05), high (0.001 4 p 4 0.01), and very high (0.0001 4 p 4 0.001) statistical diﬀerence respectively, based on p values from an unpaired Student’s t-test. In vitro cytotoxicity studies 5 Digital photographs of the clonogenic assay of MiaPaCa-2 cells performed with the combination of free Sq-GEM and PTX, or loaded in micelles below and above the LCST for 14 consecutive days in 6-well plates (the diameter of each well is approx. 34 mm). J. Mater. Chem. B, 2018, 6, 2230--2239 \| 2237 This journal is ©The Royal Society of Chemistry 2018 View Article Online Journal of Materials Chemistry B Fig. 6 The survival fraction percentile of MiaPaCa-2 cells post treatment with free Sq-GEM and PTX, and Sq-GEM/PTX loaded micelles for 14 days. , , and ** denote moderate (0.01 4 p 4 0.05), high (0.001 4 p 4 0.01), and very high (0.0001 4 p 4 0.001) statistical diﬀerence respectively, based on p values from an unpaired Student’s t-test. Journal of Materials Chemistry B Commons Attribution 3.0 Unported Licence. Paper in the longer term. Although, there are many studies in thermo- responsive block copolymer micelles, the potential role of the actual carrier, that is the polymer, is almost systematically neglected and hence we anticipate that our study will be insightful for the design of polymer based nanomedicinal formulations for cancer, especially for formulations aimed for drug combinations where the therapeutic window is far more challenging to exploit. In addition, the proposed system could be suitable for deep tissue triggered drug release via high intensity focused ultrasound probes37,38 or other types of remotely activated39,40 therapeutic modalities. Open Access Article. Published on 28 March 2018. Downloaded on 7/11/2018 3:28:19 PM. This article is licensed under a Creative Commons Attribution 3.0 Unported Licence. Acknowledgements We acknowledge the Engineering and Physical Sciences Research Council (G. P., EPSRC EP/M014649/1) and University College London (G. P., Excellence Fellowship) for financial support. We also thank Ms Mina Emamzadeh for training and helping with the in vitro studies. cell survival assay determines the long-term cytotoxic effects of drugs by measuring their ability to propagate from a single cell to a clone and produce a viable colony.36 Initially MiaPaCa-2 cells were grown on 25 cm2 T-flask which were then treated with different concentrations of both drugs in the formulated or non- formulated form, below and above the polymers’ LCST. The survival fraction was calculated 14 days post treatment in order to probe the cytotoxic properties of the micelles in the longer term. As expected, it was found that the survival fraction decreased as the concentration of the drugs increased (Fig. 6). Interestingly, the micelles below the LCST, were found to decrease the survival fraction induced by the drugs at concentrations from 0.001 mM to 0.1 mM. However, above the LCST, the survival fraction was still more diminished and was very close to the values for the free drugs (note that for 0.001 mM Sq-GEM/PTX it was statistically improved) (Fig. 6). This longer term effect of the polymer micelles further supports our proposed mechanism of action as previously discussed, especially given the fact that the release of the drugs is controlled only for a few hours by temperature (Fig. 3); therefore, a longer term mechanism as evidenced by the cell membrane interaction studies could indeed explain the longer term effects of temperature on the cytotoxicity profiles of the drugs. cell survival assay determines the long-term cytotoxic effects of drugs by measuring their ability to propagate from a single cell to a clone and produce a viable colony.36 Initially MiaPaCa-2 cells were grown on 25 cm2 T-flask which were then treated with different concentrations of both drugs in the formulated or non- formulated form, below and above the polymers’ LCST. The survival fraction was calculated 14 days post treatment in order to probe the cytotoxic properties of the micelles in the longer term. As expected, it was found that the survival fraction decreased as the concentration of the drugs increased (Fig. 6). Interestingly, the micelles below the LCST, were found to decrease the survival fraction induced by the drugs at concentrations from 0.001 mM to 0.1 mM. 2238 \| J. Mater. Chem. B, 2018, 6, 2230--2239 Open Access Article. Published on 28 March 2018. Downloaded on 7/11/2018 3:28:19 PM. This article is licensed under a Creative Commons Attribution 3.0 Unported Licence. 28 D. Sobot, S. Mura, S. O. Yesylevskyy, L. Dalbin, F. Cayre, G. Bort, J. Mougin, D. Desmae¨le, S. Lepetre-Mouelhi, G. Pieters, B. Andreiuk, A. S. Klymchenko, J. L. Paul, C. Ramseyer and P. Couvreur, , 2017, 8, 15678. 12 G. Pasparakis and M. Vamvakaki, Polym. Chem., 2011, 2, 1234–1248. J. Mougin, D. Desmae¨le, S. Lepetre-Mouelhi, G. Pieters, B. Andreiuk, A. S. Klymchenko, J. L. Paul, C. Ramseyer and P. Couvreur, , 2017, 8, 15678. B. Andreiuk, A. S. Klymchenko, J. L. Paul, C. Ramseyer and 13 C. d. l. H. Alarcon, S. Pennadam and C. Alexander, Chem. Soc. Rev., 2005, 34, 276–285. P. Couvreur, , 2017, 8, 15678. P. Couvreur, , 2017, 8, 15678. 29 S. Re´jiba, L. H. Reddy, C. Bigand, C. Parmentier, P. Couvreur and A. Hajri, Nanotechnol. Biol. Med., 2011, 7, 841–849. 14 X. Guo, D. Li, G. Yang, C. Shi, Z. Tang, J. Wang and S. Zhou, ACS Appl. Mater. Interfaces, 2014, 6, 8549–8559. 30 J. Caron, A. Maksimenko, J. Mougin, P. Couvreur and D. Desmaele, Polym. Chem., 2014, 5, 1662–1673. 15 A. K. Iyer, G. Khaled, J. Fang and H. Maeda, Drug Discovery Today, 2006, 11, 812–818. 31 Y. Lin, F. M. Richards, B. F. Krippendorﬀ, J. L. Bramhall, J. A. Harrington, T. E. Bapiro, A. Robertson, D. Zheleva and 31 Y. Lin, F. M. Richards, B. F. Krippendorﬀ, J. L. Bramhall, J. A. Harrington, T. E. Bapiro, A. Robertson, D. Zheleva and D. I. Jodrell, Br. J. Cancer, 2012, 107, 1692–1701. 16 H. Maeda, J. Wu, T. Sawa, Y. Matsumura and K. Hori, J. Controlled Release, 2000, 65, 271–284. J g , p , , D. I. Jodrell, Br. J. Cancer, 2012, 107, 1692–1701. 17 T. Hosoki, Am. J. Roentgenol., 1983, 140, 959–965. 32 T. C. Chou, Cancer Res., 2010, 70, 440–446. 33 S. R. Abulateefeh, S. G. Spain, K. J. Thurecht, J. W. Aylott, W. C. Chan, M. C. Garnett and C. Alexander, Biomater. Sci., 2013, 1, 434–442. 18 A. Sofuni, H. Iijima, F. Moriyasu, D. Nakayama, M. Shimizu, K. Nakamura, F. Itokawa and T. Itoi, J. Gastroenterol., 2005, 40, 518–525. 34 J. Akimoto, M. Nakayama, K. Sakai and T. Okano, Mol. Pharmaceutics, 2010, 7, 926–935. 19 K. N. Sugahara, T. Teesalu, P. P. Karmali, V. R. Kotamraju, L. Agemy, O. M. Girard, D. Hanahan, R. F. Mattrey and E. Ruoslahti, Cancer Cell, 2009, 16, 510–520. 35 Y. Hiruta, M. Shimamura, M. Conclusions This study introduces a new thermoresponsive block copolymer that self assembles in sub-50 nm micelles and can be co-loaded with two potent anticancer drugs, namely, PTX, and Sq-GEM. We demonstrated that the polymer micelles could modulate the release profiles of two drugs simultaneously and elicit synergistic cytotoxicity against a model pancreatic cell line. More importantly, it was experimentally demonstrated that the polymer actively interacted with the cell membrane which in turn thermally modulated and augmented the synergistic activity of the drugs 8 D. D. Von Hoﬀ, T. Ervin, F. P. Arena, E. G. Chiorean, J. Infante, M. Moore, T. Seay, S. A. Tjulandin, W. W. Ma, M. N. Saleh, M. Harris, M. Reni, S. Dowden, D. Laheru, N. Bahary, R. K. Ramanathan, J. Tabernero, M. Hidalgo, D. Goldstein, E. Van Cutsem, X. Wei, J. Iglesias and M. F. Renschler, N. Engl. J. Med., 2013, 369, 1691–1703. 8 D. D. Von Hoﬀ, T. Ervin, F. P. Arena, E. G. Chiorean, J. Infante, M. Moore, T. Seay, S. A. Tjulandin, W. W. Ma, M. N. Saleh, M. Harris, M. Reni, S. Dowden, D. Laheru, N. Bahary, R. K. Ramanathan, J. Tabernero, M. Hidalgo, D. Goldstein, E. Van Cutsem, X. Wei, J. Iglesias and M. F. Renschler, N. Engl. J. Med., 2013, 369, 1691–1703. 9 K. K. Frese, A. Neesse, N. Cook, T. E. Bapiro, M. P. Lolkema, D. I. Jodrell and D. A. Tuveson, Cancer Discovery, 2012, 2, 260–269. 2238 \| J. Mater. Chem. B, 2018, 6, 2230--2239 This journal is ©The Royal Society of Chemistry 2018 View Article Online Paper Journal of Materials Chemistry B 10 R. Alvarez, M. Musteanu, E. Garcia-Garcia, P. P. Lopez-Casas, 25 J. F. Lutz, O¨. Akdemir and A. Hoth, J. Am. Chem. Soc., 2006, 128, 13046–13047. 10 R. Alvarez, M. Musteanu, E. Garcia-Garcia, P. P. Lopez-Casas, D. Megias, C. Guerra, M. Munoz, Y. Quijano, A. Cubillo, J. Rodriguez-Pascual, C. Plaza, E. de Vicente, S. Prados, S. Tabernero, M. Barbacid, F. Lopez-Rios and M. Hidalgo, Br. J. Cancer, 2013, 109, 926–933. g , , , Q j , , J. Rodriguez-Pascual, C. Plaza, E. de Vicente, S. Prados, 26 D. E. Owens and N. A. Peppas, Int. J. Pharm., 2006, 307, 93–102. S. Tabernero, M. Barbacid, F. Lopez-Rios and M. Hidalgo, Br. J. Cancer, 2013, 109, 926–933. 27 X. Zhao, Z. Poon, A. C. Engler, D. K. Bonner and P. T. Hammond, Biomacromolecules, 2012, 13, 1315–1322. J. Mater. Chem. B, 2018, 6, 2230--2239 \| 2239 This journal is ©The Royal Society of Chemistry 2018 Conclusions 11 A. K. Singla, A. Garg and D. Aggarwal, Int. J. Pharm., 2002, 235, 179–192. 28 D. Sobot, S. Mura, S. O. Yesylevskyy, L. Dalbin, F. Cayre, G. Bort, Open Access Article. Published on 28 March 2018. Downloaded on 7/11/2018 3:28:19 PM. This article is licensed under a Creative Commons Attribution 3.0 Unported Licence. Matsuura, Y. Maekawa, T. Funatsu, Y. Suzuki, E. Ayano, T. Okano and H. Kanazawa, ACS Macro Lett., 2014, 3, 281–285. 20 B. Al-Lazikani, U. Banerji and P. Workman, Nat. Biotechnol., 2012, 30, 679–692. 36 P. R. Roper and B. Drewinko, Cancer Res., 1976, 36, 2182–2188. 21 P. Couvreur, B. Stella, L. H. Reddy, H. Hillaireau, C. Dubernet, D. Desmae¨le, S. Lepeˆtre-Mouelhi, F. Rocco, N. Dereuddre- Bosquet, P. Clayette, V. Rosilio, V. Marsaud, J. M. Renoir and L. Cattel, Nano Lett., 2006, 6, 2544–2548. 37 A. Z. Zardad, Y. Choonara, L. du Toit, P. Kumar, M. Mabrouk, P. Kondiah and V. Pillay, Polymers, 2016, 8, 359. 22 M.-C. Jones and J.-C. Leroux, Eur. J. Pharm. Biopharm., 1999, 48, 101–111. 38 N. Rapoport, A. M. Kennedy, J. E. Shea, C. L. Scaife and K.-H. Nam, Mol. Pharmaceutics, 2010, 7, 22–31. 23 J. P. Magnusson, A. Khan, G. Pasparakis, A. O. Saeed, W. Wang and C. Alexander, J. Am. Chem. Soc., 2008, 130, 10852–10853. 39 Y. Wang, G. Wei, X. Zhang, F. Xu, X. Xiong and S. Zhou, Adv. Mater., 2017, 29, 1605357. 40 Y. Guo, K. Jiang, Z. Shen, G. Zheng, L. Fan, R. Zhao and J. Shao, ACS Appl. Mater. Interfaces, 2017, 9, 43508–43519. 24 G. Pasparakis and C. Alexander, Angew. Chem., 2008, 120, 4925–4928. J. Mater. Chem. B, 2018, 6, 2230--2239 \| 2239
https://openalex.org/W4387641980	https://conferenceproceedings.ump.ac.id/index.php/pssh/article/download/836/821	English	null	Using Video Describing Modification of an Electric Motorcycle to Improve Writing Descriptive Text Skills for X Class Students of SMKN 1 Kalikajar Wonosobo	Proceedings Series on Social Sciences & Humanities	2,023	cc-by	7,006	Proceedings Series on Social Sciences & Humanities, Volume 12 Proceedings of International Conference on Social Science (ICONESS) ISSN: 2808-103X Proceedings Series on Social Sciences & Humanities, Volume 12 Proceedings of International Conference on Social Science (ICONESS) ISSN: 2808-103X Keywords: Corresponding Author: Khomsatun SMKN 1 Kalikajar Wonosobo Corresponding Author: Khomsatun SMKN 1 Kalikajar Wonosobo Email: satunkhomsa@gmail.com Email: satunkhomsa@gmail.com Using Video Describing Modification of an Electric Motorcycle to Improve Writing Descriptive Text Skills for X Class Students of SMKN 1 Kalikajar Wonosobo Khomsatun SMKN 1 Kalikajar Wonosobo ARTICLE INFO Article history: DOI: 10.30595/pssh.v12i.836 Submited: May 31, 2023 Accepted: August 24, 2023 Published: October 05, 2023 Keywords: Descriptive Text, Writing, Video Description Of Electric Motorcycle Corresponding Author: Khomsatun SMKN 1 Kalikajar Wonosobo Email: satunkhomsa@gmail.com ARTICLE INFO ABSTRACT This study is to see how video descriptions of electric motorcycles might help students improve their writing skills using classroom action research. This study was carried out at SMKN 1 Kalikajar Wonosobo. The total number of students was 72, and the researcher sampled 30 of them. Data was gathered from experiments and observations made during the preliminary, pre-test, and post-test. The findings of this study show that video media descriptions of electric motorcycles are useful in boosting students' descriptive text writing skills. This is indicated by the average test results have increased. The average test in the preliminary was 52, then in the pre-test was 70, and the post-test was 85. The results showed that there were significant differences between student skills, student participation, and student interests. As a result, video descriptions of electric motorcycles can assist students in developing their descriptive writing skills. This work is licensed under a Creative Commons Attribution 4.0 International License. This work is licensed under a Creative Commons Attribution 4.0 International License. 1. INTRODUCTION This means that students need the help of interesting media to help themselves so that their quality and writing skills become better, because with the help of media it will be easier for students and interested in knowing the material better. Based on these issues, this research will look into using video descriptions of electric motorcycles to help students improve their descriptive text writing skills. One of the good strategies in teaching writing is to employ video media [2]. Students might receive ideas and write them down by viewing videos. By watching movie trailers, it can be useful in teaching writing narrative text activities [5] noted. The utilization of fairy tale films is a great technique to boost students' interest in learning writing abilities [6]. In the context of this study, descriptive text, also known as "description" in writing, is an English lesson that should be mastered not only by junior high school students, but also by seniors, college students, and anybody who appreciates English. After all, we need descriptive content to improve our English skills. A descriptive paragraph elucidates a person's or object's attributes. Its purpose is to reveal and describe a certain person, location, or object. In a broad sense, description is defined as the sensory experience of how something appears, sounds, and tastes is what description is all about [7]. It's usually about visual perception, but it can also be about other kinds of perception. Descriptive writing, according to Kane's definition, is meaningful prose that represents a sensory experience, such as what shape, sound, or taste is. The majority of descriptive writing is focused on the sense of sight. However, humans can generate descriptive language from experiences other than sight. The descriptive language used in this study was chosen since it is familiar to people's everyday lives. By using video media, students will be more enthusiastic in learning English, they will also be easier and more interested in improving their writing skills in descriptive texts. Students can also easily recall English words if they listen to them in videos and come across them frequently in class. Researchers at SMKN 1 Kalikajar Wonosobo need to use video descriptions of electric motorcycles as a medium to address difficulties in order to boost learners' enthusiasm for the English language learning process. 1. INTRODUCTION Writing is one of the language skills that must be learned and mastered as a means of communication [1]. Writing must also be learned by students at all levels, not least by high school students. Writing, which may be a strong tool for finding meaning, not only to convey many sorts of ideas, but also to translate existing notions, is one way to investigate educational subjects. In the author's thoughts We can convey our feelings, ideas, imagination, and all of our thoughts by writing. Some people, however, believe that writing is the hardest ability to perfect. Some people claim that they have lots of ideas, but have a hard time putting them in writing. Thus, writing skills must continue to be trained and accustomed. g "Writing is undoubtedly one of the abilities that demands hard training because writing requires more time to consider than other skills," writes [2]. This is one of the reasons why students should study writing on a regular basis. It is true that having good writing skills the first time you do it is quite difficult, because there are various steps that must be learned before writing becomes good from day to day. "To train children who have less aptitude in the ability of writing skill writing abilities, teachers should not only in teaching grammar, vocabulary, spelling, and writing theory, but also must teach writing practice as often as possible," according to [3]. However, based on observations at SMKN 1 Kalikajar Wonosobo, in the process of mastering writing skills, students face a variety of challenges, such as the use of incorrect grammar, incorrect sentence structure, Proceedings homepage: https://conferenceproceedings.ump.ac.id/index.php/pssh/issue/view/25 485 ISSN: 2808-103X poor writing content or writing, many errors in writing, and improper use of punctuation. Teachers must use these issues to identify solutions to increase students' writing skills to a higher level than before. Some researchers employ the media as a means of resolving these issues. Each poor writing content or writing, many errors in writing, and improper use of punctuation. Teachers must use these issues to identify solutions to increase students' writing skills to a higher level than before. Some researchers employ the media as a means of resolving these issues. Each student actually has the ability to write with different levels of quality [4]. 1. INTRODUCTION As a result, the researchers attempted to use a description video of an electric motorcycle as a medium that may aid students in improving their descriptive text writing skills and creating enjoyable group learning activities. p g g j y g p g This study will also look at how video can be used to teach writing and how it affects students' writing skills. The author uses this medium to investigate which aspects of writing improve the most when video is used to teach descriptive writing abilities. Text Every sentence in a text has a relationship between text and context, hence every text has meaning. Text and context are linked in a text from both the cultural and situational contexts. The link between writings and their surroundings is complex. Because interactions can only be understood by viewing them, they are crucial to this framework in opposition to their social environment Text is any expanse of language that is united cohesively through meaning [12]. Whether a language stretch is text or not has nothing to do with its size or shape. This relates to the meanings of the expanse of language that work together as a unified whole. The term text is used in linguistics to describe: (1) The actual words of something written, printed, or spoken, as opposed to a summary or paraphrase. (2) A logical swath of language that can be subjected to rigorous examination. A text is traditionally defined as a piece of written or material in its most basic form (as opposed to a paraphrase or summary). Any stretch of language that can be understood in context is referred to as a text. It could be as simple as a couple of words (like a stop sign) or as complicated as a novel. A text is defined as a collection of sentences that belong together. Genre or type of text, is a sort of material that does not follow standard literary divisions and instead serves a social purpose [13]. In school, we study a variety of textbook genres. The story genre and the fact genre are two different types of text genres. In the story genre, there are six sorts of text: stories, news items, models, anecdotes, narratives, and parodies. The actual genre consists of procedures, explanations, reports, explanations, discussions, explanations, reviews, news articles, and explanations. In this case, the writer takes descriptive text as the reading genre. From the definitions above can be concluded that the text is a meaningful combination of sentences. Therefore, before studying descriptive text, it is necessary to know in advance the definition of the text, whether it is included or not. Types of Writing Skills The writing system used in the mother tongue has a significant impact on how easy it is for students to learn to write [11]. (1) Factual or Practical: There are two types of writing: formal and informal. Writing Facts are the focus of this form of writing. It can be found in letter writing and summaries by authors. This piece falls under the area of nonfiction writing; (2) Imaginary Creative Writing. This type of writing usually exists in literary works or works of fiction. Examples of imaginary writing are poetry, novels, romance, fantasy, science fiction, adventure, etc. The sort of writing assigned to kids will be determined by their age, interests, and skill level. Beginners, for example, could be asked to compose short sentences. When a teacher sets tasks to young children, the teacher ensures that they have enough words to complete them, and the same is true for intermediate and advanced students. Proceedings homepage: https://conferenceproceedings.ump.ac.id/index.php/pssh/issue/view/25 2. THEORETICAL FRAMEWORK Writing Skills Writing, which is one of the language skills, is an element of learning English as a foreign language. However, there are variances between the two; for example, the difference between writing and speaking is mostly in the result produced. The product in the form of an oral text can be a monologue or a dialogue, in other words. In writing, on the other hand, the primary product is written text. Writing abilities are a set of skills that allow writers to transform their ideas into meaningful words and psychologically connect with the message. Writing is defined differently by different linguists. Writing is part of a larger activity, such as language practice, acting out, or speaking, in which the focus is on something else [8]. Writing is a two-step process. The first stage is to figure out what the meaning is, and the second is to put that meaning into words [9]. What we believe is reflected in our writing. Because the writing process reflects thoughts that have remained in the mind, this is the case. Students who are afraid to write down their ideas often suffer as a result of this practice. For further, Brown said that the first step is to determine the meaning, followed by the translation of that meaning into language. What we imagine is embodied in writing. It's because writing represents what's on one's mind. Students who are hesitant to put down their thoughts sometimes struggle with this task. When pupils begin to hunt for reasons to write and produce written sentences, they run into challenges. When children begin to hunt for reasons to write and produce written sentences, they encounter obstacles. Writing is one of the productive language skills that is closely related to the way messages are conveyed using graphic or written symbols." Writing is a process of locating, organizing, and expressing ideas, putting them on paper, and reshaping and reworking them [10]. Students will be able to assess their English proficiency by writing, as it is strongly tied to their understanding of grammar and vocabulary. Writing is a kind of activity in which students communicate their ideas by using letters, words, phrases, and clauses from a series of interrelated sentences. Writing is the process of transmitting ideas and messages through Proceedings homepage: https://conferenceproceedings.ump.ac.id/index.php/pssh/issue/view/25 486 ISSN: 2808-103X written language, according to this definition. 2. THEORETICAL FRAMEWORK Writing Skills In other words, writing is the use of intelligible printed symbols to communicate between a writer and a reader. written language, according to this definition. In other words, writing is the use of intelligible printed symbols to communicate between a writer and a reader. Aspects of Writing Aspects of Writing In a nutshell, topic, composition, vocabulary, usage, and mechanics are the five key parts of writin can, however, be broken down into three categories: ideas, grammatical traits, and organization. In a nutshell, topic, composition, vocabulary, usage, and mechanics are the five key parts of writing. They can, however, be broken down into three categories: ideas, grammatical traits, and organization. Based on the previous description, I believe that students will be able to improve their writing skills if they have a strong comprehension of these components. Because they are fundamental skills for beginning writers. Types of Writing Skills Structure of Descriptive Text In descriptive text, the parts that need to be written are: (1) Identification. The first paragraph of descriptive text is identification, which aims to identify the object being discussed. This section does not include the details of the object. (2) Description. The next paragraph is a description. This section describes the characteristics of objects that have been described at the beginning of the text. In this section, the object must be described in as much detail as possible so that the characteristics of the object can be clearly described in the reader's mind [16]. To help the reader visualize the thing's features, the object must be described in as much detail as possible [17]. Learning Videos g Videos are audio and visual media that incorporate learning resources such as concepts, principles, methods, and knowledge application theories that assist students understand the topic being taught [18]. Learning videos are media that provide audio and visuals containing learning materials that contain concepts, principles, procedures, knowledge application theories to help understand the learning material being taught. Video itself is an audio-visual learning material that can be used in conveying messages/lesson materials. Audio-visual aids (AVA) media, or media that can be seen and heard, can be used to learn utilizing videos. The use of video in the application of learning can help students who are actively studying via visual media overcome their restrictions or gaps. "Intermediary" or "introduction" is what the word media signifies. It is defined by the National Education Association (NEA) as a tool that can be manipulated, seen, heard, read, or discussed, as well as instruments that are correctly employed in teaching and learning activities. The researcher chose in this study was adapted to the interests and majors of the students, namely Using an Electric Motorcycle Modified Video, which is closely related to students' interest in the Department of Automotive Vehicle Engineering. This is expected to increase students' interest in learning English, because it is not only related to English but also something that is their passion according to the major they choose at SMKN 1 Kalikajar Wonosobos 1 Kalikajar Wonosobos 1 Kalikajar Wonosobos Characteristics of descriptive text There are several characteristics that need to be considered in writing descriptive text: (1) Simple present. When describing something, usually the sentence is not bound by a time reference. So, the tenses used are simple present tense, namely tenses that use the first form of the verb (V1). (2) Nouns and noun phrases. When describing something, there will generally be a noun and a noun phrase. Nouns generally consist of one word, while noun phrases are nouns and their modifiers. (3) Adjectives. To explain or describe objects, we can use adjectives or adjectives. These adjectives can also be used as modifiers in noun phrases. (4) Relating verbs. A relative verb (copula) is a word that is positioned to become a verb and serves to connect the subject and an explanation of the subject. Examples of commonly used relating verbs are is, are, have, has, seem, appear, mean, become and others. (5) Action verbs. Action verbs are verbs that show action. Usually this type of verb is used when describing living things. Examples of action verbs are run, walk, see, listen, eat, and others. (6) Adverbials. Adverbial is a word or phrase that describes a verb. Usually the adverbial is used when the object being described is a living thing. Descriptive Text Understanding Descriptive Text g p To describe someone or something implies to give specific details about their appearance. As a result, descriptive text is a sort of English text that includes a detailed description or illustration of an object. It is descriptive, as the name implies, and is derived from the word describe. According to the Macmillan Dictionary, the definition of describe is "to give specifics about what someone or something is like." As a result, descriptive text is a sort of English text that includes a detailed description or illustration of an object. Inanimate objects, places, as well as individuals and other living things, are frequently described. An example of a descriptive text could be a text that tells a tourist place or about our pets. The purpose of writing a descriptive text is to describe and reveal a particular person, place, or thing. The purpose of descriptive text is to describe and indicate a person, place, or thing [15]. The purpose or social function of descriptive text is to describe a person, place or thing in such a way that a picture is formed in the reader's mind. This type of text is used to describe people, places or things in such a way that an image is formed in the mind of the reader. Teaching Writing When teaching English, a teacher needs know how to teach the four types of language skills to students. These abilities should be taught in English lessons, particularly when teaching English as a Foreign Language (TEFL). In this scenario, the teacher must be able to instruct students on how to write. There are various reasons why teachers should teach writing [2]. Writing is a process that involves a number of steps, including planning what to write and structuring it into sentences and paragraphs [14]. Reinforcement, language growth, learning style, and writing as a skill are the reasons." In addition, the following reasons shall be stated: (1) Reinforcement. Language reinforcement is linked to visual demonstrations of language creation, which are extremely useful while learning a new language. One of the reasons why teachers should teach pupils to write is because language growth is so quick. If the actual writing process is often carried out by students, this will help students learn to systematically express the ideas that are in their minds. (2) Writing Learning Style. Is appropriate for students that prefer non-rushed learning styles. It's a slow, contemplative pastime that can obstruct face-to-face interpersonal communication. Students who enjoy writing believe that generating language in a thorough and slow manner is extremely beneficial. (3) Writing as a skill. An important reason for teaching writing is that writing is a basic language skill. When writing is often done, students will get used to express their ideas smoothly. Skills only require a routine to practice, so that without realizing it students will be proficient in writing. Proceedings homepage: https://conferenceproceedings.ump.ac.id/index.php/pssh/issue/view/25 ISSN: 2808-103X 487 Descriptive Text Understanding Descriptive Text Data Collection Methods The following are the data collection procedures used in this study: a. Assigning students to experimental and control groups. g g p g p b. Teaching descriptive text writing skills using an electric motorcycle description video in the experimental class and teaching descriptive text writing skills without using an electric motorcycle description video in the control class. Details of teaching procedures are explained in the lesson plans. b. Teaching descriptive text writing skills using an electric motorcycle description video in the experimental class and teaching descriptive text writing skills without using an electric motorcycle description video in the control class. Details of teaching procedures are explained in the lesson plans. c. At the most recent meeting, students were given a written test in which they had to select the correct answer in the close test based on descriptive text for the experimental and control classes. Th P d f D t A l i c. At the most recent meeting, students were given a written test in which they had to select the correct answer in the close test based on descriptive text for the experimental and control classes. The Procedure of Data Analysis To analyze the data, two approaches were applied. The first is a central tendency index. It's used to see if the variable X1 is different from the variable X2. The second option is to apply the T-test to assess whether the variables X1 and X2 have a significant difference. 3. RESEARCH METHODS In this study, the experimental approach was used as the research method. This method examines whether the video utilized can assist a teacher in helping students enhance their capacity to write descriptive language. In this study, the subject groups are divided into two categories. The first is the experimental group, which receives research treatment, while the second is the control group, which receives the standard way of instruction, which is photos in textbooks without videos. Proceedings homepage: https://conferenceproceedings.ump.ac.id/index.php/pssh/issue/view/25 Proceedings homepage: https://conferenceproceedings.ump.ac.id/index.php/pssh/issue/view/25 488 ISSN: 2808-103X As shown by the nonequivalent-groups design, the study has two groups: experimental and control. Both groups have the same level, but the teaching methods used are different [19]. Because the two groups were not chosen at random, this study focuses on a nonequivalent-groups design. Experimental study is research that examines the effect of some variables on other variables under controlled conditions [20]. Population and Sample The population in this study were students of class X SMKN 1 Kalikajar Wonosobo majoring in Automotive Vehicle Engineering. Total population is 72 students. As a sample the researchers took 2 classes, namely X TKRO 1 and X TKRO 2 which consisted of 60 students. This sample was divided into 2 groups. Group A is a X TKRO 1 consisting of 30 students. Group A is the group that was given the video description treatment of electric motorcycles, this is the experimental group. Group B are students of class X TKRO 2 which consists of 30 students. Group B is the group that is treated without using a video description of the electric motorcycle is the control group. Both groups were given a post- test. 4. RESULT AND DISCUSSION Based on the analysis of all meetings, it can be shown that the three cycles used, namely the pre-test, Post- test, resulted in some major improvements. The graph below shows the progress in students' writing skills, their degree of participation, and their enthusiasm in creating descriptive texts: Graph 1. Comparison of Pre-Test, Post-Test I, and Post-Test II Results. Graph 1. Comparison of Pre-Test, Post-Test I, and Post-Test II Results. Graph 1. Comparison of Pre-Test, Post-Test I, and Post-Test II Results. Based on graph 1, it is clear that in the preliminary study, all students worked on the problems, with an average score of 52 The majority of students are not prepared to read and analyze in this exercise Only 6 (20%) 0 20 40 60 80 100 Priliminary Cycle-1 Cycle-2 Average result Students Good Score Students Fair Score Graph 1. Comparison of Pre-Test, Post-Test I, and Post-Test II Results. Based on graph 1, it is clear that in the preliminary study, all students worked on the problems, with an average score of 52. The majority of students are not prepared to read and analyze in this exercise. Only 6 (20%) students received a score of 70 or higher, while 24 (80%) students received a score of less than 70. The preliminary investigation showed a rise in the Post-Test 1. The overall score is 68. There are 21 students who receive a score of 70 or higher (70%) and 9 students who do not receive a score of 70 (30%). The average student score in the Post-Test 2 is 85. This suggests that since the pre-test, there has been a rise. This suggests that student achievement has risen significantly. There are 28 students who receive a score of 70 (93%) and two students who receive a score of less than 70 (7%). Furthermore, using a visual organizer 0 20 40 60 80 100 Priliminary Cycle-1 Cycle-2 Average result Students Good Score Students Fair Score 0 20 40 60 80 100 Priliminary Cycle-1 Cycle-2 Average result Students Good Score Students Fair Score Average result Students Good Score Students Fair Score Cycle-1 Based on graph 1, it is clear that in the preliminary study, all students worked on the problems, with an average score of 52. The majority of students are not prepared to read and analyze in this exercise. 4. RESULT AND DISCUSSION Only 6 (20%) students received a score of 70 or higher, while 24 (80%) students received a score of less than 70. The preliminary investigation showed a rise in the Post-Test 1. The overall score is 68. There are 21 students who receive a score of 70 or higher (70%) and 9 students who do not receive a score of 70 (30%). The average student score in the Post-Test 2 is 85. This suggests that since the pre-test, there has been a rise. This suggests that student achievement has risen significantly. There are 28 students who receive a score of 70 (93%) and two students who receive a score of less than 70 (7%). Furthermore, using a visual organizer Based on graph 1, it is clear that in the preliminary study, all students worked on the problems, with an average score of 52. The majority of students are not prepared to read and analyze in this exercise. Only 6 (20%) students received a score of 70 or higher, while 24 (80%) students received a score of less than 70. The preliminary investigation showed a rise in the Post-Test 1. The overall score is 68. There are 21 students who receive a score of 70 or higher (70%) and 9 students who do not receive a score of 70 (30%). The average student score in the Post-Test 2 is 85. This suggests that since the pre-test, there has been a rise. This suggests that student achievement has risen significantly. There are 28 students who receive a score of 70 (93%) and two students who receive a score of less than 70 (7%). Furthermore, using a visual organizer Based on graph 1, it is clear that in the preliminary study, all students worked on the problems, with an average score of 52. The majority of students are not prepared to read and analyze in this exercise. Only 6 (20%) students received a score of 70 or higher, while 24 (80%) students received a score of less than 70. Based on graph 1, it is clear that in the preliminary study, all students worked on the problems, with an average score of 52. The majority of students are not prepared to read and analyze in this exercise. Only 6 (20%) students received a score of 70 or higher, while 24 (80%) students received a score of less than 70. Instruments This study's instruments are 10 close-test questions and 5 essay questions. A post-test was also administered to determine the overall outcomes of the learning activities. This test was used to compare the average scores of two classes: the experimental class, which used an electric motorcycle description video, and the control class, which did not use an electric motorcycle description video. 4. RESULT AND DISCUSSION g The preliminary investigation showed a rise in the Post-Test 1. The overall score is 68. Ther students who receive a score of 70 or higher (70%) and 9 students who do not receive a score of 70 (30% g The average student score in the Post-Test 2 is 85. This suggests that since the pre-test, there has been a rise. This suggests that student achievement has risen significantly. There are 28 students who receive a score of 70 (93%) and two students who receive a score of less than 70 (7%). Furthermore, using a visual organizer Proceedings homepage: https://conferenceproceedings.ump.ac.id/index.php/pssh/issue/view/25 489 ISSN: 2808-103X technique as a learning aid is quite beneficial in the teaching and learning of descriptive text writing. It used T- test to know the result. technique as a learning aid is quite beneficial in the teaching and learning of descriptive text writing. It used T- test to know the result. Graph 2. Comparison of Student Participation Results from Pre-Test and Post-Test Graph 2. Comparison of Student Participation Results from Pre-Test and Post-Test Based on Graph 2 above, it can be concluded that there was a significant increase in student participation from pre-test and post-test. The activeness of students in the first cycle was 43% and in the second cycle increased by 87%. This means that it has increased to 44%. Student learning time is 13% in the first cycle and 100% in the second cycle. This is an 87% increase. In addition, student participation in asking questions was 40% in the pre-test and 77% in the post-test. That's an increase of 77%. In answering several questions, student participation in the pre-test was 27% and the post-test was 73%. Th t h th i i f 46% Fi ll t d t ti i ti i i i i i 33% i th Graph 2. Comparison of Student Participation Results from Pre-Test and Post-Test Based on Graph 2 above, it can be concluded that there was a significant increase in student participation from pre-test and post-test. The activeness of students in the first cycle was 43% and in the second cycle increased by 87%. This means that it has increased to 44%. Student learning time is 13% in the first cycle and 100% in the second cycle. This is an 87% increase. 4. RESULT AND DISCUSSION In addition, student participation in asking questions was 40% in the pre-test and 77% in the post-test. That's an increase of 77%. In answering several questions, student participation in the pre-test was 27% and the post-test was 73%. That shows there is an increase of 46%. Finally, student participation in giving opinions was 33% in the pre-test and 77% in the post-test. This is an increase of 44%. Finally, the use of a graphic organizer strategy as a learning aid is very helpful and interacts with students in the teaching and learning process of reading descriptive text. It used T-test to know the result. Graph 3. Comparison of Student Interest Results Before and After Treatment Graph 3. Comparison of Student Interest Results Before and After Treatment Based on graph 3 above, it was concluded that there was a significant increase in student interest from before and after treatment. There are 6 (20%) students who say they enjoy learning English. Conversely, 27 (90 percent) of the students said they enjoyed learning English after receiving treatment. 11 students (37%) said they were involved in the English teaching and learning process, and 27 students (90%) said they were involved after receiving treatment. Then there were 7 (23%) students who always read English material at home, with a sharp increase to 27 (90%) students who said they always read English material after treatment. Furthermore, 5 (17%) pupils claimed they loved their teacher's teaching style. Meanwhile, after treatment, there was a 29 percent rise in pupils who said they appreciated the way the teacher taught. Finally, 12 (40%) students said they frequently took notes when learning English, with an increase of 28 (93%) students saying they frequently took notes while teaching and learning English after receiving treatment. Finally, using video descriptions of electric motorcycles as learning aids boosts students' enthusiasm for the teaching and learning of descriptive text writing. Graph 3. Comparison of Student Interest Results Before and After Treatment Based on graph 3 above, it was concluded that there was a significant increase in student interest from before and after treatment. There are 6 (20%) students who say they enjoy learning English. Conversely, 27 (90 percent) of the students said they enjoyed learning English after receiving treatment. 11 students (37%) said they were involved in the English teaching and learning process, and 27 students (90%) said they were involved after receiving treatment. 4. RESULT AND DISCUSSION Then there were 7 (23%) students who always read English material at home, with a sharp increase to 27 (90%) students who said they always read English material after treatment. Furthermore, 5 (17%) pupils claimed they loved their teacher's teaching style. Meanwhile, after treatment, there was a 29 percent rise in pupils who said they appreciated the way the teacher taught. Finally, 12 (40%) students said they frequently took notes when learning English, with an increase of 28 (93%) students saying they frequently took notes while teaching and learning English after receiving treatment. Finally, using video descriptions of electric motorcycles as learning aids boosts students' enthusiasm for the teaching and learning of descriptive text writing. Based on graph 3 above, it was concluded that there was a significant increase in student interest from before and after treatment. There are 6 (20%) students who say they enjoy learning English. Conversely, 27 (90 percent) of the students said they enjoyed learning English after receiving treatment. 11 students (37%) said they were involved in the English teaching and learning process, and 27 students (90%) said they were involved after receiving treatment. Then there were 7 (23%) students who always read English material at home, with a sharp increase to 27 (90%) students who said they always read English material after treatment. Furthermore, 5 (17%) pupils claimed they loved their teacher's teaching style. Meanwhile, after treatment, there was a 29 percent rise in pupils who said they appreciated the way the teacher taught. Finally, 12 (40%) students said they frequently took notes when learning English, with an increase of 28 (93%) students saying they frequently took notes while teaching and learning English after receiving treatment. Finally, using video descriptions of electric motorcycles as learning aids boosts students' enthusiasm for the teaching and learning of descriptive text writing. Proceedings homepage: https://conferenceproceedings.ump.ac.id/index.php/pssh/issue/view/25 490 ISSN: 2808-103X According to the findings of this study, there was a substantial difference in the skills, student involvement, and interests of students who were taught using an electric motorbike description video to students who were not taught using an electric motorcycle description video. Those who were taught to use an electric motorcycle description video performed better than those who were taught before using an electric motorcycle description video. This means that video descriptions of electric motorcycles can help students improve their skills in writing descriptive texts. 5. CONCLUSIONS The use of video descriptions of electric motorcycles as a medium to improve students' writing skills in descriptive text is an alternative method that is very helpful. The use of video descriptions of electric motorcycles is very useful in the teaching and learning process of English. The use of video descriptions of electric motorcycles is highly recommended because it can make students more interested in writing descriptive text material. By using video descriptions of electric motorcycles, teachers can stimulate students to compose descriptive texts. Based on the results of research that has been carried out in two cycles, namely in cycle I and cycle II, it can be concluded that there is a significant increase. Students' writing skills in descriptive texts increased sharply after being taught using a video description of an electric motorcycle. The application of video descriptions of electric motorcycles as learning aids to improve students' writing skills in descriptive texts is very effective. This is supported by the significance of the pre-cycle mean score 53, Cycle I 68, and Cycle II 85. The use of video descriptions of electric motorcycles in learning English is very interesting for students; This can be seen from the results of observations and test results. The researcher uses a video description of an electric motorcycle as a tool because it can motivate students to learn to write descriptive text in English more fun and easily. 4. RESULT AND DISCUSSION It can also help teachers to increase students' participation and interest in writing descriptive texts. Teaching writing by using video descriptions of electric motorcycles can stimulate students to improve their descriptive text writing skills. So, the learning achievement of students who are taught to write descriptive text using a video description of an electric motorcycle is better than those who are taught to write a descriptive text without using a video description of an electric motorcycle in class X SMKN 1 Kalikajar Wonosobo majoring in Automotive Vehicle Engineering. 6. SUGGESTIONS Based on the result of this study, the author makes many suggestion such as teachers can make the teaching and learning process more engaging by providing video descriptions of electric motorcycles to help students grasp English more easily. Teachers can also assess how well students comprehend the topic at each meeting. Based on the result of this study, the author makes many suggestion such as teachers can make the teaching and learning process more engaging by providing video descriptions of electric motorcycles to help students grasp English more easily. Teachers can also assess how well students comprehend the topic at each meeting. The researcher believes that students will be able to identify ways to develop their skills, particularly their capacity to write descriptive texts in a variety of ways, such as using a video description of an electric motorcycle as an example. y p Other researchers who want to perform similar research can use the findings of this study as a g future investigations on different topics or with different expertise. REFERENCES REFERENCES [1] Suwartono, T.: FbG as an Around The Clock Support for an Essay Writing Course Amidst the Pandemic. EduLite Journal of English Education, Literature, and Culture Vol. 6, No. 2. August 2021, pp. 238-255. E-ISSN: 2528-4479, P-ISSN: 2477-5304 ttp://jurnal.unissula.ac.id/index.php/edulite DOI: http://dx.doi.org/10.30659/e.6.2.238-255 (2021) [1] Suwartono, T.: FbG as an Around The Clock Support for an Essay Writing Course Amidst the Pandemic. EduLite Journal of English Education, Literature, and Culture Vol. 6, No. 2. August 2021, pp. 238-255. E-ISSN: 2528-4479, P-ISSN: 2477-5304 ttp://jurnal.unissula.ac.id/index.php/edulite DOI: http://dx.doi.org/10.30659/e.6.2.238-255 (2021) p g ( ) [2] Harmer, J.: Practice of English language teaching. Pearson Education, Ltd: English (2001) [3] Lestari, F.: The development of writing skills through collaborative activities. Bandung: UPI (2010 [4] Permono, F. J.: The efectiveness of using flashcard to Improve students’ ability in writing procedural text. Bandung: UPI (2010) [5] Suhartini, D.: Teaching students’ writing skill on narrative text through film trailer. Surakarta: [6] Nurmilah, I.: The use of fairy tale movies to increase students’ writing skill. Semarang: IAIN Walisongo (2010) ( ) [7] Kane, T. S.: The Oxford Essential Guide to Writing. New York: Berkley Books (2000) [8] Harmer, J.: The practice of English language teaching. Malaysia: Pearson Education Limited (20 ] Brown, H. D.: Teaching by principle: and interactive approach to language pedagogy. 2nd ed. New A Pearson Education Company Cambridge: Cambridge University Press (2001) [9] Brown, H. D.: Teaching by principle: and interactive approach to language pedagogy. 2nd ed. New York: A Pearson Education Company. Cambridge: Cambridge University Press (2001) [9] Brown, H. D.: Teaching by principle: and interactive approach to language pedagogy A Pearson Education Company. Cambridge: Cambridge University Press (2001) [ ] , g y p p pp g g p g gy A Pearson Education Company. Cambridge: Cambridge University Press (2001) p y g g y ( ) [10] Meyers, A.: Gateways to academic writing. effective sentences, paragraphs and essays.New York & London: Longman (2005) p y g g y ( ) [10] Meyers, A.: Gateways to academic writing. effective sentences, paragraphs and essays.New York & London: Longman (2005) Proceedings homepage: https://conferenceproceedings.ump.ac.id/index.php/pssh/issue/view/25 Proceedings homepage: https://conferenceproceedings.ump.ac.id/index.php/pssh/issue/view/25 ISSN: 2808-103X 491 [11] Javed, M. Juan, W.X., Nazli, S.: A Study of Students’ Assessment in Writing Skills of the English Language. International Journal of Instruction. Vol.6, No.2 e-ISSN: 1308-1470 (2013) g g ( ) [12] Wilson, A.: What is a Text? Proceedings homepage: https://conferenceproceedings.ump.ac.id/index.php/pssh/issue/view/25 REFERENCES Studies in History and Philosophy of Science (nama jurnal) Doi:10.1016/j.shpsa.2011.12.027 (2012). [13] Mongot, A.: English revolution. Jepara: Mawas Press (2008) [14] Brookes, A and Grundy, P.: Beginning to Write: Writing activities for Elementary and Intermediate Learners. Cambridge: Cambridge University Press (2008) [15] Dirgeyasa, I. W.: English Language Teaching; Vol. 9, No. 9; 2016 ISSN 1916-4742 E-ISSN 1916-4750. Canadian Center of Science and Education 45 Genre-Based Approach: What and How to Teach and to Learn Writing (2017) [16] Wardani, I., Basri, H., Waris, A.: Improving The Ability in Writing Descriptive Text Through Guided- Questions Technique. e-Journal of English Language Teaching Society (ELTS) Vol. 2 No. 1 2014 t ISSN 2331-1841 (2014) [17] Guswindari, R.: Descriptive Text: Definition, Purpose, and Structure.https://www.kompas.com/skola/read/2020/10/26/163836869/descriptive-text-definition- purpose-dan-structure?page=all (2022) [17] Guswindari, R.: Descriptive Text: Definition, Purpose, Structure.https://www.kompas.com/skola/read/2020/10/26/163836869/descriptive-text-definition- [17] Guswindari, R.: Descriptive Text: Definition, Purpose, and S h // k / k l / d/2020/10/26/163836869/d i i d fi i i [18] Pratama,R. A.; Ulfa, S.; Kuswandi, D.: The Utilization of Learning Videos in Ips Lessons in Small School. Proceedings of TEP & PDs. Transformation of 21st Century Education Theme: 1 Number: 13. ISBN: 978-602-71836-6 -7) (2017) [19] Prasetya, E. P. and Komarudain: The Use of Cooking Video to Improve Students’ Writing Skill on Procedure Text. English Journal Vol. 20, No. 1; March 2017, pp. 25-33 (2017) [19] Prasetya, E. P. and Komarudain: The Use of Cooking Video to Improve Students Writing Skill on Procedure Text. English Journal Vol. 20, No. 1; March 2017, pp. 25-33 (2017) [20] S i St ti tik U t k P liti B d J b Alf b t (2010) [20] Sugiyono.: Statistika Untuk Penelitian. Bandung, Jabar: Alfabeta, cv. (2010) Proceedings homepage: https://conferenceproceedings.ump.ac.id/index.php/pssh/issue/view/25
https://openalex.org/W4254973413	https://digitalcommons.unl.edu/cgi/viewcontent.cgi?article=1011&context=publichealthresources	English	null	Alcohol research: A lifespan perspective	PsycEXTRA Dataset	2,008	public-domain	3,875	DigitalCommons@University of Nebraska - Lincoln. It has been accepted for inclusion in Public Health Resources by an authorized administrator of DigitalCommons@University of Nebraska - Lincoln. University of Nebraska - Lincoln University of Nebraska - Lincoln DigitalCommons@University of Nebraska - Lincoln DigitalCommons@University of Nebraska - Lincoln Public Health Resources U.S. Department of Health & Human Services National Institutes of Health National Institute on Alcohol Abuse and Alcoholism January 2008 Number 74 Number 74 VE ALCOHOL RESEARCH: A LIFESPAN PERSPECTI Alcohol use and the risk for alcohol-related problems change over the lifespan. College students and young adults, who often drink large quantities of alcohol at one time, are more likely to experience problems such as alcohol poisoning, drunk-driving crashes, and assaults; whereas, older individuals who drink even moderately while taking certain medications run the risk of harmful drug interactions. Additionally, patterns of alcohol use may differ across the human lifespan—for example, adolescents who begin drinking prior to age 14 are more likely to develop a serious problem with alcohol later in life. Understanding how alcohol influences people across different life stages is important, especially when designing effective approaches for diagnosing, treating, and preventing alcohol abuse and dependence and their related problems. Alcohol Research: A Lifespan Perspective Alcohol Research: A Lifespan Perspective National Institute on Alcohol Abuse and Alcoholism Follow this and additional works at: https://digitalcommons.unl.edu/publichealthresources Part of the Public Health Commons Follow this and additional works at: https://digitalcommons.unl.edu/publichealthresources Part of the Public Health Commons Part of the Public Health Commons National Institute on Alcohol Abuse and Alcoholism, "Alcohol Research: A Lifespan Perspective" (2008). Public Health Resources. 6. htt //di it l l d / bli h lth /6 https://digitalcommons.unl.edu/publichealthresources/6 https://digitalcommons.unl.edu/publichealthresources/6 This Article is brought to you for free and open access by the Public Health Resources at This Article is brought to you for free and open access by the Public Health Resources at DigitalCommons@University of Nebraska - Lincoln. It has been accepted for inclusion in Public Health Resources by an authorized administrator of DigitalCommons@University of Nebraska - Lincoln. DigitalCommons@University of Nebraska - Lincoln. It has been accepted for inclusion in Public Heal an authorized administrator of DigitalCommons@University of Nebraska - Lincoln. g @ y p an authorized administrator of DigitalCommons@University of Nebraska - Lincoln. 1 From Fetal Alcohol Syndrome: Diagnosis, Epidemiology, Prevention, and Treatment. Institute of Medicine, 1996. (http:// www.nap.edu/openbook.php?record_id=4991&page=1). The Embryo and Fetus: Focus on Fetal Alcohol Spectrum Disorder (FASD) and Fetal Alcohol Syndrome (FAS) What we know—Alcohol is a leading preventable cause of birth defects with mental deficiency. Prenatal alcohol exposure can cause a variety of problems known collectively as Fetal Alcohol Spectrum Disorder (FASD). The most severe form of FASD, Fetal Alcohol Syndrome (FAS), often is characterized by certain physical traits, such as a flattened midface, wide-set eyes and slow growth, nervous system impairments, and a range of learning and behavioral problems. Additionally, peo ple exposed to alcohol prenatally are at higher risk of develop ing an alcohol and other drug use disorder later in life. In 2006, the National Institute on Alcohol Abuse and Alcoholism (NIAAA) published the Five-Year Strategic Plan, The NIAAA Strategic Plan for Research. The Five-Year Plan introduces a new organizing principle for alcohol research studies: A Lifespan Perspective. This new perspective gives researchers a framework within which to examine how alcohol affects people at different stages of development and how different stages of development affect drinking behaviors. The Five-Year Plan examines the current state of alcohol research—what we know about alcohol-related issues— within a Lifespan Perspective, and suggests opportunities for new research and outreach based on these findings. Since 2006, the Five-Year Plan has been revised once, and NIAAA will continue to provide updates to reflect new and emerging research opportunities. This Alcohol Alert presents some of the findings and opportunities outlined in the latest version of the Five-Year Plan. It has been reported that up to 1 in 100 children in the United States are born with FASD. Additionally, 0.5 to 3.0 children out of 1,000 are diagnosed with FAS.1 Research suggests that other factors, such as the mother’s hormone status, nutrition, age, the number of children she has had previously, and the length of time she has been drinking, as well as genetic factors including those affecting the way the body breaks down alcohol, also may contribute to the development of FASD. FASD can be very difficult to recog nize. The symptoms are sometimes subtle and cover a wide spectrum of 2 According to the Surgeon General’s Call to Action To Prevent and Reduce Underage Drinking (NIAAA 2006), most college drinking studies define binge drinking as “5 or more drinks in a row for men and 4 or more drinks in a row for women” (NIAAA National Advisory Council). However, in 2004, the National Advisory Council revised the definition of binge drinking as follows: “a pat tern of drinking alcohol that brings blood alcohol concentration (BAC) to 0.08 gram percent or above. For the typical adult, this pattern corresponds to consuming 5 or more drinks (male) or 4 or more drinks (female) in about two hours” (NIAAA National Advisory Council). “Patterns of alcohol use may differ across the human lifespan.” Five Year Strategic Plan Prevalence of Past-year DSM-IV Alc ohol Dependence by Age in the Uni ted States problems. Symptoms also may change with age—for example, facial features often associated with FAS may be less prominent as a person ages. Most importantly, there is no “biomarker” for FASD—that is, there is no characteristic or distinctive biological feature that can be used to identify the disorder in every case. A major factor influencing risk for FAS and FASD is the timing of alcohol exposure during critical periods of development. Research opportunities—Researchers are work ing to diagnose FASD earlier and more accu rately using complex, detailed facial imaging to identify subtle facial abnormalities in infants. These abnormalities may serve as biomarkers for FASD. Another promising biomarker is a brain structure called the corpus callosum: the size of the corpus callosum is much more vari able in the brains of children and adults who have experienced prenatal exposure to alcohol. When prenatal alcohol exposure already has occurred (i.e., if a woman drinks before she realizes that she is pregnant), or if alcohol exposure is ongoing (i.e., a woman continues to drink during pregnancy), pharmacological intervention may be helpful in countering some of the harmful effects of alcohol. Although these interventions have not yet been studied in humans, animal studies have shown intriguing results. Antioxidants, anti-inflammatory agents (such as prostaglandin inhibitors), and the nutritional co-factor choline are some of the agents that may prove useful in reducing alcohol- induced fetal injury. For example, when young rats with prenatal exposure to alcohol were given supplements of choline, they became less hyperactive and showed decreased memory loss. memory, the hippocampus, is smaller in adolescents who begin drinking at an early age. In addition, studies of adoles cents who were receiving treatment for alcohol withdrawal showed that they were more likely to have memory problems than adolescents who did not drink. Adolescents tend to drink differently than adults. They are more likely to engage in risky behaviors such as heavy episodic (or “binge”) drinking.2 Researchers believe these risky behaviors are the result of certain social factors, such as a greater independence and pressure from peers, as well as biological factors. Adolescents tend to be less sensitive to negative effects of alcohol, such as increased sleepiness and lack of coordination. This may explain why they are able to drink so much alcohol at one time. “Patterns of alcohol use may differ across the human lifespan.” Five Year Strategic Plan On the other hand, adolescents are more likely to have trouble with complex tasks, such as driving a motor vehicle, making adolescent alcohol use especially dangerous. Researchers also are exploring ways of treating children with FASD. Again, studies in animals suggest that behavioral inter ventions might improve some of the symptoms of FASD. When adult rats with prenatal exposure to alcohol were given 20 days of motor skills training, their performance improved, suggesting that brain cells damaged by alcohol may be able to partially recover. Human studies using behavioral intervention are in the early stages but also are encouraging. Researchers also are exploring ways of treating children with FASD. Again, studies in animals suggest that behavioral inter ventions might improve some of the symptoms of FASD. According to the National Household Survey on Drug Abuse, 1.47 million adolescents ages 12–17 (5.9 percent of adoles cents in this age-group) met the criteria for alcohol depend ence or alcohol abuse in 2003. However, the survey also showed that only 15 percent of these respondents received any treatment for alcohol problems. Youth and Adolescence What we know—Adolescence is the period between 12 and 17 years of age. This is a time of dramatic physical, psychological, and social change. The brain continues to develop and mature throughout adolescence and into the mid-20s, and studies suggest that consuming alcohol during this time may have lasting effects on brain development. For example, a region of the brain involved in learning and 2 2 of increased risk for problems with alcohol. The youngest segment of this population—young adults ages 18–24—are most at risk for alcohol problems, compared with other age- groups (see Figure 1). This group is most likely to drink heav ily, regardless of their gender, ethnicity, and school or work sta tus—that is, whether they attend college or are employed full time (Figure 2). Despite increased attention in recent years, the problem of young adult drinking continues to escalate: alcohol-related deaths rose 5 percent for 18- to 24-year-olds between 1998 and 2001. According to the Five-Year Plan, preventing alcohol problems in adolescents poses unique challenges. Research suggests that they may not respond well to traditional treatment methods (e.g., alcoholism treatment programs and Alcoholics Anonymous) and that brief, alternative intervention meth ods that are targeted specifically toward young people are more effective. Research opportunities—Researchers continue to explore how the social and biological changes associated with ado lescence contribute to alcohol use, as well as how alcohol can affect the maturing brain. Researchers also are working to refine current definitions of alcohol abuse and dependence to better diagnose and screen dangerous drinking behaviors in young people, because adolescent drinking patterns tend to differ from those of adults. Research continues to iden tify factors, such as certain behaviors or a family history of alcoholism, which may put young people at risk for later alcohol problems. Research also continues to explore the complex relationship between drinking and other biological factors, such as reproductive and stress hormones and sex differences. For college students, brief interventions that target high-risk populations (e.g., freshmen, Greek organization members, athletes, students mandated to receive treatment) and Driving Under the Influence (DUI) prevention campaigns have shown promise. For young adults in the military, few programs have been formally evaluated; however, current strategies to prevent alcohol-related problems include regulating the availability and pricing of alcohol, attempting to deglamorize alcohol use, and promoting personal responsibility and good health. Youth and Adolescence Nonstudent, nonmilitary personnel may be more likely to continue dangerous drinking patterns into adulthood. This population does not have access to institutionally based programs that typically serve college students and military personnel. Additionally, this population may not have access to mental health services, making them vulnerable to psychiatric conditions, such as depression and anxiety, often associated with dangerous drinking patterns. Midlife: Focus on Detrimental Effects of Drinking and on Treatment Options In addition, it has been shown that people who drink are less likely to get tested for HIV. When they do test posi tive they are less likely to seek treatment and are less compliant with treatment. Drinking also is associated with risk-taking behaviors that might put people at increased risk for contracting HIV. What we know—The Five-Year Plan considers midlife as spanning the ages of 30–59. During this period the conse quences of heavy drinking often become evident. Alcoholic liver disease, alcohol pancreatitis, several types of cancer, disorders of the heart and circulatory system, alcohol-related brain disorders, and other adverse effects upon the endocrine and immune system are most likely to emerge during this time. For people in midlife, research often has focused on how alcohol damages body tissues, as well as methods for better tailoring treatments and interventions to this segment of the population. Clearly, alcohol use and HIV/AIDS are closely intertwined. Additional research is needed to learn more about this complex relationship. Longitudinal studies also may prove helpful, providing a better understanding of how alcohol and HIV can interact and affect individuals over time. help people with alcohol dependence who do not respond well to existing medications. Several promising agents, including baclofen, rimonabant, and memantine, now are being evaluated for safety and efficacy in treating alcohol use disorders. Current research also is exploring the possibility of preventing or treating organ injury through the use of phar maceuticals, as well as through the use of antioxidant- and cannabinoids-related agents. Individuals in midlife are more likely to seek treatment for alcohol dependence. Three drugs—disulfiram, naltrexone, and acamprosate—have been approved for use in the United States to treat alcoholism. Although some patients have responded well to these medications, others have not. Research shows that a variety of factors—both biological and social— influence an individual’s response to therapy. Individuals in midlife are more likely to seek treatment for alcohol dependence. Three drugs—disulfiram, naltrexone, and acamprosate—have been approved for use in the United States to treat alcoholism. Although some patients have responded well to these medications, others have not. Research shows that a variety of factors—both biological and social— influence an individual’s response to therapy. Alcohol metabolism has been shown to have an important role in organ damage. This is another area of research that is receiving attention. Young Adults Studies also can help to determine how the maturing brain affects a young adult’s decision to drink and how best to inter vene with college students, military personnel, and young adults who are not enlisted in college or in the military. The relationship between HIV and alcohol is complex and needs further study. Drinking might make it easier for the HIV virus to establish an infection; additionally, alcohol use might actually accelerate the course of the disease. Alcohol also may change the way the body breaks down HIV medications, interfering with their effectiveness. Although most young adults drink less as they transition into midlife, some continue to drink heavily. Research con tinues to explore the risk factors associated with continued drinking into midlife. Young Adults What we know—Young adulthood is the period between the ages of 18 and 29 years. During this period many young people pursue postsecondary education, enlist in the mili tary, or enter the workforce. This is a time of transition and Percentage of Individuals Exceeding the Daily Drinking Limit for Ages 18-20 and 21-24, by Gender, Race-Ethnicity, and College Enrollment Status Percentage of Individuals Exceeding the Daily Drinking Limit for Ages 18-20 and 21-24, by Gender, Race-Ethnicity, and College Enrollment Status Percentage of Individuals Exceeding the Daily Drinking Limit for Ages 18-20 and 21-24, by Gender, Race-Ethnicity, and College Enrollment Status Source: NIAAA 2001–2002 NESARC data. AIAN = American Indian/Alaska Native, * NHOPI = Native Hawaiian/Other Pacific Islander Source: NIAAA 2001–2002 NESARC data. AIAN = American Indian/Alaska Native, * NHOPI = Native Hawaiian/Other Pacific Islander 3 3 Research opportunities—As noted above, researchers now believe that the brain continues to develop into the mid-20s. Scientists are continuing to explore alcohol’s potentially dam aging effects on the young adult brain. Using noninvasive brain scans, such as the electroencephalogram (EEG), which measures and records the electrical activity of the brain, researchers can study changes that occur with drinking and determine whether the brain can return to a normal level of function as young adults begin to “age out” of heavy drinking. Studies also can help to determine how the maturing brain affects a young adult’s decision to drink and how best to inter vene with college students, military personnel, and young adults who are not enlisted in college or in the military. Alcohol and HIV/AIDS HIV (Human Immunodeficiency Virus), the virus that causes AIDS (Acquired Immunodeficiency Syndrome), is epidemic in the United States. As many as 950,000 people in the United States may be infected with the HIV virus, and more than 500,000 people have died from the infection. Eighty percent of people infected with HIV drink alcohol, and between 30 and 60 percent have been diagnosed with an alcohol use disorder. y researchers can study changes that occur with drinking and determine whether the brain can return to a normal level of function as young adults begin to “age out” of heavy drinking. Midlife: Focus on Detrimental Effects of Drinking and on Treatment Options Other studies are examining treat ment approaches that help people to change their drinking behavior and identifying biological and social factors that encourage drinking behaviors as well as those that increase the chances of recovery. of factors— Research opportunities—Some of the most exciting alcohol research focuses on developing new medications for alco holism, targeting multiple sites. New medications could “R “Research shows that a variety of factors— both biological and social—influence an individual’s response to therapy.” Moderate drinking among adults is another chief area of research. Although heavy alcohol use is asso ciated with the development of 4 studied, some research shows that medications for alcoholism may not work as well in older adults. many diseases, some research suggests that moderate alcohol consumption may protect individuals from some of the health problems that tend to strike during midlife, including coronary artery disease, type 2 diabetes, dementia, and ischemic stroke. More research is needed to determine whether these potential benefits outweigh the risks of drinking. According to the current literature, the most beneficial treatment for alcohol use disorders in older adults may be education; many seniors lack information on the dangers of alcohol use. The age at which they begin drinking also is important. Older adults who began problem drinking earlier in life tend to have worse treatment outcomes than those who began drinking later in life. Senior Adults and Alcohol: A National Health Issue What we know—Senior adults tend to drink less than other age-groups (see Figure 1). However, research suggests that alcohol problems in older adults soon may become a national health issue. Senior adult drinking is on the rise; as people live longer, the number of people who drink will increase. Research also shows that people born in recent years tend to drink more than older generations, suggesting that as the current population ages, these individuals will continue to drink more. Research opportunities—Although older adults are the most rapidly growing segment of the population today, very little alcohol-related research has been conducted among this age-group. Researchers continue to explore how the physiological and social changes that come with aging inter act with alcohol use. New research is examining the effec tiveness of current medications in seniors with alcohol use disorders, as well as developing medications that might work better in older individuals. Researchers are examining how alcohol use may impact the development of diseases such as Alzheimer’s, type 2 diabetes, and other age-related health problems. Ongoing research will reveal how changing demographics may affect the use, impact, and consequences of older adult drinking. Older adults are at particular risk for alcohol-related problems. As individuals age they metabolize alcohol more slowly; as a result, alcohol remains in the body longer. Older adults are more likely to have health conditions that can be exacer bated by alcohol, including stroke, hypertension, neurode generation, memory loss, mood disorders, and cognitive or emotional problems. Additionally, older adults are more likely than younger people to take medications, putting them at risk for interactions that can be dangerous or even life-threatening. Alcohol also may decrease effectiveness of some medications. Conclusion NIAAA’s Five-Year Plan gives researchers a framework within which to explore how alcohol affects individuals at different stages of development and how different stages of develop ment affect drinking behaviors. This Lifespan Perspective offers a new approach to the age-old question: Why do some people drink too much? It enables researchers to conceptualize the complex and evolving relationship between human devel opment and alcohol and to explore how alcohol intersects with biology and the environment. New insights gained through the Lifespan Perspective will result in more effective and better-targeted interventions that take into account the changing needs of individuals at every stage of life. Research shows that treatment can be effective in older indi viduals. They tend to respond better to treatment that takes place in groups of people in their same age range. Cognitive behavioral therapy has been shown to be effective in older patients. Group family therapy also has been shown to be successful, perhaps because family therapy engages support systems that might have been lacking or even exacerbating the patient’s alcohol use. Although it has not been well “This Lifespan Perspective offers a new approach to an age-old question: Why do some people drink too much?” “This Lifespan Perspective offers a new approach to an age-old question: Why do some people drink too much?” 5 5 R e s o u r c e s Source material for this Alcohol Alert originally appeared in the National Institute on Alcohol Abuse and Alcoholism (NIAAA) Five-Year Strategic Plan. The Five-Year Plan highlights new research avenues that will address alcohol use across the human lifespan, emphasizing how drinking behavior evolves over an individual’s lifetime in response to biological, psychological, and social and environmental changes. Over time the Plan will be updated to reflect new and emerging research opportunities. The Plan is available online at http://pubs.niaaa.nih.gov/ publications/StrategicPlan/NIAAASTRATEGICPLAN.htm Full text of this publication is available on NIAAA’s World Wide Web site at www.niaaa.nih.gov. All material contained in the Alcohol Alert is in the public domain and may be used or reproduced without permission from NIAAA. Citation of the source is appreciated. Copies of the Alcohol Alert are available free of charge from the National Institute on Alcohol Abuse and Alcoholism Publications Distribution Center, P.O. Box 10686, Rockville, MD 20849–0686. R e s o u r c e s Source material for this Alcohol Alert originally appeared in the National Institute on Alcohol Abuse and Alcoholism (NIAAA) Five-Year Strategic Plan. The Five-Year Plan highlights new research avenues that will address alcohol use across the human lifespan, emphasizing how drinking behavior evolves over an individual’s lifetime in response to biological, psychological, and social and environmental changes. Over time the Plan will be updated to reflect new and emerging research opportunities. The Plan is available online at http://pubs.niaaa.nih.gov/ publications/StrategicPlan/NIAAASTRATEGICPLAN.htm Full text of this publication is available on NIAAA’s World Wide Web site at www.niaaa.nih.gov. All material contained in the Alcohol Alert is in the public domain and may be used or reproduced without permission from NIAAA. Citation of the source is appreciated. Copies of the Alcohol Alert are available free of charge from the National Institute on Alcohol Abuse and Alcoholism Publications Distribution Center, P.O. Box 10686, Rockville, MD 20849–0686. U.S. DEPARTMENT OF HEALTH AND HUMAN SERVICES NIAAA Publications Distribution Center Attn.: Alcohol Alert P.O. Box 10686 Rockville, MD 20849–0686 Official Business Penalty for Private Use $300
https://openalex.org/W4293761186	https://lexikos.journals.ac.za/pub/article/download/1713/1003	English	null	Corpus-based Headword Selection Procedures for LSP Word Lists and LSP Dictionaries	Lexikos	2,022	cc-by	10,689	http://lexikos.journals.ac.za; https://doi.org/10.5788/32-1-1713 (Article) http://lexikos.journals.ac.za; https://doi.org/10.5788/32-1-1713 (Article) Corpus-based Headword Selection Procedures for LSP Word Lists and LSP Dictionaries Milica Vuković-Stamatović, Faculty of Philology, University of Montenegro, Nikšić, Montenegro (vmilica@ucg.ac.me) and Milica Vuković-Stamatović, Faculty of Philology, University of Montenegro, Nikšić, Montenegro (vmilica@ucg.ac.me) and Branka Živković, Faculty of Philology, University of Montenegro, Nikšić, Montenegro (brankaz@ucg.ac.me) and Branka Živković, Faculty of Philology, University of Montenegro, Nikšić, Montenegro (brankaz@ucg.ac.me) Abstract: In compiling both Language for Specific Purposes (LSP) word lists for foreign lan- guage learners and LSP dictionaries, the headword-selection process is of paramount importance. LSP word lists and LSP dictionaries will function effectively if they contain appropriate terms and register items, i.e. the lexical items that end users need. In this paper, we first present corpus-based LSP word lists, with special emphasis on how they were compiled. In the process, the make-up and size of the specialised corpus are important, as is the choice of the headword selection methods used. Among the possible criteria are word frequency, keyness, specialised occurrence, range, and dispersion, as well as some non-corpus linguistic methods that are more rarely applied. A greater variety of methods is used for compiling headword lists for LSP dictionaries, and of the corpus linguistic methods, frequency is typically solely applied. The article compares headword selection procedures for LSP word lists and LSP dictionaries before discussing how they can mutually inform one another. Keywords: LANGUAGE FOR SPECIFIC PURPOSES, LSP WORD LIST, LSP DICTIONARY, CORPUS LINGUISTICS, HEADWORD LIST, TERMS, HEADWORD SELECTION 1. Introduction Word lists have many purposes in the process of teaching and learning a for- eign language: they can be used as resources for vocabulary learning (Khani and Tazik 2013; Yang 2015), guidelines for designing curricula and courses, as well as for selecting reading and listening materials (Wang, Liang and Ge 2008; Jin et al. 2013), and guidelines for teachers in organising their explicit vocabu- lary teaching (Khani and Tazik 2013). The selection of headwords for inclusion in certain word lists has become an important strand of applied research in the field of foreign language teaching and learning in general, and language for specific purposes (LSP) in particular. As vocabulary sizes attained by native speakers are never attained by a vast majority of foreign language learners, the rationale guiding this type of research is to produce word lists of the sizes which are manageable for them to learn from. Word lists should provide lan- guage learners with the most useful words they need for a particular language function they are pursuing, for instance, attending university studies in a for- eign language or reading research articles from a particular specialist field in a foreign language. Some of these functions are related to LSP contexts and for them, consequently, LSP word lists are produced. Most of them are, in fact, English for Specific Purposes (ESP) word lists, given that English is the lan- guage which is most widely taught as a foreign language around the world. g g y g g g g In the past, both general and LSP word lists used to be compiled manually, typically relying on the compiler's intuition and, more rarely, on an authentic corpus of a very limited size by today's standards (West 1953; cf. Gilner 2011). However, over the past two decades, they have principally been derived from vast authentic corpora of general or specialised texts, which are carefully con- structed having particular types of foreign language learners in mind, and then scanned for words meeting certain criteria or a combination of criteria, such as the frequency of occurrence, distribution, range, or keyness (Coxhead 2000; Coxhead and Hirsh 2007; Brezina and Gablasova 2013; Browne et al. 2013a, Gardner and Davies 2014, etc.). hulle mekaar wedersyds van inligting kan voorsien. hulle mekaar wedersyds van inligting kan voorsien. hulle mekaar wedersyds van inligting kan voorsien. Keywords: LANGUAGE FOR SPECIFIC PURPOSES, LSP WORD LIST, LSP DICTIONARY, CORPUS LINGUISTICS, HEADWORD LIST, TERMS, HEADWORD SELECTION Lexikos 32 (AFRILEX-reeks/series 32: 2022): 141-161 Opsomming: Korpusgebaseerde lemmaseleksieprosedures vir TSD-woorde- lyste en -woordeboeke. In die samestelling van beide Taal vir Spesifieke Doeleindes-(TSD-)woor- delyste vir vreemdetaalleerders en TSD-woordeboeke is die lemmaseleksieproses van kardinale belang. TSD-woordelyste en TSD-woordeboeke sal effektief funksioneer indien hulle toepaslike terme en registeritems, m.a.w. die leksikale items wat eindgebruikers benodig, bevat. In hierdie artikel word korpusgebaseerde TSD-woordelyste eerste bespreek, met besondere klem op hul samestelling. In hierdie proses is die samestelling en grootte van die gespesialiseerde korpus, asook die keuse van die lemmaseleksiemetodes wat gebruik word, belangrik. Onder die moontlike kriteria is woordfrekwensie, sleutelstatus, gespesialiseerde voorkoms, omvang en verspreiding, asook enkele nie-korpus-linguistiese metodes wat minder gereeld toegepas word. 'n Groter verskeiden- heid metodes is gebruik vir die samestelling van lemmalyste vir TSD-woordeboeke, en van die korpuslinguistiese metodes is slegs frekwensie tipies toegepas. Lemmaseleksieprosedures vir TSD- woordelyste en TSD-woordeboeke word in die artikel vergelyk voordat daar bespreek word hoe http://lexikos.journals.ac.za; https://doi.org/10.5788/32-1-1713 (Article) http://lexikos.journals.ac.za; https://doi.org/10.5788/32-1-1713 (Article) 142 Milica Vuković-Stamatović and Branka Živković Corpus-based Headword Selection Procedures for LSP Word Lists & Dictionaries Baño 2008), i.e. what should be taken into account are different types of users, user situations and user needs (Tarp 2008), according to the theory of lexico- graphic functions (Bergenholtz and Tarp 1995; Tarp 2008). In principle, there are four main methods of selecting headwords for dictionaries — these assume relying on the existing dictionaries, grammar and etymology, canonical literary texts, or corpora (Esandi-Baztan and Fuertes-Olivera 2020). The fourth method, compiling headword lists based on corpora, has been an option for the past few decades and is now widely used in the process of making general dictionaries. However, as Bowker (2010: 166) notes, the use of corpus linguistic methods has been rather slow to take hold in the creation of specialised dictionaries. When it comes to the methods and procedures of compiling corpora for the purpose of creating LSP dictionaries as a type of specialised dictionaries, one may only rarely find detailed accounts regarding this issue (cf. Khumalo 2015; Đurović 2021; Kruse and Heid 2021). Also, typically, few details are also presented relating to the corpus-linguistic procedures employed as part of the process of selecting headwords from specialised corpora — most studies only briefly note that it is the frequency criterion that was applied (cf. Rundell and Kilgarriff 2011), with- out delving into the type of details that are provided by various specialised word-list compilers (cf. Lei and Liu 2016; Todd 2017; Dang 2018, etc.). In addi- tion, in these accounts, further corpus-linguistic procedures for headword selection beyond simple frequency are only sometimes mentioned in LSP dic- tionary research and projects (cf. Khumalo 2015; Đurović 2021; Kruse and Heid 2021). In this paper we compare corpus-based headword selection procedures used for producing LSP word lists and LSP dictionaries, bearing in mind that there are some similarities (although, also, important differences) between these two types of lexicographic products. We focus on the steps in headword selection that are based on corpus bearing in mind the important place that corpora currently have in their creation. The premise from which we depart is that the two fields can mutually inform and contribute to one another in terms of the corpus-based headword selection procedures. We will first present an overview of word lists, with a special focus on LSP word lists and how they are produced (section 2), after which we discuss LSP dic- tionaries and how headwords are selected for them (section 3). Corpus-based Headword Selection Procedures for LSP Word Lists & Dictionaries Section 4 com- pares headword selection for LSP world lists and LSP dictionaries. 1. Introduction The choice of the criteria and the related "cut- off" points (for instance, how frequent a word has to be to be included in a cer- tain word list) are informed by the target users' needs and involve a number of decisions during the compilation of the list. As corpora and software solutions evolve, so do the different methods for selecting those words. In this paper we will discuss various word lists intended for LSP learning, with a focus on how they were compiled. Selection of headwords for any dictionary, including specialised dictionaries, is also governed by the needs of its end users (Fuertes-Olivera and Arribas- http://lexikos.journals.ac.za; https://doi.org/10.5788/32-1-1713 (Article) 143 Milica Vuković-Stamatović and Branka Živković Milica Vuković-Stamatović and Branka Živković Milica Vuković-Stamatović and Branka Živković Although West's list was not generated using computer software, it was based on an authentic word corpus of 5 million words representing General English. About 2,000 word families1 were manually extracted and suggested to be the first words to be learned by any English language learner (they were mostly cho- sen according to the frequency criterion). This word list was very influential in English Language Teaching (ELT) and was used widely for decades (Nation 2013; Coxhead 2018). The emergence of the computer solutions providing data on a word's frequency and coverage in a corpus showed why — it turned out that West's list covered about 80% of the words used in most general English texts, or 4 in every 5 words. As English has about 70,000 word families (Nagy and Anderson 1984; Nation 2013), this word list proved to be a very useful resource (Coxhead 2000; Nation 2013). In the ensuing decades, other English words were built too (for instance, Campion and Elley 1971; Praninskas 1972; Lynn 1973; Ghadessy 1979; Xue and Nation 1984, etc.), however, the next word list which can match the influence of the GSL, the Academic Word List (AWL), came only in 2000 (Coxhead 2000). Its influence lies not only in how widely it was used in ELT, but the methodology of its compilation also set standards for many of the ensuing word lists (among them, Fraser 2007; Konstantikis 2007; Wang, Liang and Ge 2008; Khani and Tazik 2013; Valipouri and Nassaji 2013; Hsu 2013; Minshall 2013; Hsu 2014; Liu and Han 2015; Yang 2015; Lei and Liu 2016, etc.). The AWL contains 570 word families which are common in academic writing. To produce the list, Coxhead compiled a corpus of 3.5 million words of academic texts. The words were extracted according to the following criteria: (1) specialised occurrence (the words had to be outside high-frequency general words (outside the GSL in this case)), (2) frequency, (3) dispersion (the words had to occur in all the corpus's subsec- tions while featuring a certain frequency in all of them, and they also had to occur in at least half of the academic disciplines involved in the corpus) (Coxhead 2000). 2. This section first provides a brief overview of general and academic word lists, after which the focus is narrowed down to discipline-specific or LSP word lists. Reviews of word lists used for the purposes of foreign language teaching and learning typically start by presenting West's General Service List (GSL) (1953) (cf. Coxhead 2000; Coxhead and Hirsh 2007; Gardner and Davies 2014; Dang and Webb 2016; Dang, Coxhead and Webb 2017; Dang 2018; McQuillan 2020, etc.). http://lexikos.journals.ac.za; https://doi.org/10.5788/32-1-1713 (Article) 144 Corpus-based Headword Selection Procedures for LSP Word Lists & Dictionaries Corpus-based Headword Selection Procedures for LSP Word Lists & Dictionaries lion words. The lemmas from each of the 4 corpora were selected based on the criterion of the Average Reduced Frequency (this measure is obtained from the absolute frequency of the word and its distribution in the corpus (Savický and Hlavácová 2002)), and then the 4 lists were compared for overlaps — the shared items entered the New GSL. The same year, Browne, Culligan and Phil- lips (2013a) used a 273-million-word section of the Cambridge English Corpus to derive their list of about 2,800 lemmas based on the frequency criterion. Both lists outperform the old GSL in modern corpora, typically by a few percentage points. lion words. The lemmas from each of the 4 corpora were selected based on the criterion of the Average Reduced Frequency (this measure is obtained from the absolute frequency of the word and its distribution in the corpus (Savický and Hlavácová 2002)), and then the 4 lists were compared for overlaps — the shared items entered the New GSL. The same year, Browne, Culligan and Phil- lips (2013a) used a 273-million-word section of the Cambridge English Corpus to derive their list of about 2,800 lemmas based on the frequency criterion. Both lists outperform the old GSL in modern corpora, typically by a few percentage points. p p yp y y p g p Browne, Culligan and Phillips (2013b) also created the New AWL, con- taining 963 lemmas, by excluding the words already contained in the NGSL. Another replacement for the AWL was offered by Gardner and Davies (2014), who used a 120-million-word corpus (an academic subsection from COCA), to produce a list of about 3,000 lemmas (the Academic Vocabulary List, or the AVL). They did not exclude any group of words, but employed the keyness criterion solely: the authors took into account the ratio of words in their academic cor- pus, compared to a non-academic corpus. Newman (2016) and Hernandez (2017) found that the AVL outperforms the old AWL, while not much data is avail- able on how the NAWL performs against other similar lists. p g Other researchers have investigated whether lists such as the AWL might be created for other languages. Corpus-based Headword Selection Procedures for LSP Word Lists & Dictionaries Cobb and Horst (2004) studied the vocabulary profile of French and determined that the high-frequency vocabulary of this language is in fact more frequent than the high-frequency vocabulary of Eng- lish (2,000 most frequent French words reach a 90% coverage in most texts they examined), which excludes the need for creating additional lists for learners as these would reach very small coverages. Such results for French did not dis- courage other researchers to pursue creating corpus-based academic word lists for other languages, however. A Nordic joint-research project resulted in the crea- tion of the academic word lists for Swedish, Norwegian and Danish (Kokkinakis et al. 2012; Jansson et al. 2012; Ribeck et al. 2014; Johannessen et al. 2016). Two more independent lists have also been created for Danish — a word list of gen- eral, high-frequency items (2,000 words), as well as a word list of academic vocabulary (402 words) (Jakobsen et al. 2018). An Academic Vocabulary List in Russian has also been compiled recently (Talalakina et al. 2020). The develop- ment of all these word lists heavily relied on the English word-list projects pre- sented above. The word lists mentioned so far include general and non-discipline specific academic word lists. Unlike these, other word lists are much more specialised and these are the focus of this paper. They and the methods used for compiling them will be presented in the following section. Milica Vuković-Stamatović and Branka Živković These carefully weighed and strict criteria ensured that the word list would have a substantial coverage in any academic corpus, not just in the one it was derived from (Coxhead 2000). Indeed, the AWL's coverage of 10% in the corpus of its origin held strongly in many other academic corpora compiled later — for instance, it featured 10.07% in the academic medical corpus (Chen and Ge 2007), 11.17% in the academic applied linguistics corpus (Vongpumivitch et al. 2009), 9.96% in academic chemistry corpus (Valipouri and Nassaji 2013), etc. These impressive results confirmed that any future word list would have to be care- fully made, so as to be as useful as possible in a variety of similar language con- texts. One of the rare issues that may be contended against the AWL is the rela- tively small corpus it was derived from taking into account that it aims to be a general academic word list, an issue which the ensuing general lists have been trying to overcome. The dated GSL needed to be replaced and two new GSLs were offered to both research and instructional purposes in 2013. Brezina and Gablasova (2013) based their New GSL, containing about 2,500 lemmas, on a combined corpus of samples from 4 different corpora, together making 12 bil- http://lexikos.journals.ac.za; https://doi.org/10.5788/32-1-1713 (Article) 145 Milica Vuković-Stamatović and Branka Živković Milica Vuković-Stamatović and Branka Živković Milica Vuković-Stamatović and Branka Živković teaching LSP began to be strongly differentiated from teaching General Foreign Language in the 1960's. LSP teachers and researchers realised that taking the learners' specific needs into account, particularly their vocabulary needs, led to more effective teaching of the specialised language that they needed. With the rise of the ITC industry, corpus-based discipline-specific word lists, produced with the use of computers and from vast corpora, began to emerge at the turn of this century. y An overview of recent LSP word lists, along with the details of the corpora from which they were derived and the methods used for their creation, is given in the Appendix (while not entirely exhaustive, the table presents most of the word lists which have been described in scholarly papers). As was the case with general and academic word lists, the field of researching and compiling LSP word lists is almost exclusively related to the English language and, con- sequently, English word lists dominate the literature (as can be seen in the Appendix). Many of these lists follow in the AWL's footsteps given that they rely or build on the criteria used by Coxhead (2000) (see Section 2). Here we will provide a generalised description of the corpora and methods typically used to create LSP word lists. The texts for LSP corpora are chosen bearing the LSP word list's target users in mind. The corpora from which word lists are produced are typically custom-made, which makes their creation challenging and time-consuming. They also need to be of a relevant size. The corpora from which the LSP word lists were made vary widely in terms of their size — most of the word lists were developed from a specialised corpus of 1–2 million words (Mudraya 2006; Coxhead and Hirsh 2007; Wang, Liang and Ge 2008; Vongpumivitch, Huang and Chang 2009; Khani and Tazik 2013; Yang 2015; Todd 2017; Kwary and Artha 2017; Tongpoon-Patanasorn 2018; Đurović 2021). However, a recent ten- dency is to use larger corpora — most of the corpora from the last decade fea- tured 4 or more million words (Valipouri and Nassaji 2013; Hsu 2013; Hsu 2014; Lei and Liu 2016; Moini and Islamizadeh 2016; Dang 2018; Khany and Kalantari 2021; Kamrotov et al. 2022). 2.1 Corpus-based headword selection procedures for LSP word lists Realising the importance of the role of the communicative contexts in which certain foreign language learners will typically find themselves (Miller 2014: 305), http://lexikos.journals.ac.za; https://doi.org/10.5788/32-1-1713 (Article) 146 Corpus-based Headword Selection Procedures for LSP Word Lists & Dictionaries The sizes of LSP word lists also vary widely — from 92 (Martínez, Beck and Panza 2009) to 1,595 headwords (Dang 2018) and, again, the needs of the end users are taken into account when determining the list's size, as is the case with dictionaries. The criteria used for the selection of words for various recent LSP word lists can be summarised as follows: 1. frequency (the number is set depending on how large a list is wanted), 1. frequency (the number is set depending on how large a list is wanted), 2. specialised occurrence (being outside the most frequent 2,000 or 3,000 words, so as to avoid general high-frequency words; additionally, being outside the most frequent academic words (as represented by a chosen academic word list); finally, this also assumes the exclusion of proper nouns, symbols, abbreviations, numbers, non-words, etc.), y 3. dispersion (typically, occurrence in at least half the disciplines/subsec- tions which make the corpus, or being below some dispersion value (different methods for determining these are available)), 4. keyness (being found in the specialised corpus more frequently than in a reference corpus), 5. expert opinion (experts use rating scales and assign more points to more technical words), 6. cross-comparison with specialised dictionaries. The first four are purely corpus-linguistic methods and assume automatic extrac- tion of words based on the word-list compiler's decisions regarding the thresholds applied, while the last two depend on consulting either experts or specialised dictionaries, and are much more time-consuming. The final two steps have been generally avoided in developing most LSP word lists; having applied several corpus-linguistic filters, the word-list compilers found them unnecessary. Experts and dictionaries were consulted in the creation of just four out twenty-four LSP word lists presented in the Appendix (Wang, Liang and Ge 2008; Valipouri and Nassaji 2013; Jin et al. 2013; Tongpoon-Patanasorn 2018). j gp It should be added that the finalised LSP word lists are also typically vali- dated in one or several independent corpora (following Coxhead 2000) and, if their expected coverages hold in new corpora, such word lists are assumed to be truly representative. Milica Vuković-Stamatović and Branka Živković The biggest corpus used is the most recent one — a corpus of almost 30 million words of accounting research articles, which was used to obtain a list of the most frequent 658 accounting words (Khany and Kalantari 2021). The LSP word-list compilers who intend to apply the word selection crite- ria of range and dispersion need to think carefully about the make-up of their corpora as they generally need to have equal subsections of texts from various subfields. These corpora thus need to be well-structured and balanced; even though this is a challenging task, some researchers were able to produce sig- nificantly large and at the same time well-structured corpora — for instance, such is the English Hard Science Spoken Corpus of 6.5 million words, produced by Dang (2018), which features 12 subsections representing 12 hard science disciplines. This size is all the more impressive bearing in mind that this is a corpus of spoken language. http://lexikos.journals.ac.za; https://doi.org/10.5788/32-1-1713 (Article) 147 Corpus-based Headword Selection Procedures for LSP Word Lists & Dictionaries Corpus-based Headword Selection Procedures for LSP Word Lists & Dictionaries 3. LSP dictionaries As Bowker (2010) explains, LSP dictionaries belong to specialised dictionaries, i.e. dictionaries which treat specialised fields. They are also seen as a type of restricted dictionaries (Burkhanov 1998), where the term restricted does not imply their smaller size but reflects the fact that they focus on specific and pre- cise vocabulary (Mihindou 2004). LSP dictionaries exist in many fields of knowledge (Landau 2001), while developing the metalexicography related to them is in full swing (Fuertes-Olivera and Arribas-Baño 2008). g While the Anglo-Saxon strand in lexicography is mostly atheoretical (as was the case with the field of compiling word lists), the strand influenced by German and Nordic scholars advocates for developing lexicographical theories for guiding dictionary research and compilation (Fuertes-Olivera et al. 2013). As mentioned earlier, what is taken into account in the process of compiling any dictionary, including a specialised one, are the different types of users, user situations and user needs related to them, in line with the theory of lexico- graphic functions (Bergenholtz and Tarp 1995; Tarp 2008). This is one of the lexicographic theories which is very influential in pedagogical lexicography, including specialised pedagogical lexicography. g p p g g g p y As for users, specialised dictionaries have a more limited target audience than general dictionaries. According to Bergenholtz and Tarp (1995), their user type is decided based on user's mother language, level of encyclopedic knowl- edge, and native- and foreign-language competence. Applying these criteria, the authors identify four major user types for specialised dictionaries: experts with a high level of encyclopedic and foreign language competence, experts with a high level of encyclopedic competence and low level of foreign language competence, laypersons with a low level of encyclopedic competence and for- eign language competence, and layperson with a low level of encyclopedic competence and a high level of foreign language competence. Some more types are added by Fuertes-Olivera and Arribas-Baño (2008), who, among these user types, identify the following: experts from the specific field, semi-experts, experts from related or other fields, interested laypeople who would like to read some books or periodicals from the field, LSP students, translators, inter- preters, etc. Milica Vuković-Stamatović and Branka Živković Milica Vuković-Stamatović and Branka Živković atheoretical. So far, in the literature, there have not been any proposals to in- troduce a theory which would support the field. atheoretical. So far, in the literature, there have not been any proposals to in- troduce a theory which would support the field. Corpus-based Headword Selection Procedures for LSP Word Lists & Dictionaries y Few studies, typically those early ones or those using a vast corpus, used just one word-selection criterion (typically, frequency or keyness) (Mudraya 2006), while most of the studies employed a combined approach by using several of the methods — most often, following Coxhead's method (2000) (the first three steps above). None of the studies applied all the six methods combined. As can be seen, the field of producing and investigating word lists devel- oped as part of applied linguistics by Anglo-Saxon scholars, who, despite the fact that there are now many authors in it who are not Anglo-Saxon, still domi- nate it to a large extent. Most of the word lists are in fact English word lists. The creation of word lists is guided by pragmatic principles and the field remains http://lexikos.journals.ac.za; https://doi.org/10.5788/32-1-1713 (Article) 148 Corpus-based Headword Selection Procedures for LSP Word Lists & Dictionaries The mentioned user types have different needs in the mentioned different types of situations. These needs can be primary or function-related needs, which are the needs for information necessary to gain knowledge or solve a problem through using a dictionary, or they can be secondary or usage-related needs, which includes the need to know something about a specific dictionary and to know how to use it (Tarp 2008). There are different classifications of LSP dictionaries but we will briefly mention two which are relevant for our paper. Based on their size, there are two basic types — maximising LSP dictionaries, which attempt at covering as much of a field's terminology as possible, and minimising LSP dictionaries, in which a portion of the terminology is covered, typically only the most frequent items (Bergenholtz and Tarp 1995). Another possible classification recognises LSP dictionaries containing field-specific terms only, as opposed to general words, and hybrid LSP dictionaries, which combine both specialist and general words (Campoy Camillo 2002; Bowker 2010). LSP dictionaries for learners are a subtype of specialised dictionaries which are intended to assist users in learning about the terms and concepts used in a specific field, in one or more languages (Bowker 2010). Their purpose is to serve as auxiliary tools in the process of teaching and learning the lan- guage for specific purposes (Fuertes-Olivera and Arribas-Baño 2008). According to the mentioned theory of lexicographic functions, they are utility tools which assist learners in the process of learning LSP. 3. LSP dictionaries Tarp (2010) argues that there are many situations in which learners can benefit from specialised dictionaries — cognitive situations include systematic study of the specialised subject field and of problems related to the translation of specialised texts; communicative situations include reception and produc- tion of specialised texts in the mother tongue and in a foreign language, as well translation of specialised texts, while practical situations refer to various operative and interpretive situations. http://lexikos.journals.ac.za; https://doi.org/10.5788/32-1-1713 (Article) 149 Milica Vuković-Stamatović and Branka Živković Milica Vuković-Stamatović and Branka Živković general dictionaries. Bowker (2010) cites the example of the specialised diction- ary Dictionnaire d'apprentissage du français des affaires (DAFA) as a commendable example, given that it was based on a corpus of 25 million words. Taking into account the latest technological developments, recently, the compilation of such, relatively large, corpora has become much less of an issue. y g p The mentioned theory of lexicographic functions (Bergenholtz and Tarp 1995; Tarp 2008) suggests that headwords should be selected according to user's needs. When selecting headwords based on corpora, this, among other things, practically means that it is the user needs which govern the selection of texts which will enter such corpora. To illustrate how this can work in practice, we will briefly note how headwords for a Spanish accounting dictionary were selected (Fuertes-Olivera et al. 2013). Thus, following the mentioned function theory and the principle of relevance, the authors created a list of around 6,000 accounting texts, based on which three experts in accounting and one lexicog- rapher derived a stock of around 3,000 terms. Of the corpus-linguistic methods applied in this processing of the corpus, the authors calculated the word fre- quencies in their corpus, to inform their decisions of which terms to include in their specialised dictionary. They also used the Internet as a corpus and per- formed Google searches using particular word strings to extract additional 1,000 terms. Finally, 2,000 more terms were added through intensive reading of basic accounting texts. Such a hybrid approach was applied so as to ensure that the principle of relevance is adhered to. The authors argue and add that future updates of the term stock will be done by additionally analysing the log-files related to the online use of this dictionary (Fuertes-Olivera et al. 2013). y Other authors, too, mention applying the principle of frequency as one of the key steps taken in the process of selecting headwords for dictionaries (cf. Campoy Cubillo 2002; Hanks 2012; Rundell and Kilgarriff 2011). This criterion provides "solid empirical evidence for the occurrence of a word in actual language" (Xue and Tarp 2018). At the same time, they also argue that frequency may be mis- leading in some specialised fields which are updated constantly, such as accounting (Fuertes-Olivera and Nielsen 2011). 3.1 Corpus-based headword selection procedures for LSP dictionaries The process of headword selection is central in learner's lexicography (Xue and Tarp 2018), given that "dictionaries only function if they contain appropriate data," Nielsen (2018: 79). In this process, the three main questions that need to be posed refer to the size of the headword list, criteria and principles guiding their selection, and the empirical basis that their selection relies on (Tarp 2008). Tarp (2008) further suggests that headwords can be selected based on three sources, i.e. by means of introspection, using available descriptions in various publications (dictionaries, textbooks, etc.), and based on corpora. Building cor- pora as part of the preparatory stage for headword selection for LSP dictionaries is significant (Nkomo 2008: 105). Having compared corpus-based and intuition- based approaches, Verlinde and Selva (2001: 597) argue that it is the corpus- based lexicography that gives the "strong and necessary empirical evidence to the lexicographer's personal intuition", but they also note that intuition still remains helpful in filling in the gaps in cases when corpora are not balanced. As said earlier, Bowker (2010: 166) argues that the use of corpus linguistic methods has been rather slow to take hold in the creation of specialised dic- tionaries, on account of the fact that not so many specialised corpora are avail- able. Specialised corpora used for making dictionaries also tend to be relatively small, especially in comparison with the mega-corpora used for producing http://lexikos.journals.ac.za; https://doi.org/10.5788/32-1-1713 (Article) 150 Corpus-based Headword Selection Procedures for LSP Word Lists & Dictionaries Tarp too (2018). However, Tarp (2008) warns against the exalted status given to corpora and corpus-linguistic methods by certain lexicographers, arguing that corpora, however large they may be, can still be unrepresentative, and that the criteria of relevance and systematicity also need to be taken into account. Tarp too (2018). However, Tarp (2008) warns against the exalted status given to corpora and corpus-linguistic methods by certain lexicographers, arguing that corpora, however large they may be, can still be unrepresentative, and that the criteria of relevance and systematicity also need to be taken into account. What may be deduced from these various accounts is that corpora play an important role when selecting headwords for specialised dictionaries, and that word frequencies in a corpus can significantly inform the process of headword selection. Milica Vuković-Stamatović and Branka Živković Rundell and Kilgarriff (2011) rightly mention the fact that frequency is not a good selection criterion for extracting multiword items as candidates for headword lists. Likewise, Nielsen (2018: 81-82) suggests that frequency solely cannot guarantee that all relevant words will be selected, but that it should be used as a basis for the further selection process. p In some LSP dictionary compilation projects, similar to the methodology used in the production of LSP word lists, frequency is combined with addi- tional corpus-linguistic methods — thus, for instance, Khumalo (2015) and Đurović (2021) also use keyness; however, they do not ensure that the corpus contains equal shares of various subdisciplines of the field which it represents and, consequently, they do not apply the range filter. Some LSP dictionary compilers additionally use a more innovative, pattern-based approach (Kruse and Heid 2021). Frequency and relevance are suggested as two major criteria in Xue and http://lexikos.journals.ac.za; https://doi.org/10.5788/32-1-1713 (Article) 151 152 Milica Vuković-Stamatović and Branka Živković Milica Vuković-Stamatović and Branka Živković Both compilers of LSP word lists and compilers LSP dictionaries use fre- quency as a major criterion for deciding which words should enter their prod- ucts. In the process of producing LSP word lists, compilers typically either follow the cut-off points used in seminal research (such as Coxhead 2000) or, more frequently nowadays, the cut-off points are governed by the coverage achieved with the obtained word list, a coverage that allows for a certain threshold of reading or listening comprehension to be met. As for LSP dictionaries, in the literature we have not encountered detailed arguments around the chosen thresholds. The size of LSP dictionaries, in the- ory, should be governed by the user needs (even though there are always prac- tical and financial constraints to LSP dictionary projects) (Tarp 2008). However, so far, no method of quantifying them has been developed yet (and might not be, given the complexities involved). Research and projects involving LSP dictionaries frequently mention that frequency cannot be the sole criterion for selecting headwords, usually citing relevance as another major criterion to be applied, which, however, is much more difficult to define and employ. Likewise, as we have seen in the LSP word-list research, the criterion of simple frequency is also never applied as the sole criterion. Additional criteria may be applied as well, although these are also based on frequency to some extent. Thus, an important criterion for selecting headwords for LSP word lists is that of specialised occurrence, as pre- sented earlier, applied by excluding words which are highly frequent in gen- eral, reference corpora (typically 2,000 to 3,000 most frequent words in the case of English). Academic words can also be excluded, to ensure more technicality. Another criterion is that of range — applying this filter ensures that a word appears in a sufficient number of a discipline's subfields, so that it is equally valuable across that discipline, and not more valuable for some subspecialisa- tions and less valuable for others. To apply this criterion, however, one needs a corpus with equal subsections from the various subfields, as argued above. If the required structure of the corpus is not achieved, various dispersion thresholds can be applied. These criteria for guiding term extraction are rarely used when compiling headword lists for LSP dictionaries. http://lexikos.journals.ac.za; https://doi.org/10.5788/32-1-1713 (Article) http://lexikos.journals.ac.za; https://doi.org/10.5788/32-1-1713 (Article) 152 152 Milica Vuković-Stamatović and Branka Živković 4. Comparison of corpus-based headword selection procedures for LSP word lists and LSP dictionaries As we have seen, headword selection procedures for both LSP word lists and LSP dictionaries are guided by the needs of their users. The chief users of LSP word lists are LSP learners. LSP word lists are also used by LSP teachers and LSP material developers but, again, to the benefit of their end users — LSP learners. When it comes to the users of LSP dictionaries, as noted earlier, LSP learners make up an important category among them, however, many more categories of users are possible as well (e.g. translators, semi-experts, experts from other fields, etc.). This basic distinction in the types of users of the two products — LSP word lists and LSP dictionaries, has implications for how headwords are selected as part of their compilation procedures. When comparing corpus-based headword selection procedures for LSP word lists and LSP dictionaries, we can see that the former are compiled using corpus-linguistics methods almost exclusively, whereas a greater complexity of methods is used for the latter. A significant part of this difference may be explained by the respective homogeneity and heterogeneity of the end users of the two products, as explained above. p p The corpora from which LSP word lists are derived are rather large and typically well-structured and balanced, as we have seen. The details regarding their make-up are usually presented very precisely and transparently in the scholarly papers on LSP word lists, as well as given central prominence in them. On the other hand, the descriptions of corpora used for developing head- word lists for LSP dictionaries are usually not presented in such details and, typically, in the papers describing these projects relatively little space is devoted to the process of term extraction. In addition, equal representation of various subfields is rarely ensured in them. LSP word lists compilers argue that this is a good practice which allows that the frequencies of the terms obtained to reflect all subfields equally, and we tend to agree here. An implication from this comparison is that LSP lexicographers might invest this type of effort into compiling corpora from which they intend to extract terms. Moreover, given that many useful and balanced corpora have already been produced as part of LSP word-list research, some of these could be used for making LSP dictionaries as well. Corpus-based Headword Selection Procedures for LSP Word Lists & Dictionaries Corpus-based Headword Selection Procedures for LSP Word Lists & Dictionaries ever, applying at least some of the forementioned filters could help facilitate and automate that process. The mentioned filters used for obtaining LSP word lists have been found deficient, however, when it comes to extracting multi-word units and colloca- tions and, in fact, none of the word lists presented here contain such items. This is a major drawback to LSP word lists in general and a limitation that should be borne in mind if one were to apply some of the said methods for selecting pre- liminary headword lists for LSP dictionaries. Still, the ease with which most of the presented filters can be applied certainly recommends them for use in com- bination with other methods. Once an LSP word list is obtained via corpus linguistic methods, the work of the LSP word list compiler is either completed or almost completed in most cases, whereas much more work remains for a lexicographer compiling a head- word list for their LSP dictionary. The principle of systematicity is hardly ever applied to the LSP word lists obtained via corpus-linguistic methods. For instance, the Science List (Coxhead and Hirsh 2007) contains names of some common chemical elements (such as oxygen, potassium, etc.), while the names of other common elements are not mentioned (such as sulfur, for instance); it is debateable whether the word sul- fur is less useful to a science student learning English than the word potassium, for instance. Moreover, the Science List includes the word chloride, however, it does not include the name of the chemical element whose negatively charged ionic form it represents — chlorine. Thus, in general, word-list makers rely, per- haps too much, on automated procedures and avoid discussing these types of issues. As opposed to that, in LSP dictionary research and projects, systematicity is one of the central principles guiding the creation of headword lists. Observ- ing this principle when developing LSP word lists, we argue, could improve them, as the illogicalities of the types exemplified above typically stem from the imperfections of the corpus (in this case, the over-presence of texts mentioning the names of some particular chemical elements) and ought to be corrected when noticed. We would argue that, however large, well-structured and bal- anced a corpus may be, it will always suffer from some imperfections and can- not be trusted entirely. 152 Milica Vuković-Stamatović and Branka Živković One more criterion frequently mentioned when compiling LSP word lists is that of keyness, which is relatively easy to apply as no special make-up of the corpus is needed for it. As explained earlier, the frequency of the words in a specialised corpus is compared against that featured in a reference corpus and so the words found to be much more frequent in that specialised corpus are identified as terms. As we have seen, this criterion is sometimes used when extracting terms for LSP dictionaries as well. Very often, the mentioned additional criteria are used in combination when compiling LSP word lists. LSP word list compilers argue that applying them, in addition to simple frequency, ensures that the headwords selected are indeed relevant. The notion of relevance is more difficult to define for a product such as an LSP dictionary given its rather heterogeneous target audience; how- http://lexikos.journals.ac.za; https://doi.org/10.5788/32-1-1713 (Article) 153 Corpus-based Headword Selection Procedures for LSP Word Lists & Dictionaries y When finalised, LSP word lists are sometimes subjected to validation in additional corpora (not the ones they were derived from), to test how much coverage they would have in new texts. Validation, although effort- and time- consuming, is a commendable step to be taken, in our opinion. The frequency of preliminary, candidate headword lists for LSP dictionaries, could also be checked in additional specialised corpora, so as to, perhaps, rule out some can- didate terms which in validation corpora feature significantly lower frequen- cies as opposed to that from the first corpus. In developing LSP word lists, experts from the specialist fields are almost never involved, while they are always involved in compiling LSP dictionaries. This step is usually skipped in the making of modern word lists, given that http://lexikos.journals.ac.za; https://doi.org/10.5788/32-1-1713 (Article) 154 Milica Vuković-Stamatović and Branka Živković several automatic corpus-linguistic filters have already been applied. Although this is a demanding step, involving experts in the creation of any LSP product is advisable. several automatic corpus-linguistic filters have already been applied. Although this is a demanding step, involving experts in the creation of any LSP product is advisable. Endnote 1. A word family includes the headword with all its inflected and derived forms (for instance, suggest, suggests, suggested, suggesting, suggestion, suggestions). 5. Conclusion In this paper, we presented most modern LSP word lists and commented on how they were created. We also discussed corpus-based headword selection procedures for LSP dictionaries. A number of both similarities and differences were found in the two selection procedures and it was noted that both of them could, in some ways, benefit from being mutually informed. y g y On the one hand, more effort could be invested in the creation of LSP cor- pora, in terms of their size, make-up and balance, and also more corpus-lin- guistic selection procedures could be applied when compiling headword lists for LSP dictionaries than is currently typically the case, to facilitate the process. More transparency and precision when reporting on the corpora used and the corpus-linguistic methods applied for compiling headwords lists for LSP dic- tionaries is also advised. Lists obtained should also be validated in additional corpora, when possible. On the other hand, the creation of LSP word lists could be improved by applying additional non-corpus linguistic methods in their compilation, which is necessary to eliminate the illogicalities stemming from imperfectly balanced corpora, as well as to add the necessary multi-word units to them. p y Another observation that imposes itself from the comparison made in this paper is that the compilation and study of word lists remain atheoretical, while at least one strand of LSP dictionaries research has strong theoretical founda- tions. As we conclude this paper, we will ask the reader and ourselves if, per- haps, the moment has arrived that the field of word-list compilation and research be supported by a theory similar to that of the theory of lexicographic functions. Bergenholtz, H. and S. Tarp (Eds.). 1995. Manual of Specialised Lexicography: The Preparation of Spe- cialised Dictionaries. Amsterdam: John Benjamins. Bowker, L. 2010. The Contribution of Corpus Linguistics to the Development of Specialised Dic- tionaries for Learners. Fuertes-Olivera, P. (Ed.). 2010. Specialised Dictionaries for Learners: 155-168. Berlin: De Gruyter. 1. A word family includes the headword with all its inflected and derived forms (for instance, suggest, suggests, suggested, suggesting, suggestion, suggestions). Bergenholtz, H. and S. Tarp (Eds.). 1995. Manual of Specialised Lexicography: The Preparation of Spe- cialised Dictionaries. Amsterdam: John Benjamins. Bowker, L. 2010. The Contribution of Corpus Linguistics to the Development of Specialised Dic- Bowker, L. 2010. The Contribution of Corpus Linguistics to the Development of Specialised Dic- tionaries for Learners. Fuertes-Olivera, P. (Ed.). 2010. Specialised Dictionaries for Learners: 155-168. Berlin: De Gruyter. Corpus-based Headword Selection Procedures for LSP Word Lists & Dictionaries 155 Brezina, V. and D. Gablasova. 2013. Is There a Core General Vocabulary? Introducing the New General Service List. Applied Linguistics 36(1): 1-22. Browne, C., B. Culligan and J. Phillips. 2013a. New General Service List Project. http://www. newgeneralservicelist.org/. Accessed 1 September 2019. Browne, C., B. Culligan and J. Phillips. 2013b. New Academic World List. http://www newgeneralservicelist.org/nawl-new-academic-word-list/. Accessed 1 September 2019. Burkhanov, I.Y. 1998. Lexicography: A Dictionary of Basic Terminology. Rzeszów: Wydawnictwo. Wyższej Szkoły Pedagogicznej w Rzeszowie. Campion, M.E. and W.B. Elley. 1971. An Academic Vocabulary List. Wellington: New Zealand Council for Educational Research. Campoy Cubillo, M.C. 2002. Dictionary Use and Dictionary Needs of ESP Students: An Experi- mental Approach. International Journal of Lexicography 15(3): 206-228. Chen, Q. and G. Ge. 2007. A Corpus-based Lexical Study on Frequency and Distribution of Coxhead's AWL Word Families in Medical Research Articles. English for Specific Purposes 26: 502-514. Coxhead, A. 2000. A New Academic Word List. TESOL Quarterly 34(2): 213-238. Coxhead, A. 2018. Vocabulary and English for Specific Purposes Research: Quantitative and Qualitative Perspectives. London: Routledge. Coxhead, A. and D. Hirsch. 2007. A Pilot Science-specific Word List. Revue Française de Linguistique Appliquée 12(2): 65-78. Dang, T.N.Y. 2018. A Hard Science Spoken Word List. ITL — International Journal of Applied Linguis- tics 169(1): 44-71. Dang, T.N.Y., A. Coxhead and S. Webb. 2017. The Academic Spoken Word List. Language Learning 67(4): 959-997. Dang, T.N.Y. and S. Webb. 2016. Evaluating Lists of High-frequency Words. ITL — International Journal of Applied Linguistics 167(2): 132-158. Đurović, Z. 2021. Corpus Linguistics Methods for Building ESP Word Lists, Glossaries and Diction- aries on the Example of a Marine Engineering Word List. Lexikos 31: 259-282. Esandi-Baztan, M.A. and P.A. Fuertes-Olivera. 2020. Selecting an Initial Lemma List in Specialized Lexicography: A Case Study in the Field of Graphic Engineering. Lexikos 30: 57-89. Fraser, S. 2007. Providing ESP Learners with the Vocabulary They Need: Corpora and the Creation of Specialized Word Lists. Hiroshima Studies in Language and Language Education 10: 127-143. Fuertes-Olivera, P.A. and A. Arribas-Baño. 2008. Pedagogical Specialised Lexicography: The Represen- tation of Meaning in English and Spanish Business Dictionaries. Amsterdam/Philadelphia: John Benjamins. Fuertes-Olivera, P.A. and S. Nielsen. 2011. Online Dictionaries for Assisting Translators of LSP Texts: The Accounting Dictionaries. International Journal of Lexicography 25(2): 191-215. Fuertes-Olivera, P.A., H. Bergenholtz, P. Gordo-Gómez, S. Nielsen, M. Niño-Amo, Á. de los Ríos, A. Sastre-Ruano and M. Velasco-Sacristán. 2013. References Bergenholtz, H. and S. Tarp (Eds.). 1995. Manual of Specialised Lexicography: The Preparation of Spe- cialised Dictionaries. Amsterdam: John Benjamins. http://lexikos.journals.ac.za; https://doi.org/10.5788/32-1-1713 (Article) Milica Vuković-Stamatović and Branka Živković Milica Vuković-Stamatović and Branka Živković Hernandez, M.M. 2017. Comparing the AWL and AVL in Textbooks from an Intensive English Program. M.A. thesis. Provo, USA: Brigham Young University. Hsu, W. 2013. Bridging the Vocabulary Gap for EFL Medical Undergraduates: The Establishment of a Medical Word List. Language Teaching Research 17(4): 454-484. Hsu, W. 2014. Measuring the Vocabulary Load of Engineering Textbooks for EFL Undergraduates. English for Specific Purposes 33: 54-65. Jansson, H., S. Johansson Kokkinakis, J. Ribeck and E. Sköldberg. 2012. A Swedish Academic Word List: Methods and Data. Vatvedt Fjeld, R. and J.M. Torjusen (Eds.). 2012. Proceedings of the 15th EURALEX International Congress 7–11 August, 2012, Oslo, Norway: 955-960. Oslo: Department of Linguistics and Scandinavian Studies, University of Oslo. Jin, N.Y. et al. 2013. Development of the Engineering Technology Word List for Vocational Schools in Malaysia. International Education Research 1(1): 43-49. Johannessen, J.M., A. Saidi and K. Hagen. 2016. Constructing a Norwegian Academic Wordlist. Calzolari, N. et al. (Eds.). 2016. Proceedings of the Tenth International Conference on Language Resources and Evaluation, May 23–28, 2016, Portorož, Slovenia (LREC'16): 1457-1462. Portorož: European Language Resources Association. Kamrotov, M., E. Talalakina and D. Stukal. 2022. Technical Vocabulary in Languages for Special Purposes: The Corpus-based Russian Economics Word List. Lingua 273: 103326. Khani, R. and K. Tazik. 2013. Towards the Development of an Academic Word List for Applied Linguistics Research Articles. RELC Journal 44(2): 209-232. Khany, R. and B. Kalantari. 2021. Accounting Academic Word List (AAWL): A Corpus-based Study. Journal of Foreign Language Teaching and Translation Studies 6(1): 35-58. Khumalo, L. 2015. Semi-automatic Term Extraction for an isiZulu Linguistic Terms Dictionary. Lexikos 25: 495-506. Kokkinakis, S.J., E. Sköldberg, B. Henriksen, K. Kinn and J.B. Johannessen. 2012. Developing Academic Word Lists for Swedish, Norwegian and Danish — A Joint Research Project. Vatvedt Fjeld, R. and J.M. Torjusen (Eds.). 2012. Proceedings of the 15th EURALEX International Con- gress 7–11 August, 2012, Oslo, Norway: 563-569. Oslo: Department of Linguistics and Scandi- navian Studies, University of Oslo. Konstantakis, N. 2007. Creating a Business Word List for Teaching Business English. ELIA 7: 79-102. Kruse, T. and U. Heid. 2021. Lemma Selection and Microstructure: Definitions and Semantic Relations of a Domain-Specific e-Dictionary of the Mathematical Field of Graph Theory. Gavriilidou, Z., M. Mitsiaki and A. Fliatouras (Eds.). 2021. Proceedings of the XIX Euralex International Congress, Alexandroupolis, Greece, 7–11 September 2021. Volume 1: 227-233. Komotini, Greece: European Association for Lexicography. Kwary, D.A. http://lexikos.journals.ac.za; https://doi.org/10.5788/32-1-1713 (Article) http://lexikos.journals.ac.za; https://doi.org/10.5788/32-1-1713 (Article) 156 Corpus-based Headword Selection Procedures for LSP Word Lists & Dictionaries From Theory to Practice: The Selection of Spanish Lemmas in the Accounting Dictionaries. Fachsprache: International Journal of Specialized Communication 35(1–2): 25-42. Gardner, D. and M. Davies. 2014. A New Academic Vocabulary List. Applied Linguistics 35(3): 305-327. Gardner, D. and M. Davies. 2014. A New Academic Vocabulary List. Applied Linguistics 35(3): 305-327. Ghadessy, M. 1979. Frequency Counts, Wordlists, and Material Preparation: A New Approach. Gardner, D. and M. Davies. 2014. A New Academic Vocabulary List. Applied Linguistics 35(3): 305-327. Ghadessy, M. 1979. Frequency Counts, Wordlists, and Material Preparation: A New Approach. English Teaching Forum 17(1): 24-27. Ghadessy, M. 1979. Frequency Counts, Wordlists, and Material Preparation: A New Approach. English Teaching Forum 17(1): 24-27. Gilner, L. 2011. A Primer on the General Service List. Reading in a Foreign Language 23(1): 65-83. Corpus-based Headword Selection Procedures for LSP Word Lists & Dictionaries 157 Martínez, I.A., S.C. Beck and C.B. Panza. 2009. Academic Vocabulary in Agriculture Research Articles: A Corpus-based Study. English for Specific Purposes 28: 183-198. p y g f p f p McQuillan, J. 2020. Harry Potter and the Prisoners of Vocabulary Instruction: Acquiring Academic Language at Hogwarts. Reading in a Foreign Language 32(2): 122-142. Mihindou, G.-R. 2004. Some Features of Monolingual LSP Dictionaries. Lexikos 14: 118-136. Miller, L. 2014. English for Science and Technology. Bhatia, V. and S. Bremner (Eds.). 2014. The Routledge Handbook of Language and Professional Communication: 332-468. Oxon/New York: Routledge. Minshall, D.E. 2013. A Computer Science Word List. M.A. thesis. Swansea, Wales: Swansea University. Moini, R. and Z. Islamizadeh. 2016. Do We Need Discipline-specific Academic Word Lists? Lin- guistics Academic Word List (LAWL). Journal of Teaching Language Skills 35(3): 65-90. Mudraya, O. 2006. Engineering English: A Lexical Frequency Instructional Model. English for Spe- cific Purposes 25(2): 235-256. Nagy, W.E. and R.C. Anderson. 1984. How Many Words are There in Printed School English? Reading Research Quarterly: 304-330. Nation, I.S.P. 2013. Learning Vocabulary in Another Language. Second edition. Cambridge: Cam- bridge University Press. Newman, J.A. 2016. A Corpus-based Comparison of the Academic Word List and the Academic Vocabulary Newman, J.A. 2016. A Corpus-based Comparison of the Academic Wor List. M.A. thesis. Provo, USA: Brigham Young University. Newman, J.A. 2016. A Corpus-based Comparison of the Academic Word List and the Academic Vocabul List. M.A. thesis. Provo, USA: Brigham Young University. Newman, J.A. 2016. A Corpus based Comparison of the Academic Word List and the Academic Vocabulary List. M.A. thesis. Provo, USA: Brigham Young University. List. M.A. thesis. Provo, USA: Brigham Young University Nielsen, S. 2018. LSP Lexicography and Typology of Specialised Dictionaries. Humbley, J., G. Budin and C. Laurén (Eds.). 2018. Languages for Special Purposes: An International Handbook: 71-95. Berlin/Boston: De Gruyter. Nkomo, D. 2008. Towards a Theoretical Model for LSP Lexicography in Ndebele with Special Reference to a Dictionary of Linguistic and Literary Terms. M.A. thesis. Stellenbosch: Stellenbosch University. Praninskas, J. 1972. American University Word List. London: Longman. aninskas, J. 1972. American University Word List. London: Lon Ribeck, J., H. Jansson and E. Sköldberg. 2014. Från aspekt till övergripande — en ordlista över svensk akademisk vokabulär. Nordiske studier i leksikografi 12: 370-384. Rundell, M. and A. Kilgarriff. 2011. Automating the Creation of Dictionaries: Where Will it All End? Meunier, F. et al. (Eds.). 2011. A Taste for Corpora. Milica Vuković-Stamatović and Branka Živković and A.F. Artha. 2017. The Academic Article Word List for Social Sciences. MEXTESOL 41(4): 1-11. Landau, S.I. 2001. Dictionaries: The Art and Craft of Lexicography. Second edition. New York/Cam- bridge: Cambridge University Press. Lei, L. and D. Liu. 2016. A New Medical Academic Word List: A Corpus-based Study with Enhanced Methodology. English for Academic Purposes 22: 42-53. Liu, J. and L. Han. 2015. A Corpus-based Environmental Academic Word List Building and Its Validity Test. English for Specific Purposes 39: 1-11. Lynn, R.W. 1973. Preparing Word Lists: A Suggested Method. RELC Journal 4: 25-32. http://lexikos.journals.ac.za; https://doi.org/10.5788/32-1-1713 (Article) Corpus-based Headword Selection Procedures for LSP Word Lists & Dictionaries In Honour of Sylviane Granger: 257-282. Amsterdam/Philadelphia: John Benjamins. Savický, P. and J. Hlavácová. 2002. Measures of Word Commonness. Journal of Quantitative Linguis- tics 9(3): 215-231. Talalakina, E., D. Stukal and M. Kamrotov. 2020. Developing and Validating an Academic Vocabulary List in Russian: A Computational Approach. Modern Language Journal 104(3): 618-646. Tarp, S. 2008. Lexicography in the Borderland between Knowledge and Non-knowledge: General Lexico- graphical Theory with Particular Focus on Learner's Lexicography. Berlin/New York: Max Nie- meyer. Tarp, S. 2010. Functions of Specialised Learners' Dictionaries. Fuertes-Olivera, P. (Ed.). 2010. Spe- cialised Dictionaries for Learners: 39-53. Berlin/New York: De Gruyter. Todd, R.W. 2017. An Opaque Engineering Word List: Which Words Should a Teacher Focus on? English for Specific Purposes 45: 31-39. Tongpoon-Patanasorn, A. 2018. Developing a Frequent Technical Words List for Finance: A Hybrid Approach. English for Specific Purposes 51: 45-54. http://lexikos.journals.ac.za; https://doi.org/10.5788/32-1-1713 (Article) 158 Milica Vuković-Stamatović and Branka Živković Milica Vuković-Stamatović and Branka Živković Valipouri, L. and H. Nassaji. 2013. A Corpus-based Study of Academic Vocabulary in Chemistry Research Articles. Journal of English for Academic Purposes 12(4): 248-263. Verlinde, S. and T. Selva. 2001. Corpus-based Versus Intuition-based Lexicography: Defining a Word List for a French Learners' Dictionary. Rayson, P., A. Wilson, T. McEnery, A. Hardie and S. Khoja (Eds.). 2001. Proceedings of the Corpus Linguistics 2001 Conference: 594-598. Lan- caster: University Centre for Computer Corpus Research on Language, Lancaster University. and S. Khoja (Eds.). 2001. Proceedings of the Corpus Linguistics 2001 Conference: 594-598. Lan- caster: University Centre for Computer Corpus Research on Language, Lancaster University. Vongpumivitch, V., J.-Y. Huang and Y.-C. Chang. 2009. Frequency Analysis of the Words in the Academic Word List (AWL) and Non-AWL Content Words in Applied Linguistics Research Papers. English for Specific Purposes 28: 33-41. Wang, J., S. Liang and G. Ge. 2008. Establishment of a Medical Academic Word List. English for Specific Purposes 27(4): 442-458. Ward, J. 2009. A Basic Engineering English Word List for Less Proficient Foundation Engineering Undergraduates. English for Specific Purposes 28(3): 170-182. West, M. 1953. A General Service List of English Words. London: Longman, Green & Co. Xue, G. and P. Nation. 1984. A University Wordlist. Language Learning and Communication 3(2): 215-229. Xue, M. and S. Tarp. 2018. Towards Chinese Learner's Dictionaries for Foreigners Living in China: S P bl R l d L S l L ik 28 384 404 Xue, M. and S. Tarp. 2018. Towards Chinese Learner's Dictionaries for Some Problems Related to Lemma Selection. Lexikos 28: 384-404. Xue, M. and S. Tarp. 2018. Towards Chinese Learner s Dictionaries for F Some Problems Related to Lemma Selection. Lexikos 28: 384-404. Yang, M.-N. 2015. A Nursing Academic Word List. English for Specific Purposes 37: 27-38. http://lexikos.journals.ac.za; https://doi.org/10.5788/32-1-1713 (Article) Corpus-based Headword Selection Procedures for LSP Word Lists & Dictionaries 160 Milica Vuković-Stamatović and Branka Živković 160 Vongpumi- vitch, Huang and Chang (2009) Applied linguistics word list 475 AWL word forms and 128 non-AWL content word forms 1.5 mill. (200 RAs from 5 journals) / 8.6% for the 475 AWL words + 2.8% for the non-AWL words — frequency of at least 50 in the cor- pus — occurrence of at least 5 times in at least 5 journals — exclusion of the GSL, function words and abbreviations Jin et al. (2013) Engineering technology word list (ETWL) 313 word lists 124,584 words (2 textbooks) / 8.7% — exclusion of the GSL and the AWL words — to be defined as technical, a word had to appear as an entry in an online engineering dictionary — cross-checked by two experts from the field to make sure the words were technical and semi- technical Khani and Tazik (2013) Applied linguistics academic word list 773 word families 1,553,450 words (240 RAs from 12 journals) / 12.48% — exclusion of the GSL words — frequency of at least 50 in the cor- pus — occurrence of at least 4 times in at least half the journals Valipouri and Nassaji (2013) Chemistry Academic word list (CAWL) 1,400 word families 4 mill. (1,185 RAs from 38 journals) / 81.18% — frequency of at least 114 in the corpus — frequency of at least 10 in all sub- sections — exclusion of abbreviations and function words — three chemistry professors excluded technical words using a rating scale Hsu (2013) Medical word list (MWL) 595 word families 15 mill. (155 medi- cal textbooks) / 10.72% — exclusion of the BNC 3000 — occurrence in more than half the subsections — frequency of at least 863 in the corpus Minshall (2013) Computer science word list 433 word families 3,661,337 tokens (RAs and confer- ence proceedings from 10 subdisci- plines) / 6% — outside the GSL and the AWL — occurrence in at least half the sub- disciplines — occurrence of at least 80 in the corpus Hsu (2014) Engineering English word list (EEWL) 729 word families 4.57 mill. Corpus-based Headword Selection Procedures for LSP Word Lists & Dictionaries 159 p Appendix Author List Number of words Corpus/ Coverage in the corpus Methodology Mudraya (2006) Engineering English word list 1,260 word families 2 mill. (13 text- books) / not given — frequency of the word of at least 100 Coxhead and Hirsh (2007) Pilot science list 318 word families 1,761,380 words (14 disciplines) / 3.79% — exclusion of the GSL & the AWL words — occurrence in at least half the disciplines — frequency of at least 50 in the corpus — the dispersion factor of at least 35 — exclusion of proper nouns, sym- bols and abbreviations Fraser (2007) Pharmacology word list (PWI) 601 word families 185,000 (51 research articles (RAs)) — exclusion of the GSL and the AWL words — exclusion of proper names, nationalities, numbers, abbrevia- tions and acronyms — occurrence in at least 6 RAs — frequency of 10 or more and occurrence in at least 2 RAs Konstantakis (2007) Business word list for under- graduates (BWL) 560 word families 600,000 (Published Materials Corpus, compiled by Nelson (2000); 33 course books) / 2.79% — exclusion of the GSL and the AWL words — occurrence in at least 5 course books — exclusion of proper names, numericals, Latin words, nationali- ties, acronyms, interjections and abbreviations — frequency of at least 10 in the cor- pus Wang, Liang and Ge (2008) Medical aca- demic word list (MAWL) 623 word families 1,093,000 words (288 research arti- cles from 96 jour- nals) / 12.24% — exclusion of the GSL words — occurrence in at least half the disciplines — frequency of at least 30 in the cor- pus — two English for Medicine profes- sors consulted for differentiating between technical and academic vocabulary Ward (2009) Basic Engi- neering list (BEL) 299 word types 271,000 words (25 textbooks in engineering) / 16.4% — exclusion of function words — definition of word as a word type — frequency of at least 5 in each of the 5 subsections Martínez, Beck and Panza (2009) A reduced AWL for agriculture 92 word families 826,416 (218 RAs) / coverage not given — inclusion of the academic words above the mean for academic words http://lexikos.journals.ac.za; https://doi.org/10.5788/32-1-1713 (Article) 160 Milica Vuković-Stamatović and Branka Živković (100 engineering textbooks) / 14.3% — exclusion of the BNC 2000 — occurrence in all subsections — occurrence in at least 95 out 100 textbooks — frequency of at least 288 times in the corpus — exclusion of exclamations, inter- jections and proper names Liu and Han (2015) Environmental academic word list (EAWL) 458 word families 862,242 tokens (200 RAs) / 15.43% — exclusion of the GSL words — frequency of at least 30 in the cor- pus — occurrence in at least 8 out 10 sub- sections http://lexikos.journals.ac.za; https://doi.org/10.5788/32-1-1713 (Article) Corpus-based Headword Selection Procedures for LSP Word Lists & Dictionaries 161 Yang (2015) Nursing academic word list 676 word families 1,006,934 words (252 RAs) / 13.64% — exclusion of the GSL words range — occurrence in at least half the sub- sections — frequency of at least 33 in the cor- pus Lei and Liu (2016) Medical Academic Vocabulary List (MAVL) 819 lem- mas 6.2 mill. (760 medi- cal RAs from 38 journals (MAEC corpus + 1 three- volume textbook (MAET corpus)), 19.44% in MAEC and 20.18% in MAET — frequency of 28.57 per 1 mill — ratio of at least 1.5 (at least 50% higher frequency in the academic corpus than in a non-academic corpus) — occurrence of 20% of the expected frequency in at least half the sub- sections — dispersion of at least 0.5 (Jullian's D) — no lemma should occur more than 3 times the expected frequency in more than any 3 out 21 subsec- tions — special meaning criterion checked via 2 medical dictionaries Moini and Islamizadeh (2016) Linguistics academic word list 1,263 word families 4 mill. (700 RAs from 4 subdisci- plines)/ not given — a frequency of at least 114 in the corpus — occurrence of at least 10 times in each subdiscipline Todd (2017) Opaque engineering word list 186 items 1.15 mill. (27 text- books) / not given — exclusion of non-words, abbrevia- tions, and function words — occurrence in at least 15 out 27 textbooks — similar word types combined under one entry Kwary and Artha (2017) Academic article word list for social sciences 350 word families 1,040,259 tokens (122 RAs) — outside the GSL — occurrence in all the 6 subdisci- plines — frequency Dang (2018) Hard science spoken word list 1,595 word families 6.5 mill. 160 Milica Vuković-Stamatović and Branka Živković words of spoken language from 12 disciplines — occurrence in at least half the disciplines and both subsections of the corpus — frequency of at least 175 in the corpus — dispersion (DP value below 0.6) Tongpoon- Patanasorn (2018) Frequent technical word list for finance 569 word families 2 mill. words (books, journals, websites and news- papers) — keyness — rating scale, completed by experts Khany and Kalantari (2021) Accounting academic word list 658 word families 29.5 mill. words (2,098 accounting RAs) — outside the GSL — frequency of at least 839 in the corpus and at least 50 in each journal Kamrotov, Talalakina and Stukal (2022) Russian economics word list 507 lemmas 10.5 million words (economics RAs and mass media economics articles) — keyness (1.5 more frequent in the economics corpus than in other corpora) — degree of dispersion over 0.25 — minimum frequency of 10 in the corpus Corpus-based Headword Selection Procedures for LSP Word Lists & Dictionaries 161
https://openalex.org/W2797161962	https://zenodo.org/records/1487885/files/paper58_cameraready.pdf	English	null	An Open Access User Generated Video Dataset from 2016 Edinburgh Festival	null	2,017	cc-by	1,520	Farzad Toutounchi, Ebroul Izquierdo Farzad Toutounchi, Ebroul Izquierdo Multimedia and Vision Research Group Queen Mary University of London Mile End Road, E1 4NS, London, UK Multimedia and Vision Research Group Queen Mary University of London Mile End Road, E1 4NS, London, UK Keywords: Dataset, user generated content, quality assess- ment, video enhancement. The rest of this paper is structured as follows. The descrip- tion of the dataset is listed in Section 2. In Section 3 the char- acteristics of the UGC dataset is presented in terms of spatio- temporal diversity and complexity of the existing textures and features in the videos, and ﬁnally in Section 4 the terms of use and guidelines for users are explained. 2 Description of the Collection The dataset includes 7 videos recorded at the Edinburgh festi- val in August 2016, using a Samsung Galaxy S5 smart phone. The videos mainly cover the Edinburgh streets and the festi- val atmosphere, and do not cover any performances. Speciﬁc descriptions of the videos are provided in Table 1 and screen- shots are shown in Figure 1. The dataset includes videos with UHD (3840×2160), Full-HD (1920×1080), HD (1280×720), and VGA (640×480) spatial resolutions. One of the Full-HD sequences is captured at 60 fps frame rate and the rest are at 30 fps. The bit depth for all videos is 8. All except one of the videos are recorded in landscape mode. 1 Introduction With the explosion of social networks and media sharing plat- forms, users are getting more involved in capturing and record- ing of moments and highlights in different public events such as festivals and sport events. The huge amount of content recorded by the smart phones creates an opportunity for pro- fessional broadcasters to expand their coverage by using the user generated content (UGC) from particular events. It is how- ever of major importance to analyze and post-process the UGC before using it in the production to ensure a high quality of ex- perience for the viewers. Design and implementation of such tools for UGC processing requires proper test sets to validate the performance of the research modules. All videos are provided in their native format as MP4 ﬁles, as produced in the mobile device, with H.264/AVC video en- coding [2] and AAC audio encoding [3]. All original metadata are retained; only the ﬁle names have been changed. Abstract This paper presents a user generated video dataset captured during the 2016 Edinburgh festival. The provided dataset was collected using a smart phone and is available with no post- processing. The dataset can be used for evaluation of vari- ous research tools, such as video quality assessment and en- hancement. The details and characteristics of each video in the dataset are listed, and usage guidelines for the researchers are also stated. An Open Access User Generated Video Dataset from 2016 Edinburgh Festival Farzad Toutounchi, Ebroul Izquierdo 3 Analysis of the Content We analyzed the sequences from different perspectives to eval- uate their suitability to be used in the context of research ac- tivities. In particular, as this dataset is intended mainly for validation of quality assessment, quality enhancement, and compression algorithms, it is important to investigate the spa- tial and temporal diversity of the videos. In this regard, we computed the spatial and temporal indexes of each video ac- cording to the ITU-R speciﬁcations [4]. Moreover we re- encoded the sequences to evaluate the rate-distortion perfor- mance of each video. For re-encoding, we decoded the videos using FFmpeg library [5] ﬁrst, and then encoded the videos in H.265/HEVC [6] using the open source HEVC Turing codec [7] in random access conﬁguration with four QPs of 22, 27, 32, and 37. The COGNITUS project aims to combine the advances in UHD broadcasting technologies with the explosion of UGC in order to create new interactive, immersive modes of produc- tion. The project is intended to optimize how UHD content is produced and distributed, by capitalizing upon the knowledge of professional producers, the ubiquity of UGC and the power of interactive networked social creativity. [1] In the context of this project, and given that public festivals, and in particular the annual Edinburgh festival, are amongst the targeted events in COGNITUS, a collection of 7 user generated videos were created in the 2016 Edinburgh festival for evalu- ation of different UGC quality assessment and enhancement components. The public dataset can be used by researchers to validate their work on UGC processing and media synchroniza- tion. Table 2 summarizes the spatial and temporal indexes of each video, while Figure 2 shows the diversity of the spatio- temporal activity in the dataset using the spatial and tempo- EdFest2016 1 EdFest2016 2 EdFest2016 3 EdFest2016 4 EdFest2016 5 EdFest2016 6 EdFest2016 7 re 1. Screenshots of the videos. EdFest2016 1 EdFest2016 2 EdFest2016 2 EdFest2016 1 EdFest2016 1 EdFest2016 3 EdFest2016 2 EdFest2016 3 EdFest2016 5 EdFest2016 4 EdFest2016 4 EdFest2016 5 EdFest2016 6 EdFest2016 7 EdFest2016 7 EdFest2016 6 Figure 1. Screenshots of the videos. EdFest2016 6 EdFest2016 7 Figure 1. Screenshots of the videos. File name Length [sec] Resolution Frame rate [fps] Orientation Short description EdFest2016 1.mp4 19 1920×1080 30 Landscape Street scene with passersby. EdFest2016 2.mp4 19 1280×720 30 Portrait Street scene with passersby. EdFest2016 3.mp4 14 1280×720 30 Landscape Street scene with passersby. 1https://creativecommons.org/licenses/by-nc-nd/4.0/ 2https://doi.org/10.5281/zenodo.840833 References [1] COGNITUS team. Horizon 2020 COGNITUS Project. http://cognitus-h2020.eu, 2016–2018. Table 2. Spatial and temporal indexes of the dataset. [2] T. Wiegand, G. J. Sullivan, G. Bjontegaard, and A. Luthra. Overview of the H.264/AVC Video Coding Standard. IEEE Trans. Cir. and Sys. for Video Technol., 13:560–576, 2003. Sequence name Bitrate [Mbps] PSNR [dB] EdFest2016 1 5.79 37.42 EdFest2016 2 11.55 32.33 EdFest2016 3 2.76 37.22 EdFest2016 4 2.24 37.97 EdFest2016 5 16.34 38.54 EdFest2016 6 0.84 37.22 EdFest2016 7 8.01 37.88 Table 3. Bitrate and PSNR results of re-encoding the dataset. Sequence name Bitrate [Mbps] PSNR [dB] [3] ISO. ISO/IEC 13818-7:2006 - Information technology – Generic coding of moving pictures and associated audio information – Part 7: Advanced Audio Coding (AAC). Technical report, International Organization for Standard- ization, 2006. [4] ITU-R. ITU-R Recommendation P.910: Subjective video quality assessment methods for multimedia applications. Technical report, International Telecommunication Union, 1999. Table 3. Bitrate and PSNR results of re-encoding the dataset. ral indexes. Table 3 summarizes the re-encoding results of the database, where the bitrates and PSNR values represent the av- erage over the four QPs. Moreover Figure 3 illustrates the rate- distortion curves of each video in one plot, where we can see a signiﬁcant variation in the results. [5] FFmpeg team. FFmpeg. https://www.ffmpeg.org, 2000– 2017. [6] Gary J. Sullivan, Jens-Rainer Ohm, Woojin Han, and Thomas Wiegand. Overview of the High Efﬁciency Video Coding (HEVC) Standard. IEEE Transactions on Cir- cuits and Systems for Video Technology, pages 1649–1668, 2012. Acknowledgements Sequence name Spatial index Temporal index The work described in this paper has been conducted within the project COGNITUS. This project has received funding from the European Unions Horizon 2020 research and innovation program under grant agreement No 687605. 3 Analysis of the Content EdFest2016 4.mp4 30 1280×720 30 Landscape Trafﬁc junction and buildings. EdFest2016 5.mp4 45 3840×2160 30 Landscape View of the city. EdFest2016 6.mp4 36 640×480 30 Landscape View of the city. EdFest2016 7.mp4 53 1920×1080 60 Landscape Trafﬁc and buildings. File name Length [sec] Resolution Frame rate [fps] Orientation Short description Table 1. Description of the dataset. Sequence name Spatial index Temporal index EdFest2016 1 12.51 64.49 EdFest2016 2 37.21 40.65 EdFest2016 3 15.62 66.99 EdFest2016 4 12.91 51.78 EdFest2016 5 13.11 42.43 EdFest2016 6 16.51 38.27 EdFest2016 7 12.92 36.40 Table 2. Spatial and temporal indexes of the dataset. Sequence name Bitrate [Mbps] PSNR [dB] EdFest2016 1 5.79 37.42 EdFest2016 2 11.55 32.33 EdFest2016 3 2.76 37.22 EdFest2016 4 2.24 37.97 EdFest2016 5 16.34 38.54 EdFest2016 6 0.84 37.22 EdFest2016 7 8.01 37.88 Table 3. Bitrate and PSNR results of re-encoding the dataset. Sequence name Spatial index Temporal index EdFest2016 1 12.51 64.49 EdFest2016 2 37.21 40.65 EdFest2016 3 15.62 66.99 EdFest2016 4 12.91 51.78 EdFest2016 5 13.11 42.43 EdFest2016 6 16.51 38.27 EdFest2016 7 12.92 36.40 Table 2. Spatial and temporal indexes of the dataset. Acknowledgements 4 Usage Guidelines The dataset is owned by Queen Mary University of Lon- don c⃝2017, and is provided under Creative Commons Attribution-NonCommercial-NoDerivatives1 license. The dataset is accessible through the Zenodo research data repos- itory2. Any queries need to be forwarded to the authors of this paper. [7] Turing team. HEVC Turing Codec. http://turingcodec.org. [8] Rajitha Weerakkody. An Open Access Dataset of User Generated Videos from Edinburgh Festival 2016. https://doi.org/10.5281/zenodo.582605, May 2017. It is highly recommended to the users, to use this dataset along with other publicly available datasets from 2016 Edin- burgh festival, in particular [8], where there are overlapping videos covering the same instances as in our dataset. Figure 2. Spatio-temporal index space of the dataset. Figure 2. Spatio-temporal index space of the dataset. Figure 2. Spatio-temporal index space of the dataset. Figure 3. Rate-distortion results of re-encoding the dataset. Figure 3. Rate-distortion results of re-encoding the dataset. Figure 3. Rate-distortion results of re-encoding the dataset.
https://openalex.org/W2900717987	https://jneuroinflammation.biomedcentral.com/track/pdf/10.1186/s12974-018-1366-3	English	null	High neopterin and IP-10 levels in cerebrospinal fluid are associated with neurotoxic tryptophan metabolites in acute central nervous system infections	Journal of neuroinflammation	2,018	cc-by	10,941	Quist-Paulsen et al. Journal of Neuroinflammation (2018) 15:327 https://doi.org/10.1186/s12974-018-1366-3 Quist-Paulsen et al. Journal of Neuroinflammation (2018) 15:327 https://doi.org/10.1186/s12974-018-1366-3 RESEARCH Open Access High neopterin and IP-10 levels in cerebrospinal fluid are associated with neurotoxic tryptophan metabolites in acute central nervous system infections Else Quist-Paulsen1,2* , Pål Aukrust3,4,5,6,7, Anne-Marte Bakken Kran9,2,5, Oona Dunlop10, Vidar Ormaasen1, Birgitte Stiksrud1,2, Øivind Midttun8, Thor Ueland3,5,6,7, Per Magne Ueland8, Tom Eirik Mollnes2,6,7,11,13,14,15 and Anne Ma Dyrhol-Riise1,2,6,12 RESEARCH Open Access High neopterin and IP-10 levels in cerebrospinal fluid are associated with neurotoxic tryptophan metabolites in acute central nervous system infections Else Quist-Paulsen1,2* , Pål Aukrust3,4,5,6,7, Anne-Marte Bakken Kran9,2,5, Oona Dunlop10, Vidar Ormaasen1, Birgitte Stiksrud1,2, Øivind Midttun8, Thor Ueland3,5,6,7, Per Magne Ueland8, Tom Eirik Mollnes2,6,7,11,13,14,15 and Anne Ma Dyrhol-Riise1,2,6,12 High neopterin and IP-10 levels in cerebrospinal fluid are associated with neurotoxic tryptophan metabolites in acute central nervous system infections Else Quist-Paulsen1,2* , Pål Aukrust3,4,5,6,7, Anne-Marte Bakken Kran9,2,5, Oona Dunlop10, Vidar Ormaasen1, Birgitte Stiksrud1,2, Øivind Midttun8, Thor Ueland3,5,6,7, Per Magne Ueland8, Tom Eirik Mollnes2,6,7,11,13,14,15 and Anne Ma Dyrhol-Riise1,2,6,12 Abstract Background: The host response to intruders in the central nervous system (CNS) may be beneficial but could also be harmful and responsible for neurologic symptoms and sequelae in CNS infections. This immune response induces the activation of the kynurenine pathway (KP) with the production of neuroactive metabolites. Herein, we explored cytokine and KP responses in cerebrospinal fluid (CSF) and serum in patients with encephalitis, aseptic, and bacterial meningitis. Methods: Cytokines were measured in CSF and serum by multiplex assay in adult patients with encephalitis of infectious, autoimmune or unknown etiology (n = 10), aseptic meningitis (ASM, n = 25), acute bacterial meningitis (ABM, n = 6), and disease control patients with similar symptoms but without pleocytosis in CSF (n = 42). Liquid chromatography-tandem mass spectrometry (LC-MS/ MS) was used to measure KP metabolites in CSF and serum. Results: A characteristic pattern of increasing cytokine levels and KP metabolites was found in CSF from encephalitis to ASM, with the highest levels in ABM. In ASM and ABM, most inflammatory mediators, including IL-6, IL-8, and IFN- inducible protein-10 (IP-10), showed markedly elevated levels in CSF compared with serum, indicating production within the CNS. In contrast to most mediators, the highest level of IP-10 was found in the ASM group, suggesting a potential role for IP-10 in aseptic/viral meningitis. Neopterin and IP-10 were associated with marked changes in KP metabolites in CSF with increasing kynurenine/tryptophan ratio reflecting indoleamine 2,3-dioxygenase activity. Neopterin, a marker of IFN-γ activity, was associated with an unfavorable balance between neuroprotective and neurotoxic tryptophan metabolites. Conclusion: We show that parenchymal and meningeal inflammations in CNS share a characteristic cytokine profile with a general immune response in the CSF with limited influence from the systemic circulation. IFN-γ activity, assessed by neopterin and IP-10 levels, may play a role in the activation of the KP pathway in these patients, potentially mediating neurotoxic effects. Keywords: Encephalitis, Aseptic meningitis, Bacterial meningitis, Cytokines, Chemokines, Kynurenine tryptophan pathway, Indoleamine 2,3-dioxygenase, Neopterin * Correspondence: e.q.paulsen@studmed.uio.no 1Department of Infectious Diseases, Oslo University Hospital, Ullevaal Hospital, P. O. Box 4956 Nydalen, N-0450 Oslo, Norway 2Institute of Clinical Medicine, University of Oslo, Oslo, Norway Full list of author information is available at the end of the article Background indoleamine 2,3-dioxygenase (IDO) which is upregu- lated by inflammatory cytokines, mainly by interferon gamma (IFN-γ) [20], linking T cell activation to the regulation of the KP. g The host inflammatory response to intruders to the central nerve system (CNS) plays an important role for neuronal injury in encephalitis and meningitis. The cytokine profiles of aseptic meningitis (ASM) and acute bacterial meningitis (ABM) have been investi- gated in several studies, in general showing increased levels of inflammatory mediators [1–8]. However, for encephalitis, inflammatory responses have mainly been evaluated for patients with herpes simplex virus (HSV) infection [9–13]. Thus, comparison of cytokine levels in encephalitis, ASM and ABM and control pa- tients are scarce. Moreover, most studies have focused on a limited number of inflammatory markers, and few studies have examined parallel samples of serum and cerebrospinal fluid (CSF) from the same patients. It is known that the inflammation activates the kynurenine pathway (KP) resulting in the depletion of tryptophan (TRP) and formation of metabolites with potential neurotoxic (e.g., quinolinic acid [QA], 3-hydroxykynurenine [3-HK]) and neuroprotective (e.g., kynurenic acid [KYNA]) effects (Fig. 1) [14, 15]. The activation of the KP seems to also have immune modulating effects, resulting in inhibition of TH1 cells, activation of regulatory T cells (Tregs) and inhib- ition of natural killer (NK) cells [16–19]. In the CNS, the rate-limiting enzyme for TRP catabolism is The host inflammatory response to intruders to the central nerve system (CNS) plays an important role for neuronal injury in encephalitis and meningitis. The cytokine profiles of aseptic meningitis (ASM) and acute bacterial meningitis (ABM) have been investi- gated in several studies, in general showing increased levels of inflammatory mediators [1–8]. However, for encephalitis, inflammatory responses have mainly been evaluated for patients with herpes simplex virus (HSV) infection [9–13]. Thus, comparison of cytokine levels in encephalitis, ASM and ABM and control pa- tients are scarce. Moreover, most studies have focused on a limited number of inflammatory markers, and few studies have examined parallel samples of serum and cerebrospinal fluid (CSF) from the same patients. © The Author(s). 2018 Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated. Page 2 of 14 Quist-Paulsen et al. Journal of Neuroinflammation (2018) 15:327 Page 2 of 14 Quist-Paulsen et al. Journal of Neuroinflammation (2018) 15:327 Background Altered cytokine levels and associated imbalance of neurotoxic and neuroprotective metabolites in the KP have been suggested to contribute to the pathogenesis of several chronic conditions in the CNS, such as schizo- phrenia [21, 22], bipolar disorders [23, 24], Parkinson’s disease [25], Alzheimer’s [26] and Huntington’s disease [27], and AIDS-related dementia [28, 29], as well as in traumatic brain injury [30]. However, there is limited data on the role of KP in acute infections like encephal- itis and meningitis, although increased levels of KYNA have been shown in patients with HSV encephalitis, Lyme borreliosis, tick-borne encephalitis, and bacterial meningitis [4, 31–33]. Moreover, altered tryptophan me- tabolism has been linked to disease severity in tubercu- lous meningitis [34]. It is known that the inflammation activates the kynurenine pathway (KP) resulting in the depletion of tryptophan (TRP) and formation of metabolites with potential neurotoxic (e.g., quinolinic acid [QA], 3-hydroxykynurenine [3-HK]) and neuroprotective (e.g., kynurenic acid [KYNA]) effects (Fig. 1) [14, 15]. The activation of the KP seems to also have immune modulating effects, resulting in inhibition of TH1 cells, activation of regulatory T cells (Tregs) and inhib- ition of natural killer (NK) cells [16–19]. In the CNS, the rate-limiting enzyme for TRP catabolism is The aim of the present study was to elucidate the in- flammatory network and KP metabolites in ASM and ABM, characterized by meningeal inflammation, and in encephalitis, characterized by brain parenchymal inflam- mation. Parallel samples of serum and CSF were exam- ined in patient groups and control patients, i.e., patients with similar symptoms but without CSF pleocytosis or other signs of CNS infection. Fig. 1 Schematic presentation of the KP pathway. IDO is the main enzyme responsible for the TRP catabolism in CNS. KYN is further degraded into the neuroprotective NMDAr antagonist KYNA by KAT, or by KMO and KYNU into the neurotoxic metabolites of 3-HK and QA. QA is an agonist of the NMDA receptor. Abbreviations: AA, anthranilic acid; 3-HAA, 3-hydroxyanthranilic acid; HAO, 3-hydroxyanthranilic acid oxidase; 3-HK, 3-hydroxykynurenine; IDO, indoleamine-2,3-dioxygenase; KAT, kynurenine aminotransferase; KMO, kynurenine 3-monooxygenase; KYN, kynurenine; KYNA, kynurenic acid; KYNU, kynureninase; QA, quinolinic acid; PIC, picolinic acid; TRP, tryptophan; XA, xanthurenic acid. Bold box indicates neuroprotective metabolite, dashed boxes indicate neurotoxic metabolites in the KP pathway Fig. 1 Schematic presentation of the KP pathway. IDO is the main enzyme responsible for the TRP catabolism in CNS. Study participants and study design Study participants and study design y y This cross-sectional study was performed at the Oslo Uni- versity Hospital (OUS), Ullevaal, a regional hospital for 2.7 million people. Patients were eligible for inclusion if they had (1) onset of symptoms of CNS infection within less than 7 days and (2) clinical indication for a diagnostic lumbar puncture (LP). Patients were included between January 2014 and December 2015. CSF leukocyte counts ≥5 × 106 /L was found in 68 patients. Of these, 23 did not fulfill the case definition of encephalitis of viral, auto- immune or unknown cause, aseptic meningitis (ASM), or bacterial meningitis (ABM) (Additional file 1: Table S1). In four patients, the time from CSF sampling until centri- fugation was > 10 h, rendering a total of 41 patients with CNS infection (Additional file 2: Figure S1). The control group consisted of age- and gender-matched patients with similar symptoms, but without signs of CNS inflamma- tion, i.e., no pleocytosis and no microbiological agent detected in their CSF (n = 42). In the control group, pa- tients with delirium, chronic or acute psychiatric disease, Parkinson’s disease, Huntington’s disease, CNS malig- nancy, dementia, epilepsy or seizures, cerebral vascular disease, transient global amnesia, and septicemia were not included. For detailed case definitions, see Additional file 1: Table S1. Flowchart of the study population and the over- view of analyses performed for the various groups are shown in Additional file 2: Figure S1. Sampling of CSF and serum CSF samples were collected in endotoxin-free polypro- pylene tubes and stored at 4 °C until centrifugation at 2000×g for 10 min. Serum samples were collected in endotoxin-free tubes without any additives, allowed to clot at room temperature, and centrifuged at 3000×g for 10 min. Supernatants from CSF and serum were centri- fuged within 10 h after collection and immediately fro- zen in triplicates of approximately 700 μL each at −80 ° C. All analyses were performed on previously unthawed samples. For six patients, no serum was available. Mass spectrometry analyses of tryptophan metabolites in CSF and serum All patients, or their next of kin when the patient was not able to consent, gave written informed consent to participate in the study. The study was approved by The Regional Committees for Medical and Health Research Ethics (REC South East, reference number 2011/2578) and the ethical council of the hospital. Tryptophan and kynurenine metabolites were measured only in the patients with encephalitis, ASM with con- firmed viral etiology (viral meningitis, VM), ABM, and controls (Additional file 2: Figure S1). Concentrations of TRP, kynurenine [KYN], anthranilic acid [AA], KYNA, 3-HK, 3-hydroxyanthranilic acid [3-HAA], xanthurenic acid [XA], QA, picolinic acid [PIC], and neopterin were analyzed in CSF and serum by liquid chromatography-tandem mass spectrometry (LC-MS/ MS) by Bevital AS [35, 36]. For TRP, nine patients had levels in CSF below the lower limit of detection (LOD). As this may represent a finding rather than a limitation by the analysis, these were set equal to the LOD (0.4 μM) in the statistical analyses and in the calculation of the KYN/TRP ratio (KYN (nmol)/TRP (μmol)) as an index of IDO activity. XA was not detected in CSF for 46 of the 50 patients and was not included in the analysis. Background KYN is further degraded into the neuroprotective NMDAr antagonist KYNA by KAT, or by KMO and KYNU into the neurotoxic metabolites of 3-HK and QA. QA is an agonist of the NMDA receptor. Abbreviations: AA, anthranilic acid; 3-HAA, 3-hydroxyanthranilic acid; HAO, 3-hydroxyanthranilic acid oxidase; 3-HK, 3-hydroxykynurenine; IDO, indoleamine-2,3-dioxygenase; KAT, kynurenine aminotransferase; KMO, kynurenine 3-monooxygenase; KYN, kynurenine; KYNA, kynurenic acid; KYNU, kynureninase; QA, quinolinic acid; PIC, picolinic acid; TRP, tryptophan; XA, xanthurenic acid. Bold box indicates neuroprotective metabolite, dashed boxes indicate neurotoxic metabolites in the KP pathway Fig. 1 Schematic presentation of the KP pathway. IDO is the main enzyme responsible for the TRP catabolism in CNS. KYN is further degraded into the neuroprotective NMDAr antagonist KYNA by KAT, or by KMO and KYNU into the neurotoxic metabolites of 3-HK and QA. QA is an agonist of the NMDA receptor. Abbreviations: AA, anthranilic acid; 3-HAA, 3-hydroxyanthranilic acid; HAO, 3-hydroxyanthranilic acid oxidase; 3-HK, 3-hydroxykynurenine; IDO, indoleamine-2,3-dioxygenase; KAT, kynurenine aminotransferase; KMO, kynurenine 3-monooxygenase; KYN, kynurenine; KYNA, kynurenic acid; KYNU, kynureninase; QA, quinolinic acid; PIC, picolinic acid; TRP, tryptophan; XA, xanthurenic acid. Bold box indicates neuroprotective metabolite, dashed boxes indicate neurotoxic metabolites in the KP pathway Page 3 of 14 Quist-Paulsen et al. Journal of Neuroinflammation (2018) 15:327 Microbiological diagnostics For all included patients, CSF leukocyte counts (CSF WBC), CSF protein, and CSF glucose were measured. Bacterial culture of CSF and analyses for identification of causative agent were performed in all individuals by a panel of specific PCR for common neurotropic virus and bacteria. Methods Multiplex analyses of soluble markers in CSF and serum A multiplex cytokine assay (Bio-Plex Pro Human Cytokine 27-plex Panel; Bio-Rad laboratories Inc. Hercules, CA) was used to measure the concentrations of 27 different cy- tokines: tumor necrosis factor (TNF), IFN-γ, interleukin (IL)-1β, IL-1 receptor antagonist (IL-1Ra), IL-2, IL-4, IL-5, IL-6, IL-7, IL-8/CXCL8, IL-9, IL-10, IL-12(p70), IL-13, IL-15, IL-17A, monocyte chemoattractant protein (MCP)-1/CCL2, IFN-inducible protein-10 (IP-10)/ CXCL10, eotaxin/CCL11, macrophage inflammatory protein-1α and -1β (MIP-1α/CCL3, MIP-1β/CCL4), regu- lated on activation, normal T cell expressed and secreted (RANTES/CCL5), granulocyte-colony stimulating factor (G-CSF), granulocyte-macrophage CSF (GM-CSF), basic fibroblast growth factor (FGF), platelet-derived growth factor (PDGF), and vascular endothelial growth factor (VEGF). The assay was performed using the instructions of the manufacturer. CSF samples were tested undiluted. Only cytokines with less than 20% missing values were included in further analyses. Undetectable levels were assigned the lowest detectable level (LDL) measured in the cohort for the respective marker. In the CSF samples, IFN-γ, IL-5, PDGF, Basic FGF, and RANTES, and for serum, IL-2, IL-5, IL-7, IL-15, G-CSF, GM-CSF, and FGF were excluded from further analyses based on the criteria stated above. Multiplex analyses of soluble markers in CSF and serum Statistical methods Continuous data are presented as median (IQR, inter- quartile range). Due to lack of normal distribution, ana- lysis of variance (ANOVA) with the Kruskal-Wallis test for multiple groups was used. If the Kruskal-Wallis test revealed significant differences, the Mann-Whitney U Page 4 of 14 Page 4 of 14 Quist-Paulsen et al. Journal of Neuroinflammation (2018) 15:327 Quist-Paulsen et al. Journal of Neuroinflammation are presented in Table 1. There were no significant differences in gender or age between the patient groups. In the CNS infection group, four patients re- ported a history of depression, but only two of these received antidepressant drugs. The etiology of enceph- alitis was known for four patients (40%), three viral (adenovirus, HSV1, varicella-zoster virus [VZV]) and one N-methyl-D-aspartate receptor [NMDAr] enceph- alitis. Of the 25 patients with ASM, eight were diag- nosed with enterovirus in CSF, six patients suffered from HSV2 meningitis, one patient seroconverted and had positive IgM in CSF for Toscana virus, and for one patient, intrathecal antibody production of IgG against Borrelia burgdorferi was detected. For patients with ABM, Streptococcus pneumoniae (n = 2), Staphylococcus aureus (n = 2), Neisseria meningitidis (n = 1), and Haemophilus influenzae (n = 1) were de- tected in CSF by growth and/or PCR, and for all test was used to compare pairs of groups. To limit type II statistical errors, no correction for multiple compari- sons was made in this explorative study. P values < 0.05 were considered statistically significant. Categorical vari- ables are expressed as counts (percentages) and analyzed by Pearson’s chi-square test. Correlations were analyzed using Spearman’s rank correlation coefficient. All data analyses were performed in SPSS version 24 (IBM Corp. Armonk, NY, USA) and graphs generated by GraphPad Prism 7 (GraphPad, San Diego, USA). test was used to compare pairs of groups. To limit type II statistical errors, no correction for multiple compari- sons was made in this explorative study. P values < 0.05 were considered statistically significant. Categorical vari- ables are expressed as counts (percentages) and analyzed by Pearson’s chi-square test. Correlations were analyzed using Spearman’s rank correlation coefficient. All data analyses were performed in SPSS version 24 (IBM Corp. Armonk, NY, USA) and graphs generated by GraphPad Prism 7 (GraphPad, San Diego, USA). Study participant characteristics Ten patients had encephalitis of viral, autoimmune, or unknown etiology according to the case definition (Additional file 1: Table S1), 25 patients were diag- nosed with ASM, six patients with ABM, and 42 were control patients. Characteristics of the study group Table 1 Patient characteristics and clinical presentation P t E h liti ( 10) A Table 1 Patient characteristics and clinical presentation Parameter Encephalitis (n = 10) ASM (n = 25) ABM (n = 6) Controls (n = 42) p valuea Gender, males (%) 4 (40) 10 (40) 4 (67) 13 (31) 0.385 Age, years 43.5 (30, 72) 35 (28, 48) 52 (41, 68) 31 (22, 41) 0.054 Hospital stay, days 19 (11, 42)b 3 (1.5, 5.5)c 19 (14, 33)b, d 2.0 (1.0, 4.0) < 0.001 Comorbidity (%) Immunodeficiencye 2 (30)b 1 (4)c 1 (17)b – 0.029 Psychiatric disorder 2 (20)b 2 (8) – – 0.046 Etiology known (%)f 4/10 (40) 16/25 (64) 6/6 (100) – < 0.001 Headache (%) 7/10 (70)b 24/25 (96)c 3/4 (75) 39/41 (95) 0.037 Neck stiffness (%)g 2/10 (20) 12/25 (48) 3/5 (60) 12/42 (29) 0.175 Objective fever (%)h 7/10 (70) 17/25 (68) 5/6 (83) 20/42 (48) 0.168 Focal neurology (%) 5/9 (56)b 2/19 (10)c 1/6 (17) 1/24 (4) 0.003 GCS ≤14 (%) 10/10 (100)b 2/25 (8)c 5/6 (83)b,d 6/42 (14) < 0.001 CSF WBC (× 106/L) 25 (9.5, 92)b 179 (26, 271)b, c 212 (91, 1434)b, c 1.0 (1.0, 2.0) < 0.001 CSF protein (g/L) 0.57 (0.4, 0.9)b 0.59 (0.4, 0.8)b 2.2 (0.8, 5.9)b, c, d 0.26 (0.2, 0.3) < 0.001 CSF glucose (mmol/L) 3.6 (3.3, 4.4) 3.5 (3.0, 3.7) 3.6 (0.4, 6.3) 3.5 (3.2, 3.7) 0.341 Glucose ratio 0.6 (0.5, 0.7) 0.6 (0.5, 0.6)b 0.4 (0.1, 0.6)b 0.6 (0.6, 0.7) 0.009 Albumin ratio 8.4 (6.5,14)b 9.4 (6.1, 13)b 79 (36, 137)b, c, d 4.0 (2.6, 5.0) < 0.001 Blood WBC (× 109/L) 8.6 (7.4, 11) 8.2 (6.1, 11) 14 (8.5, 18) 10 (6.7, 12) 0.173 CRP, serum (mg/L) 6 (0.9, 86) 3.5 (1.3, 12)b 169 (96, 446)b, c, d 16 (2.1, 77) 0.001 CSF WBC white blood cell count in CSF, glucose ratio CSF glucose/serum glucose, albumin ratio CSF albumin/serum albumin. Significant p-values are marked in bold. CSF WBC white blood cell count in CSF, glucose ratio CSF glucose/serum glucose, albumin ratio CSF albumin/serum albumin. Significa in bold. Cytokine profiles in CSF and serum Cytokines analyzed in parallel in CSF and serum showed distinct patterns for the different patient groups. Overall, the highest levels of CSF cytokines were found in patients with ABM, including the prototypical inflammatory cytokines TNF, IL-1β and IL-6, inflammatory chemokines (e.g., IL-8, MCP-1, MIP-1α and MIP-1β), cytokines with potent effect on T cell function (e.g., IP-10, IL-7, IL-9 and IL-15), and growth factors (e.g., VEGF and G-CSF) (Fig. 2, Additional file 3: Table S2). The typical cytokine pat- tern in CSF was an increase from disease controls without CNS infection to patients with encephalitis and ASM with the highest levels in those with ABM. In contrast to this pattern, the CXC chemokine IP-10 showed the highest median level in the ASM group (Fig. 2). Although lower than in patients with ABM and ASM, patients with encephalitis had higher levels for most cytokines in CSF compared to the control group, with no significant difference in the levels between encephalitis cases with or without known eti- ology (data not shown). In order to relate the cytokine profile to tryptophan metabolism (Fig. 1), metabolites of the KP in the CSF were examined in patients with encephalitis (n = 10), those in the ASM group with verified viral cause (VM, n = 12), ABM (n = 6), and controls (n = 22). The median levels of most KP metabolites were higher in patients with encephalitis, VM, and ABM compared with the controls, with the highest median levels observed in the ABM group (Fig. 4, Additional file 5: Table S4). The only exception was TRP, which was lowest in the VM group (Fig. 4). g The KYN/TRP ratio was higher in all CNS infections compared to the controls (p = 0.009), indicating an in- creased conversion of TRP (Fig. 4). Moreover, to evalu- ate the relationship between putatively neuroprotective (KYNA) and neurotoxic (3-HK and QA) KP metabolites, we calculated the KYNA/(3-HK + QA) ratio showing de- creased levels for patients with encephalitis indicating a net neurotoxic effect of TRP metabolites in these pa- tients (Fig. 4). Most KP metabolites were present at higher concentra- tions in serum (Additional file 5: Table S4, Additional file 6: Figure S2) than in CSF (Fig. 2), but with less significant differences between the groups. Study participant characteristics Data shown are median (IQR) or numbers/n (%) Data shown are median (IQR) or numbers/n (%) ap values for one way analysis of variance (Kruskal-Wallis) b p y y bp < 0.05 for analysis with Mann-Whitney U test (MWU) in comparison with the control group pp g g fCausing agents for encephalitis were 3 viral (adenovirus, HSV1, VZV) and 1 NMDAr encephalitis. ABM; Streptococcus pneumonia (n = 2), Staphylococcus aureus (n = 2), Neisseria meningitides (n = 1), and Haemophilus influenzae (n = 1). ASM; HSV2 (n = 6), enterovirus (n = 8), Toscana virus (n = 1), and 1 patient had intrathecal antibody production of IgG against Borrelia burgdorferi gNeck stiffness was assessed by a physician before LP fCausing agents for encephalitis were 3 viral (adenovirus, HSV1, VZV) and 1 NMDAr encephalitis. ABM; Streptococcus pneumonia (n = 2), Staphylococcus aureus (n = 2), Neisseria meningitides (n = 1), and Haemophilus influenzae (n = 1). ASM; HSV2 (n = 6), enterovirus (n = 8), Toscana virus (n = 1), and 1 patient had intrathecal antibody production of IgG against Borrelia burgdorferi gNeck stiffness was assessed by a physician before LP y p y hFever was defined as either ≥38 °C upon admission or within 24 h after admission, or measured to ≥38 °C by the patient prior to admission Page 5 of 14 Quist-Paulsen et al. Journal of Neuroinflammation (2018) 15:327 Page 5 of 14 Quist-Paulsen et al. Journal of Neuroinflammation the examined mediators (Additional file 4: Table S3). Finally, except for MCP-1, all cytokines in CSF corre- lated positively with the CSF leukocyte counts, and all, except for MCP-1 and IP-10, correlated with CSF/ serum albumin ratio. Collectively, these data under- score the limited information obtained from a serum sample in contrast to that obtained from CSF when examining the inflammatory network in infectious CNS diseases. these patients, the causative bacteria was also de- tected in blood culture. Patients with encephalitis pre- sented with less headache and more focal neurology compared to the other groups, and impairment of consciousness was observed in significantly fewer patients with ASM and in the control group. Import- antly, the majority of the control group had fever, headache, and neck stiffness, similar to most of the patients with CNS infection. Cytokine profiles in CSF and serum For the total group with CNS infection (encephalitis, VM and ABM), CSF levels correlated with serum levels for 3-HK (Rho 0.6, p < 0.001), QA (0.4, p = 0.04), and PIC (0.7, p < 0.001), for all other KP metabolites, there was no significant correlation be- tween CSF and serum levels. Furthermore, all CSF/serum ratios for KP metabolites, except for TRP, KYNA, and PIC ratio, were positively correlated with CSF WBC count, while only the ratio of PIC was correlated with CSF albu- min/serum ratio (Additional file 7: Table S5). The encephalitis group and the control group gener- ally showed lower cytokine levels in CSF than in serum, with exceptions for IL-6 and IL-8 in the encephalitis group and MCP-1 and IP-10 for both groups (Figs. 2 and 3, Additional file 3: Table S2). In contrast, the groups with meningeal involvement and in particular the ABM group displayed markedly higher CSF levels than serum levels for most cytokines. For some of the cyto- kines, the CSF levels were more than tenfold higher than in serum (e.g., IL-6, IL-8, IP-10). In general, serum levels did not display the striking and significant differences between the patient groups and controls as seen in CSF (Fig. 3). In fact, although all patients in the ABM group had a positive blood culture, only TNF, IL-6, IL-8, IL-1Ra, and MIP-1α demonstrated higher serum levels than in the control group. Markers of IFN-γ activation correlate with activation of the KP When we analyzed the CNS infections all together (encephalitis, ASM, and ABM), a significant correl- ation between serum and CSF levels were found only for TNF (Rho 0.4, p = 0.03), IL-1Ra (Rho 0.4, p = 0.03), IL-6 (Rho 0.5, p = 0.004), and MCP-1 (Rho 0.3, p = 0.04) suggesting intrathecal production of most of IFN-γ is known to have a major influence on the KP [37, 38]. In the present study, IFN-γ levels in CSF were below detection level in 58% of the cases. How- ever, neopterin, being a reliable and stable marker of IFN-γ activity [39], was markedly elevated in CSF in all groups of patients with neuroinflammation Quist-Paulsen et al. Journal of Neuroinflammation (2018) 15:327 Page 6 of 14 Fig. 2 (See legend on next page.) Fig. 2 (See legend on next page.) Fig. 2 (See legend on next page.) (See figure on previous page.) Fig. 2 Cytokines in CSF in patients with encephalitis (Enc, n = 10), aseptic meningitis (ASM, n = 25), bacterial meningitis (ABM, n = 6) in comparison with controls (Ctr, n = 42). Data shown are medians with IQR and all were significant by the Kruskal-Wallis test. Comparisons of two groups were analyzed by using Mann-Whitney U test. Asterisks above patient groups indicate significant difference vs controls, asterisks above horizontal lines indicate significant differences between individual groups (Mann-Whitney U test): p < 0.05, p < 0.01, and *p < 0.001. Values below the detection limit were set to the lowest detectable level for that analyte Quist-Paulsen et al. Journal of Neuroinflammation (2018) 15:327 Page 7 of 14 Quist-Paulsen et al. Journal of Neuroinflammation (2018) 15:327 Page 7 of 14 (See figure on previous page.) Fig. 2 Cytokines in CSF in patients with encephalitis (Enc, n = 10), aseptic meningitis (ASM, n = 25), bacterial meningitis (ABM, n = 6) in comparison with controls (Ctr, n = 42). Data shown are medians with IQR and all were significant by the Kruskal-Wallis test. Comparisons of two groups were analyzed by using Mann-Whitney U test. Asterisks above patient groups indicate significant difference vs controls, asterisks above horizontal lines indicate significant differences between individual groups (Mann-Whitney U test): p < 0.05, p < 0.01, and p < 0.001. Values below the detection limit were set to the lowest detectable level for that analyte compared with controls (Fig. 4). Markers of IFN-γ activation correlate with activation of the KP Notably, when all CNS infections were analyzed together, CSF level of neopterin was strongly positively correlated with IDO (measured as the KYN/TRP ratio) and inversely cor- related with the KYNA/(3-HK + QA) ratio (Fig. 5). This suggests an association between high IFN-γ activity and net neurotoxic effects of KP metabolites. A positive correlation was also seen between IDO and IP-10, another IFN- related cytokine, but no significant correlation was found between IP-10 and KYNA/(3-HK + QA) ratio (Fig. 5). catabolism of TRP. Interestingly, patients with encephalitis and viral meningitis had an unfavorable balance between neuroprotective and neurotoxic TRP metabolites. These changes in KP metabolites were associated with CSF levels of neopterin, and for the KYN/TRP ratio also with IP-10, suggesting a link between IFN-γ and altered KP metabo- lites in these patients. The influx of inflammatory cells and the resulting dysregulation of cytokine networks may be detrimen- tal during CNS infection and is thought to contribute to neurological complications [12, 42–44]. CNS inflammation is a complex process, depending on anatomical site (parenchyma vs meninges), cell type being involved (e.g., infiltrating leukocytes in acute in- fection vs brain-derived cells), and causing agent (e.g., viral vs bacterial). In this study, we describe findings from patients with acute inflammation in the CNS, mainly of infectious cause. The levels of cytokines in CSF were generally higher for all infectious groups compared to control, and strongly correlated with each other, demonstrating a general inflammatory re- sponse in patients with acute CNS infections. The pattern of increasing levels of inflammatory markers from encephalitis to aseptic meningitis and finally to bacterial meningitis indicates increased inflammation in all subgroup of patients, but with a more modest CSF inflammation in patients with encephalitis. Not surprisingly, the highest levels of most of the media- tors were found in patients with bacterial meningitis and there was a CSF/serum ratio > 1 of most media- tors in patients with ABM, despite bacteremia. Fur- thermore, most cytokines were correlated with CSF WBC count, but not with corresponding serum levels, which implies the inflammatory response in the CNS to be independent of a peripheral immune response with no or minimal influx or efflux between the two compartments. These results indicate a predominately intrathecal production of these mediators in CNS in- fections. Markers of IFN-γ activation correlate with activation of the KP IL-6, IL-8, and TNF in CSF have in several studies been found to be elevated in meningitis and have even been proposed as biomarkers in CSF for bacterial meningitis [6, 45]. We found a general in- crease for most of the inflammatory mediators in both ASM and ABM, and our study does not support the use of one particular cytokine as a diagnostic marker to distinguish ABM from other CNS condi- tions, including ASM. Association of cytokines and metabolites in the KP and clinical variables Abnormal regulation of the KP metabolites has been re- lated to depression [40, 41]. However, excluding the two patients in the encephalitis group who reported existing or recent depression from the correlation analyses did not influence the findings (data not shown). Although the Glasgow Coma Scale (GCS) is a crude tool, it reflects the severity of critically ill patients. Among patients with aseptic meningitis, all but two had a GCS at 15; thus, only patients with encephalitis and bacterial meningitis were included in correlation analyses with GCS. We did not find any significant correlations between cytokine levels or KP metabolites in CSF and GCS in these patients (data not shown). However, the number of pa- tients was too low to make any firm conclusion. Discussion 3 Cytokines in serum in patients with encephalitis (Enc, n = 10), aseptic meningitis (ASM, n = 20), bacterial meningitis (ABM, n = 6) in comparison with controls (Ctr, n = 41). Data shown are medians with IQR. Asterisks above patient groups indicate significant difference vs controls, asterisks above horizontal lines indicate significant differences between individual groups (Mann-Whitney U test): p < 0.05, p < 0.01, and p < 0.001. Values below the detection limit were set to the lowest detectable level for that analyte detectable TRP levels. The strong correlation of KYN/ TRP ratio with neopterin and IP-10 indicates that this IDO activity is driven by IFN-γ. IDO has been found to exhibit an immunosuppressive effect by upregulation of Tregs and downregulation of Th17 cells, which could be relevant in CNS infections, especially in ASM where T cells are of particular importance. In contrast to the as- sociation with IDO activity, neopterin, but not IP-10, was associated with a “neurotoxic” KYNA/(3-HK + QA) ratio in patients with CNS infection. The lack of correl- ation of IP-10 to KYNA/(3-HK + QA) could be related to low statistical power. However, whereas IP-10 is in- duced by IFN- γ in several cell types including mono- cytes, stromal cells, and endothelial cells [59], neopterin is almost selectively produced in monocytes/macro- phages, and through its relation to tetrahydrobiopterin, neopterin may be more closely related to tryptophan metabolism than IP-10 [60, 61]. In contrast to most of the inflammatory markers, IP-10 levels were comparable in patients with ASM and ABM. IP-10 is secreted by several cell types in response to IFN-γ, and experimental studies have shown that IP-10 is highly induced in the CNS (e.g., West Nile in- fections [46], HIV encephalitis [47], HSV encephalitis [48], enteroviral encephalitis [3, 49], Japanese encephal- itis [50], and tick-borne encephalitis [TBE] [51]). In- creased levels have also been reported in patients suffering from TBE [52], neuroborreliosis [53], entero- virus infection [3], HSV meningitis, and HSV encephal- itis [9]. IP-10 is a chemoattractant for monocytes and macrophages and promotes T cell recruitment, especially of CD8+ T cells [46, 54, 55]. Although IP-10 has been linked to viral clearance in the CNS [48, 55], IP-10 has also been described to promote apoptosis of neurons and excessive production has been associated with more severe outcome [47, 50, 51]. Discussion In the present study, we found an association between IP-10 and the KYN/TRP ratio, indicating increased IDO activity and catabolism of TRP with potential harmful effects on CNS. Thus, our present data may further support a potential role for this chemokine in CNS infections, particularly in ASM. This should be further investigated. Finally, when looking at the CSF/serum ratios of KP metabolites, we found very narrow ranges in the control group, suggesting strict control of KP in healthy subjects. Studies of the inflammatory profile in human CNS infections including both meningeal and parenchymal infections are relatively scarce [62–65], and comparisons between different studies are hampered by the diversity of causing agents and divergent inclusion criteria. Never- theless, knowledge of immunological mechanism is piv- otal in order to develop better diagnostic and potentially therapeutic tools for these patients. To our knowledge, the measurement of a large panel of metabolites in the KP in both serum and CSF, with parallel analyses of a wide range of cytokines and related mediators, have not been reported for patients with these conditions, espe- cially not for encephalitis. Moreover, the correlation in present study of IP-10 and neopterin with the KP metab- olites and the decreased KYNA/(3-HK + QA) ratio in encephalitis are interesting findings that, as far as we are aware of, have not been reported in these clinical rele- vant CNS infections. However, the present study also has some limitations. The subgroups of patients, and in particular patients with ABM and encephalitis, were small and too small for sub-analyses on causing agents. Moreover, the etiology was unknown for 60% of patients with encephalitis with few patients with confirmed viral cause. Besides, due to lack of reliable measures of sever- ity and outcome, together with the relatively low number of patients with CNS infection, we cannot make any Neurologic dysfunction is common in patients with encephalitis and bacterial meningitis, and dysregulation of the KP has been shown in syndromes and disorders with certain overlap in symptoms [21, 30, 32, 56, 57]. Herein, we compared these metabolites in patients with a stringent definition of etiology; encephalitis, confirmed viral meningitis, and bacterial meningitis. Recent litera- ture has shown that in HSV encephalitis, a low level of the neuroprotective metabolite KYNA was associated with more severe outcome [31]. In TB meningitis, low TRP was associated with a better survival rate [34]. Discussion In the present study, we demonstrate a gradual increase in levels of a wide range of cytokines, including chemokines, interleukins, interferons, and growth factors from enceph- alitis to ASM and finally to ABM as compared with con- trols, reflecting the level of CSF inflammation seen in patients with these CNS infections. Moreover, in ASM and ABM, the levels were much higher in CSF than in serum for most of the mediators, even though the patients with ABM had positive blood cultures for the actual pathogen. Notably, in contrast to most of the mediators, IP-10 levels in CSF had the highest median value in the ASM group, indicating a potential role for this chemokine in aseptic meningitis. Finally, these changes in inflamma- tory mediators were associated with marked changes in KP metabolites in CSF. In particular, all groups of CNS in- fections had increased IDO activity assessed by KYN/TRP ratio compared with controls, indicating an increased Quist-Paulsen et al. Journal of Neuroinflammation (2018) 15:327 Page 8 of 14 Quist-Paulsen et al. Journal of Neuroinflammation (2018) 15:327 Page 8 of 14 Fig. 3 (See legend on next page.) Fig. 3 (See legend on next page.) Fig. 3 (See legend on next page.) Quist-Paulsen et al. Journal of Neuroinflammation (2018) 15:327 Page 9 of 14 (See figure on previous page.) Fig. 3 Cytokines in serum in patients with encephalitis (Enc, n = 10), aseptic meningitis (ASM, n = 20), bacterial meningitis (ABM, n = 6) in comparison with controls (Ctr, n = 41). Data shown are medians with IQR. Asterisks above patient groups indicate significant difference vs controls, asterisks above horizontal lines indicate significant differences between individual groups (Mann-Whitney U test): p < 0.05, p < 0.01, and p < 0.001. Values below the detection limit were set to the lowest detectable level for that analyte (See figure on previous page.) Fig. 3 Cytokines in serum in patients with encephalitis (Enc, n = 10), aseptic meningitis (ASM, n = 20), bacterial meningitis (ABM, n = 6) in comparison with controls (Ctr, n = 41). Data shown are medians with IQR. Asterisks above patient groups indicate significant difference vs controls, asterisks above horizontal lines indicate significant differences between individual groups (Mann-Whitney U test): p < 0.05, p < 0.01, and p < 0.001. Values below the detection limit were set to the lowest detectable level for that analyte (See figure on previous page.) Fig. Discussion In our study, the ratio of the neuroprotective metabolite KYNA to the sum of neurotoxic metabolites 3-HK and QA was lower for patients with encephalitis compared to the other groups, which could indicate a stronger ac- tivation of the KMO branch in encephalitis. In fact, studies are ongoing regarding the inhibition of several steps in the KP, including studies on centrally available KMO inhibitors [58]. In the present study, we observed very low levels of TRP for several patients with CNS infection, especially in patients with ASM. We hypothesize that this results from increased IDO activity, as these patients had a sig- nificantly higher level of KYN compared to patients with Quist-Paulsen et al. Journal of Neuroinflammation (2018) 15:327 Page 10 of 14 conclusion regarding the use of these cytokines and me- tabolites as prognostic markers The control group in and no pleocytosis, but many of these patients suffered from systemic infections as reflected by elevated serum Fig. 4 Neopterin, kynurenine metabolites, and ratios in patients with encephalitis (Enc, n = 10), viral meningitis (VM, n = 12), and bacterial meningitis (ABM, n = 6) in comparison with controls (Ctr, n = 22). Data shown are median with IQR, and all were significant in the analysis of variance with the Kruskal-Wallis test. Comparisons of two groups were analyzed by Mann-Whitney U test. Asterisks above patient groups indicate significant difference vs controls, asterisks above horizontal lines indicate significant differences between individual groups (Mann-Whitney U test): p < 0.05, p < 0.01, and p < 0.001. a: TRP levels below the lower level of detection (LOD) for 9 patients with CNS infection were adjusted to this value (0.4 μM) for calculation of the KYN/TRP ratio as the expression of IDO activity (KYN (nmol)/TRP(μmol)) urenine metabolites, and ratios in patients with encephalitis (Enc, n = 10), viral meningitis (VM, n = 12), and bacterial Fig. 4 Neopterin, kynurenine metabolites, and ratios in patients with encephalitis (Enc, n = 10), viral meningitis (VM, n = 12), and bacterial meningitis (ABM, n = 6) in comparison with controls (Ctr, n = 22). Data shown are median with IQR, and all were significant in the analysis of variance with the Kruskal-Wallis test. Comparisons of two groups were analyzed by Mann-Whitney U test. Discussion Asterisks above patient groups indicate significant difference vs controls, asterisks above horizontal lines indicate significant differences between individual groups (Mann-Whitney U test): p < 0.05, p < 0.01, and p < 0.001. a: TRP levels below the lower level of detection (LOD) for 9 patients with CNS infection were adjusted to this value (0.4 μM) for calculation of the KYN/TRP ratio as the expression of IDO activity (KYN (nmol)/TRP(μmol)) Fig. 4 Neopterin, kynurenine metabolites, and ratios in patients with encephalitis (Enc, n = 10), viral meningitis (VM, n = 12), and bacterial meningitis (ABM, n = 6) in comparison with controls (Ctr, n = 22). Data shown are median with IQR, and all were significant in the analysis of variance with the Kruskal-Wallis test. Comparisons of two groups were analyzed by Mann-Whitney U test. Asterisks above patient groups indicate significant difference vs controls, asterisks above horizontal lines indicate significant differences between individual groups (Mann-Whitney U test): p < 0.05, p < 0.01, and *p < 0.001. a: TRP levels below the lower level of detection (LOD) for 9 patients with CNS infection were adjusted to this value (0.4 μM) for calculation of the KYN/TRP ratio as the expression of IDO activity (KYN (nmol)/TRP(μmol)) and no pleocytosis, but many of these patients suffered from systemic infections, as reflected by elevated serum neopterin levels. This may have camouflaged significant conclusion regarding the use of these cytokines and me- tabolites as prognostic markers. The control group in our study consisted of patients with similar presentation Quist-Paulsen et al. Journal of Neuroinflammation (2018) 15:327 Page 11 of 14 A B C D Fig. 5 Correlations of neopterin and IP-10, as markers of IFN-ƴ activity, with activation of the KP. a Neopterin vs KYN/ TRP ratio (IDO) (n = 50; Rho 0.9, p < 0.001). b IP-10 vs KYN/TRP ratio (IDO) (n = 50; Rho 0.8, p < 0.001). c Neopterin vs KYNA/(3-HK + QA) ratio (n = 50; Rho −0.7, p < 0.001). d IP-10 vs KYNA/(3-HK + QA) ratio (n = 50, Rho −0.5, p < 0.001). Data shown are obtained by Spearman’s rank correlation A B B A B C D C D D Fig. 5 Correlations of neopterin and IP-10, as markers of IFN-ƴ activity, with activation of the KP. a Neopterin vs KYN/ TRP ratio (IDO) (n = 50; Rho 0.9, p < 0.001). Conclusions In conclusion, we found a marked increase in a wide range of inflammatory mediators in CSF in aseptic and bacterial meningitis with a more modest increase in encephalitis. The markedly higher levels in CNS than in serum for most of the mediators indicate compartmentalization of immune responses in the CSF. Our data may also suggest that increased IFN-γ activity, as assessed by neopterin and IP-10, may contribute to neurotoxicity through enhanced TRP catabolism. In particular, dysregulation of the KP with signs of an increased formation of neurotoxic QA in encephalitis should be explored further in these conditions. Additional file 6: Figure S2. Serum levels of KP metabolites. (PDF 81 kb) Additional file 6: Figure S2. Serum levels of KP metabolites. (PDF 81 kb) Additional file 7: Table S5. CSF/serum ratios of KP metabolites and the correlation with CSF WBC and albumin ratio. (PDF 194 kb) Additional file 7: Table S5. CSF/serum ratios of KP metabolites and the correlation with CSF WBC and albumin ratio. (PDF 194 kb) Additional files findings in serum profiles. On the other hand, these controls may be more clinically relevant than healthy controls without any disease symptoms like fever and headache. Finally, correlations do not necessarily mean any causal relationship and more mechanistic studies as well as larger studies with samples also taken during follow-up are needed to make firmer conclusions. Additional file 1: Table S1. Case definitions of encephalitis, aseptic and viral meningitis and bacterial meningitis. (PDF 179 kb) Additional file 2: Figure S1. Flowchart of inclusion of patients and overview of various analyses performed in the study population. (PDF 342 kb) Additional file 3: Table S2. Cytokine levels in CSF and serum. (PDF 215 kb) Additional file 4: Table S3. Correlations of cytokines in CSF with serum levels, CSF WBC, albumin ratio and KYN/TRP ratio. (PDF 197 kb) Additional file 5: Table S4. KP metabolites in CSF and serum. (PDF 198 kb) Additional file 6: Figure S2. Serum levels of KP metabolites. (PDF 81 kb) Additional file 7: Table S5. CSF/serum ratios of KP metabolites and the correlation with CSF WBC and albumin ratio. (PDF 194 kb) Additional file 1: Table S1. Case definitions of encephalitis, aseptic and viral meningitis and bacterial meningitis. (PDF 179 kb) Additional file 2: Figure S1. Flowchart of inclusion of patients and overview of various analyses performed in the study population. (PDF 342 kb) Additional file 3: Table S2. Cytokine levels in CSF and serum. (PDF 215 kb) Additional file 4: Table S3. Correlations of cytokines in CSF with serum levels, CSF WBC, albumin ratio and KYN/TRP ratio. (PDF 197 kb) Additional file 5: Table S4. KP metabolites in CSF and serum. (PDF 198 kb) Additional file 6: Figure S2. Serum levels of KP metabolites. (PDF 81 kb) Additional file 7: Table S5. CSF/serum ratios of KP metabolites and the correlation with CSF WBC and albumin ratio. (PDF 194 kb) Additional file 4: Table S3. Correlations of cytokines in CSF with serum levels, CSF WBC, albumin ratio and KYN/TRP ratio. (PDF 197 kb) Additional file 5: Table S4. KP metabolites in CSF and serum. (PDF 198 kb) Discussion b IP-10 vs KYN/TRP ratio (IDO) (n = 50; Rho 0.8, p < 0.001). c Neopterin vs KYNA/(3-HK + QA) ratio (n = 50; Rho −0.7, p < 0.001). d IP-10 vs KYNA/(3-HK + QA) ratio (n = 50, Rho −0.5, p < 0.001). Data shown are obtained by Spearman’s rank correlation 3-HAA: 3-Hydroxyanthranilic acid; 3-HK: 3-Hydroxykynurenine; AA: Anthranilic acid; ABM: Acute bacterial meningitis; Albumin ratio: CSF albumin/serum albumin; ASM: Aseptic meningitis; Glucose ratio: CSF glucose/serum glucose; HAO: 3-Hydroxyanthranilic acid oxidase; HSV1: Herpes simplex 1 virus; IDO: Indoleamine 2,3-dioxygenase; KAT: Kynurenine aminotransferase; KMO: Kynurenine 3-monooxygenase; KP: Kynurenine pathway of tryptophan metabolism; KYN: Kynurenine; KYNA: Kynurenic acid; KYNU: Kynureninase; LDL: The lowest detectable level; LOD: Lower limit of detection; NMDAr: N- Abbreviations 3-HAA: 3-Hydroxyanthranilic acid; 3-HK: 3-Hydroxykynurenine; AA: Anthranilic acid; ABM: Acute bacterial meningitis; Albumin ratio: CSF albumin/serum albumin; ASM: Aseptic meningitis; Glucose ratio: CSF glucose/serum glucose; HAO: 3-Hydroxyanthranilic acid oxidase; HSV1: Herpes simplex 1 virus; IDO: Indoleamine 2,3-dioxygenase; KAT: Kynurenine aminotransferase; KMO: Kynurenine 3-monooxygenase; KP: Kynurenine pathway of tryptophan metabolism; KYN: Kynurenine; KYNA: Kynurenic acid; KYNU: Kynureninase; LDL: The lowest detectable level; LOD: Lower limit of detection; NMDAr: N- Page 12 of 14 (2018) 15:327 Quist-Paulsen et al. Journal of Neuroinflammation (2018) 15:327 Received: 14 May 2018 Accepted: 11 November 2018 methyl-D-aspartate receptor; QA: Quinolinic acid; TRP: Tryptophan; VM: Viral meningitis; VZV: Varicella-zoster virus; WBC: White blood cell count; XA: Xanthurenic acid Received: 14 May 2018 Accepted: 11 November 2018 Received: 14 May 2018 Accepted: 11 November 2018 methyl-D-aspartate receptor; QA: Quinolinic acid; TRP: Tryptophan; VM: Viral meningitis; VZV: Varicella-zoster virus; WBC: White blood cell count; XA: Xanthurenic acid methyl-D-aspartate receptor; QA: Quinolinic acid; TRP: Tryptophan; VM: Viral meningitis; VZV: Varicella-zoster virus; WBC: White blood cell count; XA: Xanthurenic acid Competing interests Competing interests The authors declare that they have no competing interests. The authors declare that they have no competing interests. 14. Schwarcz R, Bruno JP, Muchowski PJ, Wu HQ. Kynurenines in the mammalian brain: when physiology meets pathology. Nat Rev Neurosci. 2012;13:465–77. Funding 5. Ichiyama T, Maeba S, Suenaga N, Saito K, Matsubara T, Furukawa S. Analysis of cytokine levels in cerebrospinal fluid in mumps meningitis: comparison with echovirus type 30 meningitis. Cytokine. 2005;30:243–7. Publisher’s Note 15. Wang Q, Liu D, Song P, Zou MH. Tryptophan-kynurenine pathway is dysregulated in inflammation, and immune activation. Front Biosci (Landmark Ed). 2015;20:1116–43. Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. Availability of data and materials 6. Prasad R, Kapoor R, Srivastava R, Mishra OP, Singh TB. Cerebrospinal fluid TNF-alpha, IL-6, and IL-8 in children with bacterial meningitis. Pediatr Neurol. 2014;50:60–5. 6. Prasad R, Kapoor R, Srivastava R, Mishra OP, Singh TB. Cerebrospinal fluid TNF-alpha, IL-6, and IL-8 in children with bacterial meningitis. Pediatr Neurol. 2014;50:60–5. The dataset used during the current study is stored in a secured research data server at Oslo University Hospital. The datasets used are available from the corresponding author on reasonable request. 7. Grandgirard D, Gaumann R, Coulibaly B, Dangy JP, Sie A, Junghanss T, Schudel H, Pluschke G, Leib SL. The causative pathogen determines the inflammatory profile in cerebrospinal fluid and outcome in patients with bacterial meningitis. Mediat Inflamm. 2013;2013:312476. 7. Grandgirard D, Gaumann R, Coulibaly B, Dangy JP, Sie A, Junghanss T, Schudel H, Pluschke G, Leib SL. The causative pathogen determines the inflammatory profile in cerebrospinal fluid and outcome in patients with bacterial meningitis. Mediat Inflamm. 2013;2013:312476. Authors’ contributions VO, EQP, PA, BS, OD, AMBK, and AMDR contributed to the study design. EQP, VO, and OD contributed to the inclusion and evaluation of patients. TEM, ØM, and PMU contributed to the laboratory analyses. EQP, PA, TU, ØM, PMU, TEM, BS, and AMDR contributed to the data analysis and interpretation. EQP, TU, and TEM contributed to the statistical analyses. EQP, PA, and AMDR contributed to the writing of the manuscript. All authors contributed to the revision of the manuscript and approved the final version. 8. Shapiro S, Miller A, Lahat N, Sobel E, Lerner A. Expression of matrix metalloproteinases, sICAM-1 and IL-8 in CSF from children with meningitis. J Neurol Sci. 2003;206:43–8. 8. Shapiro S, Miller A, Lahat N, Sobel E, Lerner A. Expression of matrix metalloproteinases, sICAM-1 and IL-8 in CSF from children with meningitis. J Neurol Sci. 2003;206:43–8. 9. Lind L, Studahl M, Persson Berg L, Eriksson K. CXCL11 production in cerebrospinal fluid distinguishes herpes simplex meningitis from herpes simplex encephalitis. J Neuroinflammation. 2017;14:134. 9. Lind L, Studahl M, Persson Berg L, Eriksson K. CXCL11 production in cerebrospinal fluid distinguishes herpes simplex meningitis from herpes simplex encephalitis. J Neuroinflammation. 2017;14:134. 10. Asaoka K, Shoji H, Nishizaka S, Ayabe M, Abe T, Ohori N, Ichiyama T, Eizuru Y. Non-herpetic acute limbic encephalitis: cerebrospinal fluid cytokines and magnetic resonance imaging findings. Intern Med. 2004;43:42–8. Acknowledgements 1. Asano T, Ichiki K, Koizumi S, Kaizu K, Hatori T, Fujino O, Mashiko K, Sakamoto Y, Miyasho T, Fukunaga Y. Enhanced expression of cytokines/chemokines in cerebrospinal fluids in mumps meningitis in children. Pediatr Int. 2011;53: 143–6. 1. Asano T, Ichiki K, Koizumi S, Kaizu K, Hatori T, Fujino O, Mashiko K, Sakamoto Y, Miyasho T, Fukunaga Y. Enhanced expression of cytokines/chemokines in cerebrospinal fluids in mumps meningitis in children. Pediatr Int. 2011;53: 143–6. The Department of Infectious Diseases at Ullevaal Hospital represented by head of department professor Dag Kvale (MD, PhD) provided salary and office facilities to EQP during the study period. We would also like to thank laboratory staff at Bevital, Bergen Norway and the Research Laboratory, Nordland Hospital, Bodø, Norway for excellent service. The authors are especially grateful to nurses and physicians at the Emergency Room, the Medical Intensive Care Unit, the Department of Internal Medicine and the Department of Neurology at Oslo University Hospital, Ullevaal for their enthusiastic and friendly help in including patients during the study period. Finally, we thank the patients and next of kin who gave consent to participate in this study. 2. Asano T, Ichiki K, Koizumi S, Kaizu K, Hatori T, Fujino O, Mashiko K, Sakamoto Y, Miyasho T, Fukunaga Y. IL-17 is elevated in cerebrospinal fluids in bacterial meningitis in children. Cytokine. 2010;51:101–6. 3. Cavcic A, Tesovic G, Gorenec L, Grgic I, Benic B, Lepej SZ. Concentration gradient of CXCL10 and CXCL11 between the cerebrospinal fluid and plasma in children with enteroviral aseptic meningitis. Eur J Paediatr Neurol. 2011;15:502–7. 4. Coutinho LG, Christen S, Bellac CL, Fontes FL, Souza FR, Grandgirard D, Leib SL, Agnez-Lima LF. The kynurenine pathway is involved in bacterial meningitis. J Neuroinflammation. 2014;11:169. 4. Coutinho LG, Christen S, Bellac CL, Fontes FL, Souza FR, Grandgirard D, Leib SL, Agnez-Lima LF. The kynurenine pathway is involved in bacterial meningitis. J Neuroinflammation. 2014;11:169. Ethics approval and consent to participate The Regional Committees for Medical and Health Research Ethics (REC South East, reference number 2011/2578) and the ethical council of the hospital approved the study protocol. Written informed consent was obtained from all individual patients or their next of kin included in the study. 11. Bociaga-Jasik M, Ciesla A, Kalinowska-Nowak A, Skwara P, Garlicki A, Mach T. Role of IL-6 and neopterin in the pathogenesis of herpetic encephalitis. Pharmacol Rep. 2011;63:1203–9. 11. Bociaga-Jasik M, Ciesla A, Kalinowska-Nowak A, Skwara P, Garlicki A, Mach T. Role of IL-6 and neopterin in the pathogenesis of herpetic encephalitis. Pharmacol Rep. 2011;63:1203–9. 12. Kamei S, Taira N, Ishihara M, Sekizawa T, Morita A, Miki K, Shiota H, Kanno A, Suzuki Y, Mizutani T, et al. Prognostic value of cerebrospinal fluid cytokine changes in herpes simplex virus encephalitis. Cytokine. 2009;46:187–93. 12. Kamei S, Taira N, Ishihara M, Sekizawa T, Morita A, Miki K, Shiota H, Kanno A, Suzuki Y, Mizutani T, et al. Prognostic value of cerebrospinal fluid cytokine changes in herpes simplex virus encephalitis. Cytokine. 2009;46:187–93. Consent for publication N li bl 13. Michael BD, Griffiths MJ, Granerod J, Brown D, Davies NW, Borrow R, Solomon T. Characteristic cytokine and chemokine profiles in encephalitis of infectious, immune-mediated, and unknown aetiology. PLoS One. 2016;11: e0146288. Author details 1D f Erhardt S, Blennow K, Nordin C, Skogh E, Lindstrom LH, Engberg G. Kynurenic acid levels are elevated in the cerebrospinal fluid of patients with schizophrenia. Neurosci Lett. 2001;313:96–8. in neuropsychiatric symptoms. Biol Psychiatry. 2011;70:175–82. 42. Fitch MT, Silver J. CNS injury, glial scars, and inflammation: inhibitory extracellular matrices and regeneration failure. Exp Neurol. 2008;209:294–301. 22. Linderholm KR, Skogh E, Olsson SK, Dahl ML, Holtze M, Engberg G, Samuelsson M, Erhardt S. Increased levels of kynurenine and kynurenic acid in the CSF of patients with schizophrenia. Schizophr Bull. 2012;38:426–32. 43. Skoldenberg B, Aurelius E, Hjalmarsson A, Sabri F, Forsgren M, Andersson B, Linde A, Strannegard O, Studahl M, Hagberg L, Rosengren L. Incidence and pathogenesis of clinical relapse after herpes simplex encephalitis in adults. J Neurol. 2006;253:163–70. 23. Savitz J, Dantzer R, Wurfel BE, Victor TA, Ford BN, Bodurka J, Bellgowan PS, Teague TK, Drevets WC. Neuroprotective kynurenine metabolite indices are abnormally reduced and positively associated with hippocampal and amygdalar volume in bipolar disorder. Psychoneuroendocrinology. 2015;52:200–11. 44. Koedel U, Scheld WM, Pfister HW. Pathogenesis and pathophysiology of pneumococcal meningitis. Lancet Infect Dis. 2002;2:721–36. 45. Pinto Junior VL, Rebelo MC, Gomes RN, Assis EF, Castro-Faria-Neto HC, Boia MN. IL-6 and IL-8 in cerebrospinal fluid from patients with aseptic meningitis and bacterial meningitis: their potential role as a marker for differential diagnosis. Braz J Infect Dis. 2011;15:156–8. 24. Olsson SK, Samuelsson M, Saetre P, Lindstrom L, Jonsson EG, Nordin C, Engberg G, Erhardt S, Landen M. Elevated levels of kynurenic acid in the cerebrospinal fluid of patients with bipolar disorder. J Psychiatry Neurosci. 2010;35:195–9. 46. Klein RS, Lin E, Zhang B, Luster AD, Tollett J, Samuel MA, Engle M, Diamond MS. Neuronal CXCL10 directs CD8+ T-cell recruitment and control of West Nile virus encephalitis. J Virol. 2005;79:11457–66. 25. Widner B, Leblhuber F, Fuchs D. Increased neopterin production and tryptophan degradation in advanced Parkinson's disease. J Neural Transm (Vienna). 2002;109:181–9. 47. Sui Y, Potula R, Dhillon N, Pinson D, Li S, Nath A, Anderson C, Turchan J, Kolson D, Narayan O, Buch S. Neuronal apoptosis is mediated by CXCL10 overexpression in simian human immunodeficiency virus encephalitis. Am J Pathol. 2004;164:1557–66. 26. Giil LM, Midttun O, Refsum H, Ulvik A, Advani R, Smith AD, Ueland PM. Kynurenine pathway metabolites in Alzheimer's disease. J Alzheimers Dis. 2017;60:495–504. 27. Guidetti P, Luthi-Carter RE, Augood SJ, Schwarcz R. Author details 1D f Neuroactive kynurenines in Lyme borreliosis. Neurology. 1992;42:43–50. 33. Holtze M, Mickiene A, Atlas A, Lindquist L, Schwieler L. Elevated cerebrospinal fluid kynurenic acid levels in patients with tick-borne encephalitis. J Intern Med. 2012;272:394–401. 54. Zhang B, Chan YK, Lu B, Diamond MS, Klein RS. CXCR3 mediates region- specific antiviral T cell trafficking within the central nervous system during West Nile virus encephalitis. J Immunol. 2008;180:2641–9. 34. van Laarhoven A, Dian S, Aguirre-Gamboa R, Avila-Pacheco J, Ricano-Ponce I, Ruesen C, Annisa J, Koeken V, Chaidir L, Li Y, et al. Cerebral tryptophan metabolism and outcome of tuberculous meningitis: an observational cohort study. Lancet Infect Dis. 2018;18:526–35. 55. Dufour JH, Dziejman M, Liu MT, Leung JH, Lane TE, Luster AD. IFN-gamma- inducible protein 10 (IP-10; CXCL10)-deficient mice reveal a role for IP-10 in effector T cell generation and trafficking. J Immunol. 2002;168:3195–204. 35. Midttun O, Hustad S, Ueland PM. Quantitative profiling of biomarkers related to B-vitamin status, tryptophan metabolism and inflammation in human plasma by liquid chromatography/tandem mass spectrometry. Rapid Commun Mass Spectrom. 2009;23:1371–9. 56. Cuartero MI, de la Parra J, Garcia-Culebras A, Ballesteros I, Lizasoain I, Moro MA. The kynurenine pathway in the acute and chronic phases of cerebral ischemia. Curr Pharm Des. 2016;22:1060–73. 57. Holmberg D, Franzen-Rohl E, Idro R, Opoka RO, Bangirana P, Sellgren CM, Wickstrom R, Farnert A, Schwieler L, Engberg G, John CC. Cerebrospinal fluid kynurenine and kynurenic acid concentrations are associated with coma duration and long-term neurocognitive impairment in Ugandan children with cerebral malaria. Malar J. 2017;16:303. 36. Bevital AS [http://www.bevital.no]. Accessed 19 Nov 2018. 37. Yamada A, Akimoto H, Kagawa S, Guillemin GJ, Takikawa O. Proinflammatory cytokine interferon-gamma increases induction of indoleamine 2,3-dioxygenase in monocytic cells primed with amyloid beta peptide 1-42: implications for the pathogenesis of Alzheimer's disease. J Neurochem. 2009;110:791–800. 58. Vecsei L, Szalardy L, Fulop F, Toldi J. Kynurenines in the CNS: recent advances and new questions. Nat Rev Drug Discov. 2013;12:64–82. 59. Liu M, Guo S, Hibbert JM, Jain V, Singh N, Wilson NO, Stiles JK. CXCL10/IP-10 in infectious diseases pathogenesis and potential therapeutic implications. Cytokine Growth Factor Rev. 2011;22:121–30. 38. Guillemin GJ, Kerr SJ, Pemberton LA, Smith DG, Smythe GA, Armati PJ, Brew BJ. IFN-beta1b induces kynurenine pathway metabolism in human macrophages: potential implications for multiple sclerosis treatment. J Interf Cytokine Res. 2001;21:1097–101. 60. Vancassel S, Capuron L, Castanon N. Author details 1D f 16. Frumento G, Rotondo R, Tonetti M, Damonte G, Benatti U, Ferrara GB. Tryptophan-derived catabolites are responsible for inhibition of T and natural killer cell proliferation induced by indoleamine 2,3-dioxygenase. J Exp Med. 2002;196:459–68. 1Department of Infectious Diseases, Oslo University Hospital, Ullevaal Hospital, P. O. Box 4956 Nydalen, N-0450 Oslo, Norway. 2Institute of Clinical Medicine, University of Oslo, Oslo, Norway. 3Research Institute of Internal Medicine, Oslo University Hospital Rikshospitalet, Oslo, Norway. 4Section of Clinical Immunology and Infectious Diseases, Oslo University Hospital Rikshospitalet, Oslo, Norway. 5Faculty of Medicine, University of Oslo, Oslo, Norway. 6K.G. Jebsen Inflammatory Research Center, University of Oslo, Oslo, Norway. 7K.G. Jebsen Thrombosis Research and Expertise Center, Tromsø, Norway. 8Bevital A/S, Bergen, Norway. 9Department of Microbiology, Oslo University Hospital, Ullevaal, Oslo, Norway. 10Department of Acute Medicine, Oslo University Hospital, Ullevaal, Oslo, Norway. 11Department of Immunology, Oslo University Hospital, Oslo, Norway. 12Department of Clinical Science, University of Bergen, Bergen, Norway. 13Research Laboratory, Nordland Hospital, Bodø, Norway. 14Faculty of Health Sciences, University of Tromsø, Tromsø, Norway. 15Centre of Molecular Inflammation Research, Norwegian University of Science and Technology, Trondheim, Norway. 17. Terness P, Bauer TM, Rose L, Dufter C, Watzlik A, Simon H, Opelz G. Inhibition of allogeneic T cell proliferation by indoleamine 2,3-dioxygenase- expressing dendritic cells: mediation of suppression by tryptophan metabolites. J Exp Med. 2002;196:447–57. 18. Fallarino F, Grohmann U, You S, McGrath BC, Cavener DR, Vacca C, Orabona C, Bianchi R, Belladonna ML, Volpi C, et al. The combined effects of tryptophan starvation and tryptophan catabolites down-regulate T cell receptor zeta-chain and induce a regulatory phenotype in naive T cells. J Immunol. 2006;176:6752–61. 19. Pallotta MT, Orabona C, Volpi C, Vacca C, Belladonna ML, Bianchi R, Servillo G, Brunacci C, Calvitti M, Bicciato S, et al. Indoleamine 2,3-dioxygenase is a signaling protein in long-term tolerance by dendritic cells. Nat Immunol. 2011;12:870–8. Page 13 of 14 Page 13 of 14 Page 13 of 14 Quist-Paulsen et al. Journal of Neuroinflammation (2018) 15:327 20. Guillemin GJ, Smythe G, Takikawa O, Brew BJ. Expression of indoleamine 2,3-dioxygenase and production of quinolinic acid by human microglia, astrocytes, and neurons. Glia. 2005;49:15–23. 41. Capuron L, Schroecksnadel S, Feart C, Aubert A, Higueret D, Barberger- Gateau P, Laye S, Fuchs D. Chronic low-grade inflammation in elderly persons is associated with altered tryptophan and tyrosine metabolism: role in neuropsychiatric symptoms. Biol Psychiatry. 2011;70:175–82. y 21. Author details 1D f Neostriatal and cortical quinolinate levels are increased in early grade Huntington's disease. Neurobiol Dis. 2004;17:455–61. 48. Lokensgard JR, Hu S, Sheng W, vanOijen M, Cox D, Cheeran MC, Peterson PK. Robust expression of TNF-alpha, IL-1beta, RANTES, and IP-10 by human microglial cells during nonproductive infection with herpes simplex virus. J Neuro-Oncol. 2001;7:208–19. 28. Heyes MP, Brew BJ, Martin A, Price RW, Salazar AM, Sidtis JJ, Yergey JA, Mouradian MM, Sadler AE, Keilp J, et al. Quinolinic acid in cerebrospinal fluid and serum in HIV-1 infection: relationship to clinical and neurological status. Ann Neurol. 1991;29:202–9. 49. Kothur K, Wienholt L, Mohammad SS, Tantsis EM, Pillai S, Britton PN, Jones CA, Angiti RR, Barnes EH, Schlub T, et al. Utility of CSF cytokine/chemokines as markers of active intrathecal inflammation: comparison of demyelinating, anti-NMDAR and enteroviral encephalitis. PLoS One. 2016;11:e0161656. 29. Valle M, Price RW, Nilsson A, Heyes M, Verotta D. CSF quinolinic acid levels are determined by local HIV infection: cross-sectional analysis and modelling of dynamics following antiretroviral therapy. Brain. 2004;127: 1047–60. 50. Bhowmick S, Duseja R, Das S, Appaiahgiri MB, Vrati S, Basu A. Induction of IP-10 (CXCL10) in astrocytes following Japanese encephalitis. Neurosci Lett. 2007;414:45–50. 30. Yan EB, Frugier T, Lim CK, Heng B, Sundaram G, Tan M, Rosenfeld JV, Walker DW, Guillemin GJ, Morganti-Kossmann MC. Activation of the kynurenine pathway and increased production of the excitotoxin quinolinic acid following traumatic brain injury in humans. J Neuroinflammation. 2015;12:110. 51. Palus M, Vojtiskova J, Salat J, Kopecky J, Grubhoffer L, Lipoldova M, Demant P, Ruzek D. Mice with different susceptibility to tick-borne encephalitis virus infection show selective neutralizing antibody response and inflammatory reaction in the central nervous system. J Neuroinflammation. 2013;10:77. 52. Lepej SZ, Misic-Majerus L, Jeren T, Rode OD, Remenar A, Sporec V, Vince A. Chemokines CXCL10 and CXCL11 in the cerebrospinal fluid of patients with tick-borne encephalitis. Acta Neurol Scand. 2007;115:109–14. 31. Atlas A, Franzen-Rohl E, Soderlund J, Jonsson EG, Samuelsson M, Schwieler L, Skoldenberg B, Engberg G. Sustained elevation of kynurenic acid in the cerebrospinal fluid of patients with herpes simplex virus type 1 encephalitis. Int J Tryptophan Res. 2013;6:89–96. 53. Lepej SZ, Rode OD, Jeren T, Vince A, Remenar A, Barsic B. Increased expression of CXCR3 and CCR5 on memory CD4+ T-cells migrating into the cerebrospinal fluid of patients with neuroborreliosis: the role of CXCL10 and CXCL11. J Neuroimmunol. 2005;163:128–34. 32. Halperin JJ, Heyes MP. 62. Kothur K, Wienholt L, Brilot F, Dale RC. CSF cytokines/chemokines as biomarkers in neuroinflammatory CNS disorders: a systematic review. Cytokine. 2015;77:227–37. Quist-Paulsen et al. Journal of Neuroinflammation (2018) 15:327 63. Azumagawa K, Suzuki S, Tanabe T, Wakamiya E, Kawamura N, Tamai H. Neopterin, biopterin, and nitric oxide concentrations in the cerebrospinal fluid of children with central nervous system infections. Brain and Development. 2003;25:200–2. 62. Kothur K, Wienholt L, Brilot F, Dale RC. CSF cytokines/chemokines as biomarkers in neuroinflammatory CNS disorders: a systematic review. Cytokine. 2015;77:227–37. 63. Azumagawa K, Suzuki S, Tanabe T, Wakamiya E, Kawamura N, Tamai H. Neopterin, biopterin, and nitric oxide concentrations in the cerebrospinal fluid of children with central nervous system infections. Brain and Development. 2003;25:200–2. 64. Ostergaard C, Benfield T. Macrophage migration inhibitory factor in cerebrospinal fluid from patients with central nervous system infection. Crit Care. 2009;13:R101. 65. Matsuzono Y, Narita M, Akutsu Y, Togashi T. Interleukin-6 in cerebrospinal fluid of patients with central nervous system infections. Acta Paediatr. 1995; 84:879–83. 64. Ostergaard C, Benfield T. Macrophage migration inhibitory factor in cerebrospinal fluid from patients with central nervous system infection. Crit Care. 2009;13:R101. Author details 1D f Brain kynurenine and BH4 pathways: relevance to the pathophysiology and treatment of inflammation-driven depressive symptoms. Front Neurosci. 2018;12:499. 39. Fuchs D, Weiss G, Wachter H. Neopterin, biochemistry and clinical use as a marker for cellular immune reactions. Int Arch Allergy Immunol. 1993;101:1–6. 61. Weiss G, Murr C, Zoller H, Haun M, Widner B, Ludescher C, Fuchs D. Modulation of neopterin formation and tryptophan degradation by Th1- and Th2-derived cytokines in human monocytic cells. Clin Exp Immunol. 1999;116:435–40. 40. Raison CL, Dantzer R, Kelley KW, Lawson MA, Woolwine BJ, Vogt G, Spivey JR, Saito K, Miller AH. CSF concentrations of brain tryptophan and kynurenines during immune stimulation with IFN-alpha: relationship to CNS immune responses and depression. Mol Psychiatry. 2010;15:393–403. Page 14 of 14 Quist-Paulsen et al. Journal of Neuroinflammation (2018) 15:327 62. Kothur K, Wienholt L, Brilot F, Dale RC. CSF cytokines/chemokines as biomarkers in neuroinflammatory CNS disorders: a systematic review. Cytokine. 2015;77:227–37. 63. Azumagawa K, Suzuki S, Tanabe T, Wakamiya E, Kawamura N, Tamai H. Neopterin, biopterin, and nitric oxide concentrations in the cerebrospinal fluid of children with central nervous system infections. Brain and Development. 2003;25:200–2. 64. Ostergaard C, Benfield T. Macrophage migration inhibitory factor in cerebrospinal fluid from patients with central nervous system infection. Crit Care. 2009;13:R101. 65. Matsuzono Y, Narita M, Akutsu Y, Togashi T. Interleukin-6 in cerebrospinal fluid of patients with central nervous system infections. Acta Paediatr. 1995; 84:879–83.
https://openalex.org/W4384655718	https://zenodo.org/records/8163050/files/Readme_DJR.pdf	English	null	Housing prices and credit constraints in competitive search	Zenodo (CERN European Organization for Nuclear Research)	2,023	cc-by	2,360	Replication Package for: “Housing Prices and Credit Constraints in Competitive Search” This replication package accompanies Díaz, Jerez and Rincón-Zapatero. (forthcoming). “Housing Prices and Credit Constraints in Competitive Search”. The Economic Journal. eplication Package for: “Housing Prices and Credit Constraints in Competitive Searc This replication package accompanies Díaz, Jerez and Rincón-Zapatero. (forthcoming). “Housing Prices and Credit Constraints in Competitive Search”. The Economic Journal. This replication package accompanies Díaz, Jerez and Rincón-Zapatero. (forthcoming). “Housing Prices and Credit Constraints in Competitive Search”. The Economic Journal. This replication package accompanies Díaz, Jerez and Rincón-Zapatero. (forthcoming). “Housing Prices and Credit Constraints in Competitive Search”. The Economic Journal. Authors: Antonia Díaz Belén Jerez Juan-Pablo Rincón-Zapatero Authors: Antonia Díaz Belén Jerez Juan-Pablo Rincón-Zapatero Instructions • Data_SCF contains the raw data and the Matlab files needed to obtain the targets from the Survey of Consumer Finances, https://www.federalreserve.gov/econres/scfindex.htm. • Codes contains the codes to replicate Figure 3 and Tables 4 to 6 in the paper. The rest of this readme file describes the contents of each folder. NOTE: Figures 1 and 2 in the paper manually created as they are illustrations of the theory and the mechanics of the main algorithm. The code does not generate them. NOTE: Figures 1 and 2 in the paper manually created as they are illustrations of the theory and the mechanics of the main algorithm. The code does not generate them. Data availability and provenance statements • ☐X I certify that the author(s) of the manuscript have legitimate access to and permission to use the data used in this manuscript. • ☐ I certify that the author(s) of the manuscript have documented permission to redistribute/publish the data contained within this replication package. Appropriate permission are documented in the LICENSE.txt file. • ☐ I certify that the author(s) of the manuscript have documented permission to redistribute/publish the data contained within this replication package. Appropriate permission are documented in the LICENSE.txt file. Summary of availability Summary of availability • ☐X All data are publicly available. • ☐ Some data cannot be made publicly available. • ☐ No data can be made publicly available. Replication Package for: “Housing Prices and Credit Constraints in Competitive Search” This replication package accompanies Díaz, Jerez and Rincón-Zapatero. (forthcoming). “Housing Prices and Credit Constraints in Competitive Search”. The Economic Journal. Replication Package for: “Housing Prices and Credit Constraints in Competitive Search” This replication package accompanies Díaz, Jerez and Rincón-Zapatero. (forthcoming). “Housing Prices and Credit Constraints in Competitive Search”. The Economic Journal. Details on data source Data of Survey of Consumer Finances were downloaded from the U.S. Board of Governors of the Federal reserve System (2019). The data are in the public domain. Data can be downloaded from https://www.federalreserve.gov/econres/scfindex.htm, select the wave, then download. Only a part of the data of each wave is needed in this paper. A copy of the data needed is provided as part of this archive in .txt format. Data of Survey of Consumer Finances were downloaded from the U.S. Board of Governors of the Federal reserve System (2019). The data are in the public domain. Data can be downloaded from https://www.federalreserve.gov/econres/scfindex.htm, select the wave, then download. Only a part of the data of each wave is needed in this paper. A copy of the data needed is provided as part of this archive in .txt format. Data of Survey of Consumer Finances were downloaded from the U.S. Board of Governors of the Federal reserve System (2019). The data are in the public domain. Data can be downloaded from https://www.federalreserve.gov/econres/scfindex.htm, select the wave, then download. Only a part of the data of each wave is needed in this paper. A copy of the data needed is provided as part of this Name Data.Files Location Provided Citation Survey of Consumer Finances 1989 Rawdata_1989.txt 3 Replication package/Data_SCF TRUE Board of Governors (2019) Survey of Consumer Finances 1992 Rawdata_1992.txt 3 Replication package/Data_SCF TRUE Board of Governors (2019) Survey of Consumer Finances 1995 Rawdata_1995.txt 3 Replication package/Data_SCF TRUE Board of Governors (2019) Survey of Consumer Finances 1998 Rawdata_1998.txt 3 Replication package/Data_SCF TRUE Board of Governors (2019) Survey of Consumer Finances 2001 Rawdata_2001.txt 3 Replication package/Data_SCF TRUE Board of Governors (2019) Survey of Consumer Finances 2004 Rawdata_2004.txt 3 Replication package/Data_SCF TRUE Board of Governors (2019) Survey of Consumer Finances 2007 Rawdata_2007.txt 3 Replication package/Data_SCF TRUE Board of Governors (2019) Consumer Price Index cpi_index.txt 3 Replication package/Data_SCF TRUE Bureau of Labor Statistics (2019) Software requirements Approximate time needed to reproduce the analyses on a standard (2022) desktop machine • ☐ <10 minutes • ☐ 10-60 minutes • ☐X 1-2 hours • ☐ 2-8 hours • ☐ 8-24 hours • ☐ 1-3 days • ☐ 3-14 days • ☐ > 14 days • ☐ Not feasible to run on a desktop machine, as described below. Computational requirements Computational requirements Software requirements Matlab (code was run with Matlab Release 2022b). It uses its main toolboxes such as • Optimization toolbox. Memory and runtime requirements Summary Approximate time needed to reproduce the analyses on a standard (2022) desktop machine: • ☐ <10 minutes • ☐ 10-60 minutes • ☐X 1-2 hours • ☐ 2-8 hours • ☐ 8-24 hours • ☐ 1-3 days • ☐ 3-14 days • ☐ > 14 days • ☐ Not feasible to run on a desktop machine, as described below. Details This code was run in an iMac 2015, 16 GB RAM, IOS 12.6.3 with Matlab version 2022b. Computation took 76 minutes. However, in a MacBook Pro 16 GB RAM, IOS 13.4.1 with Matlab version 2023b computation time fell to 32 minutes. Details This code was run in an iMac 2015, 16 GB RAM, IOS 12.6.3 with Matlab version 2022b. Computation took 76 minutes. However, in a MacBook Pro 16 GB RAM, IOS 13.4.1 with Matlab version 2023b computation time fell to 32 minutes. INSTRUCTIONS: Identifying hardware and OS can be obtained through a variety of ways: Some of these details can be found as follows: INSTRUCTIONS: Identifying hardware and OS can be obtained through a variety of ways: Some of these details can be found as follows: • (Mac) Apple-menu > “About this Mac” • (Mac) Apple-menu > “About this Mac” • (Linux) see code in tools/linux-system-info.sh` Description of program/codes • Programs in 3 Replication package/Data_SCF will use raw data and find calibration targets used in Section 4.3.2. targets used in Section 4.3.2. • Programs in 3 Replication package/Codes contains the code to simulate the benchmark and additional economies studied in the paper. • Programs in 3 Replication package/Codes contains the code to simulate t and additional economies studied in the paper. Description of the data and code in folder Data_SCF The raw data for each SCF wave is included in “rawdata_wave.txt”. The needed Matlab code to compute the targets is The raw data for each SCF wave is included in “rawdata_wave.txt”. The needed Matlab code to compute the targets is • “main_facts.m”: script file that cleans the data and extracts the data pertaining to working age households. This function calls: “ ” f ti th t l l t til • “main_facts.m”: script file that cleans the data and extracts the data pertaining to working age households. This function calls: o “per.m”: function that calculates percentiles. o “ineq.m”: function that calculates inequality statistics. households. This function calls: o “per.m”: function that calculates percentiles. The targets are written in an Excel file called “targets_paper_search”, as a csv file and as a Matlab table. The three output files are located in the folder where “main_facts.m”, “per.m” and “ineq.m” are located. The output is saved in 3 Replication package/Data_SCF/Targets_paper_search.csv 3 Replication package/Data_SCF/Targets_paper_search.xlsx o “get_buyers_sellers.m” This function gets the sum of sellers needed to satisfy the demand given by ψ$, which is the distribution of buyers over wealth levels as well as paper) and obtains Figure 3, Tables 4 to 6 of the paper. Main files called: cript file “main_code.m” loads all parameter values (whose values are shown in Tables 1, 2 Description of the code in folder Codes “get_cdfs.m” computes a finer grid of financial assets and an initial distribution of financial assets for owners and renters. It assumes that immigrants come with zero financial assets. It uses the functions: o “get_cdf_afternoon.m” gets the measure of non-owners (buyers) before the frictional housing market opens, as shown in Section (3.3). o get_cdf_afternoon.m gets the measure of non owners (buyers) before the frictional housing market opens, as shown in Section (3.3). o “get_cdf_night.m” gets the distribution of owners and renters after the frictional market has closed, as shown in Section (3.3). o “interp cdf.m” interpolates to find the cdf of a particular variable. o “get_cdf_night.m” gets the distribution of owners and renters after the frictional market has closed, as shown in Section (3.3). o “interp cdf m” interpolates to find the cdf of a particular variable ( ) o “interp_cdf.m” interpolates to find the cdf of a particular variable. • The Walrasian economy: Table 4 reports the steady state of three economies with search and matching frictions. This economy has no frictions and there is a unique housing price. The files needed to find the steady state of this economy have the same names than in the search economy but for the suffix “_walras” appended to the name of the file. Description of the code in folder Codes This folder contains the code to simulate the benchmark and additional economies studied in the paper. Specifically, this code takes as inputs the data shown in Tables 1, 2 and 3 and obtains Figure 3, Tables 4, 5 and 6. The code follows Appendix F in the paper. Figure/Table Program Location Line Number Output Notes Figure 3 Main_code.m 3 Replication package/Codes 138 Figure_3a.png Figure_3b.png Figure_3c.png Up to here, the code generates the benchmark steady state, saves it for Tables 4, 5 and 6 and plots policy functions in Figure 3. Located in 3 Replication package/Codes Table 4 Main_code.m 3 Replication package/Codes 201 Table_4.csv Located in 3 Replication package/Codes Table 5 Main_code.m 3 Replication package/Codes 251 Table_5.csv Located in 3 Replication package/Codes Table 6 Main_code.m 3 Replication package/Codes 320 Table_6.csv Located in 3 Replication package/Codes The script file “main_code.m” loads all parameter values (whose values are shown in Tables 1, 2 and 3 in the paper) and obtains Figure 3, Tables 4 to 6 of the paper. Main files called: The script file “main_code.m” loads all parameter values (whose values are shown in Tables 1, 2 and 3 in the paper) and obtains Figure 3, Tables 4 to 6 of the paper. Main files called: The script file “main_code.m” loads all parameter values (whose values are shown in Ta in the paper) and obtains Figure 3, Tables 4 to 6 of the paper. Main files called: • “find_eqm_search.m”: function that finds the equilibrium of the search economy for a particular value of the down payment parameter and price 𝑝. It uses the algorithm described in Section F.4. Its mechanics is illustrated in Figure 2. For each value for 𝑝 “find_eqm_search.m” calls the file “get_aggregates.m” finds the steady state main aggregates. It calls o “get_policies.m” finds the policy functions following the algorithm outlined in Appendix F.2. o “get_cdfs.m” finds the stationary distribution of owners, renters and buyers, as describe in Appendix B, Section F.3 and equations (8)—(9). o “get_buyers_sellers.m” This function gets the sum of sellers needed to satisfy the demand given by ψ$, which is the distribution of buyers over wealth levels as well as statistics of the distribution of equilibrium prices and vacancies and shown in Equation (9). statistics of the distribution of equilibrium prices and vacancies and shown in Equation (9). t_policies.m” solves the household’s problem as described in Section E.2.3. Description of the code in folder Codes It uses the owing functions: o “get_pol_c0_renter.m” finds the initial guess for the renter’s financial assets policy function. It assumes that renters never participate in the frictional housing market. g o “get_pol_c0_owner.m” finds the initial guess for the owner’s financial assets policy function. It assumes that owners never receive a mismatching shock. o “get_pol_p” calculates the choice of submarket that solves buyer’s problem shown i Equation (4) in Section 3.1.3. It uses “fzero” from Matlab and functions o “get_pol_p” calculates the choice of submarket that solves buyer’s problem shown in Equation (4) in Section 3.1.3. It uses “fzero” from Matlab and functions § “foc_p.m”, the First Order Condition of the problem shown in Equation (4), q ( ) § “foc_p.m”, the First Order Condition of the problem shown in Equation (4), § “eqm_prob.m” is the indifference curve of intermediaries, which gives the probabilities of selling and buying for any given price. § “eqm_prob.m” is the indifference curve of intermediari probabilities of selling and buying for any given price. o “get_pol_c_renter.m” uses the Euler equation of the problem shown in Equation (3) to find the optimal choice of financial assets for renters. It uses the optimal choice of housing price found in “get_pol_p.m”. The algorithm is described in Appendix F.2.3. o “get_pol_c_owner.m” performs the same task for owners. It solves the problem shown i Equation (2). q ( ) o “vfi_step.m” conducts the Value Function Iteration step and it is called by o “vfi_step.m” conducts the Value Function Iteration step and it is called by “get pol c renter” and “get pol c owner” o “vfi_step.m” conducts the Value Function Iter “get_pol_c_renter” and “get_pol_c_owner”. “get_pol_c_renter” and “get_pol_c_owner”. o Auxiliar functions: “felicity.m” and “der_felicity_inv.m” calculate the felicity function and its partial derivative with respect to non-housing consumption. “eqm_prob.m” calculate h b i d lli b bili f h i o Auxiliar functions: “felicity.m” and “der_felicity_inv.m” calculate the felicity function and its partial derivative with respect to non-housing consumption. “eqm_prob.m” calculates the buying and selling probability for each price 𝑝≥𝑝. y y y its partial derivative with respect to non-housing consumption. “eqm_prob.m” calculates the buying and selling probability for each price 𝑝≥𝑝. its partial derivative with respect to non housing consumption. eqm_prob.m calculates the buying and selling probability for each price 𝑝≥𝑝. the buying and selling probability for each price 𝑝≥𝑝. References Board of Governors of the Federal Reserve System (2019). `Survey of Consumer Finances', https://www.federalreserve.gov/econres/scfindex.htm. Board of Governors of the Federal Reserve System (2019). `Survey of Consumer Finances', https://www.federalreserve.gov/econres/scfindex.htm.
https://openalex.org/W2146381153	https://zenodo.org/records/2430289/files/article.pdf	English	null	Studies of the Changes occurring in Heated Soils	Journal of agricultural science	1,910	public-domain	6,904	1 Journ. Agric. Sci. n. 411, in. 33. » loc cit. p. 422. BY SPENCER TJMFREVILLE PICKERING, M.A., F.R.S. IN previous communications1 it has been shown that soils heated to temperatures from 60° to 150° exhibit an inhibitory effect ou the germination of seeds, due to the presence of some toxic substance, which must be a soluble organic, and, probably, nitrogenous, body, for the extent to which germination is retarded is roughly proportional, to both the soluble organic matter and the soluble nitrogen present. That so-called unheated soils, that is, soils which have not been heated above 20°—30°, contain some of this substance, was also probable from the fact that the results with unheated and heated soils all lie on continuous curves. It was shown, too, that treatment with antiseptics produced a chemical change in soils, closely similar to that produced by heating them to 60°—75°. As a preliminary step towards obtaining some knowledge of the nature of this deleterious substance, a study was made of the changes which it undergoes when kept under various conditions: and this was preceded by an enquiry as to the trustworthiness of determinations of soil-extracts. Such determinations, if trustworthy, should have an extensive general application in soil analysis, for the soluble organic matter in soils, at any rate, in those of a similar nature, would probably afford a valuable measure of their fertility, and the determination of it would be much simpler than that of the nitrogen. From one series of determinations already published", as well as from others which have subsequently been made, it appears that the proportionality between the nitrogen and total organic matter in such extracts is generally very close. The method adopted consisted of putting 100 grams of soil into a stoppered bottle with 1000 c.c. of water, shaking it up violently once S. U. PICKERING 259 every five minutes throughout two hours, leaving it to settle for 15 minutes, and then filtering through paper, an operation which generally occupied 8 to 12 hours. The filtrate was subsequently drawn through a Berkefeld filter, which was thoroughly cleaned by running several portions of the liquid through it first. 750 c.c. of the filtrate were then evaporated to a small bulk in porcelain, and trans- ferred to platinum for evaporation to dryness. The residue, weighed after drying at 100", and also after the strongest possible ignition over a gas blowpipe, was expressed as a percentage of soil dried at 100°. BY SPENCER TJMFREVILLE PICKERING, M.A., F.R.S. The preliminary filtering through paper can be dispensed with, though it facilitates the final filtration. One obvious source of experimental error is the difficulty in weighing a large dish containing the hygroscopic residue. The error in determining the organic matter, which depends on two such weighings, cannot be placed at much less than a possible four milligrams. There are other sources of error which interfere with the absolute value of the results: the water retained after drying at 100° will be reckoned as so much organic matter, and the apparent weight of this will be further increased by the inorganic residue losing during the ignition any carbon dioxide and chlorides which may be present. On the other hand, volatile organic matter, if present, will be lost during the evaporation, and there may be a mechanical loss in the deflagration of the residue when much nitrate is present. The compounds in the extracts are, doubtless, numerous, but three distinct features are noticeable. Some extracts, chiefly those of un- heated and not very rich soils, leave a yellow granular deposit on evaporation, which, after ignition,'is not easily removed by hydrochloric acid from the dish. In other cases a dark greasy scum separates before evaporation has proceeded far, and this adheres somewhat tenaciously to the porcelain. This is noticeable chiefly with very rich soils, and also with those which had been kept moist in closed vessels. In a third class, principally with soils heated to 125° and 150°, no solids separate on concentration, and the residue finally obtained forms a dark glassy mass. On ignition, this, and the residues from some of the less heated soils, especially from those kept moist in closed vessels, leaves an almost black deposit, which, however, contains no carbon, but dissolves readily in hydrochloric acid to form a deep brown solution, which, on heating, or on standing for some hours, becomes colourless or slightly yellow. This is probably due to the presence of manganese. The extracts, those of the last class especially, have a considerable action on platinum, the dish losing about O'OOl gram during each determination. 17—2 260 Changes in Heated Soils TABLE I. Composition of soil extracts obtained under various conditions. Ti l d TABLE I. Composition of soil extracts obtained under various conditions. Time altered. Organic matter Inorganic matter Conditions Per cent. Belative Per cent. Eelative 1. 20 minutes -0437 96 -0345 93 2. BY SPENCER TJMFREVILLE PICKERING, M.A., F.R.S. The variation of the results lead to a probable error of a single deter- mination of the organic matter being 00028 per cent., equivalent to about '002 gram, with the quantities taken, which is not greater than the probable weighing error. In the investigations of the changes occurring in soils after being heated, the heating of the soil (which was Harpenden soil) was per- formed, as in former experiments, in closed vessels, two hours at the selected temperatures being allowed. The water present in the soil was 8-8 per cent., calculated on the sample after drying at 100°, its total water capacity being 33 per cent., or 50 parts of water to 100 of dry soil. Portions of 100 grams each of these soils were then kept under two different conditions: in one case, in open glass pans, with water to the saturation point added at intervals as the soil dried up, the soil, also, being broken up occasionally, as would have obtained in ordinary cultivation: in the other case, the soil, with water to the saturation point, was kept in hermetically sealed flasks, which, besides the soil, contained about 15 c.c. of air. In both cases the soil was kept in a table hothouse at about 15°. A few supplementary experiments were made with soil in flasks, kept in the hothouse, but in the dark, and also with soil kept at a winter temperature in the light. No precautions were taken in any case to prevent the reinoculation of the heated soils by bacteria, &c, but in no cases did any moulds appear on the soils during the experiments. The water added in all cases was nearly that required for the saturation of the unheated soil, namely, 32 per cent.: but this is rather more than that necessary to saturate the soils which had been heated, for the total water capacity of the heated soils diminishes with the temperature of heating down to 274 per cent, for that heated to 150°. BY SPENCER TJMFREVILLE PICKERING, M.A., F.R.S. 40 „ -0470 102 -0:)65 99 3. 40 -0372 81 -0316 86 4. 80 -0438 97 0400 108 5. 120 -0496 108 -0380 103 6. 160 0410 90 -0395 107 7. 240 -0412 91 -0370 100 8. 320 -0512 112 -0417 113 9. 8 hours on filter -0431 95 -0333 90 Temperature altered. 5. 6-8° -0496 108 -0380* 103 10. 16-7° -0467 102 •0427* 116 11. 23-5° 0519 113 0-498* 135 Proportions altered. 12. 50 grams to 1 litre -0462 101 -0520* 141 5. 100 ,, „ -0496 108 -0380* 103 13. 200 „ , -0323* 70 -0284* 77 Mean, omitting those starred... '0459 100 -0369 100 The results of the examination of the method of analysis, in which uumanured top soil from Harpenden was used, are collected in Table I, where the relative values entered are those obtained by comparison with the mean given at the bottom of the table. In the first nine entries the proportion used was 100 grams to one litre, at a temperature of 5°—10°, the time of digestion being varied from 20 to 320 minutes. As regards both the organic and inorganic matter, no effect whatever is produced by this variation, except a very doubtful increase when the time extended to 320 minutes. In No. 9, the soil was placed on a filter, and the litre of water was drawn through it, this occupying eight hours; but the results are sensibly the same as with the other method of procedure. In the next three experiments the temperature, both of digestion and filtration, was altered, the proportions being 100 grams to the litre, and the time two hours. The organic matter dissolved does not seem to be sensibly affected by a difference of 16°, but the inorganic matter evidently is so, as it increases regularly with rise of temperature. In the last three experiments the proportion of soil to water was altered. This does not sensibly affect the organic matter till the proportion exceeds 100 grams to the litre; but the amount of inorganic matter is affected, just as it was by temperature. It is clear, therefore, that, especially as regards the dissolved organic matter, which is the point of chief importance, the method S. U. PICKERING 261 gives perfectly satisfactory results, and these are not affected by such minor variations in the conditions as are likely to occur in practice. BY SPENCER TJMFREVILLE PICKERING, M.A., F.R.S. Some supplementary experiments, however, showed that the excess which would be present in these cases was without influence on the results: thus, soils which had been heated to 125° and 150° respectively, and of which the total water capacity was 30 and 27-4 per cent., respec- tively, were kept in sealed flasks for 10 days with 32-8 and 253 per cent, of water in each case; after which the alteration in the soluble organic matter in them was found to be With deficit of water +16 per cent. + 4 „ With exoesa of water 125° +16 per cent. 150° + 9 „ 262 Changes in Heated Soils Although the heated soils can take up less water than the unheated ones, they retain it with considerably greater firmness. The weight of water added to 100 grams of the various soils in pans, to make up for the evaporation during 16 weeks, ranged from 400 c.c. in the case of the soils heated to below 100°, down to 314 c.c. with those heated to higher temperatures, though the loss did not altogether follow the temperature of heating, owing, no doubt, to irregularities in the air currents in the hothouse. Sets of the soils were analysed after 10, 44 and 112 days, and other sets were used for germination experiments after 44 and 106 days. The original soil taken was also examined by analysis and germination to start with, and again after 119 days, it having been kept during that period in its original air-dried condition in stoppered bottles at a winter temperature. These bottles were capable of holding 30—40 times the quantity of soil present, and they were occasionally opened, so that there was ample opportunity for oxidation to occur, if any could occur under such conditions. To ascertain the extent of the accidental variation which might be expected, duplicate experiments were made with soil which had been heated to 60° and to 125°, both in pans and in sealed flasks. The results were as follows:— Extractive matter in soils after 62 days. Inorganic Organic Pal 60°. 125°. Probable error of single experiment— •0019 -0023 -0040 -0030 0021 0035 A. B. A. B. Pans •U439 •0460 •0593 •0503 Flasks •0564 •0592 •0938 •0831 Pans •0294 •0387 •1182 •1038 Flasks •0487 •0581 •1967 •1883 Extractive matter in soils after 62 days. BY SPENCER TJMFREVILLE PICKERING, M.A., F.R.S. Thus, the probable error as regards the organic matter (-0035), is less than 50 per cent, greater than that determined above as the probable analytical error ("0028): as regards the inorganic matter the error is only '0021. These errors are insignificant as compared with the total change in composition occurring during the time. The results of the analyses are contained in Tables II, III and IV, and those of the germination experiments in Table V. In the latter, four sorts of seeds were used, and all the experiments were made in duplicate. The incubation period has alone been considered, the values entered being the average times required for germination, as compared with that required by the same seed in the unheated soil (i.e. heated 263 S. U. PICKERING to 30°) in the same series. The rye used in the first series had, unfortunately, such a poor germination capacity, that the results with it were of no value. TABLE II. Soluble matter in soils after- being kept watered for various lengths of time in open pans. After nsflt- ing to 30° 60° 80° 100° 125° 150° 30° 60° 80° 100° 125° 150° Percentage after Nil •0334 •0433 •0G84 0967 •17K3 •4187 0390 0117 0483 •0139 •0674 1125 10 d. •0266 0250 0342 •0705 •1362 •2688 •0443 •0481 •0380 •0434 •0451 •0716 44 d. •0471 0395 0498 •0723 •1178 •2333 112 d. •0546 •0349 •0377 •0592 •0727 •1242 P.c. alteration in 10 d. Organic -20 - 4 2 -50 -36 -24 -36 44 d. mattei + 41 - 9 -27 -25 - 3 3 -44 112 d. + 63 -20 -45 - 3 8 - 5 8 -44 Inorganic matter. •0633 •0503 •0525 •0509 •0553 •0747 •0696 •0620 •0530 •0558 •0495 •0640 + 14 + 15 - 2 1 - 1 - 3 3 -64 + 62 + 20 + 9 + 16 -18 -34 + 78 + 49 + 10 + 27 -27 - 4 3 Eelative values after Nil 100 130 205 289 528 1254 100 107 124 113 173 288 10 d. 100 94 128 265 512 1010 100 109 86 98 102 162 44 d. 100 84 106 153 250 495 100 79 83 80 87 118 112 d. 100 64 69 109 133 227 100 89 76 80 71 92 TABLE III. BY SPENCER TJMFREVILLE PICKERING, M.A., F.R.S. Soluble matter in soils after being kept moist for various lengths of time in sealed flasks. After neftt- ing to 30° 60° 80° 100° 125° 150° 30° 60° 80° 100° 125° 150° Percentage after Nil •0334 •0433 •0684 •0967 1763 •4187 •0390 •0417 •0483 •0439 •0674 •1125 10 d. 0221 •0309 •0470 •0835 •1820 •4238 0317 •0323 •0343 •0372 •0603 •1145 43 d. •0353 •0611 •0721 •1022 •1934 •4751 •0427 •0560 •0602 •0606 •0639 •1243 116 d. •0529 •0834 0935 •1055 •1904 •4531 P.O. 10 d. Organic -34 -29 -32 -14 + 3 + 1 alteration in 43 d. matter + 6 + 41 + 5 + 6 + 10 + 13 116 d. + 59 + 93 + 37 + 9 + 8 + 8 Inorganic matter. •0867 •0781 •0740 •0681 •0628 •1551 -19 -•23 -29 - 1 5 -11 + 2 + 10 + 29 + 25 + 38 — 5 + 10 + 123 + 87 + 53 + 55 - 7 + 38 Relative values after Nil 100 130 205 289 528 1254 100 107 124 113 173 288 10 d. 100 140 213 378 824 1918 100 102 106 117 190 361 43 d. 100 173 204 289 548 1346 100 131 141 142 150 292 116 d. 100 158 177 200 360 856 100 90 85 79 72 179 TABLE II. Soluble matter in soils after- being kept watered for various lengths of time in open pans. 264 Changes in Heated Soils Changes in Heated Soils TABLE IV. Soluble matter in soils kept under various conditions. After heat- ing to 30° 60° 80° 100° 150° 30° 60° 80° 100° 150° 60° 125° 60° 125° Kept dr; Ori- ginal p.o. •0334 •0433 •0684 •0976 •4187 •0390 0417 0483 •0439 •1125 — ' in bottles After 116 days P.c. •0228 •0283 •0296 •0558 •2598 •0431 •0394 •0381 •0382 •0554 — Altera- tion p.c. - 3 2 - 3 5 - 5 7 - 4 2 - 3 8 + 10 - 5 - 2 1 - 1 1 - 5 1 — Rela- tive Org 100 124 130 245 1139 Kept wet in For 43 days In light ante ?» 0353 •1022 •4751 In dark atter. •0347 •0902 •4664 Inorganic matter. 100 91 90 90 129 Org •0427 •0606 1243 •0456 •0487 1335 Diff. 1 - -0006 - -0040 - -0087 + -0029 - -0119 + -0092 For 62 days At 15° ante m •0667 •1926 At 5° atter. •0534 •1925 Inorganic matter. — •0613 •0637 •0578 •0884 Diff. - -0133 -•0001 - -0035 + -0247 sealed flasks For 116 days In light •0529 •1055 •1655 •0867 •0681 1551 In dark •0606 •1074 •1683 •0591 •0615 •1502 — Di£t. + -0077 + -0019 + -0028 - -0276 - 0066 - 0049 — —— TABLE IV. Soluble matter in soils kept under various conditions. S. U. PICKERING 265 TABLE V. Relative incubation periods of seeds in soils previously heated and kept under various conditions. Temp, of heating 30° 60° 80° 100° 125° 150° 30° 60° 80° 100° 125° 150° 30° 60° 80° 100° 125° 150° 30° G0° 80° 100° 125° 150° 30° 60° 80° 100° 125° 150° 30° 60° 80J 100° 125° 150° Incubation periods Wheat 100 148 158 152 234 220 Eye 1. 5 — Clover oil taken, 100 50 75 10ti 65 69 Mustard Oct. 19th. 100 137 149 117 156 190 Mean 100 116\ 133 120 f 143 163 186) 2. Kept in pans till Dec. 2nd, 44 days. 100 109 173 148 112 165 100 106 105 1211 146 165 100 115 114 123 161 190 100 122 115 108 112 147 100 113\ 127 127 133 167 121 3. Kept in puns till Feb. 2nd, 106' days. Changes in Heated Soils 100 85 110 93 89 97 100 105 100 120 105 100 100 119 107 154 89 102 100 104 89 121 108 122 100 104 101 122 103 105 107 4. Kepi, moist in flasks till Dec. 2nd, 44 days. 100 100 110 171 214 200 100 105 128 110 123 148 100 141 83 132 100 209 5. Kept moist in Jlasks til 100 105 106 88 124 123 100 112 144 180 136 183 100 116 116 123 98 258 6. Kept dry in bottles till 100 80 162 134 115 170 100 100 142 154 142 141 100 as 85 77 90 95 100 112 148 169 250 299 I Feb. 2nd 100 147 138 124 136 223 Feb. 15th 100 86 113 143 146 123 100 1141 117 14S 172 214. - 152 106 days. 100 120\ 126 129 124 197. 139 119 days. 100 9n 125 120^ 122 138 134J Mean B. — — — — — — — — — 114 1321 130 163 189 240; 171 106 127 \ 135 140 144 • 1 5 1 207 _ — TABLE V. Relative incubation periods of seeds in soils previously heated and kept under various conditions. Changes in Heated Soils 266 TABLE VI. Results obtained with soils kept moist in sealed flasks compared with those obtained from soils kept in open pans, repre sented by 100. After heating to 30° 60° 80° 100° 125° 150° Kept for 43—44 days Organic 75 155 147 141 164 204 Inorganic 68 111 113 119 116 166 Incubation 114 117 102 128 141 144 Kept for 106—116 days Organic 97 242 248 178 262 365 Inorganic 124 126 140 122 127 242 Incubation 106 122 184 115 140 197 TABLE VI. Results obtained with soils kept moist in sealed flasks compared with those obtained from soils kept in open pans, repre sented by 100. Charts I and II represent the results of the determinations of soluble organic and inorganic matter, respectively, the cube root of the percentages of the former being taken for plotting, so as to reduce the steepness of the curves. The lettering at the ends of the curves gives the age in days of the soils, as dated from the starting of the experiments. 1 Journ. of Agric. Soc. m. 43. In these the point of intersection was at about 80°. Other determinations were made with a series of soils in which grass was grown for 72 dayB: the curves in that case were closely similar to those here shown, though not quite so regular, and the point of intersection was 75°. Changes in Heated Soils A cursory glance at the figures is sufficient to show that changes of a very marked and complex character occur on keeping the soils, and that the alteration in the amount of soluble matter is some- times negative and sometimes positive, and may extend to over 100 per cent, of that originally present. There are, however, certain well marked features in these changes. In the first place, the initial curves, marked 0, are in all cases very regular and simple. Looking at Figs. 1 in each chart, which apply to soils kept watered in open pans, the general result is (with one exception which will be dealt with im- mediately) that with the unheated, or less highly heated soils, the soluble matter, both organic and inorganic, increases as time passes, but with the more highly heated ones, the soluble matter diminishes, so that the curves representing the results of successive examinations cut each other at some point, this point varying from 50° to 80° as regards the organic matter, and from 80" to 105° as. regards the inorganic matter; that is, soils which have been heated to these temperatures show little or no change in composition on being kept. This result is in full agreement with that of some preliminary ex- periments which have already been published1. S. U. PICKERING 267 S. U. PICKERING With the soils kept in sealed flasks (Figs. 2 on both Charts), there is a similar, though evidently more complicated, increase in the soluble matter in the case of the less heated soils, but, as regards the highly heated ones, there is no diminution, as in the other case, for the organic matter remains almost constant throughout, or, even increases Vp 15 i n I U y 10 m 112. 44 ~^ 0--" Fig.2. 116-'' 0 i o - - Fig.3. o — 116-- CHAEI //? < ^ > a a r I. Soluble organic matter >/jen • /Jans osed flasks -""""'^^ -** r'fl bottles/^ / y • * - in soils kept. / / / -: / ''7 / / / / / / ' / / .'/ - 7 / / / / / / // f * r /° •'AH 112 / « A 0.10,116 / • /lie 50° 100° 150" Temperature of heating CHAEIr I. Soluble organic matter in soils kept. 268 Changes in Heated Soils slightly, and the inorganic matter increases too, though only slightly so at first. It is clear, therefore, that the main changes occurring, or the CHART II. Soluble inorganic matter in soils kept. P. •10 Fig. 1. In open pans •05 •15 •10 •05 •10 Fig. 2. In 116 - _ _ Fig.'3. Dry, closed flasks in bottles 116 44 10 O ,0 116 50° 100° 150° Temperature of heating 100 Temperature of heating 100 Temperature of heating S. U. PICKERING 269 principal substances present, in the highly and less highly heated soils are not identical, and the results strongly indicate that, in these latter, the diminution of the soluble matter is probably due to oxidation, for oxidation would be at a maximum in the open pans, and at a minimum in the sealed flasks. The increase in the soluble matter with the less heated soils does not appear to be influenced in any definite manner by the access or absence of air. The progress of the changes occurring is, however, not quite so simple, for, the increase in the soluble matter in the less highly heated soils is preceded during the first ten days by a decrease; this is shown equally by Figs. 1 and 2 in Chart I, and Fig. 2 in Chart II, whilst Fig. S. U. PICKERING Whether there would be any decrease in the soluble matter if air was entirely absent, cannot be settled by the present experiments, for the sealed flasks contained some air to start with, but it is most probable that there would be none in such a case. That oxidation is the cause of the primary decrease in the soluble matter, is further borne out by the results with the dry soil kept in bottles for 116 days (Fig. 3, Charts I, II), where oxidation, though possible, would be less favoured thau with soils kept moist. The results with the organic matter show a reduction in the 116 days closely similar to that shown in the other cases after ten days. As regards the inorganic matter (Chart II), the results are of too doubtful a character to afford any evidence in the matter. y That an air-dried soil, on being kept for a few months, should alter in composition, is a fact which the agricultural chemist will have to bear in mind when storing soil-samples. A similar change has been previously noticed in the case of soils treated with antiseptics1. The next step is to compare the results of the analyses with those of the germination of seeds. For this purpose the former must be treated somewhat differently. The absolute values for the incubation periods of seeds can only be compared together in cases where the experiments were made at the same time, for seeds of different ages behave differently: comparative values, taking the incubation period in the unheated soil as a standard, are alone available. The results of the soil analyses expressed in the same way are entered in the last columns of the Tables, and are depicted, together with the germination results, in Charts III and IV. The unlettered curves ( - -) in Chart III give the results with the soil kept dry in bottles. g p y As regards the soils kept in pans (Chart III), it is evident that the diminution in the amount of both soluble organic matter and soluble inorganic matter in the more highly heated soils goes hand in hand with a diminution of their toxic action on the germination of seeds, the three curves for 0, 44 and 112 (or ]06) days occupying the same relative positions in all cases. S. U. PICKERING 1 in Chart II, without showing any such actual reduction, shows an irregularity tending in that direction. These results indicate that the decrease in the amount of soluble matter is common to all the soils, whatever the temperature of previous heating, but that it is only in the case of the more highly heated ones, which are rich in this oxidisable matter, that the decrease continues throughout any considerable length of time; where less of it is present, as with the less highly heated soils, the supply soon becomes exhausted, and another change, resulting in an increase of soluble matter, becomes predominant. It is probable that this other change, just like the oxidation, occurs equally in all the soils, for it is noticeable that, as regards the inorganic matter, at any rate, an increase eventually asserts itself with the highly heated soils, especially when the conditions are unfavourable for oxidation: the experiments in closed flasks (Fig. 2, Chart II) clearly show this, and there is a tendency in the same direction, even in the experiments in open pans, for the final portions of the curve for 112 days (Fig. 1, Chart II) are high in comparison with those for ten and 44 days. How far bacterial action may cooperate with atmospheric oxidation in reducing the amount of soluble matter in a heated soil when this is exposed to air is uncertain, but such a reduction does occur in the absence of all bacteria. Three samples of soil were heated in bottles to 127°: two of these were analysed and found to contain 0'7407 per cent, of soluble organic matter, whilst the third was kept in a very slow stream of dry air, under aseptic conditions, for eight weeks before analysis, and was then found to contain only 06478 per cent. The reduction in amount is not large, but the opportunities for oxidation would be very much more limited in this case, than in that of a sample of soil freely exposed to the air, and periodically moistened. 270 Changes in Heated Soils The soil was found to be perfectly sterile at the conclusion of the experiment. 1 loc. cit. in. 35. S. U. PICKERING The soil kept dry in bottles does not behave in quite so regular a manner, the analyses placing it above the curve 44, and the germination results, below it. Whilst it is evident that the soluble matter present in the highly heated soils is toxic towards germination, it would appear that that S. U. PICKERING 271 which is gradually formed in the less heated soils has no such toxic action: for with soils which have been kept for 44 and 112 days, the curves first dip (diminution of soluble matter) and then rise (increase in soluble matter), whereas the curves for the incubation periods rise CHART III. Soluble matter in soils kept in pans, and incubation periods. Relative values. 200 Fig. 1. Organic matter 100 •44 -112 .0 P.C 200 t. Fig.2. Inorganic matter 100 Tim 200 Fig.3. Incubation times 4 4 100 50° 100° 150° Temperature of heating CHART III. Soluble matter in soils kept in pans, and incubation periods. Relative values. 100 1 Temperature of heating heating 272 Changes in Heated Soils throughout1. This would be seen better by taking the results at 100° as the standard of comparison; the incubation curves pass in a uniform as the standard of comparison; the incubation curves pass in a uniform OHABT IV. Soluble matter in soils kept in flasks, and incubation periods. Relative values. 200 100 P.C 200 100 Tim 200 o.t Fig.J. Organic matter t. Fig. 2. Inorganic matter Fig. 3. Incubation times .44 /0 •112 4 4 '112 44 ,106 1 0 0 50° 100° 150" Temperature of heating 1 This, and various other facts, disprove the suggestions recently made by F. Fletche (The Cairo Scientific Journal, No. 43, Vol. iv. April 1910) that delayed germination is du to a decreased rate of inhibition, consequent on the increased amount, and not on the nature, of the soluble matter in the heated soil. OHABT IV. Soluble matter in soils kept in flasks, and incubation periods. Relative values. 100 Temperature of heating 1 This, and various other facts, disprove the suggestions recently made by F. Fletcher (The Cairo Scientific Journal, No. 43, Vol. iv. April 1910) that delayed germination is due to a decreased rate of inhibition, consequent on the increased amount, and not on the nature, of the soluble matter in the heated soil. 1 This, and various other facts, disprove the suggestions recently made by F. S. U. PICKERING Fletcher (The Cairo Scientific Journal, No. 43, Vol. iv. April 1910) that delayed germination is due to a decreased rate of inhibition, consequent on the increased amount, and not on the nature, of the soluble matter in the heated soil. S. U. PICKERING 273 direction through this point, whereas several of the analysis curves rise on both sides of it. It is specially noteworthy that the toxic effect of the heated soils has nearly entirely disappeared by the 106th day, the curve (Fig. 3, Chart III) being almost an horizontal line, whereas the soluble matter in the various soils has by no means been reduced to uniformity: evidently the toxic substance is only one of the many substances formed when the soil is heated; it may be destroyed, and yet leave in the soil much of the extra soluble matter formed by the heating. With the soils kept in closed flasks (Chart IV) there is a similar, though less exact, agreement between the analytical and germination results. The soluble organic matter in the soils heated to the highest temperature (Fig. 1), it is true, shows but little alteration with time, and the incubation periods (Fig. 3) do the same: but the three curves do not occupy quite the same relative position. It must be remem- bered, however, that the error of the germination results is considerable, and greater concordance could hardly be expected1. The incubation period for the soil heated to 125° shows a remarkable decrease on long keeping (106 days); and, on comparing this curve with those representing the soluble inorganic matter present (Fig. 2), a striking similarity between them is noticeable, showing that, in this case, the inorganic matter is a potent factor in the germination results. The germination experiments with the soils kept in open pans and in sealed flasks having been made at the same time, these may be compared together: this has been done in Chart V, the curves "Incubation" being the values for the flasks when those for the pans are represented by a horizontal line at 100. Corresponding curves for the soluble matter have been drawn in the same way, the values for all of these being given in Table VI. The figures for the more extended period (Fig. 1 The germination curves have been smoothed more than the others: the results for soil heated to 100° lie a good deal off the curves 0 and 106, and, as may be seen from Table Y, the values for this soil remained practically constant throughout. S. U. PICKERING 2) are of special interest in showing that it is the soluble matter in the highly heated soils which is the toxic substance, and not that which forms in the less heated soils on keeping, for the great rise in the soluble constituents from 100° to 150° is closely reproduced in the rise in the incubation period, whereas the great hump in the curve for the organic matter in the less heated soils produces only a very slight effect on the incubation curve. 18 Journ. of Agric. Sci. m 18 274 Changes in Heated Soils Whilst the present experiments support in every way the con- clusions previously drawn, that the toxic effect on germination is dependent on the increase in the amount of matter rendered soluble by heating the soil, the results do not show that direct proportionality between the increase in the incubation period and the increase in the CHAKT V. Behaviour of soil kept in closed flasks compared with that kept in pans represented by 100. 300 100 300 100 Fig. 1. After Fig.2. After 4-4 days 106-116 days * Organic 'Inorganic 'Incubation 'Organic 'Incubation Inorganic 50° 100° 150° Temperature of heating soluble organic matter, which was previously observed. This may well be due to differences in the composition of the soil taken in the two cases, and, as a matter of fact, judging by the extracts, the two soils were very different (cf. Table II above with Table V on p. 422, Vol. Ii). Also, the amount of soluble inorganic matter seems to have a distinct influence on the incubation period in some of the present experiments, as has already been pointed out. I i i h h iki i Fi 2 Ch IV h CHAKT V. Behaviour of soil kept in closed flasks compared with that kept in pans represented by 100. 300 100 300 100 Fig. 1. After Fig.2. After 4-4 days 106-116 days * Organic 'Inorganic 'Incubation 'Organic 'Incubation Inorganic 50° 100° 150° Temperature of heating CHAKT V. Behaviour of soil kept in closed flasks compared with that kept in pans represented by 100. soluble organic matter, which was previously observed. This may well be due to differences in the composition of the soil taken in the two cases, and, as a matter of fact, judging by the extracts, the two soils were very different (cf. Table II above with Table V on p. S. U. PICKERING 422, Vol. Ii). Also, the amount of soluble inorganic matter seems to have a distinct influence on the incubation period in some of the present experiments, as has already been pointed out. In connection with the most striking curve in Fig. 2, Chart IV, that for 112 days, it may be mentioned that the greasy scum which formed during the evaporation of the extracts in this case, was specially noticeable, both with the least heated and most heated soils, but was nearly absent with the intermediate ones, i.e. from those forming the S. U. PICKERING 275 lowest part of the graph. A lesser amount of greasiness was observed with the soils kept in flasks for 44 days, but only with those which had beeu heated least, and none was observed with these soils when kept for only ten days, nor with any of the soils kept in pans. The greasiness, therefore, is evidently connected with the substances formed gradually by processes other than oxidation. lowest part of the graph. A lesser amount of greasiness was observed with the soils kept in flasks for 44 days, but only with those which had beeu heated least, and none was observed with these soils when kept for only ten days, nor with any of the soils kept in pans. The greasiness, therefore, is evidently connected with the substances formed gradually by processes other than oxidation. The effect of light on the changes occurring in soils kept moist in sealed flasks was examined in the experiments quoted in Table IV. The differences in the soluble organic matter are all negative after 43 days, and positive after 116 days: they are probably significant, but are too small for useful discussion. The differences in the soluble inorganic matter are mostly larger, but are not regular in character. Some results with soils kept in closed flasks at an average tem- perature of 5°, instead of 15°, are entered in the lower part of the table—the values at 15° for 60 days are obtained by interpolation from the previously quoted results—but do not indicate any certain differences. S. U. PICKERING To ascertain whether there is any secular variation in the soluble matter in soils, a portion of ground in a cultivated orchard was protected from the weather by a glazed frame placed 18 inches above it, and samples of the soil were examined on about the tenth of each month throughout a year: at the same time some similar soil was kept in basins in an open shed, watered and stirred occasionally, and examined with the other. The results are given in Table VII. With TABLE VII. Soluble matter in soil.on different dates. TABLE VII. Soluble matter in soil.on different dates. Date Deo Jan. Feb. Marcl April May June July Angus Sept. Oct. Nov. t . Average... Soil in situ Organic •0026 •0483 •0308 •0366 •U417 •0520 •0526 •0688 •0744 •078U •0576 •0690 •0561 Inorganic •0895 •0587 •0541 •0490 •0618 •0470 •0579 •0749 •0748 •0820 0736 •1102 •0695 Soil kept Organic •0026 •0665 •0542 •0590 •0753 •0705 •0625 •0732 •0691 •0559 •0649 in basins Inorganic •0895 •U952 •1027 •0867 •0844 •0860 •O'J2O •0978 •1028 •0728 •0910 18—2 Changes in Heated Soils 276 the soil in basins there appears to be no variation in any definite direction, whilst, with that left in situ, there is a decrease, both in organic and inorganic matter, throughout the earlier part of the year and an increase subsequently: this, however, may be due merely to the rainfall acting through the exposed soil surrounding the protected patch. SUMMARY. The water extracts obtainable from soils are of constant composition as regards organic matter when the time allowed for the extraction varies from 20 to 320 minutes, the temperature from 7° to 23°, and the proportions from 5 to 10 of soil to 100 of water. The inorganic matter is not affected by the time, but is by the temperature and proportions. The increase in soluble matter produced by heating a soil, and the accompanying toxic qualities towards the germination of seeds in it, is gradually reduced by exposing these soils in a moist condition to the air, even under aseptic conditions, but is not reduced, when the soils are kept moist in the absence of air. The destruction of the toxic substance is probably, therefore, due to oxidation. Unheated soils, or soils heated only to a low temperature, exhibit on keeping an increase in soluble matter; this occurs whether air is admitted or not, and this change, therefore, is probably not an oxida- tion process: the substance formed, moreover, in such cases appears to have little or no toxic action on germination. This increase of soluble matter, due to the formation of a non-toxic substance, is preceded by a preliminary diminution of soluble matter, precisely similar to the diminution of toxic matter occurring continuously in the more highly heated soils: such toxic matter, therefore, appears to be present in all soils, whether heated or not, though, in the latter case, it is present in such small quantities that it soon becomes com- pletely oxidised. Air-dried soils, heated and unheated, when kept for some months show an appreciable reduction in soluble constituents, and also in toxic properties.(where such properties were originally present), closely similar to the reduction exhibited by moist soils kept in air for about ten days.
https://openalex.org/W2805897036	https://europepmc.org/articles/pmc5975416?pdf=render	English	null	Prevalence and risk factors of testicular microlithiasis in patients with hypospadias: a retrospective study	BMC pediatrics	2,018	cc-by	4,841	Prevalence and risk factors of testicular microlithiasis in patients with hypospadias: a retrospective study Michiko Nakamura1, Kimihiko Moriya1* , Yoko Nishimura1, Mutsumi Nishida2,3, Yusuke Kudo2,3, Yukiko Kanno1, Takeya Kitta1, Masafumi Kon1 and Nobuo Shinohara1 * Correspondence: k-moriya@med.hokudai.ac.jp p Full list of author information is available at the end of the article Abstract Background: It has been described that the incidence of testicular microlithiasis is high in several congenital disorders which may be associated with testicular impairment and infertility. Several reports have shown that a prepubertal or pubertal hormonal abnormality in the pituitary-gonadal axis was identified in some patients with hypospadias that is one of the most common disorders of sex development. However, exact prevalence or risk factors of testicular microlithiasis in patients with hypospadias have not reported so far. In the present study, to clarify the prevalence and risk factors of testicular microlithiasis in patients with hypospadias, a retrospective chart review was performed. Methods: Children with hypospadias who underwent testicular ultrasonography between January 2010 and April 2016 were enrolled in the present study. Severity of hypospadias was divided into mild and severe. The prevalence and risk factors of testicular microlithiasis or classic testicular microlithiasis were examined. Results: Of 121 children, mild and severe hypospadias were identified in 66 and 55, respectively. Sixteen children had undescended testis. Median age at ultrasonography evaluation was 1.7 years old. Testicular microlithiasis and classic testicular microlithiasis were documented in 17 children (14.0%) and 8 (6.6%), respectively. Logistic regression analysis revealed that presence of undescended testis was only a significant factor for testicular microlithiasis and classic testicular microlithiasis. The prevalence of testicular microlithiasis or classic testicular microlithiasis was significantly higher in children with undescended testis compared to those without undescended testis (testicular microlithiasis; 43.8% versus 9.5% (p = 0.002), classic testicular microlithiasis; 37.5% versus 1.9% (p < 0.001). Conclusions: The current study demonstrated that the presence of undescended testis was only a significant risk factor for testicular microlithiasis or classic testicular microlithiasis in patients with hypospadias. As co-existing undescended testis has been reported as a risk factor for testicular dysfunction among patients with hypospadias, the current findings suggest that testicular microlithiasis in children with hypospadias may be associated with impaired testicular function. Conversely, patients with isolated HS seem to have lower risks for testicular impairment. Further investigation with longer follow-up will be needed to clarify these findings. Keywords: Testicular microlithiasis, Hypospadias, Undescended testis, Ultrasonography * Correspondence: k-moriya@med.hokudai.ac.jp y 1Department of Renal and Genitourinary Surgery, Hokkaido University Graduate School of Medicine, North-15, West-7, Kita-Ku, Sapporo 060-8638, Japan Nakamura et al. BMC Pediatrics (2018) 18:179 https://doi.org/10.1186/s12887-018-1151-6 Nakamura et al. BMC Pediatrics (2018) 18:179 https://doi.org/10.1186/s12887-018-1151-6 Methods Medical charts of children who visited our hospital for the management or follow-up of HS between January 2010 and April 2016 were retrospectively reviewed. Among them, patients who were born between December 1999 and August 2015, and who underwent on testicular US were included in the present study. Patients with obvious disorders of sex development or with chromosomal abnormalities were excluded. To evaluate and define risk factors of TM in children with HS, the following parameters were assessed with respect to their relation to the prevalence of TM: birth weight, presence of UDT, severity of HS, testosterone adminis- tration before HS surgery, and age at US. Severity of HS was divided into mild and severe based on the necessity of transecting urethral plate for correction of chordee deformity according to Koyanagi et al. [22]. Testicular microlithiasis (TM) is characterized by mul- tiple, small, uniform-appearing echogenic foci of less than 3 mm without acoustic shadowing in the seminifer- ous tubules, which may be indicative of degeneration of the testicular parenchyma [11, 12]. Several theories about the origin or causes of TM have been reported, however, the exact etiology of TM still remains unclear [13]. Previous studies have reported an association between TM and testicular germ cell tumors and/or car- cinoma in situ [14, 15]. In addition, an association be- tween TM and infertility has been reported [15]. Although real impact of TM in children is still a matter of debate, it has been described in previous reports that the incidence of TM is high in some congenital disor- ders, such as UDT, Down’s syndrome, Klinefelter syn- drome, McCune-Albright syndrome and Peuzt-Jeghers syndrome, which may be associated with testicular im- pairment and infertility [16–20]. TM was defined as 1 or more foci measuring 1 to 3 mm in diameter on testicular US. TM was classified as limited TM (LTM, echogenic foci < 5 /field) or classic TM (CTM, echogenic foci ≥5 /field) as reported by Goede et al. [23] (Fig. 1). Among patients with TM, children with CTM in at least one testis were diagnosed with CTM, whereas others were classified as having LTM. All US evaluation was performed without sedation by sonographers. Background demonstrated that patients with associated genital anomaly, including UDT, were at higher risk for impaired testicular function [6, 9, 21], we speculated that TM may be identified at a higher rate in such patients. However, the exact preva- lence of and TM in patients with HS has not been reported so far. Hypospadias (HS) is one of the most common disorders of sex development, occurring in 0.52 to 8.2 of every 1000 live male births [1, 2]. Although the exact etiology of HS is unknown in the majority of patients, a multifactorial eti- ology including genetic, endocrine and environmental fac- tors is considered to be involved in the genesis of this disorder [3]. HS is also considered as one of the symptoms of testicular dysgenesis syndrome (TDS), which was pro- posed in 2001 [4]. It has been speculated that impaired de- velopment of fetal testes could lead to increased risk of undescended testis (UDT), HS, decreased spermatogenesis and testicular cancer [5]. Several reports have shown that a prepubertal or pubertal hormonal abnormality in the pituitary-gonadal axis was identified in some patients with HS from endocrinological point of view [6–10], which is compatible with the concept of TDS. In the present study, we retrospectively examined the prevalence and the risk factors of TM among children with HS. © The Author(s). 2018 Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated. Nakamura et al. BMC Pediatrics (2018) 18:179 Page 2 of 6 Page 2 of 6 Page 2 of 6 Side of TM and UDT (Table 5) In 11 children with bilateral UDT, unilateral and bilat- eral TM were observed in 1 and 4, respectively. Among them, CTM in at least one side was identified in 4. Of 5 with unilateral UDT, TM was identified in 2 (both with bilateral CTM). Among 105 children without UDT, uni- lateral and bilateral TM were observed in 6 and 4, re- spectively. Of those, CTM in at least one side was observed only in 2. Discussion To our knowledge, the current study represented the first report on the prevalence of TM in children with HS. TM and CTM were identified in 14.0 and 6.6%, re- spectively, of children with HS. Presence of UDT was only a risk factor for TM and CTM. y Previous studies have shown that TM is associated with several conditions, including impaired spermato- genesis, testicular cancer and carcinoma in situ [14, 15]. In asymptomatic adults, the rate of CTM varies from 0.6 to 9% [24, 25]. Recent studies revealed that the preva- lence of TM and CTM in asymptomatic boys was 4.2 and 2.4% respectively, and increased with age [23]. The prevalence of TM (14.0%) and CTM (6.6%) in children with HS in the current study seems to be relatively higher compared with that in asymptomatic boys in the previous reports. Although the etiology of HS is consid- ered to be multifactorial, the concept of TDS, which suggests that impaired development of fetal testes could lead to increased risk of HS, has been proposed as one of the causes of HS. Drut et al. proposed that TM may be related with Sertoli cell dysfunction and abnormal embryogenesis during the early stages of testicular devel- opment [12]. Wohlfahrt-Veje et al. reported that dysge- netic testes often have an irregular US pattern in which TM may also be visible [5]. The reason for the relatively high prevalence of TM may be due to that children who have such embryological causes of testis anomaly could have been included in the present study. Prevalence of TM and CTM TM and CTM were documented in 17 children (14.0%) and in 8 children (6.6%), respectively. TM was identified unilaterally in 7, and bilaterally in 10. Among 8 children with CTM, bilateral CTM was identified in 5. In the remaining 3 children, unilateral CTM alone was detected in 1, and CTM on one side and LTM on the other side in 2. No testicular tumors were detected on US in any children. Results Among 219 children who visited during the study period, 121 children (55.3%) were included in the current study. Patient characteristics are shown in Table 1. Median birth weight was 2456 g, and the number of children with low birth weight (less than 2500 g) was 62. Mild and severe HS were identified in 66 and 55 children, respectively. UDT was observed in 16 patients (bilateral in 11, unilateral in 5). Of those, 4 had mild HS, and 12 had severe HS. Regarding co-existing congenital anomaly, inguinal hernia and heart anomaly were observed in 13 and 8, respectively. Testos- terone administration before HS surgery was performed in 94 children. Median age at testicular US was 1.7 years old. Methods US evaluations were performed using a PLT-1204BT, a linear probe, 7.2–18 MHz equipped with Aplio™XG/500 (Toshiba Medical Systems Corp., Tochigi, Japan), a EUP-65, a linear probe 6–14 MHz equipped with HI VISION Avius (Hitachi- Medical, Tokyo, Japan), and a ML6–15, a linear probe 6–15 MHz equipped with Logiq Based on these previous reports, TM may be a sign of a future endocrinological abnormality in the pituitary-gonadal axis [6–10] or testicular malignancy among patients with HS as a phenotype of TDS. Therefore, we performed tes- ticular ultrasonography (US) for the screening of testicular abnormalities in patients with HS. Because several reports b a Fig. 1 Representative pictures of LTM and CTM. a Limited testicular microlithiasis (arrows) in an 11-year-old boy with hypospadias and undescended testis. Ultrasonography of right testis showed 2 small, uniform-appearing echogenic foci without acoustic shadowing. His left testis also displayed 4 echogenic foci per field. b Classic testicular microlithiasis in a 1-year-old boy with hypospadias. Ultrasonography of left testis demonstrated more than 5 echogenic foci per field. Right testis also displayed more than 5 echogenic foci per field b b Fig. 1 Representative pictures of LTM and CTM. a Limited testicular microlithiasis (arrows) in an 11-year-old boy with hypospadias and undescended testis. Ultrasonography of right testis showed 2 small, uniform-appearing echogenic foci without acoustic shadowing. His left testis also displayed 4 echogenic foci per field. b Classic testicular microlithiasis in a 1-year-old boy with hypospadias. Ultrasonography of left testis demonstrated more than 5 echogenic foci per field. Right testis also displayed more than 5 echogenic foci per field Nakamura et al. BMC Pediatrics (2018) 18:179 Page 3 of 6 Page 3 of 6 factor for TM (p = 0.006) (Table 2). Regarding CTM, pres- ence of UDT was only a risk factor on univariate analysis (p < 0.001) (Table 3). The incidence of TM or CTM was significantly higher in children with UDT compared with those without UDT (TM; 43.8% versus 9.5% (p = 0.002), CTM; 37.5% versus 1.9% (p < 0.001)) (Table 4). E9 (GE Healthcare, Amersham, UK). If multiple examina- tions were performed during follow-up, the final assess- ment findings were evaluated for prevalence and risk factors of TM. JMP®pro version 12 was used for all statistical analyses. Statistical analyses were performed using logistic regres- sion analysis and Fisher’s exact test. P < 0.05 was consid- ered significant. Risk factors of TM and CTM Although the concept of TDS included symptoms of HS, UDT and testicular cancer [5], as far as we know, there is no re- port demonstrating that the prevalence of testicular tu- mors is higher in patients with HS. Although UDT is well-known as a risk factor for testicular malignancy [32], it is obscure whether UDT is also a risk factor for testicu- lar malignancy among patients with HS. Longer follow-up will clarify the exact associations among testicular malig- nancy, TM and/or UDT in children with HS. [6, 21]. A number of reports have demonstrated the relationship between TM and impaired spermatogenesis [26–28], although this issue is still controversial [29]. Accordingly, TM in children with HS may be associated with decreased testicular function and/or impaired sperm- atogenesis. To determine the relationship between TM and testicular function/spermatogenesis in patients with HS, further follow-up with endocrinological evaluations until puberty is necessary. p y y While the prevalence of TM in patients with HS and without UDT (9.5%) in the current study was slightly higher compared to that in asymptomatic boys (4.2%) reported in the previous literature [23], the prevalence of CTM (1.9%) was almost similar to that in asymptom- atic boys (2.4%). Accordingly, patients with isolated HS seem to have lower risks for testicular impairment. On the contrary, although the presence of UDT in patients with HS was demonstrated as a risk factor for TM in the current study or impaired semen quality in the previous report [21], there has been no comparative study focus- ing on risk of TM or testicular dysfunction between pa- tients with isolated UDT and those with HS and UDT. To clarify the impact of HS in patients with UDT in terms of the risk of TM or testicular dysfunction, additional studies are necessary. There is some controversy regarding the method and duration of follow-up in patients with TM. In the guide- lines produced by the European Society of Urogenital Radiology, the consensus opinion is that the presence of TM alone in the absence of other risk factors is not an indication for regular follow-up in adults [19]. However, this guideline did not mentioned children with HS. Risk factors of TM and CTM Univariate analysis demonstrated that the severe type of HS and presence of UDT were risk factors for TM. On multivariate analyses, presence of UDT was only a risk Table 1 Patient characteristics n = 121 range Birth weight (g, median ± SD) 2456 ± 834 (unknown 1) (472–4048) Low birth weight (< 2500 g) (pts) 62 Type of HS (pts) mild: 66 / severe: 55 UDT (pts) 16 (unilateral: 5 / bilateral: 11) Testosterone administration before surgery (pts) yes: 94, no: 20 (unknown 7) topical: 67 / systemic: 10 / topical+systemic: 17 Age at USG (yrs, median ± SD) 1.7 ± 4.0 (0.5–18.2) TM (pts) (LTM / CTM) 17 (9 / 8) Table 1 Patient characteristics We demonstrated that the presence of UDT was only a risk factor for TM and CTM. There are several reports of prepubertal or pubertal hormonal abnormalities of the pituitary-gonadal axis in some patients with HS [6–10]. In addition, there are several reports on patients with both HS and UDT, which is a risk factor of TM and CTM as demonstrated in our study, who were at a higher risk for decreased testicular function or impaired spermatogenesis Nakamura et al. BMC Pediatrics (2018) 18:179 Page 4 of 6 Table 2 Risk factors for TM Univariate analysis Multivariate analysis Odds ratio (95% confidence interval) p value Odds ratio (95% confidence interval) p-value Birth weight (g, median ± SD) 0.44 (0.04–4.09) n.s. Low birth weight (< 2500 g) 0.84 (0.68–4.09) n.s. Severe type of HS 3.40 (1.17–11.35) 0.024* 2.48 (0.79–8.65) n.s. UDT 7.39 (2.23–24.55) 0.001* 5.8 (1.68–19.95) 0.006* Testosterone administration 1.57 (0.39–10.59) n.s. Age at USG 0.28 (0.04–2.52) n.s. p < 0.05 TM in the asymptomatic population ranged from 2.4 to 5.6% [24, 25], which is much higher than the prevalence of the lifetime risk of testicular cancer in the general population. Nowadays, it is recognized that TM in adults without known risk factors, such as previous testicular cancer, a history of UDT or testicular atrophy, seems to be a benign condition [19, 24, 25, 30]. Regarding TM de- tected in childhood, Suominen et al. found 15 patients with neoplasms among 421 pediatric patients (3.6%) by systematic review [31]. They described that TM should be considered a benign condition even in the pediatric age group, but the fact that TM is associated with tes- ticular malignancy (< 5%) cannot be ignored. Competing interests Competing interests The authors declare that they have no competing interests. Competing interests The authors declare that they have no competing interests. References 1. Levitt SB, Reda EF. Hypospadias. Pediatr Ann. 1988;17(1):48–9. 53-44, 57 2. Springer A, van den Heijkant M, Baumann S. Worldwide prevalence of hypospadias. J Pediatr Urol. 2016;12(3):152.e151–7. 1. Levitt SB, Reda EF. Hypospadias. Pediatr Ann. 1988;17(1):48–9. 53-44, 57 2. Springer A, van den Heijkant M, Baumann S. Worldwide prevalence of hypospadias. J Pediatr Urol. 2016;12(3):152.e151–7. 3. Marrocco G, Grammatico P, Vallasciani S, Gulia C, Zangari A, Marrocco F, Bateni ZH, Porrello A, Piergentili R. Environmental, parental and gestational factors that influence the occurrence of hypospadias in male patients. J Pediatr Urol. 2015;11(1):12–9. 4. Skakkebaek NE, Rajpert-De Meyts E, Main KM. Testicular dysgenesis syndrome: an increasingly common developmental disorder with environmental aspects. Hum Reprod. 2001;16(5):972–8. Abbreviations CTM Cl i Abbreviations CTM: Classic testicular microlithiasis; HS: Hypospadias; LTM: Limited testicular microlithiasis; TDS: Testicular dysgenesis syndrome; TM: Testicular microlithiasis; UDT: Undescended testis; USG: Ultrasonography Abbreviations CTM: Classic testicular microlithiasis; HS: Hypospadias; LTM: Limited testicular microlithiasis; TDS: Testicular dysgenesis syndrome; TM: Testicular microlithiasis; UDT: Undescended testis; USG: Ultrasonography 12. Drut R, Drut RM. Testicular microlithiasis: histologic and immunohistochemical findings in 11 pediatric cases. Pediatr Dev Pathol. 2002;5(6):544–50. Publisher’s Note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. Availability of data and materials 13. Shanmugasundaram R, Singh JC, Kekre NS. Testicular microlithiasis: is there an agreed protocol? Indian J Urol. 2007;23(3):234–9. 13. Shanmugasundaram R, Singh JC, Kekre NS. Testicular microlithiasis: is there an agreed protocol? Indian J Urol. 2007;23(3):234–9. Availability of data and materials The datasets used and/or analyzed during the current study available from the corresponding author on reasonable request. The datasets used and/or analyzed during the current study available from the corresponding author on reasonable request. 14. Meissner A, Mamoulakis C, de la Rosette JJ, Pes MP. Clinical update on testicular microlithiasis. Curr Opin Urol. 2009;19(6):615–8. 14. Meissner A, Mamoulakis C, de la Rosette JJ, Pes MP. Clinical update on testicular microlithiasis. Curr Opin Urol. 2009;19(6):615–8. 15. Sakamoto H, Shichizyou T, Saito K, Okumura T, Ogawa Y, Yoshida H, Kushima M. Testicular microlithiasis identified ultrasonographically in Japanese adult patients: prevalence and associated conditions. Urology. 2006;68(3):636–41. Conclusions 5. Wohlfahrt-Veje C, Main KM, Skakkebaek NE. Testicular dysgenesis syndrome: foetal origin of adult reproductive problems. Clin Endocrinol. 2009;71(4):459–65 TM and CTM were identified in roughly 14.0 and 6.6% of children with HS, respectively. The prevalence of TM and CTM was significantly higher in patients with UDT. As UDT among children with HS has been reported as a risk factor for endocrinological abnormality and/or impaired spermatogenesis, these findings suggest that TM in chil- dren with HS may be associated with impaired testicular function. In addition, the prevalence of CTM in patients with isolated HS was almost equal to the previously re- ported prevalence in asymptomatic boys. Therefore, pa- tients with isolated HS seem to have lower risks for testicular impairment. Further investigation with longer follow-up will be needed to clarify these findings. 6. Moriya K, Mitsui T, Tanaka H, Nakamura M, Nonomura K. Long-term outcome of pituitary-gonadal axis and gonadal growth in patients with hypospadias at puberty. J Urol. 2010;184(4 Suppl):1610–4. 7. Iwatsuki S, Kojima Y, Mizuno K, Kamisawa H, Umemoto Y, Sasaki S, Kohri K, Hayashi Y. Endocrine assessment of prepubertal boys with a history of cryptorchidism and/or hypospadias: a pilot study. J Urol. 2011;185(6 Suppl): 2444–50. 8. Nonomura K, Fujieda K, Sakakibara N, Terasawa K, Matsuno T, Matsuura N, Koyanagi T. Pituitary and gonadal function in prepubertal boys with hypospadias. J Urol. 1984;132(3):595–8. 9. Rey RA, Codner E, Iniguez G, Bedecarras P, Trigo R, Okuma C, Gottlieb S, Bergada I, Campo SM, Cassorla FG. Low risk of impaired testicular Sertoli and Leydig cell functions in boys with isolated hypospadias. J Clin Endocrinol Metab. 2005;90(11):6035–40. 10. Shima H, Ikoma F, Yabumoto H, Mori M, Satoh Y, Terakawa T, Fukuchi M. Gonadotropin and testosterone response in prepubertal boys with hypospadias. J Urol. 1986;135(3):539–42. 10. Shima H, Ikoma F, Yabumoto H, Mori M, Satoh Y, Terakawa T, Fukuchi M. Gonadotropin and testosterone response in prepubertal boys with hypospadias. J Urol. 1986;135(3):539–42. 11. Backus ML, Mack LA, Middleton WD, King BF, Winter TC, True LD. Testicular microlithiasis: imaging appearances and pathologic correlation. Radiol. 1994; 192(3):781–5. 11. Backus ML, Mack LA, Middleton WD, King BF, Winter TC, True LD. Testicular microlithiasis: imaging appearances and pathologic correlation. Radiol. 1994; 192(3):781–5. Risk factors of TM and CTM At this time, we believe that the follow-up protocol for pa- tients with HS and TM should be determined based on the presence or absence of UDT because the exact risk for testicular malignancy in patients with HS alone re- mains obscure. Previous studies reported that the prevalence of pri- mary testicular tumors in patients with TM ranged from 15 to 45% [15, 29]. Thus, there was some concern that TM may lead to testicular cancer at the end of the 1990’s. However, two studies revealed that the rate of Table 3 Risk factors for CTM Univariate analysis Odds ratio (95% confidence interval) p-value Birth weight (g, median ± SD) 4.26 (0.20–85.70) n.s. Low birth weight (< 2500 g) 1.61 (0.38–8.14) n.s. Severe type of HS 3.92 (0.86–27.53) n.s. UDT 30.90 (6.23–231.37) < 0.001 Age at USG 0.19 (0.01–3.62) n.s. *p < 0.05 Several limitations of the present study should be ad- dressed. First, this study was conducted in a retrospective Several limitations of the present study should be ad- dressed. First, this study was conducted in a retrospective Table 4 Prevalence of TM and CTM UDT (+) UDT (−) p-value Prevalence of TM 43.8% (7/16) 9.5% (10/105) 0.002 Prevalence of CTM 37.5% (6/16) 1.9% (2/105) < 0.001 Nakamura et al. BMC Pediatrics (2018) 18:179 Page 5 of 6 Page 5 of 6 Table 5 Sides of testicular microlithiasis and undescended testis Bilateral UDT Unilateral UDT Without UDT n = 11 n = 5 n = 105 Unilateral TM 1 0 6 CTM 1 0 0 LTM 0 0 6 Bilateral TM 4 2 4 Bilateral CTM: 2 2 1 CTM and LTM 1 0 1 Bilateral LTM 1 0 2 Table 5 Sides of testicular microlithiasis and undescended testis Bilateral UDT Unilateral UDT Without UDT n = 11 n = 5 n = 105 Table 5 Sides of testicular microlithiasis and undescended testis Ethics approval and consent to participate Ethics approval and consent to participate The present study was approved by the Institutional Review Board of Hokkaido University Hospital and the necessity to obtain written consent was waived because of its retrospective nature. Administrative permissions were received from the Institutional Review Board of Hokkaido University Hospital for accessing medical records in this study. (Approval number: 016–0095). Author details 1 1Department of Renal and Genitourinary Surgery, Hokkaido University Graduate School of Medicine, North-15, West-7, Kita-Ku, Sapporo 060-8638, Japan. 2Diagnostic Center for Sonography, Hokkaido University Hospital, Sapporo, Japan. 3Division of Laboratory and Transfusion Medicine, Hokkaido University Hospital, Sapporo, Japan. nature and relatively small number of children. Second, there was no control group such as Japanese boys who had no genital disease or isolated UDT. Third, evaluation of chromosomal abnormalities was not performed in all children. Fourth, as children included in the current study were relatively young and because TM sometimes appears later in childhood [23], the true prevalence of TM in pa- tients with HS may be higher than the prevalence in this study. Fifth, as endocrinological examination or semen analysis was not performed in the current study, testicular function could not be compared between patients with and without TM. nature and relatively small number of children. Second, there was no control group such as Japanese boys who had no genital disease or isolated UDT. Third, evaluation of chromosomal abnormalities was not performed in all children. Fourth, as children included in the current study were relatively young and because TM sometimes appears later in childhood [23], the true prevalence of TM in pa- tients with HS may be higher than the prevalence in this study. Fifth, as endocrinological examination or semen analysis was not performed in the current study, testicular function could not be compared between patients with and without TM. Received: 10 September 2017 Accepted: 18 May 2018 Abbreviations CTM: Classic testicular microlithiasis; HS: Hypospadias; LTM: Limited testicular microlithiasis; TDS: Testicular dysgenesis syndrome; TM: Testicular microlithiasis; UDT: Undescended testis; USG: Ultrasonography Nakamura et al. BMC Pediatrics (2018) 18:179 Nakamura et al. BMC Pediatrics (2018) 18:179 17. Rocher L, Moya L, Correas JM, Mutuon P, Ferlicot S, Young J, Izard V, Benoit G, Brailly-Tabard S, Bellin MF. Testis ultrasound in Klinefelter syndrome infertile men: making the diagnosis and avoiding inappropriate management. Abdom Radiol (NY). 2016;41(8):1596–603. 18. Wasniewska M, Matarazzo P, Weber G, Russo G, Zampolli M, Salzano G, Zirilli G, Bertelloni S. Function ISGfAoGaP: clinical presentation of McCune-Albright syndrome in males. J Pediatr Endocrinol Metab. 2006;19(Suppl 2):619–22. 19. Richenberg J, Belfield J, Ramchandani P, Rocher L, Freeman S, Tsili AC, Cuthbert F, Studniarek M, Bertolotto M, Turgut AT, et al. Testicular microlithiasis imaging and follow-up: guidelines of the ESUR scrotal imaging subcommittee. Eur Radiol. 2015;25(2):323–30. 20. Nishimura Y, Moriya K, Nakamura M, Nishida M, Sato M, Kudo Y, Omotehara S, Iwai T, Wakabayashi Y, Kanno Y, et al. Prevalence and chronological changes of testicular Microlithiasis in isolated congenital undescended testes operated on at less than 3 years of age. Urology. 2017;109:159–64. 21. Asklund C, Jensen TK, Main KM, Sobotka T, Skakkebaek NE, Jorgensen N. Semen quality, reproductive hormones and fertility of men operated for hypospadias. Int J Androl. 2010;33(1):80–7. y 22. Koyanagi T, Nonomura K, Yamashita T, Kanagawa K, Kakizaki H. One-stage repair of hypospadias: is there no simple method universally applicable to all types of hypospadias? J Urol. 1994;152(4):1232–7. 23. Goede J, Hack WW, Van der Voort-Doedens LM, Sijstermans K, Pierik FH. Prevalence of testicular microlithiasis in asymptomatic males 0 to 19 years old. J Urol. 2009;182(4):1516–20. 24. Peterson AC, Bauman JM, Light DE, McMann LP, Costabile RA. The prevalence of testicular microlithiasis in an asymptomatic population of men 18 to 35 years old. J Urol. 2001;166(6):2061–4. 25. Serter S, Gümüş B, Unlü M, Tunçyürek O, Tarhan S, Ayyildiz V, Pabuscu Y. Prevalence of testicular microlithiasis in an asymptomatic population. Scand J Urol Nephrol. 2006;40(3):212–4. 26. Aizenstein RI, DiDomenico D, Wilbur AC, O'Neil HK. Testicular microlithiasis: association with male infertility. J Clin Ultrasound. 1998;26(4):195–8. 27. Thomas K, Wood SJ, Thompson AJ, Pilling D, Lewis-Jones DI. The incidence and significance of testicular microlithiasis in a subfertile population. Br J Radiol. 2000;73(869):494–7. 28. Xu C, Liu M, Zhang FF, Liu JL, Jiang XZ, Teng JB, Xuan XJ, Ma JL. The association between testicular microlithiasis and semen parameters in Chinese adult men with fertility intention: experience of 226 cases. Urology. 2014;84(4):815–20. 29. Miller FN, Sidhu PS. Does testicular microlithiasis matter? Authors’ contributions MNa collected and analyzed the data and draft the manuscript. KM conceived the study and edit the manuscript. YN collected the data and edit the manuscript. MNi analyzed the data and edit the manuscript. YKu analyzed the data. YKa, TK and MK edit the manuscript. NS supervised this work. All authors have read and approved the final manuscript. MNa collected and analyzed the data and draft the manuscript. KM conceived the study and edit the manuscript. YN collected the data and edit the manuscript. MNi analyzed the data and edit the manuscript. YKu analyzed the data. YKa, TK and MK edit the manuscript. NS supervised this work. All authors have read and approved the final manuscript. 16. Goede J, Weijerman ME, Broers CJ, de Winter JP, Van der Voort-Doedens LM, Hack WW. Testicular volume and testicular microlithiasis in boys with Down syndrome. J Urol. 2012;187(3):1012–7. Page 6 of 6 Nakamura et al. BMC Pediatrics (2018) 18:179 A review. Clin Radiol. 2002;57(10):883–90. 30. Richenberg J, Brejt N. Testicular microlithiasis: is there a need for surveillance in the absence of other risk factors? Eur Radiol. 2012;22(11):2540–6. 31. Suominen JS, Jawaid WB, Losty PD. Testicular microlithiasis and associated testicular malignancies in childhood: a systematic review. Pediatr Blood Cancer. 2015;62(3):385–8. 32. Møller H, Cortes D, Engholm G, Thorup J. Risk of testicular cancer with cryptorchidism and with testicular biopsy: cohort study. BMJ. 1998; 317(7160):729.
https://openalex.org/W2991338921	https://www.biodiversitylibrary.org/itempdf/274801	Latin	null	Bahamas, 1907, [numbers 5354-6370]	null	1,907	public-domain	372	. W: JgTt>Cj, / ft. Il£~&JjL. di/tuo-ty & &£&•&■ / C> UL# lUiM+uyyf " rf~ I - /, 5 7m^o/C yUM. m ■ luJjAAf’ &r, J M, fJjHv Lu4^ &4-14: ^£1+ # ‘ &4LC J~/0 <UM,% J* hAm» %>$k4a Mb dt&Mj £u¥(Ms £ol> - >«^ ^ <% ., ^ i^iJo £/M\,Hj&aL - / '••'"• '-•' -'-^ t/ ^?/. /$y- 4ri? , , uAjlbL f-m .^itAm.pl, (tki£kMu 1 V kTJ^X £tdf&iM> &+t. ! n^'- —$’ - " JJi/-, fUf'1/st4j4 (ypv 4tjL . /4>uWk/ ^•- . ■ ' >: ff?/‘ .'Av'v'o fs&c* § w, JyfA nffy7. toliMi,odo^u4 -Vh,£jL mi 4^ s&oh ■ $$. MW- *$?& rftck- <&? i fay < ^U/ tfMiyrMiMptAA'lGu 2®+ Vi 0 &MAJ&&O ^fpL ^>- ■ ' > ;. ^6 tt Q£&oe/wuy y •( 7r, ^6~f8¥- 7" %&> LHtm. S"f S’S’, Ifa eMi4 - . \4TQ%b, '^eb/C, isjjL H G^b. tt Q£&oe/wuy y •( 7r, ^6~f8¥- 7"* %&> LHtm. S"f S’S’, Ifa eMi4 - . \4TQ%b, '^eb/C, isjjL H G^b. , dfou. dui. - %dt£, ’^tdo 2-2-, K$-&tf£tf t/feo t- t. i o £y IstCi . (MCjd V </fi 0 £-i/( Ify- (J-L . r t^e»L>\f. / 4> Jgsoab. /Ultoo WvJudm* 4jt-7. duMffU 11 " ioi'%, ^ ( JLf. &M^U4m 4gQW. l&jMMslM* ^4 U^U^Cc Is-wu, tuA L : )oihil. du cJjJ^d i£, iA //* 4 . ^ .' ‘ _ ' • * ^ - ' ! CdACt ifH ,:-^ft,d I » v / j y js* -S r // & [pO 14 /i &JM+ l J^oj, ju,l£ i^OJ^iL* ^^/(j 14MUyOL [pO 14 /i &JM+ l J^oj, ju,l£ i^OJ^iL* ^^/(j 14MUyOL JS < L I A cbuLgb ^fc£f b^k c^ . v^iU^ ) & * ^ - ft-C H ~ £ '/uuvjvsk !M'~ A 4Tmo-. Z$'kUm^J " ’ ^^4} 1/^isTioC 1 J aIh4. \«u- ■i^ iz&bu,. Tj.^ /;V^ 'y . j %$}; UtocU. - »W«» S//Tj [/''■Lo - J?ui „ ^ku4. W^eyr-' - UifL I <JO vJ$ fc ^'JY'- yutsti, JJoe° yLtSf ii^^lL»AAM^ vjpJUo V » H 4"1 (t iff tu ^ Mi. " ; iikktz%**+ $j$. fefc ■- _ • ' !lx&. C. ft. {Ijti Lrtj(ZU> Il£~&JjL. di/tuo-ty KJsX" & &£&•&■ / "Vt&ju<LC> KjUL# r conservation work performei 1987-1988. NY State Conservation/Preservation Discretionary Grant. Q. cons 1987 Con Disc
https://openalex.org/W4394691675	https://link.springer.com/content/pdf/10.1007/s12672-024-00963-9.pdf	English	null	Hypoxic microenvironment-induced exosomes confer temozolomide resistance in glioma through transfer of pyruvate kinase M2	Discover Oncology	2,024	cc-by	10,177	Discover Oncology Discover Oncology Discover Oncology Hypoxic microenvironment‑induced exosomes confer temozolomide resistance in glioma through transfer of pyruvate kinase M2 Received: 20 January 2024 / Accepted: 30 March 2024 © The Author(s) 2024 OPEN Keywords Glioma · Hypoxia · Exosome · Pyruvate kinase M2 · Temozolomide · Resistance Abstract bst act Objective Glioma, a malignant primary brain tumor, is notorious for its high incidence rate. However, the clinical applica- tion of temozolomide (TMZ) as a treatment option for glioma is often limited due to resistance, which has been linked to hypoxic glioma cell-released exosomes. In light of this, the present study aimed to investigate the role of exosomal pyruvate kinase M2 (PKM2) in glioma cells that exhibit resistance to TMZ. Methods Sensitive and TMZ-resistant glioma cells were subjected to either a normoxic or hypoxic environment, and the growth patterns and enzymatic activity of glycolysis enzymes were subsequently measured. From these cells, exosomal PKM2 was isolated and the subsequent effect on TMZ resistance was examined and characterized, with a particular focus on understanding the relevant mechanisms. Furthermore, the intercellular communication between hypoxic resistant cells and tumor-associated macrophages (TAMs) via exosomal PKM2 was also assessed. Results The adverse impact of hypoxic microenvironments on TMZ resistance in glioma cells was identified and charac- terized. Among the three glycolysis enzymes that were examined, PKM2 was found to be a critical mediator in hypoxia- triggered TMZ resistance. Upregulation of PKM2 was found to exacerbate the hypoxia-mediated TMZ resistance. Exosomal PKM2 were identified and isolated from hypoxic TMZ-resistant glioma cells, and were found to be responsible for trans- mitting TMZ resistance to sensitive glioma cells. The exosomal PKM2 also contributed towards mitigating TMZ-induced apoptosis in sensitive glioma cells, while also causing intracellular ROS accumulation. Additionally, hypoxic resistant cells also released exosomal PKM2, which facilitated TMZ resistance in tumor-associated macrophages. Conclusion In the hypoxic microenvironment glioma cells become resistant to TMZ due to the delivery of PKM2 by p g Conclusion In the hypoxic microenvironment, glioma cells become resistant to TMZ due to the delivery of PKM2 by exosomes. Targeted modulation of exosomal PKM2 may be a promising strategy for overcoming TMZ resistance in glioma. Supplementary Information The online version contains supplementary material available at https://doi.org/10.1007/s12672-024- 00963-9. * Changbin Shi, changbinshi@hotmail.com; Guofu Li, liguofu1012@hrbmu.edu.cn; Ziyu Xiong, rokaix@126.com; Ying Li, yingli2019@ hotmail.com; Cong Yan, 15004632954@163.com; Yingying Cheng, cskcyy@163.com; Yuwen Wang, eagleg7@163.com; Jingwei Li, 1019485398@qq.com; Zifeng Dai, zifengdai@hotmail.com; Dongdong Zhang, ddongzhang20@gmail.com; Wenzhong Du, dwz820827@126.com; Chunyang Men, chunymen@126.com \| 1Department of Neurosurgery, Harbin Medical University Cancer Hospital, Harbin, Heilongjiang, China. 2Department of Neurosurgery, The First Affiliated Hospital of Harbin Medical University, Harbin, Heilongjiang, China. 3Department of Ultrasound, Harbin Medical University Cancer Hospital, Harbin, Heilongjiang, China. Supplementary Information The online version contains supplementary material available at https://doi.org/10.1007/s12672-024- 00963-9. Discover Oncology (2024) 15:110 \| https://doi.org/10.1007/s12672-024-00963-9 * Changbin Shi, changbinshi@hotmail.com; Guofu Li, liguofu1012@hrbmu.edu.cn; Ziyu Xiong, rokaix@126.com; Ying Li, yingli2019@ hotmail.com; Cong Yan, 15004632954@163.com; Yingying Cheng, cskcyy@163.com; Yuwen Wang, eagleg7@163.com; Jingwei Li, 1019485398@qq.com; Zifeng Dai, zifengdai@hotmail.com; Dongdong Zhang, ddongzhang20@gmail.com; Wenzhong Du, dwz820827@126.com; Chunyang Men, chunymen@126.com \| 1Department of Neurosurgery, Harbin Medical University Cancer Hospital, Harbin, Heilongjiang, China. 2Department of Neurosurgery, The First Affiliated Hospital of Harbin Medical University, Harbin, Heilongjiang China. 3Department of Ultrasound, Harbin Medical University Cancer Hospital, Harbin, Heilongjiang, China. 1 Introduction Glioma is the most prevalent primary malignant tumor in the brain, accounting for approximately 78% of all malignant cases. It arises from glial cells [1]. Based on the 2021 WHO classification of tumors of the central nervous system, gliomas are clas- sified as different subtypes, including diffuse astrocytoma, polymorphous low-grade neuroepithelial tumor of the young, diffuse low-grade glioma, diffuse hemispheric glioma, diffuse pediatric-type high-grade glioma, and high-grade astrocytoma [2]. The incidence rate of glioma varies from 1.9 to 9.6 per 100,000 individuals among adults, which is influenced by age, gender, ethnicity, and geographical location [3]. The development of glioma is primarily associated with several risk factors, including advanced age, a family history of the condition, and exposure to radiation. Standard treatment for glioma typically involves a safe surgical resection, followed by the administration of radiotherapy and TMZ chemotherapy [4]. TMZ is an oral alkylating drug with easy penetration through the blood-brain barrier, commonly used as a first-line treatment for glioma [5–7]. However, around 50% of glioma patients experience resistance to TMZ and suffer relapse after a progression-free survival period of 7–10 months [8]. Moreover, many patients encounter severe side effects, such as bone marrow inhibition and female infertility [8]. However, the potential mechanisms involved in TMZ resistance remained largely unclear. Therefore, further elucidation of the mechanisms underlying TMZ resistance in glioma is imperative. Gliomas are characterized by pseudopalisading necrosis, a pathological feature commonly seen near collapsed blood ves- sels, surrounded by cells that survived in hypoxic regions [9]. Hypoxia promotes the Warburg effect, an abnormal glycolytic process resulting in increased glucose uptake and lactate production [10]. This process provides biosynthesis precursors of nucleotides, amino acids, and lipids essential for supporting rapid tumor growth [11]. Pyruvate kinase (PK) is a critical enzyme in glycolysis, catalyzing the final rate-limiting step by transferring phosphate to generate pyruvate and adenosine triphosphate from phosphoenolpyruvate and adenosine diphosphate [12]. Among the four human PK isoforms (M1, M2, L, and R), PKM2 is abundantly expressed in glioma cells but is marginally expressed in healthy brain tissues [13]. TMZ-induced DNA damage is linked to PKM2 because of its influence on pyruvate metabolism [14], though the role of PKM2 in TMZ resist- ance remains obscure. Exosomes, small membrane vesicles, are secreted by specific cell types. Constituting a vital element of the tumor microenvironment, exosomes encapsulate diverse biomolecules, including mRNAs, non-coding RNAs, and proteins. Abstract * Changbin Shi, changbinshi@hotmail.com; Guofu Li, liguofu1012@hrbmu.edu.cn; Ziyu Xiong, rokaix@126.com; Ying Li, yingli2019@ hotmail.com; Cong Yan, 15004632954@163.com; Yingying Cheng, cskcyy@163.com; Yuwen Wang, eagleg7@163.com; Jingwei Li, 1019485398@qq.com; Zifeng Dai, zifengdai@hotmail.com; Dongdong Zhang, ddongzhang20@gmail.com; Wenzhong Du, dwz820827@126.com; Chunyang Men, chunymen@126.com \| 1Department of Neurosurgery, Harbin Medical University Cancer Hospital, Harbin, Heilongjiang, China. 2Department of Neurosurgery, The First Affiliated Hospital of Harbin Medical University, Harbin, Heilongjiang, China. 3Department of Ultrasound, Harbin Medical University Cancer Hospital, Harbin, Heilongjiang, China. \| https://doi.org/10.1007/s12672-024-00963-9 \| https://doi.org/10.1007/s12672-024-00963-9 Discover Oncology (2024) 15:110 Discover Oncology (2024) 15:110 \| https://doi.org/10.1007/s12672-024-00963-9 Discover Oncology (2024) 15:110 Research Abbreviations TMZ Temozolomide PKM2 Pyruvate kinase M2 BBB Blood–brain barrier FBS Fetal bovine serum CCK-8 Cell Counting Kit-8 OE-PKM2 Overexpressing PKM2 ROS Reactive oxygen species TAMs Tumor-associated macrophages 1 Introduction They play a pivotal role in intercellular communication by transmitting their contents [15]. Research has shown that hypoxia modifies the release and contents of exosomes, impacting intercellular communication in glioma [16]. In particular, exosomal circ_0072083 induced by the Warburg effect enhances TMZ resistance in glioma [16], and hypoxic glioma-released exosomes containing microRNA-1246 induce M2 macrophage polarization [17]. Exosomes delivered by hypoxic glioma stem cells containing Linc01060 promote glioma progression [18]. All of these research indicates that exosomes have an significant relationship with gliomas, and it is very possible that exosomes are involved in the process of resistance to TMZ in gliomas. While there is limited evidence suggesting the release of PKM2-containing exosomes by tumor cells, its specific role in TMZ-resistant glioma remains unclear. This study predominantly centers on exploring exosomes transported by the hypoxic microenvironment, elucidating their role in conferring TMZ resistance to glioma cells through the delivery of PKM2. This investigation aims to unveil a novel mechanism of TMZ resistance in glioma, proposing a potential metabolic blocker for effective anti-glioma therapy. Vol:.(1234567890) Discover Oncology (2024) 15:110 \| https://doi.org/10.1007/s12672-024-00963-9 Research Discover Oncology Discover Oncology (2024) 15:110 2.5 Glucose uptake assay 2-Deoxyglucose uptake assay reagent (Sigma-Aldrich, USA) was performed. Specifically, cells were inoculated into six confluent plates at a cell density of 7 × 104 cells per tool. Cells were washed with PBS and incubated for 1 h in a glucose-free incubator, then treated with 1 mm 2-deoxyglucose for 20 min. Glucose uptake was determined by measuring absorbance at 540 nm. 2.3 Cell viability assay Cell viability experiments were performed using the Cell Counting Kit-8 (CCK-8; Solarbio, China). For each well of a 96-well plate, a uniform seeding of 2 × 103 cells was carried out. Following cell adhesion, 10 μL of CCK-8 reagent was added to each well. The cells were then incubated in a standard cell culture incubator for 2 h, allowing the CCK-8 reagent to interact with cellular dehydrogenases and formazan crystals to develop. After the incubation period, the absorbance value of the formazan product was measured at 450 nm using a microplate reader. This wavelength was selected as it corresponds to the peak absorbance of the formazan product. These steps were consistently applied across all experimental conditions to ensure the reliability and reproducibility of the cell viability assessments. 2.1 Cell culture and administration U251 cells (Procell, Wuhan, China) were used to contain 10% fetal bovine serum (FBS; HyClone, USA) and 1% penicil- lin/streptomycin in Dulbecco’s modified eagle medium (DMEM) at 37 °C. U251/TR resistant cell lines are obtained by exposing U251 cells to increasing concentrations of TMZ until cells are resistant to TMZ at a concentration of 50 μg/ mL [16]. These cells are typically cultured in a normoxic (21% O2) or hypoxic (1% O2) environment. 2.4 Colony formation assay The ability of the cells to form colonies was assessed by a colony formation assay. 500 cells were inoculated in each well of a 6-well plate separately and cultured for 10 days. The resulting colonies were then fixed and stained with 0.5% crystal violet (Sigma-Aldrich, USA), then images of colony formation were taken and the number of colonies was quantified. 2.2 Tumor‑associated macrophages (TAMs) induction U937cells (Procell, Wuhan, China) are inoculated into 10-cm culture dishes (5 × 10 5 cells/well) with RPMI-1640 (10% FBS, 100 ng/mL of PMA). After 24 h of culture, cells were washed with phosphate-buffered saline (PBS, pH 7.2) and re- culture with 10%-FBS RPMI, 20 ng/ml IL-4 (Gibco, Thermo) were added for co-incubation. Then cells were washed with PBS and incubated with serum-free 1640 for 24 h. After 24 h serum starvation, cells were collected for M2-polarized TAMs use. Cell morphology was observing for successful induction demonstration which the cell morphology was changed from round shape to long spindle shape. Some of the cells having “tentacles” protruding out, consistent with differentiation into macrophages. Exosomes at a concentration of 1 μg/mL were added to the medium of recipient cells according to the established protocol. 2.6 Lactate production assay Cells were seeded at a density of 5 × 103 cells per well in 96-well plates to measure lactate production. Human lactate concentration kit (Nanjing Jiancheng Bioengineering Institute,China) was used for calculating lactate concentration. Vol.:(0123456789) Discover Oncology (2024) 15:110 \| https://doi.org/10.1007/s12672-024-00963-9 Research Discover Oncology (2024) 15:110 2.10 Plasmids and virus infection The PKM2-CopGFP fusion gene (GenePharma, China) was used to achieve overexpression of PKM2 (OE-PKM2). A complete PKM2-CopGFP series was inserted into the Lv5 lentiviral vector. At the same time, we adopted PKM2 targeting (shPKM2; GenePharma, China) to silence PKM2. The sequences targeted were: scramble shRNA, AATGCACGCTCAGCACAAGC; PKM2- shRNA1, GTTCGGAGGTTTGATGAAATC; PKM2-shRNA2, GCTGTGGCTCTAGACACTAAA. After production and purification, lenti- viral particles are stored at − 80 °C. fusion gene (GenePharma, China) was used to achieve overexpression of PKM2 (OE-PKM2). A complete PKM2-CopGFP series was inserted into the Lv5 lentiviral vector. At the same time, we adopted PKM2 targeting (shPKM2; GenePharma, China) to silence PKM2. After production and purification, lentiviral particles are stored at − 80 °C. 2.11 Isolation of tumor‑derived exosomes Following cell collection and centrifugation, the medium containing serum was aspirated, and the cells were washed twice with PBS. Subsequently, the cells were incubated with serum-free medium for 48 h. Dead cells and cell debris were then removed by centrifugation at 300 g, 2000 g and 10,000 g at 4 °C for 10 and 30 min. The resulting supernatant was filtered through a 0.22 μm filter and centrifuged at 120,000 g for 70 min at 4 °C. After discarding the supernatant, the pellet was suspended in PBS to obtain exosomes. 2.9 Immunofluorescence After 20 min immobilization with 4% paraformaldehyde (Sigma-Aldrich, USA), cells were permeated with 0.5% Triton x-100 (Sigma-Aldrich, USA) for 10 min. Rinse the solution, then seal the cells with 0.1% BSA prepared by PBS for 30 min. Subse- quently, anti-pkm2 antibody (1/100; Ab85555; Abcam, USA) cells are incubated overnight at 4 °C. Next, the cell is coupled to Alexa Fluor® 594 with secondary resistance (1/200; Ab150080; Abcam) are incubated together and finally with DAPI (ab104139; Abcam) dye. Cell fluorescence images were taken with a fluorescence microscope (Leica, Germany). 2.7 RT‑qPCR Trizol (invitrogen, USA) was used to extract a common RNA from cells with subsequent reverse transcription using applied biosystems (USA) reverse transcription set. RT-PCR was performed with the quantitative PCR mx-3000p (Agi- lent, USA), where primers are specific to the following target genes: PKM2 (Forward 5′-ATGTCGAAGCCCCATAGTGAA-3′, Reverse 5′-TGGGTGGTGAATCAATGTCCA-3′), HK2 (Forward 5′-GAGCCACCACTCACCCTACT-3′, Reverse 5′-CCAGGCATT CGGCAATGTG-3′), LDHA (Forward 5′-ATGGCAACTCTAAAGGATCAGC-3′, Reverse 5′-CCAACCCCAACAACTGTAATCT-3′), and β-actin (Forward 5′-CATGTACGTTGCTATCCAGGC-3′, Reverse 5′-CTCCTTAATGTCACGCACGAT-3′). In our study, we employed the 2-ΔΔCT method for relative quantification of gene expression. This method involves calculating the difference in cycle threshold (CT) values between the target gene and the reference gene, using the control group as a baseline. Subsequently, it calculates the relative expression levels, providing a reliable analytical framework for our experiments that takes into account the influences of experimental errors and amplification efficiency. 2.8 Immunoblotting Lysis buffer was determined by radioimmunoprecipitation and protein concentration by the bovine serum albumin method. Proteins were isolated from the lysate by SDS-PAGE and transferred to a PVDF membrane. After containment, specific pri- mary resistance to PKM2 (1/1000; ab85555; Abcam, USA), HK2 (1/1000; Ab133691), LDHA (1/5000; Ab52488), CD36 (1/1000; Ab133625), TSG101 (1/1000; Ab133586), PARP (1/1000; Ab32071), Bax (1/1000; Ab32503), bcl-2 (1/2000; Ab182858) or beta- actin (1/200; Ab115777) for incubation followed by peroxidase-coupled secondary resistance for incubation. Final detection of protein traces using an improved chemiluminescence detection system (Pierce, USA). Vol:.(1234567890) 2.16 Statistical analysis Data are expressed as mean ± standard deviation. Statistical analysis was performed using GraphPad Prism software (version 9.0.1). Half of the maximum inhibitory concentration of TMZ (CI50) was determined and two groups were com- pared using Student’s t-test, while three or more groups were compared using one-way or two-way ANOVA. Differences in p-values less than 0.05 were considered statistically significant. 2.12 Exosome uptake assay We use the red fluorescent dye PKH26 (Sigma, USA) for fluorescent labeling of exosomes. Initially, resuspension was performed using 100 μL of exogenous diluent C and mixing was performed by adding 100 μ of LPKH26 dye reagent. Vol:.(1234567890) Discover Oncology (2024) 15:110 \| https://doi.org/10.1007/s12672-024-00963-9 Discover Oncology (2024) 15:110 Discover Oncology Research After 3 min of incubation, 200 μL of serum was added and washed with PBS. Finally, the exosome was cocultured with the recipient cell and photographed. 2.14 Intracellular reactive oxygen species (ROS) measurement ROS levels were monitored using a fluorescent determination of 2’, 7’-dichloroflucéine diacetate (dcfh-da), according to the specifications provided by the manufacturer (D6470; Solarbio, China). 2.13 Flow cytometry To assess apoptotic cells, the Accuri C6 flow cytometer (BD, USA) was employed, following labeling with Annexin V-FITC and propidium iodide (PI) (CA1020; Solarbio, China). Cells were harvested, washed with PBS, and suspended in a bind- ing buffer for Annexin V-FITC and PI staining, as per the manufacturer’s instructions. After incubation, the labeled cell suspension underwent flow cytometry analysis using appropriate laser and filter settings to detect Annexin V-FITC and PI fluorescence. Data were analyzed using FlowJo software to quantify apoptotic and necrotic cell populations. 2.15 Transwell Cell invasion experiments were conducted using a Matrigel-coated invasion chamber (BD, USA). Cells were inoculated into the upper chamber of the device, and a medium containing 20% FBS was added to the lower chamber. After 24 h of culture, cells that had penetrated the bottom surface through the filter were fixed and stained with 0.5% crystal violet. 3.1 The hypoxic microenvironment deteriorates TMZ resistance in glioma cells In this study, we aimed to investigate the impact of hypoxic conditions on TMZ resistance in glioma cells. To accomplish this, we established a TMZ-resistant U251 cell line (U251/TR) and found that it showed higher resistance to TMZ compared to the parental U251 cells, as evidenced by IC50 curves (Fig. 1A, B). We then investigated the effect of normoxic and hypoxic conditions on the proliferative capacity of U251 cells but did not observe any significant differences between the two conditions (Fig. 1C, D). Interestingly, however, colony formation assays revealed a significant increase in the number of colonies formed by U251/TR cells under hypoxic conditions compared to those under normoxic conditions (Fig. 1E, F), implying an association between hypoxia and TMZ resistance in glioma cells. Furthermore, we observed that U251/TR cells had enhanced hypoxic tolerance compared with U251 cells, suggesting that hypoxia is common in TMZ-resistant glioma tumors (Fig. 1G). Notably, when U251 and U251/TR cells were treated with differing doses of TMZ under normoxic or hypoxic conditions for 48 h, the hypoxic microenvironment exacerbated the resistance of U251/TR cells to TMZ but did not have any effect on the response of U251 cells to the drug (Fig. 1H, I). Taken together, our results suggest that the hypoxic microenvironment contributes to the development of TMZ resistance in glioma cells. 3.2 PKM2 is a crucial mediator in hypoxia‑triggered TMZ resistance of glioma cells There is evidence suggesting that the Warburg effect, which is characterized by increased glucose uptake and lactate production, contributes to TMZ resistance in glioma cells [16]. To delve deeper into this phenomenon, we assessed the Warburg effect in U251 cells exposed to normoxia for 48 h and U251/TR cells subjected to either normoxic or hypoxic conditions for the same duration. Our results revealed that U251/TR cells exhibited significantly higher glucose uptake Vol.:(0123456789) .:(0123456789) Discover Oncology (2024) 15:110 \| https://doi.org/10.1007/s12672-024-00963-9 \| https://doi.org/10.1007/s12672-024-00963-9 Discover Oncology (2024) 15:110 Research Fig. 1 Hypoxia deteriorates TMZ resistance in glioma cells. A, B IC50 curves of U251 and U251/TR cells exposed to distinct doses of TMZ for 48 h. C–F Number of colonies of U251 and U251/TR cells with normoxic and hypoxic conditions. G Cell viability of U251 and U251/TR cells with hypoxic conditions for distinct time periods. H, I Cell viability of U251 and U251/TR cells with normoxic and hypoxic conditions and administrated with distinct doses of TMZ for 48 h. Ns: no significance; p < 0.05; p < 0.01; p < 0.001; and *p < 0.0001 Fig. 1 Hypoxia deteriorates TMZ resistance in glioma cells. A, B IC50 curves of U251 and U251/TR cells exposed to distinct doses of TMZ for 48 h. C–F Number of colonies of U251 and U251/TR cells with normoxic and hypoxic conditions. G Cell viability of U251 and U251/TR cells with hypoxic conditions for distinct time periods. H, I Cell viability of U251 and U251/TR cells with normoxic and hypoxic conditions and administrated with distinct doses of TMZ for 48 h. Ns: no significance; p < 0.05; p < 0.01; p < 0.001; and **p < 0.0001 Vol:.(1234567890) Discover Oncology (2024) 15:110 \| https://doi.org/10.1007/s12672-024-00963-9 Discover Oncology (2024) 15:110 Research and lactate production compared to U251 cells under normoxic conditions (Fig. 2A, B). Furthermore, exposure to hypoxia intensified these metabolic changes in U251/TR cells. We then measured the expression levels of glycolytic enzymes, including PKM2, HK2, and LDHA. The results showed that PKM2 expression was notably increased in U251/TR cells compared to U251 cells under normoxia (Fig. 2C–I). This upregulation was further exacerbated by exposure to hypoxia. However, no significant changes were observed in the expression levels of HK2 and LDHA in either normoxic U251 cells or normoxic/hypoxic U251/TR cells. 3.2 PKM2 is a crucial mediator in hypoxia‑triggered TMZ resistance of glioma cells Additionally, we found that PKM2 activity was higher in U251/TR cells than in U251 cells, and this difference was further intensified by hypoxia (Fig. 2J, K), indicating that PKM2 may play a critical role in hypoxia-induced TMZ resistance in glioma cells. 3.5 Exosomal PKM2 released from hypoxic resistant glioma cells transmits TMZ resistance to sensitive glioma cells In this study, we explored the impact of exosomal PKM2 released by TMZ-resistant glioma cells on the sensitivity of TMZ in susceptible glioma cells. Exosomes derived from U251/TR cells markedly heightened the resistance of U251 cells to TMZ (Fig. 5A–C), which was further enhanced by exosomes from hypoxic U251/TR cells. However, treatment with U251 cell-derived exosomes had no effect on U251 cell growth. Furthermore, only exosomes derived from U251/TR cells and hypoxic U251/TR cells could increase PKM2 protein levels, but not mRNA (Fig. 5D–F). Exosomes derived from U251 cells did not affect PKM2 expression. We also examined whether exosomal PKM2 induced TMZ resistance in sensitive glioma cells. The activity of TMZ in U251 cells was evaluated, and exosomes were collected from hypoxic shNC- or shPKM2- transfected U251/TR cells (Additional file 1: Figure S2). Our data demonstrated that U251 cells treated with exosomes from hypoxic shPKM2-transfected U251/TR cells showed improved response to TMZ (Fig. 5G). Taken together, exosomes released from hypoxic TMZ-resistant glioma cells transfer PKM2 and contribute to acquired TMZ resistance in sensitive glioma cells. 3.3 PKM2 exacerbates the hypoxic microenvironment‑mediated TMZ resistance of glioma cells We investigated whether PKM2 plays a role in hypoxia-induced resistance to TMZ in glioma cells. We overexpressed PKM2 in U251 cells by transfecting them with a PKM2 plasmid (Fig. 3A–C), and stably knocked out PKM2 in U251/TR cells using shPKM2 lentivirus transfection (Fig. 3D–F). Our results showed that PKM2 overexpression promoted the growth of U251 cells, whereas its suppression inhibited the survival of U251/TR cells (Fig. 3G–L). In a hypoxic microenvironment, overexpression of PKM2 worsened the resistance of U251 cells to different concentrations of TMZ (Fig. 3M). Conversely, PKM2 knockdown enhanced the response of U251/TR cells to TMZ treatment (Fig. 3N). Furthermore, a combined treat- ment of TMZ and a PKM2 inhibitor demonstrated a more pronounced therapeutic effect on the survival of U251/TR cells compared to the individual treatments of TMZ or the PKM2 inhibitor alone (Fig. 3O). These findings suggest that PKM2 plays a crucial role in exacerbating hypoxia-induced TMZ resistance in glioma cells. 3.4 Identification of exosomal PKM2 released from hypoxic TMZ‑resistant glioma cells We investigated the role of PKM2 in cell-to-cell communication and transmission of TMZ-resistance through exosomes in glioma. Exosomes were isolated from U251 cells, U251/TR cells, and hypoxic U251/TR cells using ultracentrifugation, which were confirmed to be enriched by exosomal markers CD63 and TSG101 (Fig. 4A), indicating successful isolation of exosomes (Additional file 1: Figure S1). Among the three glycolysis enzymes—PKM2, HK2, and LDHA—PKM2 exhibited higher expression in exosomes released from U251/TR cells than in those released from U251 cells, which was further increased under hypoxia (Fig. 4B). In contrast, no significant differences in HK2 and LDHA levels were observed among exosomes from U251 cells, U251/TR cells, and hypoxic U251/TR cells (Fig. 4C, D). These data suggest that hypoxic TMZ- resistant glioma cells release exosomal PKM2. Additionally, we verified that fluorescent-labeled PKH26 exosomes were effectively internalized by U251 cells (Fig. 4E), indicating that TMZ-sensitive glioma cells can effectively acquire exosomes released from TMZ-sensitive, resistant, and hypoxic glioma cells. 3.7 Hypoxic resistant cell‑derived exosomal PKM2 motivates intracellular ROS accumulation of sensitive glioma cells Previous studies have shown that reactive oxygen species (ROS) play a role in the resistance of glioma cells to TMZ [19]. In our study, we aimed to investigate whether exosomal PKM2 released from hypoxic resistant glioma cells affects intracellular ROS accumulation in sensitive glioma cells following TMZ treatment. We found that TMZ treatment led to a significant increase in intracellular ROS levels in U251 cells treated with exosomes derived from U251 or U251/TR cells, but not in those treated with exosomes from hypoxic U251/TR cells (Fig. 7A–D). Furthermore, in the presence of TMZ treatment, exosomes derived from hypoxic U251/TR cells demonstrated a reduction in intracellular ROS levels in U251 cells compared to those from U251 and U251/TR cells. However, the administration of exosomes from hypoxic U251/TR cells, which were transfected with PKM2, intensified ROS accumulation following TMZ treatment (Fig. 7E–G). Our find- ings indicate that delivery of exosomal PKM2 from hypoxic resistant glioma cells reduces intracellular ROS generation induced by TMZ in sensitive glioma cells. 3.8 Hypoxic resistant cells transmit exosomal PKM2 to tumor‑associated macrophages (TAMs) to facilitate TMZ resistance of glioma In this study, we examined the effects of different glioma cell-derived exosomes on TAMs. TAMs were co-cultured with exosomes obtained from U251 cells, U251/TR cells, or hypoxic U251/TR cells. Results showed that compared to exosomes derived from U251 cells, higher levels of PKM2 were observed in those from U251/TR cells, and this effect was further enhanced by hypoxic U251/TR cell-exosome treatment (Fig. 8A, B). Moreover, we found that TAMs could uptake exosomes released from U251 cells, U251/TR cells, or hypoxic U251/TR cells (Fig. 8C). TAMs that received exosomes from hypoxic U251/TR cells significantly increased the proliferative and invasive abilities of U251 cells following TMZ administration (Fig. 8D–G). These findings suggest that hypoxic resistant cells can transmit exosomal PKM2 to TAMs, consequently promoting TMZ resistance in glioma. 3.6 Hypoxic resistant glioma cells mitigate TMZ‑induced apoptosis via delivering exosomal PKM2 The results demonstrated a significant increase in apoptosis in U251 cells when treated with TMZ in combination with exosomes derived from U251 or U251/TR cells (Fig. 6A–D). However, hypoxic U251/TR cell-released exosome administration did not change apoptosis levels in U251 cells under TMZ treatment. Moreover, exosomes from hypoxic shPKM2-transfected U251/TR cells were able to restore the capacity for facilitating apoptosis of U251 cells when exposed to TMZ (Fig. 6E–G). Cleaved PARP/PARP and Bax/Bcl-2 levels in U251 cells were not altered by U251 cell-derived exosomes (Fig. 6H, I). In contrast, the levels of these proteins decreased as a result of exosomes released from U251/TR cells, and this reduction was further pronounced by exosomes from hypoxic U251/TR cells. Furthermore, the effects mentioned above were reversed upon PKM2 suppression (Fig. 6J, K). Overall, exosomal PKM2 released by hypoxic resistant glioma cells reduces TMZ-induced apoptosis. 3.6 Hypoxic resistant glioma cells mitigate TMZ‑induced apoptosis via delivering exosomal PKM2 In this study, we delved into the mechanisms through which exosomes originating from hypoxic TMZ-resistant glioma cells foster acquired TMZ resistance in glioma cells. We evaluated the extent of apoptosis in U251 cells when exposed Vol.:(0123456789) Research Discover Oncology (2024) 15:110 \| https://doi.org/10.1007/s12672-024-00963-9 Fig. 2 PKM2 is a crucial mediator in hypoxia-triggered TMZ resistance of glioma cells. A, B Glucose uptake and lactate production of U251 cells with normoxia exposure for 48 h, and U251/TR cells with normoxia or hypoxia exposure for 48 h. C–E RT-qPCR of PKM2, HK2, and LDHA mRNAs in above cells. F–I Immunoblotting of PKM2, HK2, and LDHA proteins in above cells. J, K Immunofluorescence of PKM2 in above cells. Bar, 10 μm. Ns: no significance; p < 0.05; p < 0.01; p < 0.001; and **p < 0.0001 ▸ https://doi.org/10.1007/s12672-024-00963-9 Discover Oncology (2024) 15:110 Discover Oncology Research Fig. 2 PKM2 is a crucial mediator in hypoxia-triggered TMZ resistance of glioma cells. A, B Glucose uptake and lactate production of U251 cells with normoxia exposure for 48 h, and U251/TR cells with normoxia or hypoxia exposure for 48 h. C–E RT-qPCR of PKM2, HK2, and LDHA mRNAs in above cells. F–I Immunoblotting of PKM2, HK2, and LDHA proteins in above cells. J, K Immunofluorescence of PKM2 in above cells. Bar, 10 μm. Ns: no significance; p < 0.05; p < 0.01; p < 0.001; and **p < 0.0001 ▸ to exosomes released from U251, U251/TR, or hypoxic U251/TR cells. The results demonstrated a significant increase in apoptosis in U251 cells when treated with TMZ in combination with exosomes derived from U251 or U251/TR cells (Fig. 6A–D). However, hypoxic U251/TR cell-released exosome administration did not change apoptosis levels in U251 cells under TMZ treatment. Moreover, exosomes from hypoxic shPKM2-transfected U251/TR cells were able to restore the capacity for facilitating apoptosis of U251 cells when exposed to TMZ (Fig. 6E–G). Cleaved PARP/PARP and Bax/Bcl-2 levels in U251 cells were not altered by U251 cell-derived exosomes (Fig. 6H, I). In contrast, the levels of these proteins decreased as a result of exosomes released from U251/TR cells, and this reduction was further pronounced by exosomes from hypoxic U251/TR cells. Furthermore, the effects mentioned above were reversed upon PKM2 suppression (Fig. 6J, K). Overall, exosomal PKM2 released by hypoxic resistant glioma cells reduces TMZ-induced apoptosis. to exosomes released from U251, U251/TR, or hypoxic U251/TR cells. 4 Discussion As the only FDA-approved oral agent for glioma therapy, TMZ plays a critical role as a DNA-alkylating agent which causes DNA double-strand breaks, cellular cycle arrest and ultimately cell death in glioma cells [20]. Despite its widespread use, TMZ remains limited in efficacy due to various factors such as extensive drug resistance, short blood-plasma half-life, rapid hydrolysis and imprecise tumor accumulation. Furthermore, over 50% of patients do not respond to TMZ, indicating that TMZ resistance remains a major challenge in glioma treatment [21]. Hypoxic condition deprives tissues of adequate oxygen supply, which is a main microenvironmental feature of glioma [22]. Hypoxia is a typical microenvironmental feature of glioma tissues characterized by insufficient oxygen supply. Hypoxic conditions are typically caused by rapid tumor cell proliferation, depleting available cellular nutrients and oxygen due to an inadequate blood supply. Hypoxic glioma cells show higher levels of aggressiveness, resistance to TMZ, and poorer clinical outcomes among glioma patients Vol:.(1234567890) Discover Oncology (2024) 15:110 \| https://doi.org/10.1007/s12672-024-00963-9 Research ol.:(0123456789) Research Discover Oncology (2024) 15:110 \| https://doi.org/10.1007/s12672-024-00963-9 [23, 24]. As a result, hypoxia represents a major barrier that significantly reduces the therapeutic effect of TMZ in glioma treatment [25]. Our experimental evidence suggests that the hypoxic microenvironment exacerbates TMZ resistance in Fig. 3 PKM2 exacerbates the hypoxic microenvironment-mediated TMZ resistance of glioma cells. A–C PKM2 mRNA and protein levels in U251 cells with vector or OE-PKM2 plasmid transfection. D–F PKM2 mRNA and protein levels in U251/TR cells with shNC or shPKM2 len- tivirus transfection. G, H Cell viability of U251 cells with vector or OE-PKM2 plasmid transfection, and U251/TR cells with shNC or shPKM2 lentivirus transfection. I–L Number of colonies of U251 cells with vector or OE-PKM2 plasmid transfection, and U251/TR cells with shNC or shPKM2 lentivirus transfection. M, N Cell viability of vector- or OE-PKM2 plasmid-transfected U251 cells as well as shNC- or shPKM2 lentivi- rus-transfected U251/TR cells administrated with distinct doses of TMZ in the hypoxic microenvironment. O Cell viability of U251/TR cells administrated with 30 μg/mL TMZ, 3 μM PKM2 inhibitor, or their combination for distinct time periods in the hypoxic microenvironment. Ns: no significance; p < 0.05; p < 0.01; p < 0.001; and **p < 0.0001 Discover Oncology (2024) 15:110 \| https://doi.org/10.1007/s12672-024-00963-9 Discover Oncology (2024) 15:110 Research Fig. 3 PKM2 exacerbates the hypoxic microenvironment-mediated TMZ resistance of glioma cells. 4 Discussion A–C PKM2 mRNA and protein levels in U251 cells with vector or OE-PKM2 plasmid transfection. D–F PKM2 mRNA and protein levels in U251/TR cells with shNC or shPKM2 len- tivirus transfection. G, H Cell viability of U251 cells with vector or OE-PKM2 plasmid transfection, and U251/TR cells with shNC or shPKM2 lentivirus transfection. I–L Number of colonies of U251 cells with vector or OE-PKM2 plasmid transfection, and U251/TR cells with shNC or shPKM2 lentivirus transfection. M, N Cell viability of vector- or OE-PKM2 plasmid-transfected U251 cells as well as shNC- or shPKM2 lentivi- rus-transfected U251/TR cells administrated with distinct doses of TMZ in the hypoxic microenvironment. O Cell viability of U251/TR cells administrated with 30 μg/mL TMZ, 3 μM PKM2 inhibitor, or their combination for distinct time periods in the hypoxic microenvironment. Ns: no significance; p < 0.05; p < 0.01; p < 0.001; and *p < 0.0001 [23, 24]. As a result, hypoxia represents a major barrier that significantly reduces the therapeutic effect of TMZ in glioma treatment [25]. Our experimental evidence suggests that the hypoxic microenvironment exacerbates TMZ resistance in glioma cells. Moreover, PKM2 is involved in this process. [23, 24]. As a result, hypoxia represents a major barrier that significantly reduces the therapeutic effect of TMZ in glioma treatment [25]. Our experimental evidence suggests that the hypoxic microenvironment exacerbates TMZ resistance in glioma cells. Moreover, PKM2 is involved in this process. Vol:.(1234567890) Discover Oncology (2024) 15:110 \| https://doi.org/10.1007/s12672-024-00963-9 Discover Oncology (2024) 15:110 \| https://doi.org/10.1007/s12672-024-00963-9 Research Fig. 4 Identification of exosomal PKM2 released from hypoxic TMZ-resistant glioma cells. A Immunoblotting of exosomal markers CD63 as well as TSG101 in exosomes isolated from U251 cells, U251/TR cells, and hypoxic U251/TR cells. B–D RT-qPCR of PKM2, HK2, and LDHA levels in above exosomes. E Fluorescence staining of U251 cells with treatment of red fluorescent dye PKH26-labeled exosomes from U251 cells, U251/TR cells, and hypoxic U251/TR cells for evaluating exosome uptake. Bar, 10 μm. Ns: no significance; and p < 0.0001 Research Discover Oncology (2024) 15:110 Fig. 4 Identification of exosomal PKM2 released from hypoxic TMZ-resistant glioma cells. A Immunoblotting of exosomal markers CD63 as well as TSG101 in exosomes isolated from U251 cells, U251/TR cells, and hypoxic U251/TR cells. B–D RT-qPCR of PKM2, HK2, and LDHA levels in above exosomes. 4 Discussion G Viability of U251 cells treated with PBS or exosomes isolated from hypoxic shNC- or shPKM2-transfected U251/TR cells. Ns: no significance; p < 0.01; and **p < 0.0001 ▸ Fig. 5 Exosomal PKM2 released from hypoxic resistant glioma cells transmits TMZ resistance to sensitive glioma cells. A Viability of U251 cells administrated with PBS or exosomes released from U251 cells, U251/TR cells, or hypoxic U251/TR cells, followed by TMZ treatment. B, C Number of colonies of U251 cells exposed to PBS or exosomes derived from U251 cells, U251/TR cells, or hypoxic U251/TR cells. D–F Expres- sion of PKM2 in U251 cells with above treatment. G Viability of U251 cells treated with PBS or exosomes isolated from hypoxic shNC- or shPKM2-transfected U251/TR cells. Ns: no significance; p < 0.01; and **p < 0.0001 ▸ Fig. 5 Exosomal PKM2 released from hypoxic resistant glioma cells transmits TMZ resistance to sensitive glioma cells. A Viability of U251 cells administrated with PBS or exosomes released from U251 cells, U251/TR cells, or hypoxic U251/TR cells, followed by TMZ treatment. B, C Number of colonies of U251 cells exposed to PBS or exosomes derived from U251 cells, U251/TR cells, or hypoxic U251/TR cells. D–F Expres- sion of PKM2 in U251 cells with above treatment. G Viability of U251 cells treated with PBS or exosomes isolated from hypoxic shNC- or shPKM2-transfected U251/TR cells. Ns: no significance; p < 0.01; and **p < 0.0001 ▸ has been demonstrated to induce the release of tumor cell-derived exosomes [30]. In colorectal carcinoma, exosome- mediated transmission of hsa_circ_0005963 has been found to enhance the glycolytic process and induce chemoresist- ance through post-transcriptional modulation of PKM2 axis [31]. Similarly, non-small cell lung carcinoma (NSCLC) cells that are resistant to cisplatin have been shown to transmit their resistance via exosomal PKM2 to sensitive cells [32]. NSCLC progression has also been revealed to accelerate due to exosomal circSHKBP1-induced glycolytic process involv- ing PKM2 [33]. The metabolic switch of liver nonparenchymal cells has been shown to be mediated by exosomal PKM2 in activated hepatic stellate cells [34]. Additionally, it was demonstrated that exosomal PKM2 transmitted in plasma reinforced growth and motility in esophageal squamous cell carcinoma cells [35]. In prostate carcinoma, bone metas- tases were found to be motivated by exosomal PKM2 transmission into bone stroma [36]. 4 Discussion E Fluorescence staining of U251 cells with treatment of red fluorescent dye PKH26-labeled exosomes from U251 cells, U251/TR cells, and hypoxic U251/TR cells for evaluating exosome uptake. Bar, 10 μm. Ns: no significance; and *p < 0.0001 siRNA-based gene silencing has emerged as a powerful therapeutic strategy in recent years [26, 27]. Synthetic nano- particles are currently the most commonly utilized siRNA delivery system, but their use is limited by several concerns such as toxicity, clearance issues, and complex synthesis methods [28]. Therefore, researchers have turned to endogenous nanomaterials like exosomes to enhance the efficacy and safety of siRNA delivery for glioma therapy. Exosomes offer unique advantages including autologous nature, immunosurveillance friendliness, multifunctional cargo delivery, and efficient BBB traversal capability [29]. In this study, we have uncovered the release of exosomal PKM2 from glioma cells exhibiting resistance to TMZ under hypoxic conditions. Our findings further indicate that the transmission of exosomal PKM2 from resistant glioma cells to sensitive counterparts can result in the acquisition of TMZ resistance and a reduction in apoptosis induced by this drug. Moreover, intracellular ROS accumulation in sensitive glioma cells is motivated by exosomal PKM2 released from resistant glioma cells, according to our results. As previously reported in literature, PKM2 (0123456789) Research Discover Oncology (2024) 15:110 \| https://doi.org/10.1007/s12672-024-00963-9 Fig. 5 Exosomal PKM2 released from hypoxic resistant glioma cells transmits TMZ resistance to sensitive glioma cells. A Viability of U251 cells administrated with PBS or exosomes released from U251 cells, U251/TR cells, or hypoxic U251/TR cells, followed by TMZ treatment. B, C Number of colonies of U251 cells exposed to PBS or exosomes derived from U251 cells, U251/TR cells, or hypoxic U251/TR cells. D–F Expres- sion of PKM2 in U251 cells with above treatment. G Viability of U251 cells treated with PBS or exosomes isolated from hypoxic shNC- or i ▸ https://doi.org/10.1007/s12672-024-00963-9 Research Discover Oncology (2024) 15:11 Discover Oncology Fig. 5 Exosomal PKM2 released from hypoxic resistant glioma cells transmits TMZ resistance to sensitive glioma cells. A Viability of U251 cells administrated with PBS or exosomes released from U251 cells, U251/TR cells, or hypoxic U251/TR cells, followed by TMZ treatment. B, C Number of colonies of U251 cells exposed to PBS or exosomes derived from U251 cells, U251/TR cells, or hypoxic U251/TR cells. D–F Expres- sion of PKM2 in U251 cells with above treatment. 4 Discussion Furthermore, exosomes have been shown to mediate the metabolic reprogramming of cancer-associated fibroblasts (CAFs), thereby contributing to improvement of the tumor microenvironment [37]. Lastly, gastric carcinoma cells have been shown to increase NF-κB signaling activation in CAFs through exosomal PKM2 secretion. This consequently motivated dysregulated metabolism and inflammation activation [38]. As a medical expert, it is crucial to acknowledge the pivotal role played by immune cells in the progression of glioma tumors. TAMs, as the most abundant immune cells, tend to accumulate preferentially in hypoxic areas and possess the potential to interact with tumor cells, leading to reprogramming [17]. In hypoxic glioma models, released exosomes are seen to induce the polarization of M2-like macrophages [22]. Additionally, studies suggest that exosomes containing Arginase-1 derived from reprogrammed TAMs may also facilitate glioma tumor progression [39]. Our current evidence indicates that exosomal PKM2 transmitted from hypoxic-resistant glioma cells can impact TAMs and enhance drug resistance to TMZ. Therefore, targeting PKM2 may offer a novel approach for treating TMZ-resistant glioma, laying the groundwork for future clinical applications. Collectively, these findings underscore the importance of comprehending the intercellular signaling mechanisms between tumors and immune cells in the development of effective treatment strategies for resistant glioma. This study has several study limitations. One potential limitation of this study lies in the specificity of the glioma cell lines used to investigate the role of exosomal PKM2in TMZ resistance. Gliomas are a heterogeneous group of tumors with diverse genetic backgrounds, and the findings of this study may not fully capture the complexity of TMZ resistance in all glioma subtypes. Future research should consider employing a broader range of glioma cell lines and patient-derived models to enhance the generalizability of the results. Another limitation is the predominantly in vitro nature of the experimental design. While in vitro studies provide valuable insights into cellular mechanisms, the artificial conditions may not fully replicate the intricate tumor microenvironment present in vivo. Future investigations incorporating in vivo models or clinical data from glioma patients would be beneficial to validate the relevance of exosomal PKM2-mediated TMZ resistance in a more physiologically representative setting. Vol:.(1234567890) Discover Oncology (2024) 15:110 \| https://doi.org/10.1007/s12672-024-00963-9 Discover Oncology (2024) 15:110 Research Discover Oncology Discover Oncology (2024) 15:110 \| https://doi.org/10.1007/s12672-024-00963-9 Research Vol.:(0123456789) Discover Oncology (2024) 15:110 \| https://doi.org/10.1007/s12672-024-00963-9 Discover Oncology (2024) 15:110 \| https://doi.org/10.1007/s12672-024-00963-9 \| https://doi.org/10.1007/s12672-024-00963-9 Discover Oncology (2024) 15:110 \| https://doi.org/10.1007/s12672-024-00963-9 Research esearch Discover Oncology (2024) 15:110 \| https://doi.org/10.1007/s12672-024-00963- g. 6 Hypoxic resistant glioma cells mitigate TMZ-induced apoptosis via delivering exosomal PKM2. A–D Apoptotic levels of U251 cel ministrated with exosomes isolated from U251 cells, U251/TR cells, or hypoxic U251/TR cells, followed by TMZ. E–G Apoptotic lev s of U251 cells treated with hypoxic shNC- or shPKM2-transfected U251/TR cells. H, I Cleaved PARP/PARP and Bax/Bcl-2 levels in U25 lls exposed to PBS, or exosomes from U251 cells, U251/TR cells, or hypoxic U251/TR cells. J, K Cleaved PARP/PARP and Bax/Bcl-2 levels i 251 cells administrated with exosomes derived from hypoxic shNC- or shPKM2-transfected U251/TR cells. Ns: no significance; p < 0.05 Fig. 6 Hypoxic resistant glioma cells mitigate TMZ-induced apoptosis via delivering exosomal PKM2. A–D Apoptotic levels of U251 cells administrated with exosomes isolated from U251 cells, U251/TR cells, or hypoxic U251/TR cells, followed by TMZ. E–G Apoptotic lev- els of U251 cells treated with hypoxic shNC- or shPKM2-transfected U251/TR cells. H, I Cleaved PARP/PARP and Bax/Bcl-2 levels in U251 cells exposed to PBS, or exosomes from U251 cells, U251/TR cells, or hypoxic U251/TR cells. J, K Cleaved PARP/PARP and Bax/Bcl-2 levels in U251 cells administrated with exosomes derived from hypoxic shNC- or shPKM2-transfected U251/TR cells. Ns: no significance; p < 0.05; p < 0.001; and p < 0.0001 Vol:.(1234567890) Discover Oncology (2024) 15:110 \| https://doi.org/10.1007/s12672-024-00963-9 Research Fig. 7 Hypoxic resistant cell-derived exosomal PKM2 motivates intracellular ROS accumulation of sensitive glioma cells. A–D ROS positive U251 cells treated with exosomes derived from U251 cells, U251/TR cells, or hypoxic U251/TR cells, followed by TMZ. Bar, 20 μm. E–G ROS positive U251 cells administrated with exosomes from hypoxic shNC- or shPKM2-transfected U251/TR cells. Bar, 20 μm. Ns: no significance; p < 0.0001 Discover Oncology (2024) 15:110 \| https://doi.org/10.1007/s12672-024-00963-9 Research Fig. 7 Hypoxic resistant cell-derived exosomal PKM2 motivates intracellular ROS accumulation of sensitive glioma cells. A–D ROS positive U251 cells treated with exosomes derived from U251 cells, U251/TR cells, or hypoxic U251/TR cells, followed by TMZ. Bar, 20 μm. E–G ROS positive U251 cells administrated with exosomes from hypoxic shNC- or shPKM2-transfected U251/TR cells. Bar, 20 μm. Ns: no significance; p < 0.0001 Fig. 7 Hypoxic resistant cell-derived exosomal PKM2 motivates intracellular ROS accumulation of sensitive glioma cells. A–D ROS positive U251 cells treated with exosomes derived from U251 cells, U251/TR cells, or hypoxic U251/TR cells, followed by TMZ. Bar, 20 μm. \| https://doi.org/10.1007/s12672-024-00963-9 E–G ROS positive U251 cells administrated with exosomes from hypoxic shNC- or shPKM2-transfected U251/TR cells. Bar, 20 μm. Ns: no significance; *p < 0.0001 Vol.:(0123456789) Vol.:(0123456789) Discover Oncology (2024) 15:110 \| https://doi.org/10.1007/s12672-024-00963-9 Discover Oncology (2024) 15:110 Research Research Discover Oncology (2024) 15:110 \| https://doi.org/10.1007/s12672-024-00963-9 Research Discover Oncology (2024) 15:110 \| https://doi.org/10.1007/s12672 024 00963 9 Fig. 8 Hypoxic resistant cells transmit exosomal PKM2 to TAMs to facilitate TMZ resistance of glioma. A, B PKM2 expression in TAMs co- cultured with PBS, exosomes from U251 cells, U251/TR cells, or hypoxic U251/TR cells. C Red fluorescent dye PKH26 staining in TAMs co- cultured with exosomes from U251 cells, U251/TR cells, and hypoxic U251/TR cells for assessing exosome uptake. Bar, 10 μm. D, E Number of colonies of U251 cells cocultured with TAMs that were pretreated with exosomes from U251 cells, U251/TR cells, and hypoxic U251/TR cells after TMZ administration. F, G Number of invasive cells under above treatment. Bar, 100 μm. Ns: no significance; p < 0.05; *p < 0.001; and *p < 0.0001 Fig. 8 Hypoxic resistant cells transmit exosomal PKM2 to TAMs to facilitate TMZ resistance of glioma. A, B PKM2 expression in TAMs co- cultured with PBS, exosomes from U251 cells, U251/TR cells, or hypoxic U251/TR cells. C Red fluorescent dye PKH26 staining in TAMs co- cultured with exosomes from U251 cells, U251/TR cells, and hypoxic U251/TR cells for assessing exosome uptake. Bar, 10 μm. D, E Number of colonies of U251 cells cocultured with TAMs that were pretreated with exosomes from U251 cells, U251/TR cells, and hypoxic U251/TR cells after TMZ administration. F, G Number of invasive cells under above treatment. Bar, 100 μm. Ns: no significance; p < 0.05; *p < 0.001; and *p < 0.0001 Fig. 8 Hypoxic resistant cells transmit exosomal PKM2 to TAMs to facilitate TMZ resistance of glioma. A, B PKM2 expression in TAMs co- cultured with PBS, exosomes from U251 cells, U251/TR cells, or hypoxic U251/TR cells. C Red fluorescent dye PKH26 staining in TAMs co- cultured with exosomes from U251 cells, U251/TR cells, and hypoxic U251/TR cells for assessing exosome uptake. Bar, 10 μm. D, E Number of colonies of U251 cells cocultured with TAMs that were pretreated with exosomes from U251 cells, U251/TR cells, and hypoxic U251/TR cells after TMZ administration. F, G Number of invasive cells under above treatment. Bar, 100 μm. \| https://doi.org/10.1007/s12672-024-00963-9 Ns: no significance; p < 0.05; *p < 0.001; and *p < 0.0001 Fig. 8 Hypoxic resistant cells transmit exosomal PKM2 to TAMs to facilitate TMZ resistance of glioma. A, B PKM2 expression in TAMs co- cultured with PBS, exosomes from U251 cells, U251/TR cells, or hypoxic U251/TR cells. C Red fluorescent dye PKH26 staining in TAMs co- cultured with exosomes from U251 cells, U251/TR cells, and hypoxic U251/TR cells for assessing exosome uptake. Bar, 10 μm. D, E Number of colonies of U251 cells cocultured with TAMs that were pretreated with exosomes from U251 cells, U251/TR cells, and hypoxic U251/TR cells after TMZ administration. F, G Number of invasive cells under above treatment. Bar, 100 μm. Ns: no significance; p < 0.05; *p < 0.001; and **p < 0.0001 Vol:.(1234567890) Discover Oncology (2024) 15:110 \| https://doi.org/10.1007/s12672-024-00963-9 Discover Oncology (2024) 15:110 Research Ethics approval and consent to participate Not applicable. Consent for publication Not applicable. Competing interests The authors declare no Competing interest. Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/. 5 Conclusion Our study presents a new model in glioma that demonstrates how exosomes confer TMZ resistance. Through hypoxic microenvironment-induced exosomes, TMZ resistance is delivered through a PKM2-dependent mechanism. This crosstalk with TAMs resulted in the formation of a TMZ-resistant microenvironment. Consequently, exosomal PKM2 could be an effective marker for identifying clinically TMZ-resistant glioma and enabling targeted anti-PKM2 treatment for patients. Author contributions Guofu Li, Ziyu Xiong and Changbin Shi designed experiments. Guofu Li, Ziyu Xiong, Ying Li, Cong Yan, Yingying Cheng, Yuwen Wang and Wenzhong Du conducted molecular research and conducted experiments. Jingwei Li, Zifeng Dai, Dongdong Zhang and Chunyang Men contributed to data analysis and interpretation. Guofu Li and Changbin Shi drafted the manuscript. All authors read and approved the final manuscript. Funding This work was funded by National Natural Science Foundation of China (81771276 and 82071298), Key Project of Natural Science Foundation of Heilongjiang Province of China (ZD2016015), Foundation for High-Level Returned Overseas Talents of Ministry of Human Resources and Social Security of China. ets analyzed during the current study are available from the corresponding author on reasonable request. Data availability The datasets analyzed during the current study are available from the corresponding author on reasonable request. References 1. Zhang Y, Xiao Y, Li GC, et al. Long non-coding RNAs as epigenetic mediator and predictor of glioma progression, invasiveness, and prog- nosis. Semin Cancer Biol. 2022;83:536–42. 1. Zhang Y, Xiao Y, Li GC, et al. Long non-coding RNAs as epigenetic mediator and predictor of glioma progression, invasiveness, and prog- nosis. Semin Cancer Biol. 2022;83:536–42. 1. Zhang Y, Xiao Y, Li GC, et al. Long non-coding RNAs as epigenetic mediator and predictor of glioma progression, invasiveness, and prog- nosis. Semin Cancer Biol. 2022;83:536–42. 2. Gritsch S, Batchelor TT, Gonzalez Castro LN. Diagnostic, therapeutic, and prognostic implications of the 2021 W classification of tumors of the central nervous system. Cancer. 2022;128:47–58.f i y 3. Molinaro AM, Taylor JW, Wiencke JK, Wrensch MR. Genetic and molecular epidemiology of adult diffuse glioma. Nat Rev Neurol. 2019;15:405–17. i 3. Molinaro AM, Taylor JW, Wiencke JK, Wrensch MR. Genetic and molecular epidemiology of adult diffuse glioma. Nat Rev Neurol. 2019;15:405–17. 4. McKinnon C, Nandhabalan M, Murray SA, Plaha P. Glioblastoma: clinical presentation, diagnosis, and management. BMJ. 2021;374: n1560. 4. McKinnon C, Nandhabalan M, Murray SA, Plaha P. Glioblastoma: clinical presentation, diagnosis, and managem 4. McKinnon C, Nandhabalan M, Murray SA, Plaha P. Glioblastoma: clinical presentation, diagnosis, and management. BMJ. 2021;374: n1560. 5. Herrlinger U, Tzaridis T, Mack F, et al. Lomustine-temozolomide combination therapy versus standard temozolomide therapy in patients with newly diagnosed glioblastoma with methylated MGMT promoter (CeTeG/NOA-09): a randomised, open-label, phase 3 trial. Lancet. 4. McKinnon C, Nandhabalan M, Murray SA, Plaha P. Glioblastoma: clinical presentation, diagnosis, and management. BMJ. 2021;374: n1560. 5. Herrlinger U, Tzaridis T, Mack F, et al. Lomustine-temozolomide combination therapy versus standard temozolomide therapy in patients with newly diagnosed glioblastoma with methylated MGMT promoter (CeTeG/NOA-09): a randomised, open-label, phase 3 trial. Lancet. 2019;393:678–88. 6 St R T illib t S K A t l Eff t f t t ti fi ld l i t t l id i t t l id l 4. McKinnon C, Nandhabalan M, Murray SA, Plaha P. Glioblastoma: clinical presentation, diagnosis, and management. BMJ. 2021;374: n1560. 5. Herrlinger U, Tzaridis T, Mack F, et al. Lomustine-temozolomide combination therapy versus standard temozolomide therapy in patients with newly diagnosed glioblastoma with methylated MGMT promoter (CeTeG/NOA-09): a randomised, open-label, phase 3 trial. Lancet. 2019;393:678–88. 5. Herrlinger U, Tzaridis T, Mack F, et al. Declarations Ethics approval and consent to participate Not applicable. References Lomustine-temozolomide combination therapy versus standard temozolomide therapy in patients with newly diagnosed glioblastoma with methylated MGMT promoter (CeTeG/NOA-09): a randomised, open-label, phase 3 trial. Lancet. 2019;393:678–88.fi 6. Stupp R, Taillibert S, Kanner A, et al. Effect of tumor-treating fields plus maintenance temozolomide vs mainten on survival in patients with glioblastoma: a randomized clinical trial. JAMA. 2017;318:2306–16. p g 7. van den Bent MJ, Tesileanu CMS, Wick W, et al. Adjuvant and concurrent temozolomide for 1p/19q non-co-deleted anaplastic glioma (CATNON; EORTC study 26053–22054): second interim analysis of a randomised, open-label, phase 3 study. Lancet Oncol. 2021;22:813–23. 7. van den Bent MJ, Tesileanu CMS, Wick W, et al. Adjuvant and concurrent temozolomide for 1p/19q non-co-deleted anaplastic glioma (CATNON; EORTC study 26053–22054): second interim analysis of a randomised, open-label, phase 3 study. Lancet Oncol. 2021;22:813–23. 8. Clement PMJ, Dirven L, Eoli M, et al. Impact of depatuxizumab mafodotin on health-related quality of life and neurological functioning in the phase II EORTC 1410/INTELLANCE 2 trial for EGFR-amplified recurrent glioblastoma. Eur J Cancer. 2021;147:1–12. 7. van den Bent MJ, Tesileanu CMS, Wick W, et al. Adjuvant and concurrent temozolomide for 1p/19q non-co-deleted anaplastic glioma (CATNON; EORTC study 26053–22054): second interim analysis of a randomised, open-label, phase 3 study. Lancet Oncol. 2021;22:813–23. 8. Clement PMJ, Dirven L, Eoli M, et al. Impact of depatuxizumab mafodotin on health-related quality of life and neurological functioning in the phase II EORTC 1410/INTELLANCE 2 trial for EGFR-amplified recurrent glioblastoma. Eur J Cancer. 2021;147:1–12. (CATNON; EORTC study 26053 22054): second interim analysis of a randomised, open label, phase 3 study. Lancet Oncol. 2021;22:813 23. 8. Clement PMJ, Dirven L, Eoli M, et al. Impact of depatuxizumab mafodotin on health-related quality of life and neurological functioning in the phase II EORTC 1410/INTELLANCE 2 trial for EGFR-amplified recurrent glioblastoma. Eur J Cancer. 2021;147:1–12. y y p p y 8. Clement PMJ, Dirven L, Eoli M, et al. Impact of depatuxizumab mafodotin on health-related quality of life and neurological functioning in the phase II EORTC 1410/INTELLANCE 2 trial for EGFR-amplified recurrent glioblastoma. Eur J Cancer. 2021;147:1–12. p pi g ; 9. Boyd NH, Tran AN, Bernstock JD, et al. Glioma stem cells and their roles within the hypoxic tumor microenvironment. Theranostics. 2021;11:665–83. p pi g ; 9. Boyd NH, Tran AN, Bernstock JD, et al. Glioma stem cells and their roles within the hypoxic tumor microenvironment. Theranostics. 2021;11:665–83. p ; 12. References g g p g g 28. Sun Y, Niu X, Wang G, et al. A novel lncRNA ENST00000512916 facilitates cell proliferation, migration and cell cycle progression in amelo- blastoma. Onco Target Ther. 2020;13:1519–31. 28. Sun Y, Niu X, Wang G, et al. A novel lncRNA ENST00000512916 facilitates cell proliferation, migration and cell cycle progression in amelo- blastoma. Onco Target Ther. 2020;13:1519–31. g 29. Wang R, Liang Q, Zhang X, et al. Tumor-derived exosomes reversing TMZ resistance by synergistic drug delivery for glioma-targeting treatment. Colloid Surf B Biointerfaces. 2022;215: 112505. g 29. Wang R, Liang Q, Zhang X, et al. Tumor-derived exosomes reversing TMZ resistance by synergistic drug delivery for glioma-targeting treatment. Colloid Surf B Biointerfaces. 2022;215: 112505. 30. Wei Y, Wang D, Jin F, et al. Pyruvate kinase type M2 promotes tumour cell exosome release via phosphoryla protein 23. Nat Commun. 2017;8:14041. p 31. Wang X, Zhang H, Yang H, et al. Exosome-delivered circRNA promotes glycolysis to induce chemoresistance through the miR-122-PKM2 axis in colorectal cancer. Mol Oncol. 2020;14:539–55. p 31. Wang X, Zhang H, Yang H, et al. Exosome-delivered circRNA promotes glycolysis to induce chemoresistance through the miR-122-PKM2 axis in colorectal cancer. Mol Oncol. 2020;14:539–55. axis in colorectal cancer. Mol Oncol. 2020;14:539 55. 32. Wang D, Zhao C, Xu F, et al. Cisplatin-resistant NSCLC cells induced by hypoxia transmit resistance to sensitive cells through exosomal PKM2. Theranostics. 2021;11:2860–75. 32. Wang D, Zhao C, Xu F, et al. Cisplatin-resistant NSCLC cells induced by hypoxia transmit resistance to sensitive cells through exosomal PKM2. Theranostics. 2021;11:2860–75. 32. Wang D, Zhao C, Xu F, et al. Cisplatin-resistant NSCLC cells induced by hypoxia transmit resistance to sen PKM2. Theranostics. 2021;11:2860–75. ; 33. Chen W, Tang D, Lin J, et al. Exosomal circSHKBP1 participates in non-small cell lung cancer progression through PKM2-mediated glycolysis. Mol Ther Oncolytics. 2022;24:470–85. 33. Chen W, Tang D, Lin J, et al. Exosomal circSHKBP1 participates in non-small cell lung cancer progression through PKM2-mediated glycolysis. Mol Ther Oncolytics. 2022;24:470–85. y 34. Wan L, Xia T, Du Y, et al. Exosomes from activated hepatic stellate cells contain GLUT1 and PKM2: a role for of liver nonparenchymal cells. Faseb j. 2019;33:8530–42. y j 35. Yang L, Zheng S, Liu Q, et al. Plasma-derived exosomal pyruvate kinase isoenzyme type M2 accelerates the proliferation and motility of oesophageal squamous cell carcinoma cells. Oncol Rep. 2021. https://doi.org/10.3892/or.2021.8167. y 35. Yang L, Zheng S, Liu Q, et al. References The DNA methylome of DDR genes and benefit from RT or TMZ in IDH mutant low-grade glioma treated in EORTC 22033. Acta Neuropathol. 2018;135:601–15. 20. Bady P, Kurscheid S, Delorenzi M, et al. The DNA methylome of DDR genes and benefit from RT or TMZ in IDH mutant low-grade glioma treated in EORTC 22033. Acta Neuropathol. 2018;135:601–15. p 21. Rehman FU, Liu Y, Yang Q, et al. Heme oxygenase-1 targeting exosomes for temozolomide resistant glioblastoma synergistic therapy. J Control Releas. 2022;345:696–708. p 21. Rehman FU, Liu Y, Yang Q, et al. Heme oxygenase-1 targeting exosomes for temozolomide resistant glioblastoma synergistic therapy. J Control Releas. 2022;345:696–708. 22. Xu J, Zhang J, Zhang Z, et al. Hypoxic glioma-derived exosomes promote M2-like macrophage polarization by enhancing autophagy induction. Cell Death Dis. 2021;12:373. 22. Xu J, Zhang J, Zhang Z, et al. Hypoxic glioma-derived exosomes promote M2-like macrophage polarization by enhancing autophagy induction. Cell Death Dis. 2021;12:373. Extracellular vesicles derived from hypoxic glioma stem-like cells confer temozolomide resistance on glioblastoma -3p. Theranostics. 2021;11:1763–79. 23. Yin J, Ge X, Shi Z, et al. Extracellular vesicles derived from hypoxic glioma stem-like cells confer temozolom by delivering miR-30b-3p. Theranostics. 2021;11:1763–79. y g p 24. Liu J, Gao L, Zhan N, et al. Hypoxia induced ferritin light chain (FTL) promoted epithelia mesenchymal transition and chemoresistance of glioma. J Exp Clin Cancer Res. 2020;39:137.fii y g p 24. Liu J, Gao L, Zhan N, et al. Hypoxia induced ferritin light chain (FTL) promoted epithelia mesenchymal transition and chemoresistance of glioma. J Exp Clin Cancer Res. 2020;39:137.fii g p ; 25. Meng L, Wang C, Lu Y, et al. Targeted regulation of blood-brain barrier for enhanced therapeutic efficiency of hypoxia-modifier nanopar- ticles and immune checkpoint blockade antibodies for glioblastoma. ACS Appl Mater Interfac. 2021;13:11657–71. g p 25. Meng L, Wang C, Lu Y, et al. Targeted regulation of blood-brain barrier for enhanced therapeutic efficiency of hypoxia-modifier nanopar- ticles and immune checkpoint blockade antibodies for glioblastoma. ACS Appl Mater Interfac. 2021;13:11657–71. 26. Zhou Y, Zhu F, Liu Y, et al. Blood-brain barrier-penetrating siRNA nanomedicine for alzheimer’s disease therapy. Sci Adv. 2020. https://doi. org/10.1126/sciadv.abc7031. 26. Zhou Y, Zhu F, Liu Y, et al. Blood-brain barrier-penetrating siRNA nanomedicine for alzheimer’s disease therapy. Sci Adv. 2020. https://doi. org/10.1126/sciadv.abc7031. g 27. Hu B, Zhong L, Weng Y, et al. Therapeutic siRNA: state of the art. Signal Transduct Target Ther. 2020;5:101. References Yang W, Xia Y, Hawke D, et al. PKM2 phosphorylates histone H3 and promotes gene transcription and tumorigenesis. Cell. 2012;150:685–96. p ; 12. Yang W, Xia Y, Hawke D, et al. PKM2 phosphorylates histone H3 and promotes gene transcription and tumorigenesis. Cell. 2012;150:685–96. Vol.:(0123456789) Research \| https://doi.org/10.1007/s12672-024-00963-9 Discover Oncology (2024) 15:110 Discover Oncology 13. Beinat C, Patel CB, Xie Y, Gambhir SS. Evaluation of glycolytic response to multiple classes of anti-glioblastoma drugs by noninvasive measurement of pyruvate kinase M2 using [(18)F]DASA-23. Mol Imag Biol. 2020;22:124–33. py g g 14. Park I, Mukherjee J, Ito M, et al. Changes in pyruvate metabolism detected by magnetic resonance imaging are linked to DNA damage and serve as a sensor of temozolomide response in glioblastoma cells. Cancer Res. 2014;74:7115–24. py g g 14. Park I, Mukherjee J, Ito M, et al. Changes in pyruvate metabolism detected by magnetic resonance imaging are linked to DNA damage and serve as a sensor of temozolomide response in glioblastoma cells. Cancer Res. 2014;74:7115–24. g 15. Pan Z, Zhao R, Li B, et al. EWSR1-induced circNEIL3 promotes glioma progression and exosome-mediated sive polarization via stabilizing IGF2BP3. Mol Cancer. 2022;21:16.f p g 16. Ding C, Yi X, Chen X, et al. Warburg effect-promoted exosomal circ_0072083 releasing up-regulates NANGO expression through multiple pathways and enhances temozolomide resistance in glioma. J Exp Clin Cancer Res. 2021;40:164. p y g p ; 17. Qian M, Wang S, Guo X, et al. Hypoxic glioma-derived exosomes deliver microRNA-1246 to induce M2 macrophage polarization by target- ing TERF2IP via the STAT3 and NF-κB pathways. Oncogene. 2020;39:428–42. g p y g 18. Li J, Liao T, Liu H, et al. Hypoxic glioma stem cell-derived exosomes containing linc01060 promote progression of glioma by regulating the MZF1/c-Myc/HIF1α axis. Cancer Res. 2021;81:114–28. y 18. Li J, Liao T, Liu H, et al. Hypoxic glioma stem cell-derived exosomes containing linc01060 promote progression of glioma by regulating the MZF1/c-Myc/HIF1α axis. Cancer Res. 2021;81:114–28. o CY, et al. AR ubiquitination induced by the curcumin analog suppresses growth of temozolomide-resistant glio- upting GPX4-Mediated redox homeostasis. Redox Biol. 2020;30: 101413.i y 19. Chen TC, Chuang JY, Ko CY, et al. AR ubiquitination induced by the curcumin analog suppresses growth o blastoma through disrupting GPX4-Mediated redox homeostasis. Redox Biol. 2020;30: 101413.i g p g 20. Bady P, Kurscheid S, Delorenzi M, et al. References Plasma-derived exosomal pyruvate kinase isoenzyme type M2 accelerates the proliferation and motility of oesophageal squamous cell carcinoma cells. Oncol Rep. 2021. https://doi.org/10.3892/or.2021.8167. p g q p p g 36. Dai J, Escara-Wilke J, Keller JM, et al. Primary prostate cancer educates bone stroma through exosomal pyruvate kinase M2 to promote bone metastasis. J Exp Med. 2019;216:2883–99.i p g q p p g 36. Dai J, Escara-Wilke J, Keller JM, et al. Primary prostate cancer educates bone stroma through exosomal pyruvate kinase M2 to promote bone metastasis. J Exp Med. 2019;216:2883–99. p 37. Fiori ME, Di Franco S, Villanova L, et al. Cancer-associated fibroblasts as abettors of tumor progression at the crossroads of EMT and therapy resistance. Mol Cancer. 2019;18:70. 38. Gu J, Li X, Zhao L, et al. The role of PKM2 nuclear translocation in the constant activation of the NF-κB signaling pathway in cancer- p 37. Fiori ME, Di Franco S, Villanova L, et al. Cancer-associated fibroblasts as abettors of tumor progression at the crossroads of EMT and therapy resistance. Mol Cancer. 2019;18:70. 37. Fiori ME, Di Franco S, Villanova L, et al. Cancer-associated fibroblasts as abettors of tumor progression at the crossroads of EMT and therapy resistance. Mol Cancer. 2019;18:70. assoc atedi b ob asts. Ce Deat D s. 0 ; : 9 . 39. Azambuja JH, Ludwig N, Yerneni SS, et al. Arginase-1+ exosomes from reprogrammed macrophages promote glioblastoma progression Int J Mol Sci. 2020. https://doi.org/10.3390/ijms21113990. Publisher’s Note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. Publisher’s Note Springer Nature remains neutral with regard to jurisdictional claims in published maps and instit ol:.(1234567890)
https://openalex.org/W1978929478	https://www2.ia-engineers.org/conference/index.php/icisip/icisip2013/paper/download/128/166	English	null	Breast Lesions Analysis Using Kinetic and Pyramid Spatial Features in Dynamic Contrast Enhanced MRI	null	2,013	cc-by	2,736	2. Method Recently, dynamic contrast enhanced MRI (DCE-MRI) has become an important tool in locating, identifying and characterizing breast lesions(1). In the analysis of these images, diagnostic features are extracted; for example, homogeneous or heterogeneous enhancement in a region of interest (ROI) is determined. Furthermore, morphologic Abstract characteristics such as roundness of shape or margin are also examined. Among these features, the most important feature in the DCE-MRI is kinetic, which describes how dynamic contrast affects the signal intensity overtime. These characteristics, along with 3D geometric information, are the reason why DCE-MRI is the most accurate tool in breast cancer diagnosis. Local gradient based descriptors are very popular in the computer vision world recently; however, there are few applications of these descriptors in medical image analysis. The goal of this study is to classify breast lesions using local oriented gradients and temporal information. The proposed method combines the kinetics of dynamic contrast enhanced MRIs, which were examined at 0, 90 and 275 seconds separately, and extracts the histograms of oriented gradients descriptors from the interest points of kinetics extreme. The performance of the classifier was assessed with leave-one-out cross-validation among different patients. A total of 39 breast lesions of mass from 25 patients were selected; and the features extracted from these MRIs were further trained using J48 decision tree. The experimental results demonstrated that the average sensitivity and specificity are 0.83 and 0.95, respectively. By the combination of kinetics and pyramid spatial feature, it provides a new application of modern computer vision descriptors in breast MRI analysis. Previous works on breast lesion analysis in DCE-MRI mostly focus on finding the relationship between diagnostic features and breast tissues(2). Among methods of texture-base features published(3,4,5,6,7,8), there were also kinetic characteristic analysis(4,9) of breast lesions; hybrid features of these two characteristics were also proposed(4,10). Machine learning algorithms, such as support vector machine (SVM), naïve Bayes and neural network were also applied to breast lesion classification(4,11,12). In the field of image classification, researchers have shown an increased interest in local gradient based descriptors. Descriptors, such as histograms of oriented gradients (HOG) (13), have been widely accepted and applied to many field of computer vision. However, there is still few applications of these descriptors in breast lesion analysis. By using pyramid spatial feature, which is inspired by pyramid histograms of oriented gradients (PHOG)(14,15), the proposed method combines kinetic characteristic and PHOG for breast lesions analysis. Keywords: pyramid of histograms of oriented gradients, breast lesion analysis, classification, dynamic contrast enhanced MRI. a Institute of Medical Sciences, Tzu Chi University, Taiwan, R.O.C b Breast Health Center, Taipei Medical University Hospital, Taiwan, R.O.C c Medical Imaging Center of Tungs’ Taichung Metro Harbor Hospital, Taiwan, R.O.C a Institute of Medical Sciences, Tzu Chi University, Taiwan, R.O.C b Breast Health Center, Taipei Medical University Hospital, Taiwan, R.O.C c Medical Imaging Center of Tungs’ Taichung Metro Harbor Hospital, Taiwan, R.O.C Corresponding Author: hjlee@mail.tcu.edu.tw Corresponding Author: hjlee@mail.tcu.edu.tw Proceedings of the 1st IEEE/IIAE International Conference on Intelligent Systems and Image Processing 2013 Proceedings of the 1st IEEE/IIAE International Conference on Intelligent Systems and Image Processing 2013 © 2013 The Institute of Industrial Applications Engineers, Japan. DOI: 10.12792/icisip2013.034 2.1 Materials The DCE-MRI for breast detection is performed using gadolinium-based contrast agent, and it contains at least 3 volumetric images which are examined at 3 specific time 170 Fig. 1. A breast lesion is selected in an ROI manually in DCE-MRI at 3 time points (3TP). Fig. 3. The true positive rate show decrease trend while the CDF percentage increases. Fig. 3. The true positive rate show decrease trend while the CDF percentage increases. Fig. 1. A breast lesion is selected in an ROI manually in DCE-MRI at 3 time points (3TP). where i = 0 or 1 and Ii(x, y) is the pixel value of (x, y) at time i. Second, the 2nd derivative is calculated by subtraction of consecutive 1st derivative image; Ai(x, y) = Di+1(x, y)- Di(x, y). (2) Ai(x, y) = Di+1(x, y)- Di(x, y). (2) points (3TP); these time points are before contrast injection (T0), 90 second (T1) and 275 second(T2) after contrast injection. Figure 1 shows an example of these images; from left to right are the images of T0, T1 and T2 separately. For each lesion, the ROIs are slightly different between slices; the manually selected ROIs are irregular and non-rigid. The result of the 1st and 2nd derivative are shown in Fig. 2; the 2nd derivative image is further thresholded with the value selected by cumulative distribution function (CDF) analysis; 1% of the points from the smallest pixel value are reserved. Fig. 3 shows that when the percentage of CDF increases, the true positive rate declines at the same time. When the percentage is less than 1%, the quantity of the extracted features are insufficient in some patients for classifier training, thus this experiment uses CDF percentage at 1%. 2.2 Kinetic Extreme Detection By assuming the better distinctive feature is the one extracted from the point with intensive kinetic variation, we detect these points with chosen criteria before feature extraction. First, the 1st derivative is calculated by performing subtraction of series of DCE-MRI, which is defined as: These interest points are detected since their 2nd derivative values are excessive negative. In addition, these points have positive pixel values in D0 and negative D1. Therefore, their pixel values in 2nd derivative are extremely Di(x, y) = Ii+1(x, y) - Ii(x, y) , (1) Di(x, y) = Ii+1(x, y) - Ii(x, y) , (1) (a) An example of second derivative and kinetic extreme detection result (b) Histogram of second derivative Fig. 4. The result of second derivative is further thresholded at the smallest 1% by cumulative distribution function (CDF) analysis. Fig. 2. First and second derivative are calculated by performing subtraction of DCE-MRI. Top: 1st and 2nd derivative; Bottom: zoom in on a breast lesion which is circled according to the medical report. (a) An example of second derivative and kinetic extreme detection result (b) Histogram of second derivative Fig. 4. The result of second derivative is further thresholded at the smallest 1% by cumulative distribution function (CDF) analysis. Fig. 2. First and second derivative are calculated by performing subtraction of DCE-MRI. Top: 1st and 2nd derivative; Bottom: zoom in on a breast lesion which is circled according to the medical report. (a) An example of second derivative and kinetic extreme detection result (b) Histogram of second derivative Fig. 2. First and second derivative are calculated by performing subtraction of DCE-MRI. Top: 1st and 2nd derivative; Bottom: zoom in on a breast lesion which is circled according to the medical report. (b) Histogram of second derivative Fig. 4. The result of second derivative is further thresholded at the smallest 1% by cumulative distribution function (CDF) analysis. 171 small, and that indicates severe kinetic variation and satisfies the criteria required. Figure 4 is a result of interest points detection, the picture on the right is a breast lesion ROI where the points with severe enhancement are kept. Table 1. 2.3 PHOG Feature Extraction The PHOG features are extracted from the window centered at an interest point for both 3 sets of DCE-MRI; and the pyramid spatial feature is obtained by concatenating these 3 of the PHOG features as one instance that represent certain interest point. The sample was labeled as positive or negative class, and this process was repeated after all interest points were sampled. There are two parameters of PHOG: recursions from r=0 to r= 2 and window size for 16, 32 and 64. The number of grid cells in the PHOG descriptor is decided by recursions, and it also affect the histogram representations corresponding; where the final PHOG vector is a concatenation of histograms for all cells. The window size decides the extracted ROI of the PHOG descriptor. (a) True positive rate with different PHOG features (b) False positive rate with different PHOG features Fig. 5. The relationship among PHOG features with different parameter and evaluation metrics. Top: true positive; bottom: false positive rate are shown. The parameters of PHOG in this study are recursions = 1 and window size = 64. The reason why these parameters are adopted is shown in Fig. 5, where the relationship between PHOG features and evaluation metrics is shown. The true positive rate shows an upward trend and reaches a peak at recursions = 1 and window size = 64 in Fig. 5 (a); while false positive rate provides the similar result that the lowest point produced from the same parameters in Fig. 5. (b). (a) True positive rate with different PHOG features (b) False positive rate with different PHOG features (a) True positive rate with different PHOG features 2.2 Kinetic Extreme Detection Result of leave-one-out validation fold Sensitivity Specificity FPR Precision areaROC 1 0.646 0.985 0.015 0.989 0.969 2 0.883 0.983 0.017 0.981 0.991 3 0.821 0.964 0.036 0.984 0.972 4 0.979 0.930 0.070 0.971 0.985 5 0.951 0.972 0.028 0.971 0.992 6 0.989 0.977 0.023 0.968 0.997 7 0.887 0.939 0.061 0.967 0.961 8 0.995 0.943 0.057 0.867 0.996 9 0.887 0.982 0.018 0.965 0.987 10 1.000 0.968 0.032 0.355 0.996 11 1.000 0.966 0.034 0.596 0.991 12 0.777 0.951 0.049 0.960 0.954 13 1.000 0.941 0.059 0.825 0.995 14 0.673 0.967 0.033 0.958 0.968 15 0.896 0.929 0.071 0.994 0.966 16 0.868 0.964 0.036 0.993 0.978 17 0.630 0.937 0.063 0.976 0.924 18 0.011 0.974 0.026 0.797 0.625 19 0.648 0.951 0.049 0.993 0.951 20 0.854 0.950 0.050 0.910 0.973 21 1.000 0.963 0.037 0.979 0.994 22 0.997 0.974 0.026 0.995 0.996 23 0.968 0.976 0.024 0.975 0.993 24 0.848 0.941 0.059 0.943 0.956 25 0.709 0.943 0.057 0.939 0.945 AVG 0.83668 0.9588 0.0412 0.91404 0.9622 Table 1. Result of leave-one-out validation 2.4 Classification All of the DCE-MRIs were marked as breast lesion ROI’s manually; and the features extracted from these MRIs were further trained using J48 decision tree for a 2 object classifier, and each node of the tree were bi-class label (leaf) or a numeric decision node of a dimension in the pyramid spatial feature vector. The reason we choose (b) False positive rate with different PHOG features Fig. 5. The relationship among PHOG features with Fig. 5. The relationship among PHOG features with different parameter and evaluation metrics. Top: true positive; bottom: false positive rate are shown. different parameter and evaluation metrics. Top: true positive; bottom: false positive rate are shown. 172 sensitivity is 100%, the precision rate only brings about 35%. Since the area size of breast lesion in fold 10 is the smallest among all folds, this discrepancy can be attributed to the uneven ratio of amounts between positive and negative instances. The kinetic extreme detection cannot produce sufficient quantity of features from this small foreground sample size. Even if the false positive rate is 3% and the true positive rate is 100%, the distort number of false positive samples is the cause of this inaccuracy. decision tree as the learning algorithm is that each node from root to leaf represents a dimension explained features distinctive and how is a node was classified as its label. Although there is possible of over training and sensitiveness, the evaluation metrics shows better result comparing to other machine leaerning algorithms. References (1) Hillman BJ, Harms SE, Stevens G, et al. : "Diagnostic performance of a dedicated 1.5-T breast MR imaging system", Radiology, Vol. 265, No.1, pp. 51–58, 2012. where are the number of TP (true positive), FP (false positive), TN (true negative) and FN (false negative). The terms positive and negative denote the classifier's prediction, and the terms true and false denote whether that prediction corresponds to the observation. Total of 25 tests were made; and the average senstivity (true positive rate), specificity (true negative rate), false postive rate, precision (positive predictive value) and aera under ROC curve are shown in table 1. (2) Newell D, Nie K, Chen J-H, et al. : "Selection of diagnostic features on breast MRI to differentiate between malignant and benign lesions using computer-aided diagnosis: differences in lesions presenting as mass and non-mass-like enhancement", European Radiology, Vol. 20, No.4, pp. 771–781, 2009. (3) Gibbs P, and Turnbull LW : "Textural analysis of contrast-enhanced MR images of the breast", Magnetic Resonance in Medicine, Vol. 50, No.1, pp. 92–98, 2003. 3. Results A total of 25 patients with 39 breast lesions of mass were selected from an IRB approved review. The J48 decision tree classifier is eveluated by leav-one-out validation among different patients; the training data does not contain any sample of the test patient. The evalution factors are defined as follows: 5. Conclusions In summary, the combination of kinetics and PHOG reveals a new application of modern computer vision descriptors in breast MRI analysis. The false positive rate for the interest point classifier is less than 5%, and the area under the ROC curve is 0.96. Future work will include more sample collection and combining other diagnosis features such as shape descriptors. Sensitivity = , (3) Specificity = , (4) Precision = , (5) 4. Discussion While all of the folds result in a low false positive rate and high specificity, the sensitivity rates did not remain steady. Fold 18, for example, only has 10% of true positive rate; this error may be due to the inconsistent mass patterns from different patients. In other words, the within class variation is significant in some patients, thus the classifier fails to recognize the mass tissue based on the patterns found in other patients. Another characteristic of fold 18 is that the size of the mass lesion is larger than other folds, and the enhancement pattern is more consistent. As a result, the classifier may be unsuccessful to distinguish this tissue from large vessels which are considered to be background. (4) Agner SC, Soman S, Libfeld E, et al. : "Textural Kinetics: A Novel Dynamic Contrast-Enhanced (DCE)-MRI Feature for Breast Lesion Classification", Journal of Digital Imaging, Vol. 24, No.3, pp. 446–463, 2010. (5) Chen W, Giger ML, Li H, Bick U, and Newstead GM : "Volumetric texture analysis of breast lesions on contrast-enhanced magnetic resonance images", Magnetic Resonance in Medicine, Vol. 58, No.3, pp. 562–571, 2007. (6) Nie K, Chen J-H, Yu HJ, Chu Y, Nalcioglu O, and Another error occurs in fold 10 is that although the Another error occurs in fold 10 is that although the 173 shape with a spatial pyramid kernel", Proceedings of the 6th ACM international conference on Image and video retrieval, pp. 401–408, 2007. Su M-Y : "Quantitative Analysis of Lesion Morphology and Texture Features for Diagnostic Prediction in Breast MRI", Academic Radiology, Vol. 15, No.12, pp. 1513–1525, 2008. (7) (7) Woods BJ, Clymer BD, Kurc T, et al. : "Malignant-lesion segmentation using 4D co-occurrence texture analysis applied to dynamic contrast-enhanced magnetic resonance breast image data", Journal of Magnetic Resonance Imaging, Vol. 25, No.3, pp. 495–501, 2007. (8) (8) Yao J, Chen J, and Chow C : "Breast Tumor Analysis in Dynamic Contrast Enhanced MRI Using Texture Features and Wavelet Transform", IEEE Journal of Selected Topics in Signal Processing, Vol. 3, No.1, pp. 94–100, 2009. (9) Jansen SA, Shimauchi A, Zak L, Fan X, Karczmar GS, and Newstead GM : "The diverse pathology and kinetics of mass, nonmass, and focus enhancement on MR imaging of the breast", Journal of Magnetic Resonance Imaging, Vol. 33, No.6, pp. 1382–1389, 2011. 4. Discussion (10) Zheng Y, Englander S, Schnall MD, and Shen D : "STEP: Spatial-temporal enhancement pattern, for MR-Based breast tumor diagnosis", Biomedical Imaging 2007 IEEE International Symposium on, pp. 520–523, 2007. (11) Lee SH, Kim JH, Park JS, et al. : "Computerized Segmentation and Classification of Breast Lesions Using Perfusion Volume Fractions in Dynamic Contrast-enhanced MRI", IEEE, pp. 58–62, 2008. (12) Nagarajan MB, Huber MB, Schlossbauer T, Leinsinger G, Krol A, and Wismüller A : "Classification of small lesions in dynamic breast MRI: eliminating the need for precise lesion segmentation through spatio-temporal analysis of contrast enhancement", Machine Vision and Applications, 2012. (13) Dalal N, and Triggs B : "Histograms of oriented gradients for human detection", Computer Vision and Pattern Recognition, 2005. CVPR 2005. IEEE Computer Society Conference on, pp. 886–893, 2005. (14) Lazebnik S, Schmid C, and Ponce J : "Beyond bags of features: Spatial pyramid matching for recognizing natural scene categories", Computer Vision and Pattern Recognition, 2006 IEEE Computer Society Conference on, pp. 2169–2178, 2006. (15) Bosch A, Zisserman A, and Munoz X : "Representing 174
https://openalex.org/W4254048818	https://www.qeios.com/read/GPS01G/pdf	English	null	Authorization for Release of Confidential Health Information Indicator	Definitions	2,020	cc-by	125	Authorization for Release of Confidential Health Information Indicator National Cancer Institute Qeios · Definition, February 7, 2020 Open Peer Review on Qeios Qeios ID: GPS01G · https://doi.org/10.32388/GPS01G National Cancer Institute. Authorization for Release of Confidential Health Information Indicator. NCI Thesaurus. Code C70966. National Cancer Institute. Authorization for Release of Confidential Health Information Indicator. NCI Thesaurus. Code C70966. An indicator that can be set or unset in order to signal whether a signed authorization from a patient, or his legal representative, or clinical study participant for the use or disclosure of oral, written, or electronic form of confidential health information that identifies the individual and relates to the medical history, diagnosis, treatment, or prognosis of his condition has been obtained. Qeios ID: GPS01G · https://doi.org/10.32388/GPS01G 1/1
https://openalex.org/W4318620397	https://nutritionandmetabolism.biomedcentral.com/counter/pdf/10.1186/s12986-023-00724-5	English	null	Association between weight loss and outcomes in patients undergoing atrial fibrillation ablation: a systematic review and dose–response meta-analysis	Nutrition & metabolism	2,023	cc-by	9,255	Zhao et al. Nutrition & Metabolism (2023) 20:5 https://doi.org/10.1186/s12986-023-00724-5 Zhao et al. Nutrition & Metabolism (2023) 20:5 https://doi.org/10.1186/s12986-023-00724-5 Nutrition & Metabolism RESEARCH Open Access Association between weight loss and outcomes in patients undergoing atrial fibrillation ablation: a systematic review and dose–response meta‑analysis Huilei Zhao1†, Xiaozhong Li2†, Peng Yu3†, Menglu Liu4, Jianyong Ma5, Jingfeng Wang6,7†, Wengen Zhu8* and Xiao Liu6,7* RESEARCH Open Access Association between weight loss and outcomes in patients undergoing atrial fibrillation ablation: a systematic review and dose–response meta‑analysis Huilei Zhao1†, Xiaozhong Li2†, Peng Yu3†, Menglu Liu4, Jianyong Ma5, Jingfeng Wang6,7†, Wengen Zhu8* and Xiao Liu6,7* RESEARCH Association between weight loss and outcomes in patients undergoing atrial fibrillation ablation: a systematic review and dose–response meta‑analysis 3 Department of Endocrinology, The Second Affiliated Hospital of Nanchang University, Nanchang 330006, Jiangxi, China 4 Department of Cardiology, Seventh People’s Hospital of Zhengzhou, Zhengzhou 334000, Henan, China 5 Department of Pharmacology and Systems Physiology, University of Cincinnati College of Medicine, 45267 Cincinnati, USA 6 Department of Cardiology, Sun Yat-Sen Memorial Hospital of Sun Yat- Sen University, Guangzhou 510080, Guangdong, China 7 Guangdong Province Key Laboratory of Arrhythmia and Electrophysiology, Guangzhou 510120, China 8 Department of Cardiology, The First Affiliated Hospital of Sun Yat-Sen University, Guangzhou 510080, Guangdong, China Abstract Background Obesity is an strong risk factor for atrial fibrillation (AF), and obesity can affect the prognosis of AF. However, the role of weight loss on outcomes after ablation remains unclear. Background Obesity is an strong risk factor for atrial fibrillation (AF), and obesity can affect the prognosis of AF. However, the role of weight loss on outcomes after ablation remains unclear. Objectives This study aims to determine the relationship between weight loss and outcomes in patients with AF ablation, as well as the potential dose–response relationship. Objectives This study aims to determine the relationship between weight loss and outcomes in patients with AF ablation, as well as the potential dose–response relationship. Methods The Cochrane Library, PubMed, and Embase databases were searched to identify studies that reported a relationship between weight loss and ablation up to August 17, 2021. Relative risks (RRs) were pooled using random- effects models. Methods The Cochrane Library, PubMed, and Embase databases were searched to identify studies that reported a relationship between weight loss and ablation up to August 17, 2021. Relative risks (RRs) were pooled using random- effects models. Results One randomized, open-labeled clinical trial and seven cohort studies involving 1283 patients were included. The mean body mass index of all included studies was over 30 kg/m2. The clinical trial showed a non-significant ben- efit of weight loss intervention on AF recurrence (Odd risk [OR] = 1.02, 95% confidence interval [CI] 0.70–1.47). Meta- analysis based on observational studies showed that the recurrence rate of AF after ablation was significantly reduced (RR = 0.43, 95% CI 0.22–0.81, I2 = 97%) in relatively obese patients with weight loss compared with the control group. Each 10% reduction in weight was associated with a decreased risk of AF recurrence after ablation (RR = 0.54, 95% CI 0.33–0.88) with high statistical heterogeneity (I2 = 76%). An inverse linear association (Pnon-linearity = 0.27) between AF relapse and increasing weight loss was found. Conclusions Our results first suggest an inverse dose–response association between weight loss and risk of recur- rent AF after ablation, with moderate certainty. Abstract 3 Department of Endocrinology, The Second Affiliated Hospital of Nanchang University, Nanchang 330006, Jiangxi, China 4 Department of Cardiology, Seventh People’s Hospital of Zhengzhou, Zhengzhou 334000, Henan, China 5 Department of Pharmacology and Systems Physiology, University of Cincinnati College of Medicine, 45267 Cincinnati, USA 6 Department of Cardiology, Sun Yat-Sen Memorial Hospital of Sun Yat- Sen University, Guangzhou 510080, Guangdong, China 7 Guangdong Province Key Laboratory of Arrhythmia and Electrophysiology, Guangzhou 510120, China 8 Department of Cardiology, The First Affiliated Hospital of Sun Yat-Sen University, Guangzhou 510080, Guangdong, China †Peng Yu and Jingfeng Wang are senior authors Correspondence: Wengen Zhu zhuwg6@mail.sysu.edu.cn Xiao Liu liux587@mail.sysu.edu.cn 1 Department of Anesthesiology, The Third Hospital of Nanchang, Nanchang 330006, Jiangxi, China 2 Department of Cardiology, The Second Affiliated Hospital of Nanchang University, Nanchang 330006, Jiangxi, China Xiao Liu liux587@mail.sysu.edu.cn 1 Department of Anesthesiology, The Third Hospital of Nanchang, Nanchang 330006, Jiangxi, China 2 Department of Cardiology, The Second Affiliated Hospital of Nanchang University, Nanchang 330006, Jiangxi, China © The Author(s) 2023. Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/. The Creative Commons Public Domain Dedication waiver (http://creativeco mmons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated in a credit line to the data. Page 2 of 12 Zhao et al. Nutrition & Metabolism (2023) 20:5 Zhao et al. Nutrition & Metabolism Keywords Atrial fibrillation, Recurrence, Weight loss, Ablation Graphical Abstract Keywords Atrial fibrillation, Recurrence, Weight loss, Ablation also the effective management of related comorbidities and risk factors. Introductioni Atrial fibrillation (AF) is the most common persistent arrhythmia in adults and contributes to an increased risk of stroke, dementia, and mortality. Several impor- tant risk factors have been identified in the occurrence and development of AF, such as hypertension, diabe- tes, obesity, and obstructive sleep apnea [1–5]. In the era that stresses everything from treatment to preven- tion, it is worth noting that effective management of AF requires not only the treatment of the disease itself but Catheter ablation of AF is a recognized treatment method to prevent recurrence of AF, especially for drug- refractory symptomatic AF [6]. However, long-term reports of AF ablation showed that the success rate of ablation gradually declines over time [7–9]. Although there is an obesity paradox in AF patients, however, over- weight and obesity have an adverse effect on cardiovas- cular prognosis in long-term follow-up [10, 11], and the Zhao et al. Nutrition & Metabolism (2023) 20:5 Zhao et al. Nutrition & Metabolism Page 3 of 12 And a full-text evaluation was presented after initial identification. guidelines [12] showed obesity increases the risk of AF progressively. But due to obstructive sleep apnea as a con- founding factor, the guidelines did not demonstrate that obesity is a risk factor for AF recurrence ablation-post. i Eligible studies had to fulfil the following criteria: (1) Clinical trials or observational studies; (2) Study on the relationship between weight loss and outcomes after AF ablation; (3) The patients in this study were adults (age > 18 years), diagnosed with AF, and undergo cath- eter ablation with weight management;(4) reported the relationship between weight loss and AF recurrence and other outcomes (AF severity, quality, symptoms); (5) The literature reported odds’ ratio (OR), Relative risk (RR), hazard ratio (HR), and the 95% confidence interval (CI) provided available data to calculate the estimation effect for the AF recurrence. At present, research on controlling risk factors to reduce AF recurrence after ablation has become a hot topic. Obesity is a worldwide health problem and con- tributes to AF incidence after ablation [10, 11]. Some studies have shown that weight loss can reduce the AF incidence and burden of AF in obese populations [13, 14]. Statistical analysis and bias risk assessmenth For outcomes: we did not apply any keywords for out- comes because all reported outcomes related to AF abla- tion were included, such as AF recurrence, AF severity, or quality of life. Review Manager (Version 5.1., The Nordic Cochrane Center, The Cochrane Collaboration, Copenhagen, Den- mark, 2011.) and Stata 16.0 (Stata Corp LP, College Sta- tion, TX, USA) was used for statistical analysis. The HR was converted to RR, and the supposed OR provided a good estimate of RR. RR were used to combine the esti- mated effects of random effects models. We estimate the adjusted RR’s by calculating the natural logarithm of RR (log [RR]) and its standard error (SE log [RR]). For those studies that did not provide RRs, we calculated crude RR by event and total number. In addition, we performed Additional file 1: Table S2 provides the detailed search strategy. Literature search PubMed, Cochrane Library, Embase database and con- ference articles (American Heart Association: https:// www.ahajournals.org/journal/circ, American College of Cardiology: https://www.jacc.org/ and European Society of Cardiology: https://www.escardio.org/) were searched by using the following MeSH to retrieve articles up to August 17, 2021, includes full text and conference abstracts, without language restrictions. The quality of the included studies was assessed accord- ing to the Newcastle–Ottawa evaluation scale (NOS) and modified Jada scale for cohort and trial, respectively. Scores range from 0 to 9, with NOS scores greater than 7 being considered high quality [16]. For patients: “atrial fibrillation”, “atrial flutter”, “atrial tachycardia”, and “ablation”. For exposure/intervention: “weight loss”, “weight reduction”. Introductioni Similarly, some data suggest the benefit of weight loss in the prevention of AF recurrence after ablation; however, whether weight loss improves the outcomes after AF ablation remains unclear.h Additionally, we excluded studies with: Thus, we conducted a meta-analysis to assess the rela- tionship between weight loss and outcomes, including AF recurrence, AF symptom severity, AF burden, and quality of life. (1) For multiple reports based on the same data source, we excluded studies with the shorter follow-up time or smaller sample size. (2) Case–control design due to the potential bias. Data extraction and quality assessment y Studies were reviewed by two independent authors (X.Z-L and X-L) according to the above inclusion and exclusion criteria. Disagreements were resolved by con- sensus. Data were extracted by 2 investigators (X.Z-L and X-L), including first author, publication year, coun- try, follow-up time, demographic characteristics (sam- ple size, average age, gender, body mass index (BMI), left atrial diameter, AF type, history of diabetes, history of hypertension, high history of lipemia), study design, data source, methods of weight loss and AF diagnosis, outcomes, corresponding 95% CI and estimate effect, and adjustments. Methodh This research was conducted in accordance with the guidelines of the 2020 Systematic Review and Meta- Analysis Preferred Reporting Project (PRISMA 2020) [15] (Additional file 1: Table S1).The protocol has been registered in PROSPERO (Prospective Registration for International System Evaluation.https://www.crd.york. ac.uk/prospero/display_record.php?RecordID=273682. CRD4202173680). Study selection We used Endnote X8 database, a reference manage- ment software, to organize all studies. All titles and abstracts were reviewed to consider eligible for inclusion. Zhao et al. Nutrition & Metabolism (2023) 20:5 Zhao et al. Nutrition & Metabolism Page 4 of 12 [19] = 1) were included. Overall, these studies were published between 2014 and 2021. Sample sizes for the included studies ranged from 90 to 304 with a total of 1150 patients. The average age was 58 to 65 years, and the female proportions was 27.8–44.0%. Follow-up for the entire study ranged from 6 to 60 months. The aver- age BMI of the clinical trial [21] was 34.8kg/m2, while the average BMI of the other seven observational stud- ies [19, 22–27] was greater than 30 kg/m2. The enrolled patients in the clinical trial [21]are all obese. Meanwhile, the patients in the three observational studies [19, 24, 25] included are all obese; in the observational stud- ies of Bunch et al. [26] and Lau et al. [22], the obesity proportions are 44.4% and 76.0%, respectively, and the obesity proportion in the two observational studies [23, 27] is unclear. Four studies were from the United States [19, 23, 25, 26], two were from the United Kingdom [22, 24], one was from Australia [27], and one was from Germany [21]. Eight studies reported on the recurrence of AF [19, 21–27], three studies reported the burden of AF [21, 25, 27], and one study reported on quality of life [25]. The method of weight loss in the clinical trial [21] was improved lifestyle. The method of weight loss in four observational studies [22, 24, 25, 27]was an improved lifestyle; bariatric surgery was adopted for weight loss in one observational study [19], and the method of weight loss in the other two observational studies [23, 26] was unknown. Weight intervention was used in five of the studies before ablation [19, 22, 24, 25, 27], whereas three studies reported the use of an intervention after ablation [21, 23, 26]. The blank period after ablation of six studies [19, 21, 24–27] is 3 months, and resting two studies [22, 23] were unclear. Four studies defined AF recurrence as an episode for 30 s or more [19, 21, 25, 27], and four stud- ies did not clarify the definition of AF recurrence [22–24, 26]. Weight loss and AF recurrence after ablation d d l b l d l l One randomized, open-labeled clinical trial(SORT- AF) [21] and seven cohorts reported weight loss and AF recurrence after ablation [19, 22–27]. In the clini- cal trial, the intervention group achieved a mean per- centage weight loss of 3.91% compared with 0.91% in the control group, demonstrating that weight loss did not significantly reduce the recurrence rate of AF after Study selection All AF events were detected by 12-lead ECG, Holter or other forms of dynamic monitoring. subgroup analysis stratified by gender, methods of weight loss, and the weight loss time. g For dose–response analysis, we computed summary RRs and 95% CIs for a 10% in weight loss using a random effects model. Study-specific slopes (linear trends) and 95% CIs from the natural logs of the reported RRs and CIs across categories of weight loss by using the method of Greenland and Longnecker [17]. We performed the non-linear dose–response analysis by using the robust error meta-regression method described by Xu et al. [18]. It requires known levels of weight loss and RRs with variance estimates for at least two quantitative exposure categories. If the median or mean weight loss was not provided and reported in ranges, we estimated the mid- point of each category by averaging the lower and upper boundaries of that category. If the highest or lowest cat- egory was open-ended, we assumed that the open-ended interval length was the same as the adjacent interval. We used Q statistic and I2 statistics to estimate heterogene- ity between studies. For Q statistic,a P < 0.10 was regared to indicate significant heterogeneity in the Q statistic. For I2 statistics, in I2 < 50%, I2 at 50% to 75%, I2 > 75% were regard as low heterogeneity, moderate heterogeneity and high heterogeneity [16], respectively. Egger’s, Begg’s, or Funnel plot were used to detect publication bias. P < 0.05 with two tails is considered statistically significant. Study selection As shown in Fig. 1, 236 publications and five conference abstracts were identified in the initial literature search (PubMed = 121; Cochrane Library = 9; Embase = 96; other sources = 10). After excluding duplicates and screening titles and abstracts, 32 articles remained for evaluation by full text. Two studies were based on dupli- cated populations [19, 20], and studies with a large sam- ple size were included [19]. Eight studies were finally included, and 24 studies were excluded for the following reasons: (1) reviews, comments and case reports (n = 7); (2) cross-sectional studies (n = 1); (3) did not reported targeted outcomes (n = 1); (4) insufficient data (n = 2); (5) study protocols (n = 2); (6) meta-analyses (n = 4); and (7) Literature from the same population (n = 1); (8) patients did not receive ablation therapy (n = 3), (9) The study’s purpose does not meet inclusion criteria (n = 3). Addi- tional file 1: Table S3 describes the excluded studies. It is generally considered that the quality of the lit- erature with a score of 4–7 on the modified Jada scale is high, while the score of the modified Jada scale of the SORT-AF trial is 4, so the quality of the SORT-AF trial is acceptable (Additional file 1: Table S5). The overall study quality of observational data was acceptable with NOS > 6 for all studies (Additional file 1: Table S6). Study characteristics and qualityi Table 1 and Additional file 1: Table S4 summarizes the characteristics of the included studies. One rand- omized open-labeled,clinical trial [21] with 133 (weight loss = 67, control = 66) patients and seven cohorts with 1150 patients (prospective [22–27] = 6, retrospective Page 5 of 12 Zhao et al. Nutrition & Metabolism (2023) 20:5 Zhao et al. Nutrition & Metabolism Fig. 1 Flowchart of selection of studies that investigated weight loss and outcomes in AF ablation identified in the systematic literature search on of studies that investigated weight loss and outcomes in AF ablation identified in the systematic literature search Fig. 1 Flowchart of selection of studies that investigated weight loss and outcomes in AF ablation identified i Fig. 1 Flowchart of selection of studies that investigated weight loss and outcomes in AF ablation identified in the systematic literature search articles showed a lower AF recurrence rate after abla- tion of weight loss [19, 22, 23, 27], whereas the remain- ing three reported no significant association [24–26]. The pooled results of observational studies showed ablation (OR = 1.02, 95% CI 0.70–1.47) after 12 months of follow-up. Among the observational studies, the mean BMI was over 30 kg/m2 in all studies, and six studies were based on patients with obesity [19, 22–25, 27]. Four Page 6 of 12 Zhao et al. Nutrition & Metabolism (2023) 20:5 Zhao et al. Study characteristics and qualityi Nutrition & Metabolism (2023) 20:5 dose–response association was fitted by the restricted cubic splines function, showing a significant trend of inverse association (Pnon-linearity = 0.27) of AF relapse after ablation with increasing weight loss. However, the results were merely significant at a range of 6–13% weight loss (Fig. 4). Additional file 1: Table S7 summarizes weight loss exposure doses for included studies. that weight loss in relatively obese patients was associ- ated with a significant risk of AF relapse after ablation (RR = 0.43.95% CI 0.22–0.81) with substantial hetero- geneity (Q statistic: P < 0.00001; I2 = 97%) [19, 22–27] (Fig. 2). Study characteristics and qualityi Nutrition & Metabolism Table 1 Main characteristics of the included studies in the meta‐analysis of weight loss and outcomes after AF ablation AF Atrial fibrillation, ARREST-AF Aggressive Risk factor reduction study for atrial fibrillation, CARDIO-FIT Study cardio- respiratory fitness, SORT-AF Supervised obesity reduction trial for AF, ILR Implantation loop recorder, UK United Kingdom, USA United States of America, ECG Electrocardiograph, BMI Body mass index, HbA1C Hemoglobin A1C, BS Bariatric surgery, EFV Epicardial fat volume, LA Left atrium & unclear the exactly adjusted covariate, NA not available AF burden was quantified by using the validated Atrial Fibrillation Severity Scale # The first number refers to the loss weight group and the second to the control; one column refers to the mean or median values of the overall cohort Refences (First author, Year, Country/ Region) Source of individuals Outcomes Study design AF diagnosis Methods of weight loss N Age# (year) Male# (%) BMI# Weight loss (%) RR (95% CI) Follow-up times (months) Adjusted covariates Clinical trial Gessler 2021, Germany SORT-AF trial AF recurrence AF burden Randomized, open-labeled clinical trial ECG, ILR inter- rogation Lifestyle 67/66 58.7/62.1,64/62, 34.9/34.8 4.5 1.14 (0.37–3.61) 12 NA Observational studies Pathak 2014, Australian ARREST-AF Cohort Study AF recur- rence dura- tion and symptoms of AF Prospective Cohort ECG Lifestyle 149 57.7, 63.8, 32.7 13.1 0.21 (0.09–0.49) 24 Multiple proce- dures, type of AF, poor BP control Bunch 2016, USA LDS Hospital, Intermountain Medical Center AF recurrence Prospective Cohort ECG NA 288 36.3, 59.3, 36 4.5 0.73 (0.50–1.10) 36 Multivariable adjusted& Mohanty 2017, USA St. David’s Medical Center AF recur- rence symp- tom severity of AF quality of life Prospective Cohort ECG Lifestyle 90 62.7, 72.2, 37.6 20.5 0.89 (0.52–1.53) 12 NA Donnellan 2019, USA Cleveland Clinic AF recurrence Retrospective Cohort ECG, Holter Bariatric surgery 239 64.6, 55.6, 41.1 20.6 0.14 (0.05–0.39) 60 HbA1C, present- ing rhythm at the time of ablation, EFV, BMI, LA Ding 2020, UK Liverpool Heart and Chest Hospital AF recurrence Prospective Cohort ECG, Holter Lifestyle 92 64.5, 56.5, 36.0 6 1.85 (0.70–4.90) 12 Sex, type of AF, age Y Lau 2020, UK Glasgow Royal Infirmary AF recurrence Prospective Cohort NA Lifestyle 146 NA, NA, 30.4 4.2 0.33 (0.29–0.85) 6 NA Shah 2020, USA The University of Rochester Medical Center AF recurrence Prospective Cohort NA NA 146 61, 67,31 22 0.29 (0.11–0.80) 12 Multivariable adjusted& Page 7 of 12 Zhao et al. Fig. 2 Forest plot of the study of weight loss and AF recurrence after AF ablation in a random-effects model, stratified by study design. Referents for weight loss were the individuals reporting no weight loss or stable weight within the specific study. The diamond indicates the pooled estimate. Red or blue boxes are relative to study size, and the black vertical lines indicate 95% CIs around the effect size estimate Fig. 2 Forest plot of the study of weight loss and AF recurrence after AF ablation in a random-effects model, stratified by study design. Referents for weight loss were the individuals reporting no weight loss or stable weight within the specific study. The diamond indicates the pooled estimate. Red or blue boxes are relative to study size, and the black vertical lines indicate 95% CIs around the effect size estimate Fig. 2 Forest plot of the study of weight loss and AF recurrence after AF ablation in a random-effects model, stratified by study design. Referents for weight loss were the individuals reporting no weight loss or stable weight within the specific study. The diamond indicates the pooled estimate. Red or blue boxes are relative to study size, and the black vertical lines indicate 95% CIs around the effect size estimate Fig. 3 Forest plot of the study-specific RRs for AF recurrence after ablation for every 10% weight loss. Study-specific estimates obtained by the method of Greenland and Longnecker assuming a linear relationship of the RRs to the referent in a random-effects model. Referents for weight loss were the individuals reporting no weight loss or stable weight within the specific study. The diamond indicates the pooled estimate. Red boxes are relative to study size, and the black vertical lines indicate 95% CIs around the effect size estimate Fig. 3 Forest plot of the study-specific RRs for AF recurrence after ablation for every 10% weight loss. Study-specific estimates obtained by the method of Greenland and Longnecker assuming a linear relationship of the RRs to the referent in a random-effects model. Referents for weight loss were the individuals reporting no weight loss or stable weight within the specific study. The diamond indicates the pooled estimate. Red boxes are relative to study size, and the black vertical lines indicate 95% CIs around the effect size estimate Fig. 3 Forest plot of the study-specific RRs for AF recurrence after ablation for every 10% weight loss. Subgroup analysisi A pooled analysis from seven cohort studies showed a significant association between each 10% weight loss and AF recurrence rates after ablation (RR = 0.54, 95% CI 0.33–0.88) with high statistical heterogeneity (Q statistic: P = 0.0003; I2 = 76%) [19, 22–27] (Fig. 3). The nonlinear A pre-defined subgroup analysis stratified by the weight loss method was available. The results showed that bari- atric surgery was associated with lower AF relapse after ablation (bariatric surgery, RR = 0.14, 95% CI 0.12–0.17) 2 Forest plot of the study of weight loss and AF recurrence after AF ablation in a random-effects model, stratified by study design. Referents weight loss were the individuals reporting no weight loss or stable weight within the specific study. The diamond indicates the pooled estimate. or blue boxes are relative to study size, and the black vertical lines indicate 95% CIs around the effect size estimate g. 2 Forest plot of the study of weight loss and AF recurrence after AF ablation in a random-effects model, stratified by study design. Referents weight loss were the individuals reporting no weight loss or stable weight within the specific study. The diamond indicates the pooled estimate. d bl b l i d i d h bl k i l li i di 95% CI d h ff i i Discussion Major findingshi Fig. 4 Dose–response association between weight loss and AF recurrence after AF ablation modeled using restricted cubic splines. The bold lines indicate the pooled restricted cubic spline model, and the black dashed line indicates the 95% CIs of the pooled curve Fig. 4 Dose–response association between weight loss and AF recurrence after AF ablation modeled using restricted cubic splines. The bold lines indicate the pooled restricted cubic spline model, and the black dashed line indicates the 95% CIs of the pooled curve The major findings of the present meta-analysis based on observational studies were as follows: (i) Weight loss is associated with a reduced risk of AF recurrence after ablation. (ii) A 10% reduction in weight loss is associ- ated with a 27% risk of AF recurrence after ablation. (iii) A potential dose–response relationship is noted between weight loss and AF relapse after ablation. To the best of our knowledge, this is the first meta-analysis that reported the dose-response association between weight loss and the risk of AF recurrence after ablation, provid- ing an informative finding for the improvement of out- comes of AF ablation. [19]. However, although a nonsignificant association was noted for lifestyle management (lifestyle management, RR = 0.56, 95% CI 0.18–1.69) [22, 24, 25, 27]. a significant decreasing trend was observed (Additional file 1: Figure S1A). Similar results were found for each 10% reduction in different weight loss method (bariatric surgery [19], RR = 0.39, 95% CI 0.24–0.64, lifestyle management [22, 24, 25, 27], RR = 0.58, 95%: 0.21–1.59), with substantial heterogeneity (Q statistic: P < 0.0007; I2 = 82%) (Addi- tional file 1: Figure S1B). Obesity is a well-known independent risk factor for AF. As shown by several studies, obesity is associated with an approximately 50% increase in the prevalence of AF in the general population with a 1-unit increase in BMI increasing the risk of AF by 3% to 4.7% [28, 29]. This asso- ciation persisted in patients undergoing ablation. Previ- ous studies demonstrated that overweight/obese patients receiving ablation might be associated with worse out- comes, including increased AF recurrence and poor quality of life [30, 31]. Previous observational studies, such as the landmark LEGACY study, showed that long- term sustained weight loss is associated with a significant reduction in the burden of AF and the maintenance of sinus rhythm. GRADE assessmentsh The GRADE framework indicated moderate certainty in the weight loss summary finding for AF recurrence after ablation based on observational studies (data not shown). Discussion Major findingshi A meta-analysis based on post hoc analy- sis of randomized controlled studies (RCTs) also showed similar results [14]. In the present study, we showed a potential benefit of weight loss on AF recurrence after ablation based on observational studies, adding more recent evidence on this subject. Notably, the SORT-AF trial [20] showed that weight loss based on lifestyle man- agement had a nonsignificant benefit on the AF burden or AF recurrence rate. Several reasons might explain these inconclusive results. First, the SORT-AF weight loss group achieved a mean weight reduction of 4.6 kg (3.91% of their initial body weight), which might not be sufficient to show an effect on outcomes. Regarding the LEGACY study, the results showed the greatest effect on freedom from AF in patients who lost > 10% of their body i Pre-defined subgroup analysis stratified by the weight loss time showed that weight loss was associated with lower AF relapse when patients receive weight manage- ment before ablation (RR = 0.41; 95% CI 0.19–0.92). How- ever, when patients received weight management after ablation, the weight loss did not significantly decrease AF recurrence (RR = 0.46, 95% CI 0.19–1.14) (Additional file 1: Figure S2), with no significant subgroup difference (P = 0.86) (Additional file 1: Figure S2). Study-specific estimates obtained by the method of Greenland and Longnecker assuming a linear relationship of the RRs to the referent in a random-effects model. Referents for weight loss were the individuals reporting no weight loss or stable weight within the specific study. The diamond indicates the pooled estimate. Red boxes are relative to study size, and the black vertical lines indicate 95% CIs around the effect size estimate Fig. 3 Forest plot of the study-specific RRs for AF recurrence after ablation for every 10% weight loss. Study-specific estimates obtained by the method of Greenland and Longnecker assuming a linear relationship of the RRs to the referent in a random-effects model. Referents for weight loss were the individuals reporting no weight loss or stable weight within the specific study. The diamond indicates the pooled estimate. Red boxes are relative to study size, and the black vertical lines indicate 95% CIs around the effect size estimate Zhao et al. Nutrition & Metabolism (2023) 20:5 Zhao et al. Nutrition & Metabolism Page 8 of 12 (2023) 20:5 Fig. 4 Dose–response association between weight loss and AF recurrence after AF ablation modeled using restricted cubic splines. The bold lines indicate the pooled restricted cubic spline model, and the black dashed line indicates the 95% CIs of the pooled curve from the observational study showed that weight loss was significantly associated with reduced duration and symp- toms of AF (P < 0.001) [27]. Mohanty et al. [25] showed that AF weight loss significantly improved quality of life (P < 0.002) but not the symptom severity of AF (P = 0.84). from the observational study showed that weight loss was significantly associated with reduced duration and symp- toms of AF (P < 0.001) [27]. Mohanty et al. [25] showed that AF weight loss significantly improved quality of life (P < 0.002) but not the symptom severity of AF (P = 0.84). Comparisons with previous studies Several systematic reviews have shown that weight loss decreases the risk of AF. Jones et al. [40] showed that a 5% loss in body weight was not associated with a sig- nificant change in the incidence of AF. Aldaas et al. [41] showed that patients who lost ≥ 10% of their initial body weight had a lower risk of recurrent AF, a reduction in AF burden, and an improvement in AF symptom severity. Similarly, a systematic review demonstrated that weight loss is associated with a lower long-term recurrence of AF after ablative therapy [42]. Our results expand the effect of weight loss on clinical outcomes, including AF severity, quality, symptoms, for AF ablation and further firstly clarify the potential dose–response association. The subgroup analysis of pre-ablative loss weight dem- onstrated that weight loss was caused by a healthy life- style. In this meta-analysis, weight intervention was used in five of the studies before ablation [19, 22, 24, 25, 27], whereas three studies reported the use of an interven- tion after ablation [21, 23, 26]. The results of the clinical trial reveal that weight loss did not significantly reduce the recurrence rate of AF after ablation (OR = 1.02, 95% CI 0.70–1.47) after 12 months of follow-up [21]. Other two observations studies show that loss weight did not significantly reduce the recurrence rate of AF after abla- tion (RR = 0.46, 95% CI 0.19–1.14) [23, 26]. There are studies demonstrated that pre-ablative weight loss can enhance insulin sensitivity and improve glycemic control, ameliorating epicardial fat deposition and inflammation [32, 33]. Moreover, pre-ablative weight loss can optimize the outcomes of ablation through the improved obesity- mediated structural remodeling of the atrium [34]. How- ever, our results showed that weight loss was associated with a lower AF relapse trend, but the pre-ablative and post-ablation did not showed a subgroup difference. Sensitivity analysis and publication biash The sensitivity analysis of leave-one-out methods, omit- ting crude results, and changing the random model to a fixed model generally produced results consistent with observational studies (Additional file 1: Figure S3). Publi- cation bias and meta-regression were not performed due to the limited number of included studies according to the guidelines (N < 10). Weight loss and AF severity, symptoms, and quality of life The results from SORT-AF showed that no significant difference in the AF burdens were noted among the patients undergoing weight loss intervention and controls (OR = 1.14, 95% CI 0.36–3.61, P = 0.81). [21]Evidence Page 9 of 12 Zhao et al. Nutrition & Metabolism (2023) 20:5 Zhao et al. Nutrition & Metabolism weight. Second, the number of included patients were also limited, which might make the study underpowered to detect an effect on AF burden. Third, group 1 of the SORT-AF trial had more persistent AF patients, although cox proportional hazards models had adjusted types of AF, but the type of AF in both groups was not compa- rable. This may be regarded as a significant confounder. Finally, if there had been weight loss during ablation and not afterward, the effect of weight loss might be benefi- cial. Our dose–response curve also showed a trend but no significant decreased risk of AF recurrence for weight loss < 6%, which reinforced this opinion. Thus, we sup- posed that weight loss in patients with AF ablation was associated with a lower risk of recurrent AF. recurrence after ablation than men [36, 37]. Some rea- sons may account for the sex differences in ablation effec- tiveness, including lower frequency or delayed referral for ablation, higher atrial fibrosis, older age, a more com- plex clinical profile, and a higher prevalence of non-pul- monary vein triggers in women [31, 37–39]. Underlying mechanismh There are several potential mechanisms involved in the association between weight loss and AF recurrence after AF ablation. Obesity leads to atrial structure and elec- trical remodeling, which makes patients prone to AF [43]. This observation is mainly related to the inflamma- tory response, fibrosis and oxidative stress induced by the increase in atrial fat cells and epicardial fat [1, 44]. Weight loss was shown to reduce the atrial area, improve inflammation, and reduce myocardial fibrosis, which has an important impact on slowing down the structural remodeling of the heart [45]. Furthermore, obesity is an independent risk factor for hypertension, diabetes, and obstructive sleep apnea syndrome. These comorbidities are common in patients with obesity, and weight loss can reduce blood pressure, improve insulin sensitivity, and blood sugar control, improve sleep apnea syndrome, and reduce all these comorbidities. f The ablation strategy across studies varied. Radiofre- quency ablation was used in SORT-AF trial’s [20] and two observation studies [24, 26]. Cryoballoon ablation was used in one observation study [25]. One cohort [23] reported mixed radiofrequency and Cryoballoon ablation. Ablation strategy of the resting three observa- tion studies [21, 22] is unclear. Currently, a Systematic Review and Meta-Analysis [35] showed there was no significant difference of the clinical outcomes between cryoballoon ablation and radiofrequency, and the 2020 ESC guideline [12] of AF indicated that in the first pro- cedure for paroxysmal AF, the outcomes of either strat- egies are analogous. Therefore, the heterogeneity due to different ablation strategies may not influence our result. Limitations Several limitations are noted for any observational study. First, our main findings were based on observa- tional studies, the numbers of included studies were limited, and measurement and unmeasured bias can- not be completely excluded. However, the cohort studies reduced the selection bias. Second, we did not assess the association between weight loss and hard outcomes, such as stroke, cardiovascular death, or all- cause death. As we previously described, overweight or class I obesity patients seem to have better survival outcomes than normal-weight individuals in patients with AF, a phenomenon that is known as the “obesity paradox” [16, 48]. The LEGACY results showed better survival outcomes in patients with obesity undergoing bariatric surgery. Current guidelines do not have specific comments on bariatric surgery on AF ablation. Our results showed that bariatric surgery has a significant effect on preventing AF relapse after ablation, but only one study was included. Nevertheless, considering the results of the LEGACY study, we hypothesized that bariatric surgery might be more appropriate and beneficial in those with mortal- ity obesity [46], especially those with decreased physical ability. Nevertheless, the decision for bariatric surgery before ablation should be evaluated in combination with a clinical basis; the potential benefit and risk of weight loss need to be balanced. Increasing the duration of AF has adverse effects on ablation success rates. This pro- cedure would necessitate an at least a 6-months delay in ablation, leading to the progression of AF with known worsening of success rates. In the subgroup of lifestyle management, the results were nonsignificant, but the sta- tistical power might be limited by the small sample size. However, direct evidence for the association between weight loss and AF ablation involving hard outcomes is limited. Third, significant heterogeneity was noted in the main results, which might be derived from the difference in baseline characteristics, antiarrhythmic medications, time of weight loss, and different types of weight loss interventions (lifestyle management or bariatric surgery). According to the GRADE framework, the weight loss summary finding for AF recurrence after ablation based on observational studies has moderate certainty. Thus, we recommend AF patients who are obese may loss weight to maximize the effect of AF ablation. Policy implications and further researchh Policy implications and further research The 2020 ESC Guidelines set the weight loss target at ≥ 10% weight reduction for patients with BMI > 27 kg/ m2 for the improvement of AF ablation [12]. However, whether a smaller magnitude of weight loss benefits prognosis after ablation remains unknown. Our study showed that a 10% reduction in weight is associated with a decreased risk of AF recurrence based on observational studies. Furthermore, a potential linear inverse dose– response association is noted between weight loss and Meanwhile, we must recognize that the prognosis for AF ablation is related to gender. Some studies dem- onstrate that women might have a higher risk of AF Zhao et al. Nutrition & Metabolism (2023) 20:5 Zhao et al. Nutrition & Metabolism (2023) 20:5 Page 10 of 12 Page 10 of 12 Zhao et al. Nutrition & Metabolism reduced AF recurrence. Yet, we acknowledged that the dose–response association should be considered explor- atory due to the limited study numbers and intrinsic limitations of the observational design. Further trials are needed to confirm the effect of a smaller magnitude of weight loss. Conclusion Our meta-analysis suggests that weight loss is associ- ated with a decreased recurrence rate of AF after abla- tion, with moderate certainty. The effect of weight loss on AF burden, and quality of life in patients after AF ablation needs to be further studied. f Although we showed a significant benefit of weight loss on AF recurrence, specifics regarding the severity of AF and the burden of AF of weight loss are not clear. Evidence from the SORT-AF trial did not achieve a sig- nificant benefit for the AF burden. The evidence based on observational studies is also inclusive. However, emerging evidence has shown that benefits of weight loss on blood pressure, glucose, insulin resistance, low- density lipoprotein, cardiovascular events, and mortal- ity. We hypothesize that this benefit might also persist in AF ablation, which may have been even clearer in fur- ther trials with effective designs, such as a larger group of patients. Abbreviations AF Atrial fibrillation RRs Relative risks CI Confidence interval OR Odds’ ratio HR Hazard ratio NOS Newcastle–Ottawa evaluation scale RCTs Randomized controlled study BMI Body mass index Abbreviations AF Atrial fibrillation RRs Relative risks CI Confidence interval OR Odds’ ratio HR Hazard ratio NOS Newcastle–Ottawa evaluation scale RCTs Randomized controlled study BMI Body mass index Abbreviations AF Atrial fibrillation RRs Relative risks CI Confidence interval OR Odds’ ratio HR Hazard ratio NOS Newcastle–Ottawa evaluation scale RCTs Randomized controlled study BMI Body mass index Supplementary Information The online version contains supplementary material available at https://doi. org/10.1186/s12986-023-00724-5. Additional file 1. Table S1. Article content checklist. Table S2. Detailed description of the search strategy. Table S3. Studies excluded (n=24) with reasons. Table S4. Main Clinical characteristics of the included studies in the meta‐analysis. Table S5. Quality assessment of included randomized, open-labeled clinical trial. Table S6. Quality assessment of included stud- ies. Table S7. Summary of weight loss exposure dose for the included studies. Figure S1. Subgroup analysis of loss weight on AF recurrence after ablation, stratified by method of weight loss. Figure S2. Subgroup analysis of loss weight on AF recurrence after ablation, stratified by pre- ablative. Figure S3. Sensitivity analysis of loss weight on AF recurrence after ablation. Availability of data and materials The raw data required to reproduce these findings are available from the cor- responding author. 13. Middeldorp ME, Pathak RK, Meredith M, Mehta AB, Elliott AD, Mahajan R, et al. PREVEntion and regReSsive effect of weight-loss and risk factor modification on atrial fibrillation: the REVERSE-AF study. Europace. 2018;20(12):1929–35. Ethics approval and consent to participate This study does not involve human participants, does not require discussion on Ethics Committee(s) or Institutional Board(s). 14. Pathak RK, Middeldorp ME, Meredith M, Mehta AB, Mahajan R, Wong CX, et al. Long-term effect of goal-directed weight management in an atrial fibrillation cohort: a long-term follow-up study (LEGACY). J Am Coll Cardiol. 2015;65(20):2159–69. Received: 18 May 2022 Accepted: 4 January 2023 Received: 18 May 2022 Accepted: 4 January 2023 Received: 18 May 2022 Accepted: 4 January 2023 j 18. Xu C, Doi SAR. The robust error meta-regression method for dose– response meta-analysis. Int J Evid Based Healthcare. 2018;16(3):138–44. https://doi.org/10.1097/XEB.0000000000000132. 19. Donnellan E, Wazni OM, Kanj M, Baranowski B, Cremer P, Harb S, et al. Association between pre-ablation bariatric surgery and atrial fibrillation recurrence in morbidly obese patients undergoing atrial fibrillation abla- tion. Europace. 2019;21(10):1476–83. Funding This work was supported by the National Natural Science Foundation of China (Peng Yu, 81760050, 81760048; Xiao Liu, 82100347) and the Jiangxi Provincial Natural Science Foundation for Youth Scientific Research (Peng Yu, 20192ACBL21037); as well as the Young Teachers’ Basic Scientific Research Business Expenses Project (Wengen zhu, 20ykpy72); the China Postdoctoral Science Foundation (Xiao Liu, 2021M703724); Science and Technology Projects in Guangzhou (202102010007); and the China National Postdoctoral Program for Innovative Talents (Wengen zhu, BX20200400). 11. Mohanty S, Mohanty P, Di Biase L, Bai R, Pump A, Santangeli P, et al. Impact of metabolic syndrome on procedural outcomes in patients with atrial fibrillation undergoing catheter ablation. J Am Coll Cardiol. 2012;59(14):1295–301. 12. Hindricks G, Potpara T, Dagres N, Arbelo E, Bax JJ, Blomström-Lundqvist C, et al. 2020 ESC Guidelines for the diagnosis and management of atrial fibrillation developed in collaboration with the European Association for Cardio-Thoracic Surgery (EACTS): the task force for the diagnosis and management of atrial fibrillation of the European society of cardiology (ESC) developed with the special contribution of the European Heart Rhythm Association (EHRA) of the ESC. Eur Heart J. 2021;42(5):373–498. Acknowledgements None Acknowledgements None 7. Weerasooriya R, Khairy P, Litalien J, Macle L, Hocini M, Sacher F, et al. Catheter ablation for atrial fibrillation: are results maintained at 5 years of follow-up? J Am Coll Cardiol. 2011;57(2):160–6. Author contributions 8. Ouyang F, Tilz R, Chun J, Schmidt B, Wissner E, Zerm T, et al. Long-term results of catheter ablation in paroxysmal atrial fibrillation: lessons from a 5-year follow-up. Circulation. 2010;122(23):2368–77. Guarantor of the article: X-L, W.G-Z. X-L contributed to the study concept and design and revised the draft. H.L-Z and X.Z-L performed the search strategy and contributed to database research, acquisition of data, and statistical analyses. All the authors participated in data analysis, reviewed, and approved the final manuscript. 9. Steinberg JS, Palekar R, Sichrovsky T, Arshad A, Preminger M, Musat D, et al. Very long-term outcome after initially successful catheter ablation of atrial fibrillation. Heart Rhythm. 2014;11(5):771–6. 10. Jongnarangsin K, Chugh A, Good E, Mukerji S, Dey S, Crawford T, et al. Body mass index, obstructive sleep apnea, and outcomes of catheter ablation of atrial fibrillation. J Cardiovasc Electrophysiol. 2008;19(7):668–72. Competing interests None. 17. Greenland S, Longnecker MP. Methods for trend estimation from sum- marized dose-response data with applications to meta-analysis. Am J Epidemiol. 1992;135(11):1301–9. https://doi.org/10.1093/oxfordjournals. aje.a116237. Supplementary Information The online version contains supplementary material available at https://doi. org/10.1186/s12986-023-00724-5. Finally, several issues remain unresolved. For example, men might have a better outcome than women regard- ing AF ablation [47]. Furthermore, post hoc analysis from SORT-AF showed that weight loss was associated with a more pronounced reduction in the AF recurrence rate in persistent AF patients than in paroxysmal AF patients [47]. These observations highlight the complexity of the physiology of AF, which calls for more clinical studies. Additional file 1. Table S1. Article content checklist. Table S2. Detailed description of the search strategy. Table S3. Studies excluded (n=24) with reasons. Table S4. Main Clinical characteristics of the included studies in the meta‐analysis. Table S5. Quality assessment of included randomized, open-labeled clinical trial. Table S6. Quality assessment of included stud- ies. Table S7. Summary of weight loss exposure dose for the included studies. Figure S1. Subgroup analysis of loss weight on AF recurrence after ablation, stratified by method of weight loss. Figure S2. Subgroup analysis of loss weight on AF recurrence after ablation, stratified by pre- ablative. Figure S3. Sensitivity analysis of loss weight on AF recurrence after ablation. Additional file 1. Table S1. Article content checklist. Table S2. Detailed description of the search strategy. Table S3. Studies excluded (n=24) with reasons. Table S4. Main Clinical characteristics of the included studies in the meta‐analysis. Table S5. Quality assessment of included randomized, open-labeled clinical trial. Table S6. Quality assessment of included stud- ies. Table S7. Summary of weight loss exposure dose for the included studies. Figure S1. Subgroup analysis of loss weight on AF recurrence after ablation, stratified by method of weight loss. Figure S2. Subgroup analysis of loss weight on AF recurrence after ablation, stratified by pre- ablative. Figure S3. Sensitivity analysis of loss weight on AF recurrence after ablation. Page 11 of 12 Page 11 of 12 Zhao et al. Nutrition & Metabolism (2023) 20:5 Zhao et al. Nutrition & Metabolism Consent for publication We confirm that this manuscript is not under consideration for publication elsewhere and that none of the contents have been published previously. All authors have read and approved the manuscript and agree with its submis- sion to the Nutrition & Metabolism. 15. Page MJ, McKenzie JE, Bossuyt PM, Boutron I, Hoffmann TC, Mulrow CD, et al. The PRISMA 2020 statement: an updated guideline for reporting systematic reviews. Int J Surg. 2021;88:105906. 16. Liu X, Guo L, Xiao K, Zhu W, Liu M, Wan R, et al. The obesity paradox for outcomes in atrial fibrillation: evidence from an exposure-effect analysis of prospective studies. Obes Rev. 2020;21(3):e12970. References 1. Aune D, Feng T, Schlesinger S, Janszky I, Norat T, Riboli E. Diabetes mellitus, blood glucose and the risk of atrial fibrillation: a system- atic review and meta-analysis of cohort studies. J Diabetes Complic. 2018;32(5):501–11. 1. Aune D, Feng T, Schlesinger S, Janszky I, Norat T, Riboli E. Diabetes mellitus, blood glucose and the risk of atrial fibrillation: a system- atic review and meta-analysis of cohort studies. J Diabetes Complic. 2018;32(5):501–11. 20. Donnellan E, Wazni O, Kanj M, Hussein A, Baranowski B, Lindsay B, et al. Outcomes of atrial fibrillation ablation in morbidly obese patients follow- ing bariatric surgery compared with a nonobese cohort. Circ Arrhythm Electrophysiol. 2019;12(10):e007598. 2. Cadby G, McArdle N, Briffa T, Hillman DR, Simpson L, Knuiman M, et al. Severity of OSA is an independent predictor of incident atrial fibrillation hospitalization in a large sleep-clinic cohort. Chest. 2015;148(4):945–52. 21. Gessler N, Willems S, Steven D, Aberle J, Akbulak RO, Gosau N, et al. Supervised obesity reduction trial for AF ablation patients: results from the SORT-AF trial. Europace. 2021;23(10):1548–58. 3. Hobbelt AH, Siland JE, Geelhoed B, Van Der Harst P, Hillege HL, Van Gelder IC, et al. Clinical, biomarker, and genetic predictors of specific types of atrial fibrillation in a community-based cohort: data of the PREVEND study. Europace. 2017;19(2):226–32. 22. Lau Y, Mackey W, Connelly D. P347 Physician-led, weight loss advice prior to atrial fibrillation ablation, overall weight change and clinical outcomes. EP Europace. 2020;22(Supplement_1):euaa162. 340. 4. Nalliah CJ, Sanders P, Kalman JM. The impact of diet and lifestyle on atrial fibrillation. Curr Cardiol Rep. 2018;20(12):137. 23. Shah F, Sherazi S, Hall B, McNitt S, Rosero S, Huang D, et al., editors. Rela- tion of weight change to atrial fibrillation recurrence following catheter ablation. J Cardiovasc Electrophysiol; 2020: WILEY 111 RIVER ST, HOBOKEN 07030–5774, NJ USA. 5. Lip GYH, Coca A, Kahan T, Boriani G, Manolis AS, Olsen MH, et al. Hyper- tension and cardiac arrhythmias: a consensus document from the Euro- pean Heart Rhythm Association (EHRA) and ESC council on hypertension, endorsed by the Heart Rhythm Society (HRS), Asia-Pacific Heart Rhythm Society (APHRS) and Sociedad Latinoamericana de Estimulación Cardíaca y Electrofisiología (SOLEACE). Europace. 2017;19(6):891–911. 24. Yew Ding W, Kozhuharov N, Hao Chin S, Shaw M, Snowdon R, Lip GYH, et al. Feasibility of weight loss in obese atrial fibrillation patients attend- ing a specialist arrhythmia clinic and its impact on ablation outcomes. J Arrhythm. 2020;36(6):984–90. i 6. Calkins H, Reynolds MR, Spector P, Sondhi M, Xu Y, Martin A, et al. Treat- ment of atrial fibrillation with antiarrhythmic drugs or radiofrequency ablation: two systematic literature reviews and meta-analyses. Circ Arrhythm Electrophysiol. 2009;2(4):349–61. References 6. Calkins H, Reynolds MR, Spector P, Sondhi M, Xu Y, Martin A, et al. Treat- ment of atrial fibrillation with antiarrhythmic drugs or radiofrequency ablation: two systematic literature reviews and meta-analyses. Circ Arrhythm Electrophysiol. 2009;2(4):349–61. 25. Mohanty S, Mohanty P, Natale V, Trivedi C, Gianni C, Burkhardt JD, et al. Impact of weight loss on ablation outcome in obese patients with Page 12 of 12 Page 12 of 12 Zhao et al. Nutrition & Metabolism (2023) 20:5 Zhao et al. Nutrition & Metabolism Zhao et al. Nutrition & Metabolism (2023) 20:5 47. Staerk L, Sherer JA, Ko D, Benjamin EJ, Helm RH. Atrial fibrillation: epidemiology, pathophysiology, and clinical outcomes. Circ Res. 2017;120(9):1501–17. longstanding persistent atrial fibrillation. J Cardiovasc Electrophysiol. 2018;29(2):246–53. longstanding persistent atrial fibrillation. J Cardiovasc Electrophysiol. 2018;29(2):246–53. 26. Bunch TJ, May HT, Bair TL, Crandall BG, Cutler MJ, Jacobs V, et al. Long- term influence of body mass index on cardiovascular events after atrial fibrillation ablation. J Interv Card Electrophysiol. 2016;46(3):259–65. 48. Lavie CJ, Arena R, Alpert MA, Milani RV, Ventura HO. Management of cardiovascular diseases in patients with obesity. Nat Rev Cardiol. 2018;15(1):45–56. 27. Pathak RK, Middeldorp ME, Lau DH, Mehta AB, Mahajan R, Twomey D, et al. Aggressive risk factor reduction study for atrial fibrillation and implications for the outcome of ablation: the ARREST-AF cohort study. J Am Coll Cardiol. 2014;64(21):2222–31. Publisher’s Note Springer Nature remains neutral with regard to jurisdictional claims in pub- lished maps and institutional affiliations. 28. Wanahita N, Messerli FH, Bangalore S, Gami AS, Somers VK, Steinberg JS. Atrial fibrillation and obesity–results of a meta-analysis. Am Heart J. 2008;155(2):310–5. 29. Wang TJ, Parise H, Levy D, D’Agostino RB Sr, Wolf PA, Vasan RS, et al. Obesity and the risk of new-onset atrial fibrillation. JAMA. 2004;292(20):2471–7. 30. Liu M, Mei K, Liu X, Zhao Y. Impact of body mass index on procedural complications, procedure duration, and radiation dose in patients with atrial fibrillation undergoing radiofrequency ablation: a systematic review and meta-analysis. Clin Cardiol. 2020;43(10):1067–75. 31. Patel D, Mohanty P, Di Biase L, Sanchez JE, Shaheen MH, Burkhardt JD, et al. Outcomes and complications of catheter ablation for atrial fibrilla- tion in females. Heart Rhythm. 2010;7(2):167–72. 32. Clamp LD, Hume DJ, Lambert EV, Kroff J. Enhanced insulin sensitivity in successful, long-term weight loss maintainers compared with matched controls with no weight loss history. Nutr Diabetes. 2017;7(6):e282. 33. Tsalamandris S, Antonopoulos AS, Oikonomou E, Papamikroulis GA, Vogiatzi G, Papaioannou S, et al. The role of inflammation in diabetes: current concepts and future perspectives. Eur Cardiol. 2019;14(1):50–9. 34. Nalliah CJ, Sanders P, Kottkamp H, Kalman JM. The role of obesity in atrial fibrillation. Eur Heart J. 2016;37(20):1565–72. 35. Waranugraha Y, Rizal A, Yuniadi Y. A systematic review and meta-analysis of the direct comparison of second-generation cryoballoon ablation and contact force-sensing radiofrequency ablation in patients with paroxys- mal atrial fibrillation. J Pers Med. 2022;12(2):298. i 36. Cheng X, Hu Q, Gao L, Liu J, Qin S, Zhang D. Sex-related differences in catheter ablation of atrial fibrillation: a systematic review and meta- analysis. Europace. 2019;21(10):1509–18. 37. Asad ZUA, Yousif A, Khan MS, Al-Khatib SM, Stavrakis S. Catheter ablation versus medical therapy for atrial fibrillation: a systematic review and meta-analysis of randomized controlled trials. Circ Arrhythm Electrophys- iol. 2019;12(9):e007414. 38. Akoum N, Mahnkopf C, Kholmovski EG, Brachmann J, Marrouche NF. Age and sex differences in atrial fibrosis among patients with atrial fibrillation. Europace. 2018;20(7):1086–92. 39. Forleo GB, Tondo C, De Luca L, Dello Russo A, Casella M, De Sanctis V, et al. Gender-related differences in catheter ablation of atrial fibrillation. Europace. 2007;9(8):613–20. 40. Jones NR, Taylor KS, Taylor CJ, Aveyard P. Weight change and the risk of incident atrial fibrillation: a systematic review and meta-analysis. Heart. 2019;105(23):1799–805. 41. Aldaas OM, Malladi CL, Hsu JC. • fast, convenient online submission • thorough peer review by experienced researchers in your ﬁeld • rapid publication on acceptance • support for research data, including large and complex data types • gold Open Access which fosters wider collaboration and increased citations maximum visibility for your research: over 100M website views per year • At BMC, research is always in progress. Learn more biomedcentral.com/submissions Ready to submit your research Ready to submit your research ? Choose BMC and benefit from: ? Choose BMC and benefit from: Publisher’s Note Catheter ablation of atrial fibrillation in patients with heart failure. Am J Cardiol. 2019;123(1):187–95. 42. Park DY, An S, Murthi M, Kattoor AJ, Kaur A, Ravi V, et al. Effect of weight loss on recurrence of atrial fibrillation after ablative therapy: a systematic review and meta-analysis. J Interv Card Electrophysiol. 2022. https://doi. org/10.1007/s10840-022-01168-2. 43. Munger TM, Dong YX, Masaki M, Oh JK, Mankad SV, Borlaug BA, et al. Elec- trophysiological and hemodynamic characteristics associated with obe- sity in patients with atrial fibrillation. J Am Coll Cardiol. 2012;60(9):851–60. • fast, convenient online submission • thorough peer review by experienced researchers in your ﬁeld • rapid publication on acceptance • support for research data, including large and complex data types • gold Open Access which fosters wider collaboration and increased citations maximum visibility for your research: over 100M website views per year • At BMC, research is always in progress. Learn more biomedcentral.com/submissions Ready to submit your research Ready to submit your research ? Choose BMC and benefit from: ? Choose BMC and benefit from: i 44. Ariyaratnam JP, Middeldorp M, Thomas G, Noubiap JJ, Lau D, Sanders P. Risk factor management before and after atrial fibrillation ablation. Cardiac Electrophysiol Clin. 2020;12(2):141–54. 45. Abed HS, Wittert GA, Leong DP, Shirazi MG, Bahrami B, Middeldorp ME, et al. Effect of weight reduction and cardiometabolic risk factor manage- ment on symptom burden and severity in patients with atrial fibrillation: a randomized clinical trial. JAMA. 2013;310(19):2050–60. 46. Jamaly S, Carlsson L, Peltonen M, Jacobson P, Sjöström L, Karason K. Bari- atric surgery and the risk of new-onset atrial fibrillation in swedish obese subjects. J Am Coll Cardiol. 2016;68(23):2497–504.
https://openalex.org/W4254424661	http://www.scielo.cl/pdf/ssa/v17n2/0719-0948-ssa-17-02-00195.pdf	Spanish; Castilian	null	Reseña	Si somos americanos	2,017	cc-by	1,416	* Licenciado en Ciencias Políticas y Administrativas, Universidad de Concepción, Chile. Magíster en Seguridad y Defensa, Academia Nacional de Estudios Políticos y Estratégicos, Chile. Master in International Service, American University, Washington DC. Doctor (c) en Estudios Internacionales, FLACSO-Ecuador. Dirección: Calle La Pradera E7-174 y Av. Diego de Almagro, Torre 2, 6to piso. Código Postal 170518. Correo electrónico: cabrera.lester@gmail.com Afiliación institucional: investigador doctoral, FLACSO-Ecuador. Todo sobre Bolivia y la compleja disputa por el mar, de José Rodríguez Elizondo. Ediciones El Mercurio. Agosto 2016 Lester Cabrera Toledo* El “tema del mar” puede ser considerado como uno de los fenómenos más complejos que al día de hoy, aquejan a las relaciones entre países de América del Sur y, específicamente, entre Bolivia y Chile. Pero la complejidad no solamente radica en el hecho de que existe una reclamación territorial por parte de un país hacia otro, lo cual se podría circunscribir en un aspecto netamente político y jurídico, sino que la complejidad deviene de una serie de elementos que se ubican más en el plano geopolítico, social, e incluso como parte de la idiosincrasia de los países que se ven involucrados, tanto directa como indirectamente, en el mencionado proceso. Esto es, a grandes rasgos, lo que el último libro de José Rodríguez Elizondo, Todo sobre Bolivia y la compleja disputa por el mar, busca dilucidar. El texto no intenta ser un manual sobre el cual se pueda resolver un problema que posee sus orígenes en diferencias dadas desde el siglo XIX, sino que trata de reflejar las complejidades que se evidencian en un problema como lo es la reivindicación marítima boliviana al Océano Pacífico. Para esto, Rodríguez Elizondo realiza frecuentes viajes al pasado, como forma de reflejar que el problema no es nuevo, y que las soluciones que muchas veces se plantean en el presente carecen de una novedad real. Aquello lo establece con la debida claridad cuando cita los intentos de los presidentes chilenos Gabriel González Videla y Ricardo Lagos Escobar, al momento de generar eventuales soluciones al enclaustramiento boliviano. El autor no se queda solamente en eso y señala explícitamente que muchos de los aspectos que en la actualidad se discuten, tuvieron una solución pragmática en el pasado, pero que la incomprensión y desconocimiento de los reales alcances que aquellos tuvieron (y tienen) son uno de los principales problemas que se afrontan. El ejemplo más claro al respecto se da en la suscripción del Tratado de Lima en 1929 y su Protocolo Complementario. Dicho acuerdo internacional dio origen a la Si Somos Americanos. Revista de Estudios Transfronterizos 195 célebre frase del presidente boliviano Daniel Salamanca Urey: “Chile le puso el candado a la salida al mar de Bolivia y le entregó la llave al Perú”. Aquella afirmación tiene un especial significado y tratamiento a lo largo del libro. Todo sobre Bolivia y la compleja disputa por el mar, de José Rodríguez Elizondo. Ediciones El Mercurio. Agosto 2016 De acuerdo tanto a argumentos históricos como coyunturales, el problema de otorgar una eventual salida al mar a Bolivia pasa por el reconocimiento de tres grandes factores. En primer lugar, el necesario involucramiento de Perú; el posicionamiento geopolítico de Arica; y tal vez el factor más complejo, la aceptación por parte de los tomadores de decisiones en el ámbito de la política exterior de los tres países, en que la naturaleza del problema pasa por el adecuado entendimiento de los dos puntos antes señalados. Y pese a que dentro de la obra dichas perspectivas poseen un tratamiento diferenciado, una visión global de los temas es una de las fortalezas en el análisis realizado. Cuando se toma la posición sobre Perú en el tema de la aspiración marítima boliviana, se realiza sobre una perspectiva pragmática; es decir, cuando Bolivia establece una necesidad en torno a una salida soberana hacia el Océano Pacífico, particularmente por un corredor, este debe tener la cualidad de no dividir territorialmente a Chile. Y para que aquella condición se genere, solamente debe realizarse en un punto en la frontera de Chile con Perú; sin embargo, si Chile decide ceder aquel espacio territorial, debe contar con la debida autorización del Perú. Es por ello que en la obra se menciona de manera recurrente la concepción trilateral de la aspiración boliviana, pese a que se señale por las partes involucradas que aquello es estrictamente bilateral, lo cual se demuestra en amplios pasajes a lo largo del texto. Producto de aquella cualidad geopolítica es que Arica, ciudad que queda al extremo norte de Chile y que en gran parte del siglo XIX perteneció al Perú, se convierte en un punto clave para la totalidad del análisis en la obra, no solamente por su ubicación geográfica, sino por la cercanía a Bolivia, sus características portuarias y por ser un punto clave para el comercio del país altiplánico. Pero Arica no solamente es un punto clave para Bolivia, sino también para Perú, por ser parte de un espacio territorial generador de la propia identidad peruana, haciendo una alusión directa al sacrificio de héroes del otro lado de la Línea de la Concordia como Francisco Bolognesi y Alfonso Ugarte. Todo sobre Bolivia y la compleja disputa por el mar, de José Rodríguez Elizondo. Ediciones El Mercurio. Agosto 2016 El punto más importante en el libro, no obstante, es la permanente crítica que realiza José Rodríguez Elizondo a los tomadores de decisión de los tres países involucrados, al no reconocer de manera abierta la propia naturaleza del problema en cuestión. Para el autor, al ser el problema de naturaleza más geopolítica que jurídica, aquello no puede encapsularse dentro de una lógica bilateral, tomando en cuenta los múltiples intereses que poseen cada uno de los actores en cuestión. Pero al negar aquella realidad, el problema no solamente se atomiza, sino que deriva en soluciones que no dan una respuesta a la real naturaleza de la problemática, incluso al punto de aplicar nuevos elementos conflictivos al proceso. La nacionalidad chilena del autor no le resta objetividad a la obra. Incluso, es posible evidenciar una postura muy crítica con respecto a la gestión que ha realizado la Cancillería chilena en las controversias jurídicas que se han llevado a La Haya. Es así como el autor coloca un especial énfasis en las estructuras sicológicas y sociológicas en las cuales los tomadores de decisión de la política exterior chilena evidencian, como por ejemplo el estricto e inamovible apego a lo que dicta el Derecho Internacional, la falta de imaginación geopolítica, como también la carencia de “otras voces”. Estos, a juicio del autor, son los elementos que principalmente explican por qué el país (Chile) ha fallado en las negociaciones (o simplemente se ha negado a negociar), con resultados que no son los más positivos a la hora de hablar de protección y resguardo de los intereses nacionales. Es por ello que aboga, como en sus últimas cuatro obras al respecto, por una necesaria modernización de la Cancillería, la cual tenga al Derecho como un apoyo y medio, pero no como un fin. Una de las principales cualidades del autor, lo cual también se evidencia en sus anteriores obras relacionadas a la temática, es que narra de manera simple y amena los complejos acontecimientos que se dan en la relación, los cuales mezcla tanto con vivencias propias, como también con hallazgos encontrados en sus investigaciones previas. Pero lo que llama la atención es que dentro de la obra es posible encontrar una serie de hipótesis que se buscan responder y comprobar, como una manera de concebir los sucesos desde un prisma periodístico. Todo sobre Bolivia y la compleja disputa por el mar, de José Rodríguez Elizondo. Ediciones El Mercurio. Agosto 2016 Aquello otorga una visión diferente, ya que no solamente se dedica a describir sucesos, sino que les otorga un valor, con el objetivo de resolver un problema. Con esto, se desmarca de un tratamiento netamente histórico o jurídico, que tanto abundan cuando se buscan explicar los pormenores de la relación boliviano-chilena. Tomando en consideración el momento que se vive en la relación boliviano-chilena y las demandas interpuestas por ambos países en La Haya, Todo sobre Bolivia y la compleja disputa por el mar da luces sobre cómo, dentro de la complejidad y negatividad que se vislumbra en un primer momento, es posible encontrar puntos en común. Se trata de una obra propositiva, que todo estudioso de las relaciones entre Bolivia, Chile y Perú, como amante de la geopolítica, no puede dejar de leer. Si Somos Americanos. Revista de Estudios Transfronterizos 197
https://openalex.org/W2580441030	https://qmro.qmul.ac.uk/xmlui/bitstream/123456789/69827/2/Wanyonyi_Dental-Treatment-State-Funded_Published_2017.pdf	English	null	Dental Treatment in a State-Funded Primary Dental Care Facility: Contextual and Individual Predictors of Treatment Need?	PloS one	2,017	cc-by	9,461	RESEARCH ARTICLE Methods Routine clinical and demographic data were extracted from a live dental patient manage- ment system in a state-funded facility using novel methods. The data, spanning a four-year period [2008–2012] were cleaned, validated, linked by means of postcode to deprivation status, and analysed to identify factors which predict dental treatment need. The predictive relationship between patients’ individual characteristics (demography, smoking, payment status) and contextual experience (deprivation based on area of residence), with common dental treatments received was examined using unadjusted analysis and adjusted logistic regression. Additionally, multilevel modelling was used to establish the isolated influence of area of residence on treatments. Editor: Peter M. Milgrom, University of Washington, UNITED STATES Received: April 8, 2016 Accepted: December 9, 2016 Published: January 24, 2017 Copyright: © 2017 Wanyonyi et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited. Editor: Peter M. Milgrom, University of Washington, UNITED STATES Received: April 8, 2016 Accepted: December 9, 2016 Published: January 24, 2017 Copyright: © 2017 Wanyonyi et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited. Kristina L. Wanyonyi1,2, David R. Radford1,3, Jennifer E. Gallagher2 1 University of Portsmouth Dental Academy, Hampshire Terrace, Portsmouth, United Kingdom, 2 King’s College London Dental Institute, Population and Patient Health Division, London, United Kingdom, 3 King’s College London Dental Institute, Teaching Division, Guys Tower, Guys Hospital, London, United Kingdom a1111111111 a1111111111 a1111111111 a1111111111 a1111111111 a1111111111 a1111111111 a1111111111 a1111111111 a1111111111 kristina.wanyonyi@port.ac.uk OPEN ACCESS This study examined individual and contextual factors which predict the dental care received by patients in a state-funded primary dental care training facility in England. Citation: Wanyonyi KL, Radford DR, Gallagher JE (2017) Dental Treatment in a State-Funded Primary Dental Care Facility: Contextual and Individual Predictors of Treatment Need? PLoS ONE 12(1): e0169004. doi:10.1371/journal.pone.0169004 Dental Treatment in a State-Funded Primary Dental Care Facility: Contextual and Individual Predictors of Treatment Need? Kristina L. Wanyonyi1,2, David R. Radford1,3, Jennifer E. Gallagher2 Introduction The World Health Organization (WHO) is promoting Universal Health Coverage (UHC) in support of achieving the 2030 Sustainable Development Goals [1]. Research within state- funded health systems such as the National Health Service (NHS) [2] in England provides evi- dence of how opportunities to access dental services free at the point of delivery (children and vulnerable adults) or at reduced cost (adults make a co-payment), align to oral health needs. There is clear evidence that variations in dental service uptake are associated with social status [3,4], with socially deprived groups accessing care less frequently [4], and requiring more extensive services such as treatment under sedation [5], when they do access care. These pat- terns of access further contribute to increased oral health need and inequalities across the socio-economic spectrum, despite the state-funded service provision [6]. What is unknown is how routine care provision relates to need once access to dental services is gained, and how this impacts on equity of health outcomes. This is particularly important for state-funded den- tal services such as the NHS which is committed to maximising preventative care [7], as well as equity and quality [8]. The NHS serves a large proportion of the national population with seven out of 10 children, and five out of 10 adults, attending primary dental care within a 24-month period [9]. So far the analysis of NHS data has predominantly been studied to moni- tor new initiatives [10], assess value for money [11], and the longevity of treatments [12–19], with much of this research conducted under previous models of care. A more analytical evalu- ation of dental activity from contemporary NHS primary care has the potential to provide information on how encounters with health care under the current system contribute to addressing oral health needs. More specifically, dental activity records, which are routinely collected within patient man- agement systems, could answer questions regarding what happens when patients enter primary dental care. Is patients’ treatment related to known patterns of need and health behaviours such as smoking, and how does this relate to NHS provisions to improve access such as payment exemption? Finally, this could further inform understanding of how care relates to contextual level predictors of oral health need. Thus patient management systems are potentially a rich data source consisting of patient demography, and care received experience. Conclusion This is the first study to model patient management data from a state-funded dental service and show that individual and contextual factors predict common treatments received. Impli- cations of this research include the importance of making provision for our aging population and ensuring that preventative care is available to all. Further research is required to explain the interaction of organisational and system policies, practitioner and patient perspectives on care and, thus, inform effective commissioning and provision of dental services. Dental Treatment and Predictors of Need the multilevel models was a significant difference in treatment related to area of residence; adults from the most deprived quintile were more likely to receive ‘tooth extraction’ when compared with least deprived, and less likely to receive preventive ‘instruction and advice’ (p = 0.01). the multilevel models was a significant difference in treatment related to area of residence; adults from the most deprived quintile were more likely to receive ‘tooth extraction’ when compared with least deprived, and less likely to receive preventive ‘instruction and advice’ (p = 0.01). KCLDI. KLW contributed to the undergraduate dental teaching programme at KCLDI and is now on the staff of UPDA. Results Data on 6,351 dental patients extracted comprised of 147,417 treatment procedures deliv- ered across 10,371 courses of care. Individual level factors associated with the treatments were age, sex, payment exemption and smoking status and deprivation associated with area of residence was a contextual predictor of treatment. More than 50% of children (<18 years) and older adults (65 years) received preventive care in the form of ‘instruction and advice’, compared with 46% of working age adults (18–64 years); p = 0.001. The odds of receiving treatment increased with each increasing year of age amongst adults (p = 0.001): ‘partial dentures’ (7%); ‘scale and polish’ (3.7%); ‘tooth extraction’ (3%; p = 0.001), and ‘instruction and advice’ (3%; p = 0.001). Smokers had a higher likelihood of receiving all treatments; and were notably over four times more likely to receive ‘instruction and advice’ than non-smokers (OR 4.124; 95% CI: 3.088–5.508; p = 0.01). A further new finding from Data Availability Statement: All relevant data are within the paper and its Supporting Information files. Funding: KLW received funding from King’s College London Dental Institute and University of Portsmouth Dental Academy for a PhD studentship to undertake this work. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript. Competing Interests: DRR is Director of Clinical Studies for the Dental Students at UPDA, whilst JEG leads Dental Public Health teaching across PLOS ONE \| DOI:10.1371/journal.pone.0169004 January 24, 2017 1 / 17 PLOS ONE \| DOI:10.1371/journal.pone.0169004 January 24, 2017 Materials and Methods Ethical approval for this research was provided by NRES Committee Fulham REC: Reference No. 11/LO/1138 Protocol No. NTMHWMOV3 and research governance approval by NHS Portsmouth R&D Committee Reference No. SSPS/05/11. This research was conducted on patient care at the University of Portsmouth Dental Acad- emy (UPDA), a state-funded NHS primary care service and training centre, where dental stu- dents on community outreach, and UPDA dental hygiene-therapy students, learn together in practice teams. There is a strong ethos on using the skill mix of the dental team and delivering contemporary evidence-based primary care. For two out of the four years under study, patients were treated at no cost. UPDA holds a dental contract with the NHS similar to practices across England, which requires performance targets to be achieved annually. To manage patient care the centre uses a live patient management system (Clinical +) developed by Carestream Lim- ited [24,25], from which data for this study were extracted. The extracted data constituted courses of care, which had either been completed or closed within the four-year period, including both emergency and planned care. Courses of care were ‘closed’ when patients failed to return for care but could not be distinguished from ‘completed’ courses in the system. The first two years involved dental hygiene-therapy students providing care supported by general dental practitioners, whilst the latter two years involved team train- ing between dental and dental hygiene-therapy students, with care provided free at the point of delivery. The data were cleaned and validated using a combination of techniques adapted from health services research and information technology [26–30]. The process identified data inputting errors and software design problems. After data were cleaned a validation analysis was undertaken to identify outliers and inconsistency in the data when compared to other locally generated reports. Within the extracted data multiple courses of care were included and each treatment received was linked to date of completion or closing of the care plan. This was further linked to the patient’s age, sex, payment-exemption status, postcode and smoking sta- tus. Smoking status was only available for patients who had been seen at least once in the two latter years. Postcode data were transformed using the Indices of Multiple Deprivation (IMD) to provide a contextual measure; area level deprivation. Introduction Information on patient’s residence is also available, and when augmented with census data, can provide infor- mation about the patients’ deprivation at area of residence [20], which represents a contextual variable. Obtaining data from patient management systems eliminates recall, selection and PLOS ONE \| DOI:10.1371/journal.pone.0169004 January 24, 2017 2 / 17 Dental Treatment and Predictors of Need social desirability biases common to surveys; an alternative research method [21]. In the USA, dental claims data derived from patient management systems have been used to investigate pat- terns in children’s oral health [22]. In Ireland, similar data have been used to describe national trends in treatment requirements [23]. Whilst in the NHS in England, primary care data gener- ated from these systems have been mainly used to monitor and negotiate contracts with provid- ers, and only recently to predict future demand [11]. The literature to date is largely descriptive in nature; therefore, leaving the analytical potential of dental data underexplored. Thus, there is a need for research to mine routinely collected primary care data in order to understand pat- terns of dental care and advance in this field of research. The objective of this study was to investigate the relationship between dental care received and both individual (age, sex, adult payment status and smoking behaviour) and contextual factors (deprivation in the patients’ immediate living environment) using routinely collected patient management data. PLOS ONE \| DOI:10.1371/journal.pone.0169004 January 24, 2017 Materials and Methods The latter is a measure used in England to provide a relative measure of deprivation at Lower Level Super Output Area (LSOA), which is an area of 1000–1500 households describing the cumulative deprivation score of individuals in the area [31]. Payment-exemption status was used in the analysis; it has been shown to cor- relate with income deprivation at the area level, but not overall deprivation at the area level PLOS ONE \| DOI:10.1371/journal.pone.0169004 January 24, 2017 3 / 17 Dental Treatment and Predictors of Need [20]. Therefore, this was used as a measure of individual income deprivation in the analysis, together with the variable for overall contextual area level deprivation—identified as a patient’s quintile of deprivation–and derived from IMD [31]. Analysis was undertaken in five stages. First, descriptive analysis of patient demography was undertaken to provide insight to the patient base. Second, analysis of all treatments received and their rate of occurrence. Third, the five treatments with the highest volume of activity were further analysed against the socio-demography of patients using unadjusted anal- ysis and chi-square tests of significance; to examine differences in treatment rates by socio- demography. The treatments included were i] ‘tooth restoration’, ii] ‘instruction and advice’ (information on brushing teeth and diet), iii] ‘scale and polish’, iv] ‘tooth extraction’, and v] ‘partial dentures’. For this unadjusted analysis, the age variable was converted to a categorical variable of three groups (< 18 years, i.e. children), 18–64 years (working-age adults) and 65 years and over (older people). This was in order to validate the data through outlining expected and established differences in cohort effect between these three sections of the population. Fourth, adjusted logistic regression was used to examine predictors of common treatments while controlling for confounding variables. Five separate logistic regression models were undertaken. This involved five sub-sets of data, one for each treatment. Each dataset consisted of all adult patients, and the creation of a binary variable indicating whether they had or had not received specific treatments. Only adult patients were included in the logistic regression analysis to mitigate bias associated with non-fee-paying children. Age was then transformed to a continuous variable for this adjusted logistic regression analysis, as this second analysis was exploratory of adults only, and sought to investigate less established links between patient-age of adults and the range of treatment types. Participants The patient management dataset comprised 6,351 patients that had received 147,417 treatment procedures of care across 10,371 courses of care over the four-year study period [2008–2012]. All courses of care extracted were either completed or closed; the latter because the patient had not returned to complete care. The majority were adult (82%), male (52.2%), and non-smokers as shown in Table 1. The age range across the four-year period was 1–94 years and patients from the most deprived quintile (23.3%) exceeded those from the least deprived (11.9%). Materials and Methods Receiver Operator Curve (ROC) was used to vali- date the models’ predictive power [0.5–0.6 weak; 0.6–0.7 fair; 0.8-and above strong model pre- dicting power]. Fifth, and finally, multilevel modelling was undertaken. Logistic regression models which had indicated existing associations with area level measures of deprivation of p<0.05 (quintile of deprivation) were subjected to this analysis in order to establish whether there was grouping at the level of residence that predicted care. LSOAs were selected as the grouping variable, as this was the smallest aggregating variable available to be augmented to the data set [32]. This multilevel analysis was able to extend the regression analysis to a situation where data were hierarchical [33], and test the potential influence of unknown factors at the level of area of resi- dence on health. Patient related variables Frequency % Adult exemption status (n = 5185) Exempt adult 1,005 19.4 Non-exempt adult 4,180 80.6 Age groups (n = 6,351) 0–2yrs 85 1.3 3–5yrs 247 3.9 6–12yrs 541 8.5 13–17yrs 274 4.3 18–24yrs 1,211 19.1 25–34yrs 1,272 20 35–44yrss 1,008 15.9 45–54yrs 813 12.8 55–64yrs 494 7.8 65–74yrs 260 4.1 Over 75yrs 146 2.3 Quintile of deprivation (n = 6259) Most deprived 1,477 23.3 2 1,318 20.8 3 1,414 22.3 4 1,314 20.7 Least deprived 736 11.6 Sex (n = 6,351) Female 3,098 48.8 Male 3,253 52.2 Smoking status (n = 3436) Non-smoker 2,803 81.6 Smoker 633 18.4 Note n = 6,351 unless otherwise stated Age groupings are based on state-funded banding UPDA- University of Portsmouth Dental Academy doi:10.1371/journal.pone.0169004.t001 PLOS ONE \| DOI:10.1371/journal.pone.0169004 January 24, 2017 5 / 17 Table 1. Individual and social characteristic of patients with closed/completed treatment plans between 2008/09 to 2011/12 academic years at UPDA. Patient related variables Frequency % Adult exemption status (n = 5185) Exempt adult 1,005 19.4 Non-exempt adult 4,180 80.6 Age groups (n = 6,351) 0–2yrs 85 1.3 3–5yrs 247 3.9 6–12yrs 541 8.5 13–17yrs 274 4.3 18–24yrs 1,211 19.1 25–34yrs 1,272 20 35–44yrss 1,008 15.9 45–54yrs 813 12.8 55–64yrs 494 7.8 65–74yrs 260 4.1 Over 75yrs 146 2.3 Quintile of deprivation (n = 6259) Most deprived 1,477 23.3 2 1,318 20.8 3 1,414 22.3 4 1,314 20.7 Least deprived 736 11.6 Sex (n = 6,351) Female 3,098 48.8 Male 3,253 52.2 Smoking status (n = 3436) Non-smoker 2,803 81.6 Smoker 633 18.4 cteristic of patients with closed/completed treatment plans between 2008/09 to 2011/12 academic years at Treatments and socio-demography Of the 147,417 treatments delivered, the five most frequently occurring were: ‘tooth restora- tion’ (51.5%); ‘instruction and advice’ (49.2%); ‘scale and polish’ (38.7%); ‘tooth extraction’ (25.1%); and ‘partial dentures’ (5.1%). 4 / 17 PLOS ONE \| DOI:10.1371/journal.pone.0169004 January 24, 2017 Dental Treatment and Predictors of Need Unadjusted analysis (adults and children) The results of the unadjusted analysis, which included adults and children (Table 2), suggest that the proportion of patients receiving at least one ‘tooth restoration’ in the four-year period significantly differed by age, sex and smoking status, but not by social factors (deprivation or payment status). Older (65 years) and working age adults (18–64 years) had a higher rate of ‘tooth restoration’ (54.4% and 54.3% respectively), exceeding the volume amongst children and young people (<18 years) (38%); p = 0.001. A higher proportion of males had received a ‘tooth restoration’ compared with females (54% cf 49%; p = 0.001). Interestingly, smokers had a higher proportion of patients with ‘tooth restoration’ compared with non-smokers (67% cf 54%; p = 0.01). In contrast to the above, ‘instruction and advice’ had been received by a higher proportion of adults who were non-exempt from payment compared with exempt (50.3% cf. 32.1%; p = 0.01. Similarly more adult patients from the least deprived areas compared with the most deprived had received ‘instruction and advice’ (53.9 cf 46.1%; p = 0.003), whilst more smokers had received ‘instruction and advice’ than non-smokers (90.2% cf 74.5%: p = 0.001). There was no significant difference in receipt of ‘instruction and advice’ by sex. A higher proportion of smokers had received ‘scale and polish’ than non-smokers (68.2% cf 47.7%; p = 0.001), as had a higher proportion of non-exempt than exempt adults (47% cf 34%; p = 0.001). Analysis by deprivation status suggested that 47% of those from the least deprived Table 1. Individual and social characteristic of patients with closed/completed treatment plans between 2008/09 to 2011/12 academic years at UPDA. Unadjusted analysis (adults and children) The results of the unadjusted analysis, which included adults and children (Table 2), suggest that the proportion of patients receiving at least one ‘tooth restoration’ in the four-year period significantly differed by age, sex and smoking status, but not by social factors (deprivation or payment status). Older (65 years) and working age adults (18–64 years) had a higher rate of ‘tooth restoration’ (54.4% and 54.3% respectively), exceeding the volume amongst children and young people (<18 years) (38%); p = 0.001. A higher proportion of males had received a ‘tooth restoration’ compared with females (54% cf 49%; p = 0.001). Interestingly, smokers had a higher proportion of patients with ‘tooth restoration’ compared with non-smokers (67% cf 54%; p = 0.01). In contrast to the above, ‘instruction and advice’ had been received by a higher proportion of adults who were non-exempt from payment compared with exempt (50.3% cf. 32.1%; p = 0.01. Similarly more adult patients from the least deprived areas compared with the most deprived had received ‘instruction and advice’ (53.9 cf 46.1%; p = 0.003), whilst more smokers had received ‘instruction and advice’ than non-smokers (90.2% cf 74.5%: p = 0.001). There was no significant difference in receipt of ‘instruction and advice’ by sex. A higher proportion of smokers had received ‘scale and polish’ than non-smokers (68.2% cf 47.7%; p = 0.001), as had a higher proportion of non-exempt than exempt adults (47% cf 34%; p = 0.001). Analysis by deprivation status suggested that 47% of those from the least deprived PLOS ONE \| DOI:10.1371/journal.pone.0169004 January 24, 2017 5 / 17 Dental Treatment and Predictors of Need Table 2. Unadjusted model of proportion of patients who experienced a treatment by sociodemography between 2008/09 to 2011/12 academic years at UPDA. Patient related variables Never received treatment within four academic years N (%) Received treatment within four academic years N(%) P value Tooth restoration Overall 3,083 (48.5) 3,268(51.5) Adult payment status (n = 5,185) Exempt 475 (47.3) 530 (52.7) 0.236 Non-exempt 1,889 (45.2) 2291(54.8) Age groups (n = 6,351) Under 18 707 (61.6) 440 (38.4) 0.001 18–64 years 2,209 (45.7) 2,629 (54.3) Over 65 167 (45.6) 199 (54.4) Sex (n = 6,351) Female 1,576 (50.9) 1,522 (49.1) Male 1,507 (46.3) 1,746 (53.7) 0.001* Quintiles of deprivation in PCT (n = 6,259) Most deprived1 741 (50.2) 736 (49.8) 0.139 2 657 (49.8) 661 (50.2) 3 684 (48.4) 730 (51.6) 4 600 (45.7) 714 (54.3) Least deprived 5 353 (48.0) 383 (52.0) Smoking cessation signposting (n = 3436) No 1304 (46.5) 1499 (53.5) 0.001* Yes 211(33.3) 422 (66.7) Instruction and advice Overall 3,224 (50.8) 3,127(49.2) Adult payment status (n = 5,185) Exempt 682 (67.9) 283 (32.1) 0.01* Non-exempt 2,078 (49.7) 2,102 (50.3) Age groups (n = 6,351) Under 18 468 (40.8) 679 (59.2) 0.001* 18–64 years 2,597 (54.1) 2,201 (45.9) Over 65 159 (39.2) 247 (60.8) Sex (n = 6,351) Female 1,584 (51.1) 1,514 (48.9) 0.293 Male 1,640 (50.4) 1,613 (49.6) Quintiles of deprivation in PCT (n = 6,259) Most deprived 1 796 (53.9) 681 (46.1) 0.003* 2 676 (51.3) 642 (48.7) 3 721 (51.0) 693 (49.0) 4 631 (48.0) 683 (52.0) Least deprived 5 339 (46.1) 397 (53.9) Smoking status (n = 3436) No 716 (25.5 2,087 (74.5) 0.001* Yes 62 (9.8) 571 (90.2) Scale and polish Overall 3,890 (61.3) 2,461(38.7) Adult payment status (n = 5,185) Exempt 662 (65.9) 343 (34.1) 0.001* Non-exempt 2,199 (52.6) 1,981 (47.4) Age groups (n = 6,351) Under 18 1,028 (89.6) 119 (10.4) 0.001* 18–64 years 2,717 (56.2) 2,121 (43.8) Over 65 145 (39.6) 221 (60.4) Sex (n = 6,351) Female 2,390 (77.1) 708 (22.9) 0.398 Male 1,998 (61.4) 1,255 (38.6) Quintiles of deprivation in PCT (n = 6,259) Most deprived 1 1,027 (69.5) 450 (30.5) 0.001* 2 775 (58.8) 543 (41.2) 3 841 (59.5) 573 (40.5) (Continued) on of patients who experienced a treatment by sociodemography between 2008/09 to 2011/12 academic PLOS ONE \| DOI:10.1371/journal.pone.0169004 January 24, 2017 6 / 17 quintile had received a ‘scale and polish’, compared with only 31% of those from the most deprived quintile (p = 0.001). Adjusted analysis (adults only) Further analysis, involving adjusted regression modelling on adult patient data (n = 2,782; 70%), revealed an association between adult patient characteristics and treatment, controlling for other variables. The partial denture model had the strongest predictive power (ROC = 0.83) and tooth restoration the weakest (ROC 0.64), whilst the other models were fairly good (ROC = 0.7). Each of the treatments is presented in Table 3 starting with the most common, tooth restoration. Of the variables examined, ‘exemption from payment’ was the strongest predictor of ‘tooth restoration’, with exempt adults being twice as likely as non-exempt to receive a ‘tooth restora- tion’; (p = 0.001). There was a higher likelihood of receiving one or more restorations for adults in the two most deprived groups than the least deprived (65%; p = 0.01 and 37%; p = 0.012). Increasing adult age was associated with a higher likelihood of receiving any of the common treatments. With each year of increasing age, adults were more likely to receive treatment as follows: partial denture (7%; p = 0.01); ‘scale and polish’ (3.7%); ‘tooth extraction’ (3%; p = 0.001) and ‘instruction and advice’ (3%; p = 0.001). Patients identified as smokers were more likely to require one or more of the spectrum of treatments compared with non-smokers (p = 0.01); they were four times more likely to receive ‘instruction and advice’; three times more likely to receive a partial denture; twice as likely to receive a ‘tooth extraction’, and just under twice as likely to receive a ‘scale and polish’ (x1.7) and ‘tooth restoration’ (x1.5). Patients exempt from patient charges were more likely to have received one or more of the following than those who pay charges: partial dentures (x2.6); ‘tooth restoration’ (x2); ‘instruc- tion and advice’ (x2); ‘tooth extraction’ (x1.8); ‘scale and polish’ (x1.7). When compared with the least deprived quintile, the most deprived were more likely to have received the following at least once in the four-year period: tooth restoration (x1.7) and ‘tooth extraction’ (x1.5); how- ever, they were less likely to have received a ‘scale and polish’ (x0.5) and ‘instruction and advice’ (x0.3). Females were 20% less likely to receive a tooth restoration’ (p = 0.02), and a ‘tooth extraction’ (p = 0.012) than males. The influence of area deprivation on the multilevel model is presented in Table 4. Dental Treatment and Predictors of Need Table 2. (Continued) Patient related variables Never received treatment within four academic years N (%) Received treatment within four academic years N(%) P value 4 792 (60.3) 522 (39.7) Least deprived 5 388 (52.7) 348 (47.3) Smoking status (n = 3436) No 1,466(52.3) 1337 (47.7) 0.001* Yes 201 (31.8) 432 (68.2) Tooth extraction Overall 4754 (74.9) 1597 (25.1) Adult Payment status (n = 5,185) Exempt 622 (61.9)) 383 (38.1) 0.001* Non-exempt 3,115 (74.5) 1,065(25.5) Age groups (n = 6,351) Under 18 1,002 (87.4)) 145 (12.6) 0.001* 18–64 years 3,538 (73.1) 1,300 (26.9) Over 65 214 (58.5) 152 (41.5) Sex (n = 6,351) Female 2,390 (77.1) 708 (22.9) 0.001* Male 2,364 (72.7) 889 (27.3) Quintiles of deprivation in PCT (n = 6,259) Most deprived1 1,028 (69.6) 449 (30.4) 0.001* 2 982 (74.5) 336 (25.5) 3 1,086 (76.8) 328 (23.2) 4 1,014 (77.2) 300 (22.8) Least deprived 5 571 (77.6) 165 (22.4) Smoking status (n = 3,436) No 2,176 (77.6) 627 (22.4) 0.001* Yes 382 (60.3) 251 (39.7) Partial denture Overall 6,027 (94.9) 324 (5.1) Adult payment status (n = 5,185) Exempt 915(91) 90 (9) Non-exempt 3,947 (94.4) 233 (5.6) 0.001* Age groups (n = 6,351) Under 18 1147(100) 0 (0) 0.001* 18–64 years 4,611 (95.3) 227 (4.7) Over 65 269 (73.5) 97 (26.5) Sex (n = 6,351) Female 2,954 (94.5) 144 (4.6) 0.109 Male 3,073 (95.4) 180 (5.5) Quintiles of deprivation (in PCT) (n = 6,259) Most deprived 1 1,386 (93.8) 91 (6.2) 0.02* 2 1,248 (94.7) 70 (5.3) 3 1,343 (95.0) 71 (5.0) 4 1,269 (96.6) 45 (3.4) Least deprived 5 692 (94.0) 44 (6) Smoking status (n = 3,436) No 2,661 (94.9) 142 (5.1) 0.001* Yes 545 (86.1) 88 (13.9) PLOS ONE \| DOI:10.1371/journal.pone.0169004 January 24, 2017 PLOS ONE \| DOI:10.1371/journal.pone.0169004 January 24, 2017 7 / 17 Dental Treatment and Predictors of Need There was evidence of a higher rate of ‘tooth extraction’ amongst adults who were exempt payment, older (65 years), male and from areas of higher deprivation (p = 0.001); provision of this treatment showed a clear social gradient. Smokers were three times more likely to have received ‘partial dentures’ compared with non-smokers (p = 0.001) and adults exempt payment were twice as likely to have received them compared with non-exempt. Additionally, over five times more ‘older adults’ received partial dentures compared with those of working age (26.5% cf 4.7%: p = 0.001). Adjusted analysis (adults only) When individual level variables were added to the model, the co-efficient remained the same as those presented in Table 3. Therefore Table 4 presents the null model of variance independently associated with the area level variable (123 LSOAs). The model suggested that 2.8% (p = 0.01) of the variance in proportion of patients who had received a ‘tooth extraction’ can be explained by LSOA. For ‘scale and polish’ this increased to 3.6% (p = 0.01) and for ‘instruction and advice’ 7% (p = 0.04). Discussion This is the first study to model contemporary NHS patient management data from primary dental care in England in order to predict clinical care. The findings demonstrate an 8 / 17 PLOS ONE \| DOI:10.1371/journal.pone.0169004 January 24, 2017 Dental Treatment and Predictors of Need Table 4. Multilevel regression models predicting treatments (Instruction and advice, ‘tooth extraction’ and ‘scale and polish’) by area of residence within closed/ completed treatment plans between 2008/09 to 2011/12 academic years at UPDA (Null models of 123 LSOAs). Treatment Null model Variance Partition Coefﬁcient (VPC) Instruction and advice Variance = 0.062 SE 0.024 (Wald 40.415, p = 0.0001). The β_0j = -0.263(0.041) p = 0.0001. 0.07* Tooth extraction Variance = 0.097, SE 0.033 (Wald 4.229, p = 0.003). The β_0j = -0.194 (0.048); p = 0.0001 0.028* Scale and polish Variance = 0.125, SE 0.034; Wald 12.505, p = 0.0001. The β_0j = -0.317 (0.048) p = 0.04 0.036* Table 4. Multilevel regression models predicting treatments (Instruction and advice, ‘tooth extraction’ and ‘scale and polish’) by area of residence within closed/ completed treatment plans between 2008/09 to 2011/12 academic years at UPDA (Null models of 123 LSOAs). Treatment Null model Variance Partition Coefﬁcient (VPC) Instruction and advice Variance = 0.062 SE 0.024 (Wald 40.415, p = 0.0001). The β_0j = -0.263(0.041) p = 0.0001. 0.07* Tooth extraction Variance = 0.097, SE 0.033 (Wald 4.229, p = 0.003). The β_0j = -0.194 (0.048); p = 0.0001 0.028* Scale and polish Variance = 0.125, SE 0.034; Wald 12.505, p = 0.0001. The β_0j = -0.317 (0.048) p = 0.04 0.036* Note n = 2,782 * marks statistically signiﬁcant differences (p<0.05) Total number of areas 123 Lower Super Output Areas (LSOAs) doi:10.1371/journal.pone.0169004.t004 doi:10.1371/journal.pone.0169004.t004 association between treatment received and patients’ individual socio-demographic character- istics (demography, smoking, payment status) and context (deprivation at patients’ area of res- idence). These predictive relationships largely mirror population oral health needs from national epidemiological surveys [3,4], with one notable exception: ‘instruction and advice’ which relates to prevention. The study provides evidence of increasing treatment need with age, smoking, exemption from payment and deprivation status, all of which have implications for health services planning and provision. p g p This research has two important strengths. First, the use of patient management data pro- vided valid accounts of the care provided, without patient recall or selection bias [21]. Second, this is the first example of what is analytically possible with NHS patient management primary dental care data and provides insight to the care delivered within a large educational primary care facility. Dental Treatment and Predictors of Need Table 3. Adjusted logistic regression model predicting odds of treatment by patient characteristics between 2008/09 to 2011/12 academic years at UPDA. Outcome (reference category in brackets$) Predictor variable Odds ratio 95% C.I. for Odds ratio P value ROC Lower Upper Tooth restoration Adult payment exemption 2.108 1.576 2.819 0.01* 0.6 Age 1.02 1.015 1.025 0.01* (Quintile in PCT (5$) Quintile in PCT(1) 1.655 1.274 2.151 0.01* Quintile in PCT(2) 1.376 1.074 1.764 0.012* Quintile in PCT(3) 1.145 0.904 1.449 0.262 Quintile in PCT(4) 1.086 0.862 1.368 0.483 (Male$) Sex 0.832 0.712 0.971 0.02* (Non-smoker$) Smoking status 1.569 1.29 1.909 0.01* Instruction and advice Adult payment exemption 2.198 1.506 3.207 0.001* 0.7 Age 1.038 1.032 1.045 0.001* (Quintile in PCT (5$) Quintile in PCT(1) 0.371 0.256 0.536 0.001* Quintile in PCT(2) 0.48 0.332 0.692 0.001* Quintile in PCT(3) 0.54 0.376 0.776 0.001* Quintile in PCT(4) 0.608 0.421 0.879 0.009* (Male$) Sex 1.192 0.993 1.43 0.059 (Non-smoker$) Smoking status 4.124 3.088 5.508 0.001* Scale and polish Adult payment exemption 1.745 1.308 2.327 0.001* 0.7 Age 1.037 1.032 1.043 0.001* (Quintile in PCT (5$) Quintile in PCT(1) 0.512 0.379 0.692 0.001* Quintile in PCT(2) 0.754 0.56 1.016 0.063 Quintile in PCT(3) 0.72 0.538 0.964 0.027* Quintile in PCT(4) 0.818 0.608 1.101 0.185 (Male$) Sex 1.039 0.885 1.219 0.642 (Non-smoker$) Smoking status 1.737 1.421 2.124 0.001* Tooth extraction Adult payment exemption 1.815 1.38 2.388 0.001* 0.7 Age 1.033 1.028 1.039 0.001* (Quintile in PCT (5$) Quintile in PCT(1) 1.508 1.102 2.063 0.01* Quintile in PCT(2) 0.997 0.727 1.367 0.983 Quintile in PCT(3) 0.994 0.728 1.355 0.968 Quintile in PCT(4) 1.002 0.731 1.374 0.989 (Male$) Sex 0.8 0.672 0.953 0.012* (Non-smoker$) Smoking status 2.03 1.663 2.477 0.001* Partial denture Adult payment exemption 2.604 1.758 3.856 0.0001* 0.8 Age 1.075 1.065 1.085 0.0001* (Quintile in PCT (5$) Quintile in PCT(1) 1.087 0.655 1.802 0.748 Quintile in PCT(2) 0.813 0.484 1.365 0.433 Quintile in PCT(3) 0.927 0.561 1.531 0.767 Quintile in PCT(4) 0.704 0.412 1.205 0.201 (Male$) Sex 0.941 0.695 1.268 0.688 (Non-smoker$) Smoking status 3.142 2.277 4.337 0.0001* Note edicting odds of treatment by patient characteristics between 2008/09 to 2011/12 academic years at Table 3. Adjusted logistic regression model predicting odds of treatment by patient characteristics betwee UPDA 9 / 17 PLOS ONE \| DOI:10.1371/journal.pone.0169004 January 24, 2017 Furthermore, in this research, patient’s payment exemption status was used to provide an individual-level description of economic status and mitigate against bias. Payment exemp- tion provides an indication of income at the time of care [9], and in this data set has shown good correlation to area level income deprivation [20]. However, as it includes pregnant and nursing women, findings relating to this variable should be examined and interpreted by sex. The unadjusted model, which included children, confirmed differing patterns of care asso- ciated with age in line with having longer retention of teeth by adults in England [4]. This included older (65 years), and working age adults (18–64 years), having a higher rate of ‘tooth restoration’ (54.3% and 54.3% respectively), exceeding the volume amongst children and young people under 18 years (38%; p = 0.001). Equally this analysis validated the data, as expectedly children would not have denture treatment, and the results confirmed this with no children having received dentures. In addition, the relationships between common risks to oral health such as smoking, leading to increased dental need and thus the number of smokers that received all treatments exceeded the average for the whole study population. All five adjusted models relating to adult dental care showed reasonably good predictive capability. The first four, namely ‘tooth restoration’, ‘instruction and advice’, ‘scale and polish’ and ‘tooth extraction’ shared the same predictors: age, smoking status, payment exemption and deprivation status. Sex was a predictor for just two of these treatment models: ‘tooth resto- ration’ and ‘tooth extraction’. In contrast to the four most common treatments, the adjusted model involving ‘partial dentures’ only included age and payment exemption as predictors. Specifically, these models confirm that older patients, smokers, adults exempt from pay- ment and from an area of higher deprivation, in particular, are more likely to receive common treatments. These findings may be explained by the patterns of need amongst adults described by successive epidemiological surveys highlighting trends in dental caries and periodontal dis- ease by age [4,40,41], sex [4,42], and socioeconomic status [4,43,44], greater tooth loss and reli- ance on dentures in older people [4,45–47], and the contemporary approach to care within the service [48]. Socio-economic deprivation is well accepted as predicting self-reported dental need [4,49], and higher requests for tooth extraction have been demonstrated among deprived groups [50]. There were, however, a number of limitations related to the structure and amount of information that could be extracted from the patient information system which need to be addressed. First, it was not possible to obtain data on presenting complaints and baseline oral health, as this information was recorded as free text, without any form of coding. Information on presenting complaints or initial oral health status would have enriched the findings by pro- viding a full picture of the pathway towards oral health; however, data processing develop- ments in dentistry have not yet resulted in script that can extract written text [34,35]. Even so, the data obtained were useful because of the system of practice in UPDA, which involved agreement of care plans with the patient in common with NHS contracts; thus ensuring expressed and normative needs were addressed in the treatment plans. In future, the use of assessments such as International Caries Diagnosis and Assessment System (ICDAS) [36], and Bleeding on Probing Indices (BPI), whereby the scores are recorded within the patient man- agement system, would enable baseline oral health needs to be captured and prove useful for analysis. Second, the data were cross-sectional data which, as with all such studies, restricted analysis of temporality between treatments; for example, whether a patient was more likely to receive a ‘tooth restoration’ after ‘instruction and advice’ or vice versa, and limit researchers’ ability to establish causality [37]. It also limited the comparability to national reports, which identify treatment annually, or national surveys which provide evidence of treatment received during the lifecourse [38]. This limitation is a feature of how data are stored in the administra- tive system. It was, however, possible to ensure reliable accounts of receiving treatment within the cross-sectional study period as the validity was tested against a sample of clinical records. Third, and finally, the limitation of IMD to account for individual level deprivation [39]; this represents an average of the people living in an area and wealthier people may choose to live in PLOS ONE \| DOI:10.1371/journal.pone.0169004 January 24, 2017 10 / 17 Dental Treatment and Predictors of Need socially deprived areas because of convenience. However, the use of area level measures can help establish whether there are factors in a person’s environment that may impact on their health. PLOS ONE \| DOI:10.1371/journal.pone.0169004 January 24, 2017 These findings suggest that in this state-funded health service, adults access care when they are at social disadvantage. Smokers who were found to be between 1.5 and 4 times more likely to receive the treat- ments are at increased need because smoking is a major risk factor for periodontal disease [51,52], and there is evidence from other studies that smokers are more likely to attend the dentist more symptomatically than non-smokers regardless of deprivation status [53]; thus, they are more likely to require more treatment when they do attend. Differences by sex in the treatments received may be explained by health seeking behaviour [54]. In this study female patients were 20% more likely to receive ‘tooth extraction’ and ‘tooth restoration’. Additionally, the impact of payment exemption on treatment as the second stron- gest predictor of care in most of the models, to smoking status, is worth further investigation. This relationship may partially be explained by its role as a proxy for income [20], and con- firms the role of income inequality on oral health [55]. Whereas for males exemption from payments is solely income-related, for females it also covers pregnant and nursing mothers, thus questioning whether it can reliably be used as a proxy for income, an issue which should be investigated further in national data. The association between partial dentures and age can be explained by patterns of tooth loss [56], and the fact that tooth loss increases with age, older adults not having benefitted from fluoride toothpaste in their earlier years [57], and having received more surgical than PLOS ONE \| DOI:10.1371/journal.pone.0169004 January 24, 2017 11 / 17 Dental Treatment and Predictors of Need restorative care; thus, requiring dentures to replace missing teeth. The influence of payment exemption may be explained by the fact that payment removes the barrier to dentists deliver- ing, and patients receiving, these more expensive treatment items of care. This facility delivered a full range of routine NHS primary dental care which involved higher levels of preventive care in the form of ‘instruction and advice’ ‘fissure sealants’ and ‘fluoride varnish’ [58], than primary care nationally. Overall, almost half of the patients at UPDA received prevention in the form of ‘instruction and advice’ and this high rate can be explained by organisational philosophy, since UPDA is an educational institution which deliv- ers contemporary care and embraces dental team skill mix [58]. Furthermore, this approach is supported by an NHS contract [59], which includes key performance indicators including the identification and direction of smokers to tobacco cessation services [60]. The latter explains the higher receipt of ‘instruction and advice’ for smokers. In the past, prevention has been a reflection of financial incentives [61], and poorly incentivised prevention has resulted in it being relegated to lower priority [62–64]. Thus, there is evidence that ‘instruction and advice’ was prioritised amongst the adult smokers, suggesting that performance indicators may pro- vide an incentive for change. The most controversial finding from the study, however, was that adults from areas of higher deprivation were less likely to receive prevention in the form of ‘instruction and advice’ as their less deprived counterparts in the adjusted model; instead there was a clear inverse social gradient in relation to need. This is surprising as the institutional phi- losophy supported preventive care and high levels of prevention were delivered overall; how- ever, additional factors such as individual practitioner and patient attitudes to the delivery and receipt of care should be considered [62]. There is evidence that even when practitioners might wish to deliver prevention to those who need it, most of these patients present late in the disease process and require emergency care [62], with the majority failing to return for prevention. Equally there is evidence of high needs patients having poor prioritisation of health care seeking [53], and of patients attending for emergency care being less likely to receive prevention [65]. It has to be remembered that one of the limitations of this data set was the inability to differentiate between completed and closed courses of care. Perhaps patients attending UPDA attended on an emergency basis when it may be less acceptable to receive prevention and others who did not complete their prescribed course of care missed out on preventive advice; thus, suggesting that the health behaviours of patients may widen inequalities? Further research, subject to better coding being possible within the patient management system, is required to examine the differences in care between those who complete care and those who do not. This should ideally be supplemented by exploring the views of such patients on what they want and don’t want from dental care and why. PLOS ONE \| DOI:10.1371/journal.pone.0169004 January 24, 2017 S1 Dataset. (SAV) S1 Dataset. (SAV) Through the multilevel modelling this study suggests that within small residential areas (LSOAs) there are further influences which explain 7% of the variance in proportions of ‘par- tial dentures’, and 3% and 4% for ‘tooth extraction’ and ‘scale and polish’ respectively. This may relate to the influence of peers and social norms, but is worth further investigation. These findings parallel a study by Jamieson et al, who showed that variance in dental caries experi- ence was associated with residence [35]. Further research examining the influence of local environments is necessary to uncover protective and risk factors to oral health in the environment. This research has a number of implications. First, it has highlighted the importance of max- imising the use and enhancing the quality of routinely collected data and the need to improve storage mechanisms to enable longitudinal analysis to further develop our understanding of care patterns and health. This should include specific enhancement to ensure the inclusion of relevant indices and coded data to provide patient baseline data, and in-time outcomes. PLOS ONE \| DOI:10.1371/journal.pone.0169004 January 24, 2017 12 / 17 Dental Treatment and Predictors of Need Second, these results provide insight into the provision and receipt of contemporary dental care, and should inform discussions regarding performance indicators that target priority groups such as smokers and future planning for our ageing population. Third, it could be argued that despite a system of subsidised care, people from areas of high deprivation had more extractions, and experienced inequality in prevention and restoration which is likely to increase the gap in oral health between deprived and affluent sections of society. The implica- tion is that this inverse social gradient in relation to ‘instruction and advice’ must be investi- gated further to explore the balance of patient, practitioner, organisational and system issues to ensure that inequalities do not widen. There is also a need to identify factors that hamper or encourage preventive care for adults, in order to fulfil the NHS priority to make every encoun- ter count for prevention [66]. Fourth, the role of payment exemption requires further investi- gation to understand whether this variable acts as a proxy for income deprivation over time, particularly between the sexes. Fifth, and finally, further research is needed on environmental factors that may contribute to dental care. Acknowledgments This research was funded through a PhD studentship supported by King’s College London and the University of Portsmouth Dental Academy. We acknowledge the support of the Bio- statistics Team at King’s College London Dental Institute [KCLDI] and University of Ports- mouth Dental Academy [UPDA]. We also acknowledge the support of Dr Wael Sabbah in commenting on the manuscript. Conclusion This is the first study to model patient management data from a state-funded dental service and show that individual and contextual factors predict common treatments received. Implica- tions of this research include the importance of making provision for our aging population and ensuring that preventative care is available to all. Further research is required to explain the interaction of organisational and system policies, practitioner and patient perspectives on care and, thus, inform effective commissioning and provision of dental services. Author Contributions Conceptualization: KLW DRR JEG. References 1. World Health Organisation. Universal Health Coverage: Health Essential for Sustainable Develop- ment2012. Available from: http://www.who.int/universal_health_coverage/un_resolution/en/ 27/9/16 2. NHS England. The NHS Constitution 2013. Available from: http://www.nhs.uk/choiceintheNHS/ Rightsandpledges/NHSConstitution/Documents/2013/the-nhs-constitution-for-england-2013.pdf 14/ 10/2015 3. NHS digital. Child Dental Health Survey 2013. 2015. Available from: http://digital.nhs.uk/catalogue/ PUB17137. 27/9/16 4. Steele J, O’Sullivan I. Adult Dental Health Survey. In: Health Do, editor. London: The Health and Social Care Information Centre; 2011. 5. Wanyonyi K, White S, Gallagher JE. Conscious sedation: is this provision equitable? Analysis of seda- tion services provided within primary dental care in England, 2012–2014. BDJ Open. 2016 6. Steele J, Shen J, Tsakos G, Fuller E, Morris S, Watt R, et al. The Interplay between Socioeconomic Inequalities and Clinical Oral Health. Journal of Dental Research. 2015 January 1, 2015; 94(1):19–26. 7. Public Health England. Delivering better oral health: an evidence-based toolkit for prevention, Third Edi- tion 2014. Available from: https://www.gov.uk/government/publications/delivering-better-oral-health- an-evidence-based-toolkit-for-prevention 27/9/16 8. Department of Health. Equity and Excellence: Liberating the NHS. In: Health Do, editor. London 2010. 9. NHS Choices. Who is entitled to free NHS dental treatment in England? 2014. Available from: http:// www.nhs.uk/chq/Pages/1786.aspx?CategoryID=74. 04/08/2014 10. Department of Health. NHS dental contract pilots- Early findings- A report by the dental contract pilots evidence and learning reference group. 2012. 11. Guiney H, Felicia P, Whelton H, Woods N. Analysis of a Payments Database Reveals Trends in Dental Treatment Provision. Journal of Dental Research. 2013; 92(7): 63–9. 12. Burke FJT, Lucarotti PSK. Ten-year outcome of crowns placed within the General Dental Services in England and Wales. Journal of Dentistry. 2009 1//; 37(1):12–24. doi: 10.1016/j.jdent.2008.03.017 PMID: 18487003 13. Burke FJT, Lucarotti PSK. Ten-year outcome of porcelain laminate veneers placed within the general dental services in England and Wales. Journal of Dentistry. 2009 1//; 37(1):31–8. doi: 10.1016/j.jdent. 2008.03.016 PMID: 18538912 14. Burke FJT, Lucarotti PSK. Re-intervention in glass ionomer restorations: What comes next? Journal of Dentistry. 2009 1//; 37(1):39–43. doi: 10.1016/j.jdent.2008.07.008 PMID: 18819740 15. Burke FJT, Lucarotti PSK. Re-intervention on crowns: What comes next? Journal of Dentistry. 2009 1//; 37(1):25–30. doi: 10.1016/j.jdent.2008.11.007 PMID: 19100674 16. Burke FJT, Lucarotti PSK, Holder RL. Outcome of direct restorations placed within the general dental services in England and Wales (Part 2): Variation by patients’ characteristics. Journal of Dentistry. 2005 11//; 33(10):817–26. doi: 10.1016/j.jdent.2005.03.007 PMID: 16221517 17. Ferreira MC, Vieira RS. Marginal leakage in direct and indirect composite resin restorations in primary teeth: An in vitro study. Writing – review & editing: KLW DRR JEG. Writing – review & editing: KLW DRR JEG. Writing – review & editing: KLW DRR JEG. PLOS ONE \| DOI:10.1371/journal.pone.0169004 January 24, 2017 Conceptualization: KLW DRR JEG. Data curation: KLW DRR JEG. Formal analysis: KLW. Funding acquisition: JEG. Investigation: KLW DRR JEG. Investigation: KLW DRR JEG. Methodology: KLW JEG. Methodology: KLW JEG. Project administration: KLW DRR JEG. PLOS ONE \| DOI:10.1371/journal.pone.0169004 January 24, 2017 13 / 17 Dental Treatment and Predictors of Need References Journal of Dentistry. 2008 5//; 36(5):322–5. doi: 10.1016/j.jdent.2008.01.016 PMID: 18336986 18. Lucarotti PSK, Holder RL, Burke FJT. Analysis of an administrative database of half a million restora- tions over 11 years. Journal of Dentistry. 2005 11//; 33(10):791–803. doi: 10.1016/j.jdent.2005.06.011 PMID: 16214285 19. Lucarotti PSK, Holder RL, Burke FJT. Outcome of direct restorations placed within the general dental services in England and Wales (Part 3): Variation by dentist factors. Journal of Dentistry. 2005 11//; 33 (10):827–35. doi: 10.1016/j.jdent.2005.03.009 PMID: 16246480 PLOS ONE \| DOI:10.1371/journal.pone.0169004 January 24, 2017 14 / 17 Dental Treatment and Predictors of Need 20. Wanyonyi KL, Radford DR, Gallagher JE. The relationship between access to and use of dental ser- vices following expansion of a primary care service to embrace dental team training. Public health. 2013 Nov; 127(11):1028–33. doi: 10.1016/j.puhe.2013.08.010 PMID: 24210166 21. Neale J. Research methods in health and social care. London: Palgrave Macmillan; 2009. 22. Heller KE, Eklund SA, Burt BA, Briskie DM, Lawrence LM. Using Insurance Claims and Demographic Data for Surveillance of Children’s Oral Health. Journal of Public Health Dentistry. 2004; 64(1):5–13. PMID: 15078055 23. Guiney H, Felicia P, Whelton H, Woods N. Comparing epidemiologically estimated treatment need with treatment provided in two dental schemes in Ireland. BMC oral health. 2012; 12:31. doi: 10.1186/1472- 6831-12-31 PMID: 22898307 24. Radford DR, Holmes S, Woolford MJ, Dunne SM. The impact of integrated team care taught using a live NHS contract on the educational experience of final year dental students. British Dental Journal. 2014; 217(10):581–5. doi: 10.1038/sj.bdj.2014.1002 PMID: 25415016 25. Radford DR, Holmes S, Dunne SM, Woolford MJ. Outreach clinical dental education: the Portsmouth experience–a 4-year follow-up study. European Journal of Dental Education. 2015 26. Maletic JI, Marcus A. Data Cleansing: Beyond Integrity Analysis Information Quality (IQ2000) [Internet]. 2000 15/07/2013:[200–9 pp.]. Available from: http://dc-pubs.dbs.uni-leipzig.de/files/ Maletic2000DataCleansingBeyond.pdf. 27/9/16 27. Hall GC, Bryant TN, Merrett LK, Price C. Validation of the quality of The National Pain Database for pain management services in the United Kingdom. Anaesthesia. 2008; 63(11):1217–21. doi: 10.1111/j. 1365-2044.2008.05609.x PMID: 19032256 28. Kudyakov R, Bowen J, Ewen E, West SL, Daoud Y, Fleming N, et al. Electronic health record use to classify patients with newly diagnosed versus preexisting type 2 diabetes: infrastructure for comparative effectiveness research and population health management. Popul Health Manag. 2012; 15(1):3–11. doi: 10.1089/pop.2010.0084 PMID: 21877923 29. Rahm E, Hai Do H. Data cleaning: problems and current approaches 2013 17/10/2013. Available from: http://wwwiti.cs.uni-magdeburg.de/iti_db/lehre/dw/paper/data_cleaning.pdf. 27/9/16 30. PLOS ONE \| DOI:10.1371/journal.pone.0169004 January 24, 2017 References Lang IA, Gibbs SJ, Steel N, Melzer D. Neighbourhood deprivation and dental service use: a cross-sec- tional analysis of older people in England. Journal of Public Health. 2008. 30(4):472–8. doi: 10.1093/ pubmed/fdn047 PMID: 18587145 45. Unluer S, Gokalp S, Dogan BG. Oral health status of the elderly in a residential home in Turkey. Gero- dontology. 2007; 24(1):22–9. doi: 10.1111/j.1741-2358.2007.00136.x PMID: 17302927 46. Warpeha WS Jr. Expanding prosthodontic services to the older patient in dental practices. Northwest Dent. 2011; 90(6):21–4. PMID: 22372258 47. Allen PF, McKenna G, Creugers N. Prosthodontic care for elderly patients. Dental Update. 2011; 38 (7):460–2. PMID: 22046906 48. Watt RG, Steele JG, Treasure ET, White DA, Pitts NB, Murray JJ. Adult Dental Health Survey 2009: implications of findings for clinical practice and oral health policy. British Dent Journal. 2013; 214(2):71– 5. 49. Trohel G, Bertaud-Gounot V, Soler M, Chauvin P, Grimaud O. Socio-Economic Determinants of the Need for Dental Care in Adults. PLoS One. 2016; 11(7):e0158842. doi: 10.1371/journal.pone.0158842 PMID: 27441841 50. Richards W, Ameen J, Coll AM, Higgs G. Reasons for tooth extraction in four general dental practices in South Wales. British Dental Journal. 2005; 198(5):275–8. doi: 10.1038/sj.bdj.4812119 PMID: 15870746 51. Bergstrom J. Smoking rate and periodontal disease prevalence: 40-year trends in Sweden 1970–2010. Journal of clinical periodontology. 2014; 41(10):952–7. doi: 10.1111/jcpe.12293 PMID: 25039432 52. Javed F, Bashir Ahmed H, Romanos GE. Association between environmental tobacco smoke and peri- odontal disease: a systematic review. Environmental research. 2014; 133:117–22. doi: 10.1016/j. envres.2014.05.008 PMID: 24926917 53. Csikar J, Kang J, Wyborn C, Dyer TA, Marshman Z, Godson J. The Self-Reported Oral Health Status and Dental Attendance of Smokers and Non-Smokers in England. PLoS One. 2016; 11(2):e0148700. doi: 10.1371/journal.pone.0148700 PMID: 26863107 54. Green CA, Pope CR. Gender, psychosocial factors and the use of medical services: a longitudinal anal- ysis. Social science & medicine (1982). 1999; 48(10):1363–72. 55. Bhandari B, Newton JT, Bernabe E. Income Inequality and Use of Dental Services in 66 Countries. J Dent Res. 2015; 94(8):1048–54. doi: 10.1177/0022034515586960 PMID: 25994178 56. Hiltunen K, Vehkalahti MM, Mantyla P. Is prosthodontic treatment age-dependent in patients 60 years and older in Public Dental Services? Journal of oral rehabilitation. 2015; 42(6):454–9. doi: 10.1111/joor. 12267 PMID: 25545698 57. Sheiham A. Impact of dental treatment on the incidence of dental caries in children and adults. Commu- nity dentistry and oral epidemiology. 1997; 25(1):104–12. PMID: 9088699 58. Wanyonyi KL, Radford DR, Gallagher JE. References Thomas A, Zheng C, Jung H, Chang A, Kim B, Gelfond J, et al. Extracting data from electronic medical records: validation of a natural language processing program to assess prostate biopsy results. World Journal of Urology. 2014 2014/02/01; 32(1):99–103. English. doi: 10.1007/s00345-013-1040-4 PMID: 23417341 31. Department for Communities and Local Government. The English Indices of Multiple Deprivation 2010. 2011. 32. Brunton-Smith I, editor Applied Multilevel Modelling. CASS Courses in Applied Social Surveys; 2013; University of Southampton. 33. Leyland A, Groenewegen P. Multilevel modelling and public health policy. Scandinavian Journal of Pub- lic Health. 2003; 31:267–74. PMID: 15099032 34. Chapman WW, Christensen LM, Wagner MM, Haug PJ, Ivanov O, Dowling JN, et al. Classifying free- text triage chief complaints into syndromic categories with natural language processing. Artificial Intelli- gence in Medicine. 2005; 33(1):31–40. doi: 10.1016/j.artmed.2004.04.001 PMID: 15617980 35. de Lusignan S, van Weel C. The use of routinely collected computer data for research in primary care: opportunities and challenges. Family Practice. 2006 April 2006; 23(2):253–63. doi: 10.1093/fampra/ cmi106 PMID: 16368704 36. Pitts N, Amaechi B, Niederman R, Acevedo AM, Vianna R, Ganss C, et al. Global oral health inequali- ties: dental caries task group—research agenda. Advanced Dental Research. 2011; 23(2):211–20. 37. UK Data Service Census Support. Geoconvert tool. 2011. Available from: http://geoconvert.mimas.ac. uk/.11/05/2016 38. NHS digital. NHS dental statistics for England:2008–20122016. Available from: http://digital.nhs.uk/.27/ 9/16 39. Marshman Z, Porritt J, Dyer T, Wyborn C, Godson J, Baker S. What influences the use of dental ser- vices by adults in the UK? Community dentistry and oral epidemiology. 2012; 40(4):306–14. doi: 10. 1111/j.1600-0528.2012.00675.x PMID: 22409397 40. Feres M, Teles F, Teles R, Figueiredo LC, Faveri M. The subgingival periodontal microbiota of the aging mouth. Periodontol 2000. 2016. 72(1):30–53. doi: 10.1111/prd.12136 PMID: 27501490 41. Lamster IB, Asadourian L, Del Carmen T, Friedman PK. The aging mouth: differentiating normal aging from disease. Periodontol 2000. 2016. 72(1):96–107. doi: 10.1111/prd.12131 PMID: 27501493 15 / 17 PLOS ONE \| DOI:10.1371/journal.pone.0169004 January 24, 2017 Dental Treatment and Predictors of Need 42. Schulze A, Busse M. Gender Differences in Periodontal Status and Oral Hygiene of Non-Diabetic and Type 2 Diabetic Patients. The open dentistry journal. 2016; 10:287–97. doi: 10.2174/ 1874210601610010287 PMID: 27347232 43. Donaldson AN, Everitt B, Newton T, Steele J, Sherriff M, Bower E. The Effects of Social Class and Den- tal Attendance on Oral Health. Journal of Dental Research. 2008. 1; 87(1):60–4. PMID: 18096895 44. PLOS ONE \| DOI:10.1371/journal.pone.0169004 January 24, 2017 65. Brennan DS, Spencer AJ. Evaluation of service provision patterns during a public-funded dental pro- gram. Australian and New Zealand journal of public health. 1999; 23(2):140–6. PMID: 10330727 PLOS ONE \| DOI:10.1371/journal.pone.0169004 January 24, 2017 66. Department of Health. The NHS’s role in the public’s health: A report from the NHS Future Forum. 2012. Dental Treatment and Predictors of Need References Dental skill mix: a cross-sectional analysis of delegation prac- tices between dental and dental hygiene-therapy students involved in team training in the South of England. Human Resources for Health. 2014; 12(1):1–8. 59. Personal Dental Services Regulations SI 2005 [3373], (2005). 60. Milsom KM, Threlfall A, Pine K, Tickle M, Blinkhorn AS, Kearney-Mitchell P. The introduction of the new dental contract in England—a baseline qualitative assessment. British Dental Journal. 2008; 204 (2):59–62. doi: 10.1038/bdj.2008.1 PMID: 18223578 61. Tickle M, McDonald R, Franklin J, Aggarwal VR, Milsom K, Reeves D. Paying for the wrong kind of per- formance? Financial incentives and behaviour changes in National Health Service dentistry 1992– 2009. Community dentistry and oral epidemiology. 2011; 39(5):465–73. doi: 10.1111/j.1600-0528. 2011.00622.x PMID: 21668463 62. Aljafari AK, Gallagher JE, Hosey MT. Failure on all fronts: general dental practitioners’ views on promot- ing oral health in high caries risk children- a qualitative study. BMC oral health. 2015; 15(1):1–11. 63. Brocklehurst PR, Tickle M. Is skill mix profitable in the NHS dental contract? Br Dent J. 2011; 210:303– 8. doi: 10.1038/sj.bdj.2011.238 PMID: 21475274 64. Harris RV, Sun N. Translation of remuneration arrangements into incentives to delegate to English den- tal therapists. Health Policy. 2012 3//; 104(3):253–9. doi: 10.1016/j.healthpol.2011.11.013 PMID: 22177418 PLOS ONE \| DOI:10.1371/journal.pone.0169004 January 24, 2017 16 / 17 Dental Treatment and Predictors of Need 17 / 17
https://openalex.org/W2807507388	http://zbornik.gf.uns.ac.rs/doc/NS2018.52.pdf	Serbian	null	SAVREMENI PRISTUPI KANALISANJA KIŠNOG OTICAJA U URBANOJ REGENERACIJI	Zbornik radova Građevinskog fakulteta	2,018	cc-by-sa	3,847	h INTERNATIONAL CONFERENCE h INTERNATIONAL CONFERENCE 1 Magdalena Vasilevska, mast.inž.arh., Građevinsko-arhitektonski fakultet Univerziteta u Nišu, e-mail: magdalena.vasilevska@gaf.ni.ac.rs 2 Dr Ljiljana Vasilevska, red.prof., Građevinsko-arhitektonski fakultet Univerziteta u Nišu, e-mail: ljiljana.vasilevska@gaf.ni.ac.rs MODERN STORMWATER MANAGEMENT APPROACHES IN URBAN REGENERATION Magdalena Vasilevska 1 Ljiljana Vasilevska 2 Summary: In response to climate change, the negative effects of urbanization and industrialization, as well as numerous socio-economic, physical and environmental changes in urban areas and urban catchments, developed countries have generated and implemented several integrated stormwater management approaches. They are based on the principle of longer and safe stormwater retention in the urban catchment, which is achieved through the application of a number of systemic measures and technical elements designed to have minimal influence on the natural hydrological cycle. The benefits of their applications are multiple, as indicated by the experience of countries that enabled the integration modern stormwater management approaches into the process of urban planning and design, through defining the appropriate legal and institutional framework. Keywords: modern stormwater management approach, urban planning, urban design, urban regeneration, integration 2. METHODOLOGY In analyzing the basic characteristics and possibilities of applying modern stormwater management approaches and exploring the methodological and conceptual framework of their integration into the process of urban planning and design, the methodological framework is based on an analytical approach which relies on description and analysis. 6. МЕЂУНАРОДНА КОНФЕРЕНЦИЈА Савремена достигнућа у грађевинарству 20. април 2018. Суботица, СРБИЈА complexity of the practical options for solving the problems related to the quantity and quality of stormwater in the built environment, as well as the fact that activities related to the reconstruction of the existing system are most often realized within the framework of urban regeneration projects, which is even more complex process according to the goals and structure. In addition to the general characteristics of modern stormwater management approaches, the main goal of this paper is the analysis of the genesis and development of the conceptual framework for integrating modern approaches into the urban planning and design process, as well as the benefits that arise from this. The focus of the research is on urban regeneration, as one of the key methods and tool of urban planning and design. 1. INTRODUCTION In the last decades a several modern stormwater management approaches have been developed. Scientific thought has dealt with various aspects of this issue, such as the role and significance of modern approaches [1] [2], the relationship between modern approaches and urban planning and design [3], as well as recommendations for the application of individual measures and technical elements [3]. However, as current planning theory and practice made a significant shift towards implementation of modern approaches, moreover to their integration into the urban planning and design process, research of this type is important and necessary. In the unbuilt/newly-planned areas, the measures and elements of modern stormwater management approaches are selected and applied according to the characteristics of natural conditions, while in the urban environment the character and the possibility of their application are additionally dependent on the characteristics of the existing physical structures. In the second case, when the upgrading and/or reconstruction of the existing system is performed, the planning and design tasks are much more complex, due to the \| CONFERENCE PROCEEDINGS INTERNATIONAL CONFERENCE (2018) \| 525 3. ORIGIN AND BASIC CHARACTERISTICS OF MODERN STORMWATER MANAGEMENT APPROACHES Modern stormwater management approaches have been created as a result of the need to find answers to the global changes and the problems that human society has been facing for the last few decades, such as: 1) rapid urbanization; 2) industrialization; and 3) climate change. Some UN projections indicate that the potential risks of global climate change are: a) an increase in the average temperature by 3° C by 2070; b) reduction of average rainfall amounts by 20-40% by 2070; c) increasing the sea level in cooperation with storm events; and d) an increase in the frequency and intensity of the storm periods, where there are some countries and regions with varying degrees of vulnerability. Rapid urbanization, which reflects in the fact that in 2007, for the first time on a global scale, the number of urban population equaled the number of rural population and that since then, it has steadily increased by an average of about 1.5 million inhabitants every week, that by 2025. there will be about 40 cities with a population of more than 10 million, or that since 1990, the number of people living in slums increased by 33%, disrupted the natural water cycle and influenced major imbalances in the urban water cycle. Unlike the natural environment, where over 50% of the rainwater infiltrates through the soil (25% shallow and 25% deep infiltration), about 40% is regulated by evapotranspiration, while only about 10% of the rainwater is retained on the surface, in the urban environment the situation is quite different. Namely, due to the high participation of non-porous surfaces, only about 15% of the rainwater is infiltrated (10% shallow and only 5% deep infiltration), evapotranspiration participates with about 30%, while more than half of the total rainfall is about 55% . In the central, densely built urban areas, the share of surface rainfall ranges from 80% to as much as 100%, ie there is no \| ЗБОРНИК РАДОВА МЕЂУНАРОДНЕ КОНФЕРЕНЦИЈЕ (2018) \| 526 INTERNATIONAL CONFERENCE INTERNATIONAL CONFERENCE Contemporary achievements in civil engineering 20. April 2018. Subotica, SERBIA possibility of infiltration or it is reduced to a minimum. Numerous studies indicate a direct connection between the intensity of the urbanization process (expressed by increasing population density and greater participation of impermeable surfaces) with an increase in the amount of rainfall that is retained on the surface (Figure 1). Figure 1. Influence of the intensity of urbanization on the increase of surface rainfall [4] Figure 1. Influence of the intensity of urbanization on the increase of surface rainfall [4] The need to streamline the solutions to problems and challenges arising from the process of rapid urbanization in a socially sustainable way, and to adapt urban areas to climate change such as increasing the frequency and intensity of storm events, is directly related to the need to establish and implement an effective, new stormwater management approaches. In line with that, in the last decades, several new, innovative approaches and principles of stormwater management have been developed, e.g. the drainage of rain and stormwater in urban areas, which are conceptually based in a completely different way compared to traditional atmospheric water management systems. The following basic goals of new approaches are defined [2]: 1) replace and/or increase the capacity of the existing drainage system in urban catchments by mimicking nature environment; 2) solve the flooding problems; 3) solve the problems associated with stormwater quantity and quality. Although the initial researches and patterns of stormwater management were initially motivated by the specific characteristics and problems of the country in which they were created, today most of them have an integrated approach to the problem, through the tendency to deviate from the natural hydrological cycle and the conditions that are present in the urban catchment. Under the paradigm of sustainable development, modern approaches are based on additional common goals, which are simultaneously in synergy with the primary goals of urban planning and design, such as: 1) improving the characteristics of the built environment - increasing the quality of life; 2) improving the quality of water resources; 3) reduction of negative impacts of stormwaters and risk management; and 4) the preservation and improvement of urban ecosystem. 527 \| CONFERENCE PROCEEDINGS INTERNATIONAL CONFERENCE (2018) \| 6. МЕЂУНАРОДНА КОНФЕРЕНЦИЈА 6. МЕЂУНАРОДНА КОНФЕРЕНЦИЈА Савремена достигнућа у грађевинарству 20. април 2018. Суботица, СРБИЈА The most recognized and prominent modern stormwater approaches are Water Sensitive Urban Design (WSUD) in Australia, Sustainable Drainage System (SuDS) and Sustainable Urban Drainage System (SUDS) in Great Britain and Scotland, Best Management Practices (BMPs) and Low Impact Development (LID) in the United States, Alternative techniques (ATs) in French speaking countries, Source Control in Canada, etc. [5]. The most recognized and prominent modern stormwater approaches are Water Sensitive Urban Design (WSUD) in Australia, Sustainable Drainage System (SuDS) and Sustainable Urban Drainage System (SUDS) in Great Britain and Scotland, Best Management Practices (BMPs) and Low Impact Development (LID) in the United States, Alternative techniques (ATs) in French speaking countries, Source Control in Canada, etc. [5]. For example, Sustainable Drainage System (SuDS) is an stormwater management approach that takes into account the amount of water (flooding), water quality (pollution) and questions of affection, while in technical terms it presents a set of management practices, control facilities and strategies designed to efficiently and effectively drain surface water, while reducing pollution to the smallest possible extent and managing the impact of quality on local water bodies. On the other hand, WSUD is an approach based on the integration of stormwater management into the process of urban planning and design in order to reduce environmental degradation and improve its sustainability and attractiveness [6]. However, regardless of specificity and certain differences, all approaches are commonly based on the same conceptual definition (Figure 2) - replacement of the existing drainage system in urban basins or creating new drainage systems using measures and elements that imitate or support the natural environment [1]. 1 Figure 2. Evolution in stormwater management [7] 1 Figure 2. Evolution in stormwater management [7] Also, all approaches offer a set of different technologies and ways of treating rainfall. There are four basic treatments, which are applied in modern approaches, separately or in combination, and which are also representative of the evolution of contemporary approaches in relation to the traditional approach (Figure 3). These are: 1) infiltration; 2) disposal; 3) storage; and/or 4) re-use of atmospheric water. Each of the treatment modes involves the application of different systemic measures and technical elements. \| ЗБОРНИК РАДОВА МЕЂУНАРОДНЕ КОНФЕРЕНЦИЈЕ (2018) \| 528 6. МЕЂУНАРОДНА КОНФЕРЕНЦИЈА Савремена достигнућа у грађевинарству 20. април 2018. Суботица, СРБИЈА Infiltrationbased elements are primarily designed to dispose atmospheric water into the ground, with its complete removal through the drainage system from the atmospheric sewage, which also requires the use of porous substrates. Storage-based elements retain part of the flow but have limited capacity (when it reaches, excess water goes into atmospheric sewage). Some of the elements, such as bioretensions, can provide both, while many approaches offer a combination of infiltration and storage by the application and incorporation of a set of technical elements into the whole system. In addition to the above, other technical elements, such as sedimentary basins - basins for collecting deposits, tampon systems, elements for the treatment of street timber, are possible, and also the use of green roofs (Figure 4). It should be noted that green roofs are not a compulsory technical element of contemporary approaches, since adequate rainwater treatment can be realized without them. However, the green roofs are recognized through practice as an important technical and aesthetic element in the integrated stormwater management approach, which, in addition to turning and channeling the drainage from the combined sewage system to the green areas with infiltration systems (1), and the transfer of atmospheric waters from the traffic surfaces through shallow greened liner bioretentions in the lakes or dry depression in the open space (2), became the third, most well-known and applied measure in time (Figure 4). Contemporary achievements in civil engineering 20. April 2018. Subotica, SERBIA Depending on the approach, the elements are differently defined, although in general they have the same function. 2 Figure 3. Traditional vs. modern stormwater management approach [8] 2 Figure 3. Traditional vs. modern stormwater management approach [8] Thus, WSUD recognizes the following elements: 1) swales (dry or wet); 2) filter trenches; 3) sand filters; 4) bioretension systems; 5) porous paving; 6) infiltration channels; 7) infiltration basins; 8) rainwater tanks; and 9) elements of landscape architecture. In the framework of SuDS, the division of elements is as follows: 1) porous paving (pedestrian communication and other surfaces); 2) filter belts; 3) filter and infiltration trenches; 4) swales (dry or wet); 5) retention ponds; 6) underground rainwater storage tanks; 7) swamps; and 8) little water surface - lakes (Figure 4). Figure 4. Technical elements of SuDS integrated in the process of urban regeneration in residential ares of Augustenborg, Malmo [9] Figure 4. Technical elements of SuDS integrated in the process of urban regeneration in residential ares of Augustenborg, Malmo [9] \| CONFERENCE PROCEEDINGS INTERNATIONAL CONFERENCE (2018) \| 529 4. BENEFITS OF INTEGRATING MODERN STORMWATER MANAGEMENT APPROACHES IN THE PROCESS OF URBAN REGENERATION Preconditions for the development of modern approaches and their subsequent integration into the urban planning and design process originated in the 1980s. Under the paradigm of "living with water", there was a general departure from the concept of water as “urban and city life enemy” and “hidden element behind pipes and taps”, to an “element that contributes to the quality of life”. In addition to providing an opportunity to integrate modern approaches into the process of urban planning and design, these circumstances also favored a radical change in the urban development paradigm [10], primarily in terms of the evolution of the role of urban water cycle management, its conceptual framework integration and cumulative socio-economic factors (Figure 5). The basic intention of the new conceptual framework is to establish a greater harmony between water as a key resource and the social community through a process of urban planning and design in a sustainable, socially rational and responsible way. Accordingly, the goals of integrating modern approaches in the conceptual and methodological framework of urban planning and design, including urban regeneration, is to create an attractive, functional and "environmentally-friendly" urban environment. In that term, with its physical and functional structure, this environment would be adapted to future challenges of urbanization, environmental protection and climate changes. This determination certainly involves the transformation of a traditional urban approach, whose principles of planning and design and the accompanying methodological framework are often based on sectoral and ex post consideration of the problem of channeling stormwaters. Instead, a modern urban approach has been established, which is both conceptually and methodologicaly ex ante ''water sensitive'', which can have 530 \| ЗБОРНИК РАДОВА МЕЂУНАРОДНЕ КОНФЕРЕНЦИЈЕ (2018) \| Савремена достигнућа у грађевинарству 20. април 2018. Суботица, СРБИЈА On the other hand, modern approaches, in contrast to the traditional ones, offer the possibility, in addition to reducing the amount of surface waters and flooding of the area, to form or increase the degree of utilization and design potential of the urban environment, both newly planned and already existing (Figure 4). In this way, the principles of modern approaches are in correlation with the principles of current theoretical approaches to urban development and urban planning and design, such as New Urbanism, Smarth Growth and TOD - Transit Oriented Development. Modern approaches are also in close correlation with the European 2020 Strategy (EU 2020), the Green Infrastructure and Green Urbanism. Modern approaches have thus become important as a conceptual framework and as a methodological tool of urban planning and design which leads to: improving recreation and housing features, creating quality open spaces/multifunctional open spaces, increasing the level of biodiversity in the urban areas, allowing closer contact with nature, reducing the effects of the thermal island or re-use of rainwater. In addition to practical solutions related to the management of quantity and quality of stormwater, modern approaches provide a whole set of planning and design solutions and possibilities in the development of newly-planned and/or regenerated inherited sites. Because they are based on the application of measures and elements that imitate or support the natural environment, contemporary approaches provide the possibility of applying a wide range of different elements in forming or improving the usable and design potential of urban spaces, since each technical element has its own specific use and design characteristics (Figure 4). The experiences of the countries that have integrated modern stormwater management approaches in the process of urban regeneration indicate that the effects of applying their principles, measures and technical elements (at different spatial and functional levels of the organization of the urban environment) influenced a higher degree of its: 1) sustainability; 2) flexibility and adaptability; 3) the ability to respond to climate change; and 4) ecological comfort and sense of comfort [12]. The application of modern approaches, either through comprehensive or partial coverage, is particularly important in housing areas. The approaches are of particular importance for the treatment of public, semi-public and private open spaces within these areas, since the quality of life depends largely on the types of organization, typological structure, use and design potential. Contemporary achievements in civil engineering 20. April 2018. Subotica, SERBIA numerous benefits for the utilization and design potential of urban spaces and its ecological characteristics (Figure 6). numerous benefits for the utilization and design potential of urban spaces and its ecological characteristics (Figure 6). numerous benefits for the utilization and design potential of urban spaces and its ecological characteristics (Figure 6). ecological characteristics (Figure 6). Figure 5. Relations between the evolution of stormwater management, cumulative socio- political drivers, urban planning and design process and transformation of the city [10] Figure 5. Relations between the evolution of stormwater management, cumulative socio- political drivers, urban planning and design process and transformation of the city [10] Figure 6. Difference between traditional and modern approach and their influence on characteristics of urban areas in urban design and regeneration process [11] Figure 6. Difference between traditional and modern approach and their influence on characteristics of urban areas in urban design and regeneration process [11] CONFERENCE PROCEEDINGS INTERNATIONAL CONFERENCE (2018) \| 531 Contemporary achievements in civil engineering 20. April 2018. Subotica, SERBIA Contemporary achievements in civil engineering 20. April 2018. Subotica, SERBIA and 3) artful. The traditional approach does not have any additional benefits for shaping the space and quality of life in the urban environment, since the stormwater is directly driven into the separate sewage system. An integrated approach is based on incorporation of technical elements of modern approaches into the process of planning and design of open space. Artful rainwater design refers not only to channeling the stormwater, but also to the transformation of space and the creation of its identity, based on the benefits and design possibilities of the technical elements of modern approaches. Examples of good practice suggest that modern stormwater management approaches have found their adequate application within the process of urban regeneration as one of the key methods in the urban planning process and design of inherited urban structures. 5. CONCLUSION Most modern approaches are in full synergy with urban regeneration as one of the key methods of a sustainable and socially rational process of urban planning and design. Based on the conducted analysis, it can be concluded that environmental improvements, improvement of the built environment and empowerment of the local community are recognized as key benefits of implementing the modern stormwater management approaches in urban regeneration process. The analysis of current theoretical and practical experiences has shown that the benefits of applying modern approaches within urban regeneration are numerous and, among others, include: 1) reducing the risks and incidents of floods; 2) improving usage and morphologic characteristics of urban areas, especially the open spaces; 3) providing ways to supplement the water supply systems in areas that are under stress; 4) efficient rainwater treatment, 5) improving biodiversity, 6) reducing the impact of urban heat island. All these represent a contribution to mitigating the expected negative impacts of climate change. \| CONFERENCE PROCEEDINGS INTERNATIONAL CONFERENCE (2018) \| Савремена достигнућа у грађевинарству 20. април 2018. Суботица, СРБИЈА As most approaches support "environmentally friendly" housing and lifestyle, the previously mentioned technical elements (green roofs, bioretensions, rain gardens, porous paving, etc.) are in direct relation to the degree of usable and design quality of open spaces an its key characteristic: 1) ecological comfort, which is evaluated through indicators, such as physical isolation in relation to streets (a), percentage of green areas, their position and form/type (b), and percentage of increase in biodiversity (c); 2) diversification of contents and forms of open spaces; 3) the level of safety and privacy, which is evaluated through indicators, such as implemented materials for construction of open spaces surfaces and communications (a) [13], various forms of visual and physical protection and barriers (b), spatial distance and size of the area that is isolated from access to the motor traffic or communications access and views of the pedestrian communication (c) and other physical dimensions and urban and architectural design arrangements (d); 4) social sustainability. When it comes to the relations between modern approaches and landscape architecture, it is possible to identify three dominant design approaches: 1) traditional; 2) integrated; \| ЗБОРНИК РАДОВА МЕЂУНАРОДНЕ КОНФЕРЕНЦИЈЕ (2018) \| 532 th INTERNATIONAL CONFERENCE [7] https://www.slideshare.net/KatiePekarek/stormwater-and-water-quality-trading- from-10000-feet download 20.11.2017. [8] https://inhabitat.com/pierces-park-combines-art-play-and-stormwater-management- into-a-stunning-urban-oasis-in-baltimore/pierces-park-by-mahan-rykiel-23/, download 20.11.2017. [9] http://www.malmo.se/English/Sustainable-City-Development/Augustenborg-Eco- City.html, download 20.11.2017. [10] Brown Р., Keath Н., Wong Т.: Transitioning to Water Sensitive Cities: Historical, Current and Future Transition States, 11th International Conference on Urban Drainage, 2008, Edinburgh, Scotland, UK [11] https://www.sa.gov.au/__data/assets/pdf_file/0003/21756/WSUD_chapter_2.pdf, download 20.11.2017. [12] Vasilevska, Lj., Blagojević, B., Vasilevska, M.: Linijski tehnički elementi u integrisanim pristupima upravljanju atmosferskim otpadnim vodama, Zbornik radova Građevinsko-arhitektonskog fakulteta, 2014, № 29, p.p. 27-43 [13] Vasilevska, M., Vasilevska, Lj.: Porous concrete as a technical element of modern stormwater managament approaches: usable and morphologic potential in the process of urban design, Zbornik radova III Medjunarodna naučno-stručna konferencija Tehnologija betona SFERA, 2017, p.p. 136-141 REFERENCES [1] Gordon-Walker, S., Harle, T., & Naismith, I.: Using science to create a better place. Cost-benefit of SUDS retrofit in urban areas, 2007., Science Report – SC060024, Environmental Agency, Bristol [2] https://www.waternz.org.nz/Attachment?Action=Download&Attachment_id=139, download 03.05.2018. [3] Wong, T. H. F.: An Overview of Water Sensitive Urban Design Practices in Australia. Water Practice & Technology, 2006., vol 1, № 1, p.p. [4] http://greatlakesresilience.org/case-studies/infrastructure/resilient-stormwater- planning-takes-time-and-pays, download 05.03.2018. [5] Fletcher, T. D. Shuster, William, H., William F. A., Richard. Butler, D., Arthur, S., Trowsdale, S., Barraud, S., Semadeni-Davies, A., Bertrand-Krajewski, J., Steen Mikkelsen, P., Rivard, G., Uhl, M., Dagenais, D., Viklander, M.: SUDS, LID, BMPs, WSUD and more – The evolution and application of terminology surrounding urban drainage. Urban Water Journal, 2015, vol.12, № 7, p.p. 525–542 [6] http://www.rms.nsw.gov.au/documents/projects/planning-principles/urban- design/water-sensitive-urban-design-guideline.pdf, download 20.11.2017. \| CONFERENCE PROCEEDINGS INTERNATIONAL CONFERENCE (2018) \| \| CONFERENCE PROCEEDINGS INTERNATIONAL CONFERENCE (2018) \| 533 6. МЕЂУНАРОДНА КОНФЕРЕНЦИЈА Савремена достигнућа у грађевинарству 20. април 2018. Суботица, СРБИЈА САВРЕМЕНИ ПРИСТУПИ КАНАЛИСАЊА КИШНОГ ОТИЦАЈА У УРБАНОЈ РЕГЕНЕРАЦИЈИ Резиме: Као одговор на климатске промене, негативне ефекте урбанизације и индустријализације, као и бројне социо-економске, физичке и енвајронменталне промене у урбаним срединама и урбаним сливовима, у развијеним земљама су у последњих неколико деценија генерисани и имплементирани интегрисани приступи каналисања кишног отицаја, односно управљања атмосферским водама. Они се заснивају на принципу дужег а безбедног задржавања кишнице у урбаном сливу, што се постиже применом низа системских мера и техничких елемената осмишљених тако да што мање утичу на природни хидролошки циклус. Користи њихове примене су вишеструке, на шта указује искуство земаља које су кроз дефинисање одговарајућег законског и институционалног оквира омогућиле интеграцију примене савремених приступа каналисању кишног отицаја у процес урбанистичког планирања и пројектовања. Кључне речи: савремени приступи каналтисања кишног отицаја, урбанистичко планирање, урбанистичко пројектовање, урбана регенерација, интегрисање \| ЗБОРНИК РАДОВА МЕЂУНАРОДНЕ КОНФЕРЕНЦИЈЕ (2018) \| 534
https://openalex.org/W4393043902	https://www.qeios.com/read/WZEUCZ/pdf	English	null	Review of: "A Novel Variable Neighborhood Search Approach for Cell Clustering for Spatial Transcriptomics"	null	2,024	cc-by	591	Qeios, CC-BY 4.0 · Review, March 21, 2024 Qeios ID: WZEUCZ · https://doi.org/10.32388/WZEUCZ Review of: "A Novel Variable Neighborhood Search Approach for Cell Clustering for Spatial Transcriptomics" Simon de Givry Potential competing interests: No potential competing interests to declare. The paper proposes the application of a well-known metaheuristic, variable neighborhood search, to single cell clustering. The minimization criterion combines a distance between gene expressions with a spatial distance between cells. Different embeddings are evaluated to measure the gene expression distance. A background on the main classical clustering methods is given. Although interesting, it lacks an introduction to the type of clustering problem solved to understand how these clustering methods are adequate for the problem. Then different existing embedding methods are presented. Again, it is unclear if these methods are dedicated to the problem or not. The notion of centroid is exploited but not introduced (when speaking about k-means/k-medoids). The problem formulation is very close to the p-median. I don't see what is specific to cell clustering apart from a linear combination of spatial distance and gene expression similarity. Tuning the alpha hyperparameter might be complex. I expect some form of normalization for the two contributions. Looking at their Pareto front might be interesting too. The justification of using VNS is missing. Why use this metaheuristic? A comparison with a basic greedy heuristic (selecting the best centroid at each iteration until k_max selections), and a simple lower bound computed by taking the sum of second min values per gene, would help to evaluate the remaining optimality gap. Overall, the technical part of the paper is sound, and the experimental results are convincing, showing a clear improvement compared to existing clustering methods with respect to clustering similarity to ground truth solutions. Minor corrections: p.5 line 188 Title Section "Implementation" should be replaced by "Methods" What is the cosine similarity of two embeddings? What is the cosine similarity of two embeddings? Eq (2) is incorrect: it says only one i and j should be swapped. Which value was used for max_iter? Why in Table 2 is the VNS PCA time much longer than in Table 1? Qeios ID: WZEUCZ · https://doi.org/10.32388/WZEUCZ Review of: "A Novel Variable Neighborhood Search Approach for Cell Clustering for Spatial Transcriptomics" Qeios ID: WZEUCZ · https://doi.org/10.32388/WZEUCZ 1/2 Qeios, CC-BY 4.0 · Review, March 21, 2024 Qeios, CC-BY 4.0 · Review, March 21, 2024 I had to modify the default command suggested in the README; I could not get GraphST or other embeddings except X_pca: python implementation/vns.py -adata_path SS200000128TR_E2.h5ad -k 33 -max_iter 10 -p 12 -m 15 -alpha 1 -seed 1 python implementation/vns.py -adata_path SS200000128TR_E2.h5ad -k 33 -max_iter 10 -p 12 -m 15 -alpha 1 -seed 1 It lt i t f l t ! It results in out-of-memory on my laptop! It took more than 50GB to run on the large field mouse brain hemisphere E dataset with the following result of 19125.7 (far from the reported solution of 9550.0 in Table 1) Output: Output: Number of clusters: 33 Number of cells: 38811 max_iter, parameter, m, alpha, s, seed = [10, 12, 15, 1, 10, 1] Best solution: [3630,328,33812,3751,37710,34437,6667,7963,18252,18502,34176,36587,25559,16906,19028,28821,15541,19642,228 90,25378,37882,29104,6406,16972,34770,7429,6366,6142,1938,11456,71,36448,3351] Best objective function: 19125.722361663233 Elapsed time: 83.8853 seconds Number of clusters: 33 max_iter, parameter, m, alpha, s, seed = [10, 12, 15, 1, 10, 1] [3630,328,33812,3751,37710,34437,6667,7963,18252,18502,34176,36587,25559,16906,19028,28821,15541,19642,228 90,25378,37882,29104,6406,16972,34770,7429,6366,6142,1938,11456,71,36448,3351] Best objective function: 19125.722361663233 Best objective function: 19125.722361663233 Elapsed time: 83.8853 seconds For the same dataset, selecting manually a single best centroid (K=1) results in an overall score of 24878.77. The sum of n-K second minima gives a lower bound of 8433.62. 2/2
https://openalex.org/W4361896927	https://figshare.com/articles/journal_contribution/Figure_S2_from_LIF_Drives_Neural_Remodeling_in_Pancreatic_Cancer_and_Offers_a_New_Candidate_Biomarker/22411550/1/files/39857465.pdf	English	null	Figure S3 from LIF Drives Neural Remodeling in Pancreatic Cancer and Offers a New Candidate Biomarker	null	2,023	cc-by	344	Sup Fig2. Bressy et al. Sup Fig2. Bressy et al. CO-CULTURE METHOD MIGRATION ASSAY Bressy et al. secreted molecules by cells X secreted molecules by cells X+Y secreted molecules by cells Y pore 8µM supernatant (CM) medium 0% FBS cells X cells Y cells X+ Y secreted molecules by cells X secreted molecules by cells X+Y secreted molecules by cells Y 300 000 cells X 300 000 cells Y 300 000 cells X + 300 000 cells Y cell culture (24h) with 3ml of medium 0.5% cell seeding overnight with medium 10% FBS washing (PBS) supernatant (CM) mRNA or Protein extracts LIF quantifcation by ELISA assay insert Schwann cells migrated cells Counting cells supernatant (CM) A B A CO-CULTURE METHOD cells X cells Y cells X+ Y secreted molecules by cells X secreted molecules by cells X+Y secreted molecules by cells Y 300 000 cells X 300 000 cells Y 300 000 cells X + 300 000 cells Y cell culture (24h) with 3ml of medium 0.5% cell seeding overnight with medium 10% FBS washing (PBS) supernatant (CM) mRNA or Protein extracts LIF quantifcation by ELISA assay 300 000 cells X 300 000 cells Y 300 000 cells X + 300 000 cells Y cell culture (24h) with 3ml of medium 0.5% cell seeding overnight with medium 10% FBS washing (PBS) supernatant (CM) mRNA or Protein extracts LIF quantifcation by ELISA assay cells X cells Y cells X+ Y secreted molecules by cells X secreted molecules by cells X+Y secreted molecules by cells Y 300 000 cells X 300 000 cells Y 300 000 cells X + 300 000 cells Y cell culture (24h) with 3ml of medium 0.5% cell seeding overnight with medium 10% FBS washing (PBS) supernatant (CM) mRNA or Protein extracts LIF quantifcation by ELISA assay cell culture (24h) with 3ml of medium 0.5% ) s cell seeding overnight with medium 10% FBS LIF quantifcation by ELISA assay MIGRATION ASSAY insert Schwann cells migrated cells Counting cells supernatant (CM) B B MIGRATION ASSAY Schwann cells supernatant (CM) migrated cells Counting cells
https://openalex.org/W2972320385	http://thesai.org/Downloads/Volume10No8/Paper_2-A_Literature_Review_on_Medicine.pdf	English	null	A Literature Review on Medicine Recommender Systems	International journal of advanced computer science and applications/International journal of advanced computer science & applications	2,019	cc-by	6,947	I. INTRODUCTION Hospitals have access to vast amount of data about patients and their health parameters. Thus, there is a need for convenient way for medical professionals to utilize this information effectively. An example would be the access to aggregated information from existing database on a specific problem at the point of care when it is necessary. Moreover, there are more drugs, tests, and treatment recommendations (e.g. evidence-based medicine or clinical pathways) available for medical staff every day. Thus, it becomes increasingly difficult for them to decide which treatment to provide to a patient based on her symptoms, test results or previous medical history. On the other hand, all these data can be used to strive personalized healthcare which is currently on the rise and predicted to get a major disruptive trend in healthcare in the upcoming years.  Intervention: medication recommendation system  Outcome of relevance: system for medication recommendation  Experimental Design: empirical studies, systematic literature reviews, solution descriptions The intervention and outcome of relevance category are the same. Therefore, they were only included one time. Once this has been agreed on, the search algorithm was constructed. The logical operators AND as well as OR were used to combine the search terms defined in the previous step. The following synonyms were considered: Therefore, a recommendation engine for medical use could be employed to fill this gap and support decision making during therapy. Based on a patient’s current health status, prehistory, current medications, symptoms and past treatments, the engine can look for individuals with similar parameters in the database. At the end, the recommender system would suggest the drugs that were most successful for similar patients. With the help of such a system, the doctor will be able to make a better-informed decision on how to treat a patient.  Medication: Medication, drug, Drug  Recommendation: Recommendation, Recommender, recommender  System: System, framework, Framework, algorithm, Algorithm, engine, Engine This resulted in the following search algorithm: This resulted in the following search algorithm: IBM’s artificial intelligence machine Watson Health [8] is already able to find a suitable treatment for patients based on other patients’ outcome and evidence-based medicine. IBM claims that 81% of healthcare executives familiar with Watson Health agreed that it has a positive impact on their business. This demonstrates that using technology and analytics become increasingly important in healthcare. (IJACSA) International Journal of Advanced Computer Science and Applications, Vol. 10, No. 8, 2019 (IJACSA) International Journal of Advanced Computer Science and Applications, Vol. 10, No. 8, 2019 A Literature Review on Medicine Recommender Systems Benjamin Stark1, Constanze Knahl2, Mert Aydin3, Karim Elish4 Department of Computer Science, Florida Polytechnic University, Lakeland, USA In this paper, we review the existing medicine recommendation system solutions, and compare them based on various features. The goal is to demonstrate the existing solutions for the healthcare providers in order to improve the medicine selection process and select an appropriate medication for the patients. Abstract—Medicine recommender systems can assist the medical care providers with the selection of an appropriate medication for the patients. The advanced technologies available nowadays can help developing such recommendation systems which can lead to more concise decisions. Many existing medicine recommendation systems are developed based on different algorithms. Thus, it is crucial to understand the state-of-the-art developments of these systems, their advantages and disadvantages as well as areas which require more research. In this paper, we conduct a literature review on the existing solutions for medicine recommender systems, describe and compare them based on various features, and present future research directions. The rest of this paper is organized as follows: The methodology for the literature review is presented in Section 2. In Section 3, we discuss the findings. Section 4 presents the limitations. Finally, Section 5 concludes the paper and presents the future work. II. RESEARCH METHOD Keywords—Medicine; recommendation systems; healthcare; systematic review We conducted our literature review in several steps. We followed the guidelines defined in [9]. First, we defined search terms based on population, intervention, outcome of relevance and experimental design. However, we concluded that for our approach the population contains all healthcare facilities. Since this population is so comprehensive and non-specific, we excluded keywords about the population. This resulted in the following major keywords: 6 \| P a g e A. Categorization of Approaches  Logical and reasonable in results and findings regarding the domain of knowledge All the papers included in the final selection are categorized and summarized.  Clearly stated objectives, results and findings regarding the domain of knowledge 1) Ontology and rule-based medicine recommendation systems: The drug recommendation system GalenOWL [4] is based on the Greek drug guide GALINOS where doctors can search for a drug and find details on the drugs and additional information, such as interactions with other drugs. The paper describes a system that recommends drugs for a patient based on the disease of the patient, allergies and known drug interactions for the drugs in the database. To recommend the best fitting drug, rules for medications and interactions are stored in the system, which is based on ontologies, ICD-codes and other information. The application is accessible via the browser.  Well-presented and justified arguments  Well-presented and justified arguments  Reasonably tested and/or applied system  Well referenced with a minimum of ten sources Only papers which met all criteria, thus had 5 points, were included in the final selection. Only papers which met all criteria, thus had 5 points, were included in the final selection.  John Wiley Inc.  John Wiley Inc. We also agreed on using Google Scholar as a search engine as a sixth source because it provides results from a high variety of databases which we might not have included and thus, can lead to a higher quantity of relevant papers. Furthermore, for Google Scholar, there were 11 different publication venues included in the initial selection, from which 5 were included in the final selection. Five of the initial papers retrieved from Google Scholar (29%) were published in IEEE. In the final selection, 44% of the papers from Google Scholar were published by the IEEE. This leads to the conclusion that medicine recommendation systems are a widely discussed topic with many specifics which can also be recognized by the different areas of the journals, but the IEEE database seems to be the most attractive venue for publication in this area. However, in general, approaches and techniques related to medicine recommendation systems are published in a high variety of journals. Then, we agreed on inclusion and exclusion criteria which are defined as follows:  Inclusion criteria: 1) Conference Proceedings and Journals published after 1999 2) Studies focusing on medicine recommendation systems in general and/or specified for any disease 3) Studies focusing on medicine recommendation systems based on graph databases  Exclusion criteria: 1) Papers published before 2000 Specific journals, such as the Journal of Biomedical Semantics, seem to be promising for future literature research in this area, although the results retrieved via the IEEE database met by far the most the inclusion and quality criteria. Moreover, more than 50% of the documents from Google Scholar were included. This is not surprising since Google Scholar fetches documents from many different databases. Despite the strict quality criteria of our study, 25% of all initially selected papers were included in the final set of papers. This indicates a high quality of the databases searched in. 2) Manuscripts written in another language than English 3) Technical reports and white papers as well as Graduation projects, Master thesis and PhD dissertation 4) Textbooks (print and electronic) 4) Textbooks (print and electronic) 5) Studies in other domains of knowledge Finally, some quality criteria for the papers which met the inclusion criteria were defined to guarantee a selection of high- quality papers only. A scored system was used. For each of the following criteria met, a paper is assigned one point: I. INTRODUCTION {{medication OR Medication OR drug OR Drug} AND {recommendation OR Recommendation OR recommender OR Recommender} AND {system OR System OR Engine OR engine OR framework OR Framework OR algorithm OR Algorithm}}. To verify the algorithm and the terms used, we 6 \| P a g e www.ijacsa.thesai.org (IJACSA) International Journal of Advanced Computer Science and Applications, Vol. 10, No. 8, 2019 (IJACSA) International Journal of Advanced Computer Science and Applications, Vol. 10, No. 8, 2019 conducted a test for some papers we knew already. The test was successful as we could find relevant papers. gaps. Table I presents the numbers of papers for the initial and final phase as well as the rate of included papers in percent. Also, to create more transparency, we included a column per phase for the results of the search with Google Scholar. Afterwards, we chose the databases to search in based on available access which led to five databases: As shown in Table I, 52 documents were included initially. After the screening and cross-evaluation as described above, 13 documents remained. The IEEE database has the highest inclusion rate with 5 from the initial papers being included in the final set (71%). From the initial 22 papers from the ACM database 3 remained, leading to an inclusion rate of 14%. Also, it is surprising that none of the papers of ScienceDirect, Elsevier and John Wiley Inc. could be included in the final set of documents. One reason for that might be the publications about recommendation systems are not published in those databases, maybe because the editors of those publications prefer other journals and conferences.  ACM Digital Library  IEEExplore  ScienceDirect  Elsevier III. RESULTS AND DISCUSSION This section describes the final collection of papers in more detail, compares them with each other regarding different parameters, summarizes their approaches and defines research 7 \| P a g e www.ijacsa.thesai.org (IJACSA) International Journal of Advanced Computer Science and Applications, Vol. 10, No. 8, 2019 Vol. 10, No. 8, 2019 TABLE. I. NUMBER OF PAPERS FOUND IN THE RESPECTIVE SEARCH ENGINES (INCL. GOOGLE SCHOLAR RESULTS AND FINAL INCLUSION RATE) Resource Initial Total (incl. Google Scholar) Initial Google Scholar only Final Total (incl. Google Scholar) Final Google Scholar only Inclusion rate in % Total (incl. Google Scholar) ACM 22 2 3 0 14% IEEExplore 7 5 5 4 71% ScienceDirect 5 0 0 0 0% Elsevier 5 0 0 0 0% John Wiley Inc. 3 0 0 0 0% Springer 1 1 1 1 100% Americana Medical Informatics Association 1 1 1 1 100% Journal of Biomedical Semantics 2 2 2 2 100% International Journal of Environmental Research and Public Health 1 1 0 0 0% International Journal of e- Education, e-Business, e- Management and e- Learning 1 1 0 0 0% Journal of Chemical and Pharmaceutical Sciences 1 1 0 0 0% Advanced Internet of Things 1 1 1 1 100% Clinical Pharmacology & Therapeutics 1 1 0 0 0% International Journal of Computer Applications 1 1 0 0 0% Total 52 17 13 9 25% The drug-drug and drug-interaction discovery framework Panacea [5] is based on the approach GalenOWL and uses standardized medical terms and a rich knowledge base which are both modeled as rules. They used SKOS vocabulary, an ontology and reasoning engine and a medical and rules-based reasoning approach. The results show that Panacea is a promising solution, but still needs some improvement. SemMed [14] which is a medical recommendation engine based on Semantic Web Technologies, applies an ontology- based approach. It consists of an inference engine, a rules manager, a support database, and ontology manager. The core classes "Diseases", "Medicines" and "Allergies" were used to develop rules. Another solution proposed by [2] utilizes an ontology for anti-diabetic drug recommendations. However, it also includes the Multiple Criteria Decision-Making approach to compute weights and rank the drugs. It mainly utilizes laboratory data, but also considers risk and benefit factors. IRS-T2D [11] is a drug recommender system which was specifically developed to individualize patient treatment of type 2 diabetes mellitus patients. III. RESULTS AND DISCUSSION The solution combines rule- based decision making with ontologies and semantic web technologies while taking specific patient information, such as the individual HbA1c target into consideration Chen et al. [3] used semantic web rule language to describe the relationship between the rules retrieved from AACEMG. With the rules and knowledge from patient ontology and medicine ontology an inference is derived utilizing the Java Expert System Shell. The inference is then displayed in the system interface. 2) Data mining and machine learning-based medicine recommendation systems: The approach proposed by Sun et al. [15] analyzed EMR records to detect typical treatment regiments and measures (quantitatively) the effectiveness for those regimens for specific patient cohorts. The authors measure the similarity between the treatment records in the EMR, cluster similar ones to treatment regimens based on Map Reduce Enhanced Density Peaks based Clustering, extract semantically meaningful information for the doctor and estimate the treatment outcome for a patient cohort for a typical treatment regimen. The results of applying this approach in an empirical study show that the effective rate of the patient increases as well as the cure rate. Hamed et al. [7] utilized Tweets from Twitter to analyze the well-being of the Tweeter and to give recommendations about alternative medicine possibilities. Therefore, the authors get the information of the Tweets, send the Tweeter a q estionnaire to get more information abo t her state and appl VE SEARCH ENGINES (INCL. GOOGLE SCHOLAR RESULTS AND FINAL INCLUSION RATE) Chen et al. [3] used semantic web rule language to describe the relationship between the rules retrieved from AACEMG. With the rules and knowledge from patient ontology and medicine ontology an inference is derived utilizing the Java Expert System Shell. The inference is then displayed in the system interface. The drug-drug and drug-interaction discovery framework Panacea [5] is based on the approach GalenOWL and uses standardized medical terms and a rich knowledge base which are both modeled as rules. They used SKOS vocabulary, an ontology and reasoning engine and a medical and rules-based reasoning approach. The results show that Panacea is a promising solution, but still needs some improvement. 2) Data mining and machine learning-based medicine recommendation systems: The approach proposed by Sun et al. [15] analyzed EMR records to detect typical treatment regiments and measures (quantitatively) the effectiveness for those regimens for specific patient cohorts. III. RESULTS AND DISCUSSION a trained C4.5 decision tree algorithm to predict the condition of the user. Based on that, the algorithm can derive a recommendation for an alternative medical product. In Table II, the column ―Disease‖ describes whether the concept described in the study focuses on a specific disease. ―Data storage‖ summarizes the method applied to store the data. ―Interface‖ refers to the connection of the back-end modules with each other and the front-end. The column ―Data collection‖ describes the sources the data used for testing the approach, if applicable, were gathered from. DiaTrack was developed by Medvedeva et al. [12] as a drug recommendation system for type 2 diabetes and intends to give doctors a dashboard where they can see similar patient cases and their reaction to a drug or other factors. Therefore, the system compares the disease pattern of multiple patients and gives back the results in a color-coded, easy to understand graph. 1) Disease: As shown in Table II, most studies do not focus on a particular disease. This shows that most work in this field attempts to develop a general-purpose recommendation engine. Finding a recommender system that will work for all diseases would be very useful for general practitioners. However, all studies dealing with a specific disease focus on drug recommendation for diabetes. This means that this type of disease seems to be relatively important and well suited for a drug recommendation system. Since a highly-individualized treatment is required for diabetes, this is also reasonable. The approach proposed by Kushwaha et al. [10] describes a drug recommendation system based on semantic web technology and data mining algorithms. Those two methods were combined to first extract semantic data and then apply data mining algorithms on those data. Data mining algorithms were used to individualize the treatment dependent on the patient’s attributes. The system will not recommend drugs which the patient took before or that would interact with drugs the patient took before. A hybrid framework to recommend drugs by ranking is proposed in [16]. Practitioners make inquiries and order lab tests. Information about this patient is entered into the system as a new case during the process. The system will process the new data and extract patient features. A diagnosis is made based on the patient's problem. The diagnosis is matched to a specific disease category in the system to determine which symptom-drug classifier to use. III. RESULTS AND DISCUSSION Patient features in the new case are put into the classifier to predict which drug cluster/clusters to choose for this patient. Drugs in each cluster will be ranked by the ranking module to form the final recommendation list. 2) Data storage: Data storage is not widely discussed in the studies we reviewed. 5 out of 13 papers mention their data storage approach for datasets such as patient data and drug data. This shows that mostly it is preferred to focus on selected parts of the solution, such as the algorithm. For the studies that include data storage, they all have different ways to store data sets. 2) Data storage: Data storage is not widely discussed in the studies we reviewed. 5 out of 13 papers mention their data storage approach for datasets such as patient data and drug data. This shows that mostly it is preferred to focus on selected parts of the solution, such as the algorithm. For the studies that include data storage, they all have different ways to store data sets. GalenOWL [4] stores data in RDF graphs and utilizes SPARQL queries whereas the SWRL [3] leverages a software called Protégé [6] to store its data. Author in [17] utilizes cloud storage services and the IRS-T2D [11] applies ontologies and semantic web technologies. This shows that there is no standardized approach to store data although the data sets comprise similar data from the electronic medical record (EMR). Mahmoud et al. [1] investigated three different algorithms: Support Vector Machine (SVM), Back Propagation neural network, and ID3 decision tree to find out which algorithm is optimal for a drug recommendation framework. The evaluation is based on scalability, accuracy, and efficiency. Since accuracy is the most important criteria for recommending a drug, the SVM algorithm was identified as the most useful algorithm. The next steps are to implement the model along with the data preparation, visualization, and database system module. 3) Interface: Little information is provided about the interface of drug recommendation systems. The focus is on the recommendation algorithm. Two studies utilize Protégé for the interface and semantic web rule language to show the output of the result of the algorithm. On the other hand, DiaTrack [12] leverages dynamic-service middleware to provide a visualization of the output. This shows that drug recommendations are still in development. III. RESULTS AND DISCUSSION The authors measure the similarity between the treatment records in the EMR, cluster similar ones to treatment regimens based on Map Reduce Enhanced Density Peaks based Clustering, extract semantically meaningful information for the doctor and estimate the treatment outcome for a patient cohort for a typical treatment regimen. The results of applying this approach in an empirical study show that the effective rate of the patient increases as well as the cure rate. 2) Data mining and machine learning-based medicine recommendation systems: The approach proposed by Sun et al. [15] analyzed EMR records to detect typical treatment regiments and measures (quantitatively) the effectiveness for those regimens for specific patient cohorts. The authors measure the similarity between the treatment records in the EMR, cluster similar ones to treatment regimens based on Map Reduce Enhanced Density Peaks based Clustering, extract semantically meaningful information for the doctor and estimate the treatment outcome for a patient cohort for a typical treatment regimen. The results of applying this approach in an empirical study show that the effective rate of the patient increases as well as the cure rate. SemMed [14] which is a medical recommendation engine based on Semantic Web Technologies, applies an ontology- based approach. It consists of an inference engine, a rules manager, a support database, and ontology manager. The core classes "Diseases", "Medicines" and "Allergies" were used to develop rules. Another solution proposed by [2] utilizes an ontology for anti-diabetic drug recommendations. However, it also includes the Multiple Criteria Decision-Making approach to compute weights and rank the drugs. It mainly utilizes laboratory data, but also considers risk and benefit factors. IRS-T2D [11] is a drug recommender system which was specifically developed to individualize patient treatment of type 2 diabetes mellitus patients. The solution combines rule- based decision making with ontologies and semantic web technologies while taking specific patient information, such as the individual HbA1c target, into consideration. Hamed et al. [7] utilized Tweets from Twitter to analyze the well-being of the Tweeter and to give recommendations about alternative medicine possibilities. Therefore, the authors get the information of the Tweets, send the Tweeter a questionnaire to get more information about her state and apply 8 \| P a g e www.ijacsa.thesai.org (IJACSA) International Journal of Advanced Computer Science and Applications, Vol. 10, No. 8, 2019 summary and discussion of each column presented in the corresponding table. III. RESULTS AND DISCUSSION Generally, once the recommendation engine is defined, it seems like the focus is on the user experience. Moreover, for the studies that did provide information about user interfaces, Protégé is the framework mostly applied. Protégé [13] is an open-source ontology editor that provides developers with a user interface to create intelligent systems. Developed by Stanford University, this application is appealing due to its free-to-use license terms, its active community for support and its extensible environment. Another drug recommendation system is a cloud-based platform utilizing various algorithms [17]. Using the vector service model, the drug character is formatted according to the description of the drug information. Then a k-means algorithm is applied to cluster drugs. Subsequently, an evaluation using collaborative filtering leads to recommendations. Finally, tensor decomposition is applied to address sparsity and massive data, shortcomings of collaborative filtering. This multi-step process helps to make an accurate recommendation. B. Characteristics of Approaches Tables II, III and IV present the results of the literature review. The columns refer to the different dimensions we compare the studies with. In each table, there is a short 9 \| P a g e www.ijacsa.thesai.org (IJACSA) International Journal of Advanced Computer Science and Applications, Vol. 10, No. 8, 2019 TABLE. II. RESULTS OF LITERATURE REVIEW IN TERMS OF DISEASE, DATA STORAGE, INTERFACE AND DATA COLLECTION USED IN THE STUDIES Study Disease Data storage Interface Data collection Data-driven Automatic Treatment Regimen Development and Recommendation [15] Not specified No information No information 14 grade 3 hospitals in China Panacea, a semantic-enabled drug recommendations discovery framework [5] Not specified No information No information University hospital of Thessaloniki SemMed [14] Not specified No information No information No information The recommendation of medicines based on multiple criteria decision making and domain ontology [2] Diabetes (not further specified) No information No information No information T-Recs [7] Not specified (Author tested it for symptoms such as diarrhea, headache, fever, stomach ache) No information No information Twitter (~500,000 Tweets) GalenOWL [4] Not specified RDF graph SPARQL queries No information No information IRS-T2D [11] Type 2 Diabetes Knowledge base with ontologies and SWR Jess SWRL2Jess OWL2Jess Protégé-OWL API Test patients DiaTrack [12] Type 2 Diabetes Database management system Dynamic service middle-ware (similarity, reasoning and visualization components and provides professional interaction tools) No information LOD Cloud Mining for Prognosis Model [10] Not specified No information No information Various Semantic Web sources (SIDER (for side effects), Drug Bank) A framework of hybrid recommender system [16] Not specified No information No information Electronic Medical Record System (EMRS) A recommendation system based on domain ontology and SWRL for anti-diabetic drugs selection [3] Type 2 Diabetes Protégé Protégé No information An Intelligent Medicine Recommender System Framework [1] Not specified No information No information Hospital Information System (HIS) with 70% training data and 30% testing data. CADRE [17] Not specified Cloud No information Walgreens 4) Data collection: Generally, patient data are collected from hospital information systems (HIS). Data retrieved from HIS accurately resemble the data hospitals have available in practice. HIS data are usually not in an appropriate format for most ontology-based intelligent systems. Hence, it is essential to think of ways to format the data so they fit to the algorithm applied. For the drug data, various sources were leveraged. B. Characteristics of Approaches CADRE [17] used the Walgreens website whereas [10] include data from a drug bank and SIDER for side effects. Comprising different sources for drug data could lead to side effects of drugs. These differences might influence the drug recommendation from the data perspective. Furthermore, data are in different languages. For example, the data used by [15] include data in the Chinese language, since they are retrieved from Chinese hospitals. Thus, it is important to assess the semantics when dealing with multiple languages. Furthermore, sentiment data are generally not collected except for T-Recs [7], which used 500,000 tweets from Twitter. This means that the proposed recommendation systems generally do not consider patient feedback. This is an important element side effects of drugs. These differences might influence the drug recommendation from the data perspective. Furthermore, data are in different languages. For example, the data used by [15] include data in the Chinese language, since they are retrieved from Chinese hospitals. Thus, it is important to assess the semantics when dealing with multiple languages. Furthermore, sentiment data are generally not collected except for T-Recs [7], which used 500,000 tweets from Twitter. This means that the proposed recommendation systems generally do not consider patient feedback. This is an important element that is being omitted and might reduce overall patient satisfaction. side effects of drugs. These differences might influence the drug recommendation from the data perspective. Furthermore, data are in different languages. For example, the data used by [15] include data in the Chinese language, since they are retrieved from Chinese hospitals. Thus, it is important to assess the semantics when dealing with multiple languages. Furthermore, sentiment data are generally not collected except for T-Recs [7], which used 500,000 tweets from Twitter. This means that the proposed recommendation systems generally do not consider patient feedback. This is an important element that is being omitted and might reduce overall patient satisfaction. 4) Data collection: Generally, patient data are collected from hospital information systems (HIS). Data retrieved from HIS accurately resemble the data hospitals have available in practice. HIS data are usually not in an appropriate format for most ontology-based intelligent systems. Hence, it is essential to think of ways to format the data so they fit to the algorithm applied. For the drug data, various sources were leveraged. B. Characteristics of Approaches CADRE [17] used the Walgreens website whereas [10] include data from a drug bank and SIDER for side effects. Comprising different sources for drug data could lead to discrepancies in the description of benefits as well as in the 10 \| P a g e www.ijacsa.thesai.org (IJACSA) International Journal of Advanced Computer Science and Applications, Vol. 10, No. 8, 2019 studies use online services. CADRE [17] utilizes the cloud platform to give medicine recommendations based on symptoms. The LOD cloud mining study by [10] leverages semantic knowledge from the LOD cloud. The GalenOWL [4] uses semantic-enabled online services to provide drug-drug and drug-disease interaction discovery. Furthermore, T-Recs [7] utilizes Twitter to monitor tweet sentiment, create an analysis for the tweets, and the calculate recommendations. Other studies apply rule-based inference engines such as Pellet and Jena/Drools. In all 13 studies, the technology used to apply the algorithm was different. This shows that researchers do not restrict themselves to apply only one specific software tool, but utilize the various possibilities available. Despite this flexibility, scientists need to consider the costs of the technology to make it reasonable for an average hospital to purchase it. Hence, open-source software which was used to develop algorithms such as Protégé, Pellet and Jena rule engine seem to be reasonable and preferable choice. In Table III, the column ―Data preparation‖ relates to the steps applied to the raw data so they fit for the algorithm. ―Platform/Technology‖ refers to the technology and/or platform utilized for the implementation. 5) Data preparation: In Table III, we found that the studies apply different data preparation procedures, for instance regarding data formatting and cleaning. This seems reasonable if we assume that the format of data is different across the different studies. Furthermore, the table shows that most studies utilize data preparation modules to provide an acceptable format for their algorithm module. Studies such as [1], [16] or [2] use normalization techniques to uniformly scale the data across the modules. On the other hand, some authors decided to manually prepare the data. In the case of T- Recs [7], tweets were distinguished to be either relevant or irrelevant. The study in [15] shows the most relevant medicines were selected and then divided into four periods. Since most studies have information about data preparation, it shows that this aspect is essential when developing a recommendation engine. B. Characteristics of Approaches Table IV compares the studies in terms of the algorithms used, and presents future work identified by these studies. The algorithms used in the reviewed studies were described earlier in Section III. 6) Platform/Technology: With regards to the platform/technology of the recommender systems, three of the TABLE. III. RESULTS OF LITERATURE REVIEW IN TERMS OF DATA PREPARATION AND THE PLATFORM/TECHNOLOGY USED IN THE STUDIES Study Data preparation Platform/Technology Data-driven Automatic Treatment Regimen Development and Recommendation [15] Yes (select most relevant medicines (138); divide treatment record into 4 periods) Custom Panacea, a semantic-enabled drug recommendations discovery framework [5] Yes (applying the SKOS vocabulary) Querying instance and knowledge base Rule engines (Jena/Drools rule engine) SemMed [14] No information Inference engine Rules manager Support DB and ontology manager The recommendation of medicines based on multiple criteria decision making and domain ontology [2] Yes (normalization of benefit and risk factors) No information T-Recs [7] Yes (Manual distinction between relevant and irrelevant tweets; grouping of tweets) Twitter Tweet Sentiment monitor Tweet analysis and computing recommendation GalenOWL [4] Yes (ATC, ICD-10, UNII, Substances, Conditions, Indications-Contraindications) Online-service IRS-T2D [11] No information No information DiaTrack [12] No information A standard web-browser front-end for Data Entry, Research, Practice Administration and Site Administration LOD Cloud Mining for Prognosis Model [10] Yes (Queried with SPARQL) LODD cloud, queries on drug data with SPARQL 1.1 with Java IDE, database: RDF dump stored in Sesame, app uses the server of Sesame A framework of hybrid recommender system [16] Yes (Text Mining Module, Data Normalization, Drug Clustering Module) No information A recommendation system based on domain ontology and SWRL for anti-diabetic drugs selection [3] Yes (Inference engine (Pellet) transformed the format acceptable to the recommendation system) Medicine ontology was created by Protégé Inference engine (Pellet) An Intelligent Medicine Recommender System Framework [1] Yes (Data normalization using min and max functions. Correlation analysis using Chi-Square Tests.) No information CADRE [17] No information Cloud TABLE. III. RESULTS OF LITERATURE REVIEW IN TERMS OF DATA PREPARATION AND THE PLATFORM/TECH 11 \| P a g e 11 \| P a g e www.ijacsa.thesai.org (IJACSA) International Journal of Advanced Computer Science and Applications, Vol. 10, No. 8, 2019 Vol. 10, No. 8, 2019 TABLE. IV. B. Characteristics of Approaches RESULTS OF LITERATURE REVIEW IN TERMS OF ALGORITHMS AND FUTURE WORK PRESENTED IN EACH STUDY Study Algorithm Future work Data-driven Automatic Treatment Regimen Development and Recommendation [15]  Custom (not pre-defined, but used Map Reduce Enhanced Density Peaks based Clustering and decision tree)  No information Panacea, a semantic-enabled drug recommendations discovery framework [5]  Ontology  Rules  Weighing of interactions, contraindications based on severity observation, probabilistic inference  Addition of dosage recommendation  Increase sample for testing.  Automate manual work which is required to enrich rule base. SemMed [14]  Ontology  Semantic web techniques  Integration of this system with other existing ones, e.g. MEDBOLI.  Addition of functions, e.g. prediction of treatment times with neural networks. The recommendation of medicines based on multiple criteria decision making and domain ontology [2]  Ontology (created with protégé)  Multiple Criteria Decision Making  Evaluation by multiple diabetes physicians GoT-Recs [7]  Decision tree (C4.5)  Extend it to other products, e.g. beverages, snacks, etc.  Include other online data as well, e.g. "PatientsLikeMe"  Get permission from twitter  Modify architecture so it is scalable  Test different algorithms GalenOWL [4]  Ontologies  Semantic web technologies  Rules  Expansion of semantic rules  Prioritization of interactions of drugs and diseases since not all interactions have the same importance  Performance optimization (e.g. context extraction from medical knowledge) IRS-T2D [11]  Ontologies  Semantic web technologies  Test more patients  Improve patient profile to store more information  Add insulin data and rules considering insulin to the system DiaTrack [12]  Pattern comparison in data  Enhance system to not only work based previous patients but also based on general drug features (consider ontologies and semantic web technologies) LOD Cloud Mining for Prognosis Model [10]  Semantic Web techniques  Data mining algorithm (Decision trees usin  C4.5 algorithm and bagging)  Update data on drugs, diseases and interactions as needed  Extract more meaningful features like toxicity, food interaction etc. B. Characteristics of Approaches A framework of hybrid recommender system [16]  Case-based reasoning for ranking the drug clusters  Artificial neural network for Symptom- Drug Classifier module  cTAKE system for text-mining module  Implementation of the recommender system  Free text messages in EMRS need to be taken into consideration  Distinguishable features for personalization need to be extracted  The recommender must be dynamic and adaptive to assess temporal efficiency of drugs and add new drugs A recommendation system based on domain ontology and SWRL for anti-diabetic drugs selection [3]  Semantic Web Rule Language(SWRL)  Java Expert System Shell (JESS)  Ontology  Strengthen patient ontology  Test more patient data for the system  Calculate the dosage of the medicine An Intelligent Medicine Recommender System Framework [1]  SVM (Support Vector Machine)  Back-propagation neural network  ID3 decision tree  Build the recommendation model  Apply MapReduce to enlarge the ability of processing big diagnosis data CADRE [17]  Vector space model (VSM)  K-means clustering  Collaborative Filtering  Tensor decomposition  Investigate how to improve the accuracy of CADRE by considering user’s age, geography, and other factors As the table indicates, almost all of the approaches (9 out of 13) state some future work and research areas. Although some of them are rather specific to the technique presented in the paper, it is possible to derive some general fields where more research is required. The three main areas based on this literature review are:  Finding solutions for including a recommendation for the dosage of a medicine  Verifying the results, e.g. by increasing testing, especially the sample of testing  Finding solutions which are highly scalable 12 \| P a g e 12 \| P a g e www.ijacsa.thesai.org (IJACSA) International Journal of Advanced Computer Science and Applications, Vol. 10, No. 8, 2019 (IJACSA) International Journal of Advanced Computer Science and Applications, Vol. 10, No. 8, 2019 This extensive literature review shows that there are many solutions for drug recommendation systems. Most of them are based on manually constructed ontologies and use sophisticated data mining or machine learning methods. Especially the processes including manual work are very time consuming. Also, none of these approaches utilizes a graph database to model the relationships between patients and to apply an algorithm to this model, although this might be a well-suited approach. IV. LIMITATIONS Our literature review has two main limitations, namely, the paper selection and content. Out of 52 papers, only 13 were reviewed based on the strict inclusion, exclusion and quality criteria we chose. Along with the strict search criteria, the systematic review included papers from a limited number of databases. However, we used six main databases that are well known. [5] Doulaverakis, C., Nikolaidis, G., Kleontas, A., and Kompatsiaris, I. 2014. Panacea, a semantic-enabled drug recommendations discovery framework. Journal of Biomedical Semantics 5, 13. [6] Guo, W. Y. 2008. Reasoning with semantic web technologies in ubiquitous computing environment. Journal of Software 3, 8, 27–33. [7] Hamed, A. A., Roose, R., Branicki, M., and Rubin, A. 2012. T-Recs: Time-aware Twitter-based Drug Recommender System. 2012 IEEE/ACM International Conference on Advances in Social Networks Analysis and Mining. Some papers offer little detail on the exact implementation and architecture of the solutions built. This made it more difficult to assess which applications were used to build the system. Also, some papers proposed only a theoretical solution on how to recommend a drug such as [16], but did not implement the solution. On the other hand, some papers did implement the solution such as [2], but no evaluation was made on the performance. Therefore, several questions stay under investigation, such as "how accurate are these recommender systems?" and "does it reduce the symptoms patients have?" [8] IBM. 2017. IBM Watson Health. http://www.ibm.com/watson/health/. [9] Kitchenham, B. and Charters, C. 2007. Guidelines for Performing Systematic Literature Reviews in Software Engineering. 2007 Joint Report - EBSE 20 07-0 01. [10] Kushwaha, N., Goyal, R., Goel, P., Singla, S., and Vyas, O. P. 2014. LOD Cloud Mining for Prognosis Model(Case Study. Native App for Drug Recommender System). AIT 04, 03, 20–28. [11] Mahmoud, N. and Elbeh, H. 2016. IRS-T2D. Individualize Recommendation System for Type2 Diabetes Medication Based on Ontology and SWRL. In Proceedings of the 10th International Conference on Informatics and Systems - INFOS '16. ACM Press, New York, New York, USA, 203–209. DOI=10.1145/2908446.2908495. REFERENCES [1] Bao, Y. and Jiang, X. 2016. An Intelligent Medicine Recommender System Framework. 2016 IEEE 11th Conference on Industrial Electronics and Applications (ICIEA). [2] Chen, R.-C., Chiu, J. Y., and Batj, C. T. 2011. The recommendation of medicines based on multiple criteria decision making and domain ontology — An example of anti-diabetic medicines, 27–32. [3] Chen, R.-C., Huang, Y.-H., Bau, C.-T., and Chen, S.-M. 2012. A recommendation system based on domain ontology and SWRL for anti- diabetic drugs selection. Expert Systems with Applications 39, 4, 3995– 4006. [4] Doulaverakis, C., Nikolaidis, G., Kleontas, A., and Kompatsiaris, I. 2012. GalenOWL: Ontology-based drug recommendations discovery. Journal of Biomedical Semantics 3, 14. B. Characteristics of Approaches Graph databases can model the data in graphs which is a more natural way to store data than any other database offers. Medical institutions usually have many patients who can be illustrated in a graph as a network of patients. Thus, this approach may be superior to the ones discussed earlier in this paper and also addresses the last topic listed for future research. The reasons for that are the unique features of graph databases, such as high consistency and high scalability. they are highly scalable and consistent which would account for the last of the aforementioned topics for future work. Furthermore, their data model seems to be promising for recommendation systems due to their network structure and ease for querying. Hence, another direction for future research would be the creation of medicine recommendation engines based on graph database. 13 \| P a g e V. CONCLUSIONS AND FUTURE WORK This paper presented a systematic literature review for medicine recommendation engines. We reviewed 13 studies that met our strict criteria in six different databases. These studies can be split into two categories: (i) machine learning and data mining-based, and (ii) ontology and rule-based approach. The studies were summarized and evaluated across several parameters: diseases, data storage, interface, data collection, data preparation, platform/technology, algorithm, and future work. Most of the studies that did not focus on any disease, had less information about data storage, interface, data collection, data preparation, platforms and technology, and customized algorithms. [12] Medvedeva, O., Knox, T., and Paul, J. 2007. DiaTrack. Web-based application for assisted decision-making in treatment of diabetes. Journal of Computing Sciences in Colleges 23, 1, 154–161. [13] Protégé. 2016. Protégé. http://protege.stanford.edu/. Accessed 15 March 2017. [14] Rodríguez, A., Jiménez, E., Fernández, J., Eccius, M., Gómez, J. M., Alor-Hernandez, G., Posada-Gomez, R., and Laufer, C. 2009. SemMed: Applying Semantic Web to Medical Recommendation Systems. 2009 First International Conference on Intensive Applications and Services. [15] Sun, L., Liu, C., Guo, C., Xiong, H., and Xie, Y. 2016. Data-driven Automatic Treatment Regimen Development and Recommendation. KDD '16 Proceedings of the 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, 1865–1874. [16] Zhang, Q., Zhang, G., Lu, J., and Wu, D. 2015. A framework of hybrid recommender system for personalized clinical prescription. 2015 International Conference on Intelligent Systems and Knowledge Engineering. For future work, our review suggests to extend the existing solutions by adding recommendations for the dosage of drugs, as well as building highly scalable solutions. Also, based on the evaluation, we identified that none of the studies we reviewed include a graph database in their solution for a drug recommendation system. Graph database such as Neo4j seem to be very suitable for drug recommendation engines because [17] Zhang, Y., Zhang, D., Hassan, M. M., Alamri, A., and Peng, L. 2015. CADRE. Cloud-Assisted Drug REcommendation Service for Online Pharmacies. Mobile Netw Appl 20, 3, 348–355. 13 \| P a g e www.ijacsa.thesai.org
https://openalex.org/W2949042908	http://old.scielo.br/pdf/abmvz/v71n2/1678-4162-abmvz-71-02-374.pdf	English	null	Prevalence of bovine tuberculosis in the Aksu Region of Xinjiang, China, between 1985 and 2016	Arquivo Brasileiro de Medicina Veterinária e Zootecnia/Arquivo brasileiro de medicina veterinária e zootecnia	2,019	cc-by	2,721	http://dx.doi.org/10.1590/1678-4162-10550 http://dx.doi.org/10.1590/1678-4162-10550 Arq. Bras. Med. Vet. Zootec., v.71, n.2, p.374-378, 2019 ABSTRACT Prevalence of bovine tuberculosis infection in cattle in Aksu Prefecture determined by intradermal tuberculin skin test (TST), between 1985 and 2016. Cattle were analyzed according to region, feeding pattern, herds and age. A total of 890,009 cattle were tested, with overall bovine tuberculosis prevalence of 0.13% (1172/890009). Statistically significant difference was found in feeding pattern and herds. Prevalence in cows (0.19%, 615/327022) was higher than that in beeves (P< 0.01, OR= 1.903, 95% CI = 1.696 to 2.134). Significant difference (P< 0.01; OR= 2.238, 95%; CI= 1.937 to 2.585) was evident for rates for bovine tuberculosis in the peasant household (0.12%, 942/802343) and farm groups (0.26%, 230/87666). The overall prevalence of bTB was decreased in the Aksu Prefecture, especially the positive rate was under 0.1% in 2010s. We concluded that the control measures forbovine tuberculosis in the Aksu region cattle herds are effective. Keywords: prevalence, bovine tuberculosis, aksu prefecture Recebido em 27 de janeiro de 2018 Aceito em 25 de setembro de 2018 # Equal contributors Autor para correspondência (corresponding author) E-mail: zhaolidky@126.com Y.H. Liu1#, Q. Kang2#, B. Yang3#, F. Li1, 2, X.Q. Li1, L. Zhang1, L. Zhao1 Y.H. Liu1#, Q. Kang2#, B. Yang3#, F. Li1, 2, X.Q. Li1, L. Zhang1, L. Zhao1* 1College of Animal Science ˗ Tarim University ˗ Alar, Xinjiang, China 2Animal Loimia Controlling and Diagnostic Center of Aksu Region ˗ Aksu, Xinjiang, China 3Animal Disease Control Center of Ordos ˗ Ordos, China 1College of Animal Science ˗ Tarim University ˗ Alar, Xinjiang, China 2Animal Loimia Controlling and Diagnostic Center of Aksu Region ˗ Aksu, Xinjiang, China 3Animal Disease Control Center of Ordos ˗ Ordos, China 1College of Animal Science ˗ Tarim University ˗ Alar, Xinjiang, China 2Animal Loimia Controlling and Diagnostic Center of Aksu Region ˗ Aksu, Xinjiang, China 3Animal Disease Control Center of Ordos ˗ Ordos, China htt X. htt L. htt L 1College of Animal Science ˗ Tarim University ˗ Alar, Xinjiang, China 2Animal Loimia Controlling and Diagnostic Center of Aksu Region ˗ Aksu, Xinjiang, China 3Animal Disease Control Center of Ordos ˗ Ordos, China RESUMO Prevalência de infecção por tuberculose bovina em gado na prefeitura de Aksu determinada por teste cutâneo tuberculínico (TST) entre 1985 e 2016 foi avaliada. O gado foi analisado de acordo com região, padrão alimentar, rebanho e idade. Um total de 890009 animais foram testados, com prevalência de 0,13% de tuberculose bovina (1172/890009). Diferença estatisticamente significativa foi encontrada em padrão alimentar e rebanhos. Prevalência em vacas (0,19%, 615/327022) foi mais alta que em bois (P< 0,01, OR= 1,903, 95% CI = 1,696 a 2,134). Diferenças significativas (P< 0,01; OR= 2,238, 95%; CI= 1,937 a 2,585) foram evidentes em taxas para tuberculose bovina em casas de camponeses (0,12%, 942/802343) e grupos de fazendeiros (0,26%, 230/87666). A prevalência de bTB caiu na prefeitura Aksu, a taxa positiva se encontrava abaixo de 0.1% a partir de 2010. Conclui-se que as medidas de controle para tuberculose bovina na região de Aksu foram eficazes. Palavras-chave: prevalência, tuberculose bovina, aksu Prevalence of bovine tuberculosis in the Aksu Region of Xinjiang, China, between 1985 and 2016 Y H Liu Y.H. Liu1# https://orcid.org/0000-0003-4432-9433 Q. Kang2# https://orcid.org/0000-0001-5926-3946 B. Yang3# https://orcid.org/0000-0002-8701-1727 F. Li1, 2 https://orcid.org/0000-0003-0689-8811 X.Q. Li1 https://orcid.org/0000-0002-7296-7179 L. Zhang1 https://orcid.org/ 0000-0001-9739-0663 L. Zhao1* https://orcid.org/0000-0001-8538-5605 INTRODUCTION wildlife and humans, mainly due to Mycobacterium bovis (Humblet et al., 2009; Katale et al., 2013). Cattle may serve as the main host for M. bovis worldwide (Gumi et al., 2011), while many or most domestic animals such as pigs, dogs, cats, horses and sheep are considered to be spill-over hosts (Franck et al., 2005; Dejene et al., 2016). Aerosol is considered to be the primary transmission pathway of infection in animals (Gumi et al., 2011; Franck et al., 2005). Infectious diseases are important factors affecting the health of populations. Bacterial zoonoses still represent a serious medical problem (Wang et al., 2012; Karabay et al., 2004). Bovine tuberculosis (bTB) is a chronic debilitating and contagious disease of livestock, Infectious diseases are important factors affecting the health of populations. Bacterial zoonoses still represent a serious medical problem (Wang et al., 2012; Karabay et al., 2004). Bovine tuberculosis (bTB) is a chronic debilitating and contagious disease of livestock, Prevalence of bovine... Other transmission pathways of infection such as ingestion of contaminated feeds, water and fomites have been confirmed (Gumi et al., 2011). Other transmission pathways of infection such as ingestion of contaminated feeds, water and fomites have been confirmed (Gumi et al., 2011). Several surveys on bovine tuberculosis in cow were conducted in some provinces in China in recent years (Ma et al., 2012; Li et al., 2012; Xin and An, 2007; Gan, 2008; Fu et al., 2004). However, no studies have been reported on the prevalence of bovine tuberculosis in the Aksu region. The current study determined to estimate the risk factors associated with exposure to the pathogen of bovine tuberculosis infection in cow in this region. While most cattle infected with M. bovis usually exhibit a chronic disease often characterized by lesions in the lungs and the lymph nodes of the head and thoracic cavity, progressive emaciation, diarrhea and reduced milk yields can also occur. Most cases of bovine tuberculosis infection are inapparent and lack clinical symptoms. The main pathologic changes of bovine tuberculosis are tuberculous nodules and caseous necrosis or calcified nodules in variety of tissues and organs. Arq. Bras. Med. Vet. Zootec., v.71, n.2, p.374-378, 2019 RESULTS The results of the TST detections on bovine tuberculosis positive samples are shown in Table 1. Among the 890,009 cattle, 1172 (0.13%) were positive by TST. The prevalence in cows (0.19%, 615/327022) and in farm group (0.26%, 230/87666) were higher than that in beeves (AFe= 0.300; OR= 1.903, 95%; CI= 1.696 to 2.134) and in peasant household groups (AFe= 0.499; OR= 2.238, 95%; CI= 1.937 to 2.585) (Table 1). The data were analyzed (χ2-test) using the statistical software package SPSS version 20.0 (SPSS Inc., Chicago, IL, USA). The risk factors, Table 1. Prevalence of bovine tuberculosis infection in cattle in Aksu Prefecture determined by TST Variable Number of cattle (%) Positive cattle (%) single-factor analysis OR (95% CI) AFe/P Total 890009 (100) 1172 (0.13) Variety Cows 327022 (36.74) 615 (0.19) 1 Beeves 562987 (63.26) 557 (0.10) 1.903 (1.696-2.134) 0.300/<0.01 Feeding pattern Farm (≥ 20) 87666 (9.85) 230 (0.26) 1 Peasant household 802343 (90.15) 942 (0.12) 2.238 (1.937-2.585) 0.499/<0.01 Table 1. Prevalence of bovine tuberculosis infection in cattle in Aksu Prefecture determined b Table 1. Prevalence of bovine tuberculosis infection in cattle in Aksu Prefecture determined by TST Variable Number of cattle (%) Positive cattle (%) single-factor analysis OR (95% CI) AFe/P The prevalence of bovine tuberculosis infection from 1980s to 2010s was 0.04% (39/104462) in Aksu, 0.19% (156/82529) in Xayar, 0.21% (304/145964) in Kuqa, 0.10% (88/88183) in Baicheng, 0.08% (41/49426) in Xinhe, 0.12% (240/201567) in Wensu, 0.11% (72/68189) in Uqturpan, 0.10% (37/35883) in Kalpin and 0.17% (195/113806) in Awat (Table 2). The prevalence of bovine tuberculosis infection in Aksu Prefecture was 0.18% (222/120847) in 1980s, 0.12% in 1990s (208/175962), 0.19% in 2000s (687/352938) and 0.02% in 2010s (55/240262). The prevalence of bovine tuberculosis infection from 1980s to 2010s was 0.04% (39/104462) in Table 2. Liu et al. The current survey was implemented from 1985 to 2016, and 890,009 samples were collected from dairy herds and 10% beef cattle in the whole Aksu Prefecture (total approximate 6 million cattle, including cows and beef). All tasks were conducted by Animal Loimia Controlling and Diagnostic Center of Aksu Region. The experiment was performed with pure bovine tuberculin (PPD, produced by Harbin Pharmaceutical Group Bio-vaccine Co, Ltd.) skin allergy test (TST), according to the diagnostic techniques for tuberculosis of animal (GB/T18645-2002) methods. Biometric data from the cattle, including variety and feeding patterns were acquired from the animal owners or medical records. including variety and feeding pattern were noted in present survey. Identification of a risk factor required a 95% confidence level (P< 0.05) as well as a biologically plausible association between the factor and rate of allergic reaction to bovine tuberculosis. MATERIALS AND METHODS The Aksu Prefecture, which comprises 1 city (Aksu) and 8 counties (Xayar, Kuqa, Baicheng, Xinhe, Wensu, Uqturpan, Kalpin and Awat), is an important area in South Xinjiang. The district of study (Figure 1) is located at 78°03′ to 84°07′ N and 39°30′ to 42°41′ E. The Aksu Prefecture has a total land area of 132,500km2 and a human population of approximately 2.5 million. The annual rainfall in this region varies between 42.4 and 94.4mm. Currently, the livestock population consists of 354,200 head of cattle, 208,500 pigs, 4.11 million sheep and 1.35 million goats. Livestock husbandry is regarded as an important source of income in the Aksu Prefecture. Bovine tuberculosis is a serious disease with potential public health and economic costs to society, since it can significantly affect international trade of animals and their products and create productivity losses (Ayele et al., 2004; Dejene et al., 2016). Bovine tuberculosis, which is currently ranked as the 7th highest in the world (Gumi et al., 2012), is a significant public health and food safety concern and considered as a second-class animal epidemic disease in China. Figure 1. Aksu Prefecture administrative map. Figure 1. Aksu Prefecture administrative map. Figure 1. Aksu Prefecture administrative map. 375 Arq. Bras. Med. Vet. Zootec., v.71, n.2, p.374-378, 2019 Arq. Bras. Med. Vet. Zootec., v.71, n.2, p.374-378, 2019 376 Arq. Bras. Med. Vet. Zootec., v.71, n.2, p.374-378, 2019 DISCUSSION bovine tuberculosis infection was 0.26% and 0.12% in the farm and peasant household groups, respectively. Statistically significant differences between these values were obtained (P< 0.01; OR= 2.238, 95%; CI= 1.937 to 2.585). Cattle in most of the farm are pastoralism, based on transhumance, which refers to a pattern of seasonal movement between dry season and wet season pastures. Given the feeding patterns and proximity to wildlife, studies are needed to explore the disease status, dynamics in the ecosystem and risk factors for bTB infection in animals and ultimately also humans. The overall prevalence in cows in the Aksu Prefecture was 0.13% (1172/890009), comparatively higher than those observed in Hui Autonomous Prefecture of Changji in Xinjiang (0.12%) (Ma et al., 2012). However, the obtained prevalence in the Aksu Prefecture is lower than those in Hui Autonomous Prefecture of Changji at 0.27% from 2008 to 2011, Urumqi city at 0.52% from 2007 to 2011 (Li et al., 2012) and Bole city at 3.82% from 2002 to 2004 (Xin and An, 2007) in Xinjiang, and Guangxi Zhuang Autonomous Region at 0.48% (Gan, 2008), as well as those in Qinghai province at 0.18% (Fu et al., 2004). The elimination of bTB from cattle is likely to be a difficult goal, but it is an extremely important one to accomplish. In Xinjiang, bTB has been controlled through “a test-and slaughter measures” for cattle and made major progress in decreasing the positive rate. Nevertheless, bTB remains a problem in some regions where surveillance and control activities are often inadequate or unavailable possibly due to lack of funds to support the whole exercise and compensate for tested and slaughtered animals (Katale et al., 2013). In conclusion, current control measures for bovine tuberculosis of Aksu Prefecture cattle herds might be valuated as effective. The bovine tuberculosis in cattle was widespread in the 1940s, and outbreak in the 1950s, the highest infection rate of province reached 41.3% in China. In the 1970s, the epidemic of bovine tuberculosis achieved the historical high point, the infection rate of some area reached 67.4%. In the 1980s, the positive rate of tuberculin skin allergy test achieved 2.68%, and 28 P/A/M (provinces, autonomous regions and municipalities) of China detected TB positive cattle, Xinjiang is particularly severe (Gan, 2008). In the study, the high prevalence of bTB was Kuqa in the 1980s and 1990s. DISCUSSION In the 2000s, the prevalence of bTB in the Aksu Prefecture increased, except in Kuqa. The overall prevalence of bTB decreased in the Aksu Prefecture; the positive rate was under 0.1% in 2010s. ACKNOWLEDGEMENT We thank the Animal Loimia Controlling and Diagnostic Center of Aksu Prefecture for supporting this work and permit to work under their jurisdiction. This study was funded by the National Natural Science Foundation of China (No. 31460655) and the open project of Key Laboratory of Tarim Animal Husbandry Science and Technology, Xinjiang Production & Construction Corps (HS201501, HS201801). Tuberculin skin allergy test is the only recommended TB diagnosis method in the world organization for animal health (OIE). It has generally been used to determine the prevalence of bovine tuberculosis in China. The present study provides the first information on bovine tuberculosis infection in cattle in the Aksu Prefecture. In the current research, the prevalence in cows was 0.19% (615/327022), higher than that in beeves (P< 0.01, OR= 1.903, 95% CI= 1.696 to 2.134) at 0.10% (557/562987), possibly due to the susceptibility in difference between different breeds of cattle. Cows are more susceptible to TB. RESULTS Prevalence of bovine tuberculosis infection in cattle in different region in Aksu Prefecture determined by TST Area Age Aksu Xayar Kuqa Baicheng Xinhe Wensu Uqturpan Kalpin Awat Total Prevalence (%, total tested) 1980s 0.13 (12279) 0.14 (11356) 0.43 (21513) 0.05 (13975) 0.01 (6993) 0.20 (21120) 0.19 (14575) 0.05 (1954) 0.12 (17082) 0.18 (120847) 1990s 0.02 (18637) 0.03 (21333) 0.52 (24547) 0.10 (17555) 0 (10218) 0.09 (36415) 0.02 (25108) 0.09 (2151) 0.07 (19998) 0.12 (175962) 2000s 0.04 (40884) 0.45 (28368) 0.13 (59836) 0.20 (26406) 0.29 (12630) 0.15 (105405) 0.22 (15817) 0.20 (15106) 0.31 (48486) 0.19 (352938) 2010s 0.003 (32662) 0.03 (21472) 0.01 (40068) 0.03 (30247) 0.02 (19585) 0.02 (38627) 0.05 (12689) 0.02 (16672) 0.04 (28240) 0.02 (240262) Total 0.04 (104462) 0.19 (82529) 0.21 (145964) 0.10 (88183) 0.08 (49426) 0.12 (201567) 0.11 (68189) 0.10 (35883) 0.17 (113806) 0.13 (890009) Table 2. Prevalence of bovine tuberculosis infection in cattle in different region in A d i d b TST 376 Arq. Bras. Med. Vet. Zootec., v.71, n.2, p.374-378, 2019 Prevalence of bovine... Arq. Bras. Med. Vet. Zootec., v.71, n.2, p.374-378, 2019 REFERENCES AYELE W.; NEILL S.; ZINSSTAG J. et al. Bovine tuberculosis: an old disease but a new threat to Africa. Int. J. Tuberc. Lung. Dis., v.8, p.924-937, 2004. DEJENE S.; HEITKÖNIG I.; PRINS H. et al. Risk factors for bovine tuberculosis (bTB) in cattle in Ethiopia. PLoS ONE, v.11, p.1-16, 2016. Feeding pattern was another important risk factor for cow in the present survey. The prevalence of 377 Arq. Bras. Med. Vet. Zootec., v.71, n.2, p.374-378, 2019 Liu et al. FRANCK B.; MARIA L.; MARIE F. et al. Zoonotic aspects of Mycobacterium bovis and Mycobacterium avium-intracellulare complex (MAC). Vet. Res., v.36, p.411-436, 2005. KATALE B.; MBUGI E.; KARIMURIBO, E. et al. Prevalence and risk factors for infection of bovine tuberculosis in indigenous cattle in the Serengeti ecosystem, Tanzania. BMC Vet. Res., v.9, p.823-834, 2013. FU Y.; FU G.; WANG S. Investigation and prevention of cow tuberculosis in qinghai province. Chin. J. Vet. Med., v.40, p.31, 2004. LI A.; ZHAO J.; HU D. et al. Epidemiological investigation and prevention of cow tuberculosis in Urumqi. Chin. J. Anim. Quarantine, v.29, p.52-54, 2012. GAN H. Application and promotion of quarantine and detection to bovine tuberculosis in Guangxi. Guangxi Univ., 2008. MA W.; LIU T.; MA J. Investigation and prevention measures of brucellosis and tuberculosis in Chang Ji. Xinjiang Anim. Husbandry, v.8, p.50-51, 2012. GUMI B.; SCHELLING E.; BERG S. et al. Zoonotic transmission of tuberculosis between pastoralists and Their Livestock in South-East Ethiopia. Ecohealth, v.9, p.139-149, 2012. WANG Y.; BAI W.; GUO H. et al. Seroprevalence of brucellosis in sheep in the Aksu Region of Xinjiang Uygur Autonomous Region, People’s Republic of China, between 1990 and 2010. Afr. J. Microbiol. Res., v.6, p.2512-2516, 2012. GUMI B.; SCHELLING E.; FIRDESSA R. et al. Prevalence of bovine tuberculosis in pastoral cattle herds in the Oromia region, southern Ethiopia. Trop. Anim. Health Prod., v.43, p.1081-1087, 2011. XIN D.; AN N. Investigation of cow tuberculosis in bole of Xinjiang. Xinjiang Anim. Husbandry, v.1, p.38-39, 2007. HUMBLET, M.; BOSCHIROLI, M.; SAEGERMAN, C. Classification of worldwide bovine tuberculosis risk factors in cattle: a stratified approach. Vet. Res., v.40, p.50, 2009. KARABAY O.; SERIN E.; TAMER A. et al. Hepatitis B carriage and Brucella seroprevalence in urban and rural areas of Bolu province of Turkey: a prospective epidemiologic study. Turk. J. Gastroenterol., v.15, p.11-13, 2004. 378
https://openalex.org/W2085707036	https://europepmc.org/articles/pmc5200467?pdf=render	English	null	THE REFORM OF THE WEST AFRICAN MEDICAL SERVICE.	Lancet	1,909	cc-by	936	A Journal of Sciences an& hospital Hftmi flDc&tcal Sciences an& Newt Series. No. 125, Vol. V. [No. 1197, Vol. XLVL] Reform in the West Afr The Departmental Committee's report on the organisation of the West African Medical Service has just been issued, and it contains material for careful consideration on the part of medical school authorities and recently qualified medical men. In he first place, and by way of introduction to their formal recommendations, the committee express the opinion, based upon the evidence which they have eceived, that medical service in West Africa de- serves to be considered by junior medical practi- ioners as a promising and permanent career. From certain documents submitted to them, and, we may perhaps add, by common consent amongst hose who are in touch with the subject, the com- mittee feel quite certain that a different view has itherto been taken. In accordance with this pre- alent idea, young medical practitioners have so far been induced to join the service for a limited period of time only, with the implied suggestion that it would be best for them to leave West Africa before t was too late to set up in practice in England; hus, their retirement, with a gratuity of ?1,000 at he end of nine years' service, has been officially, or at least semi-officially, regarded as being in the best interests both of the public service and of the medical officers themselves. The committee dis- sent from this view. They understand that, after nine years' West African service, a medical officer hospital Hftminiatratlon. SATURDAY, JULY 17, 1909. ican Medical Service. ine years, of which some six years only have actu- y , years only lly been spent in West Africa. They consider? nd, general medical opinion will surely agree with hem?that the continued presence of medical men trained in the study of tropical disease, and ully acquainted with local conditions, is of immense mportance, especially for the elucidation of the roblems of preventive medicine in West Africa. hey point out that the lack of continuity in ad- inistration, acknowledged to be the chief defect f tropical African Government, will tend to in- rease if experienced officers are encouraged to etire early. The committee thus base their report upon the view hat medical officers should be encouraged to make he West African service their permanent career. Th - - d tropical hygiene has made such strides that t West Coast is no longer pre-eminently the whi the medical officers, the nature of whose duties quired their residence in the districts where si revalent and man' A Journal of Sciences an& hospital Hftmi romotion is at present very slow; and officers may emain for years at a salary of ?500 a year. The ommittee are convinced that the slowness of pro- otion and meagreness of remuneration which xists throughout the West African medical service, re mainly responsible for the discontent in the staff. hey recommend that the rates of pay should be They pay improved. A .still more important proposal is to the effect that a higher and better-paid rank in the West. African medical service should be instituted. This will afford opportunities for the advancement of officers of merit and enthusiasm, who up till now have been discouraged through the operation of a grotesque official misconception. The dangers to health and life said to be incident to a career in this region of the tropics seem to have been magnified unduly into an excuse for prejudicing the welfare and progress of the community, as well as of its individual medical officers. It may have been true a few years ago that after nine years' service few civil officers of the administration were likely to render valuable service by continued resi- dence : and this may have been especially true of y , is not well suited for general practice in the United Kingdom. Moreover, it seems clear that the majority of medical officers already in this service would much prefer not to begin a new mode of pro- fessional life after eight or nine years under Govern- ment employment, but to make the Colonial Medical Service their career. Further, the committee are strongly of opinion that it is not in the best interests of the West African Colonies and Protectorates that medical men of experience should be removed from the Colonial Service after so short a period as at time, say a score of years the percentage of those who The natural c years' service, offi " s y f . - y e n f years service, longer fever-stricken shadows, but men in the full tide vigour and of a ripe experience in their dutie another consequence is that the number of tho who retire after eighteen years' service to spe many years of healthy life on very comfortab pensions is steadily increasing. These recommend tions will, if carried into effect, both attract a bet class of candidate in the future, and ameliorate t lot of those who are already members of the We African Medical Service. ago, probable percentage survived a stay of eighteen years on " the Coas and lived to enjoy the full pension that could t be earned, was very small. But the new w of life which has come to be the rule instead the exception in West Africa has changed all th Temperance is now the rule instead of the rare ception : sanitation is attended to instead of tota neglected: malaria and other tropical diseases h been shorn of much of their terror: everything, a word, that is comprised within the science
https://openalex.org/W3164086783	http://eprints.uanl.mx/22446/1/22446.pdf	English	null	A global-scale screening of non-native aquatic organisms to identify potentially invasive species under current and future climate conditions	Science of the total environment	2,021	cc-by	20,183	journal homepage: www.elsevier.com/locate/scitotenv journal homepage: www.elsevier.com/locate/scitotenv A global-scale screening of non-native aquatic organisms to identify potentially invasive species under current and future climate conditions Lorenzo Vilizzi a, Gordon H. Copp a,b,c,d, Jeffrey E. Hill e, Boris Adamovich f, Luke Aislabie b, Daniel Akin g, Abbas J. Al-Faisal h, David Almeida i, M.N. Amal Azmai j, Rigers Bakiu k,l, Adriana Bellati m, Renée Bernier n, Jason M. Bies o, Gökçen Bilge p, Paulo Branco q, Thuyet D. Bui r, João Canning-Clode s,t, Henrique Anatole Cardoso Ramos u, Gustavo A. Castellanos-Galindo v,w, Nuno Castro s, Ratcha Chaichana x, Paula Chainho y,z,aa, Joleen Chan ab, Almir M. Cunico ac, Amelia Curd ad, Punyanuch Dangchana ae, Dimitriy Dashinov af, Phil I. Davison b, Mariele P. de Camargo ac, Jennifer A. Dodd ag, Allison L. Durland Donahou e,ah, Lennart Edsman ai, F. Güler Ekmekçi aj, Jessica Elphinstone-Davis ak, Tibor Erős al, Charlotte Evangelista am, Gemma Fenwick an, Árpád Ferincz ao, Teresa Ferreira ap, Eric Feunteun aq, Halit Filiz p, Sandra C. Forneck ac, Helen S. Gajduchenko ar, João Gama Monteiro s, Ignacio Gestoso s,t, Daniela Giannetto as, Allan S. Gilles Jr at, Francesca Gizzi s, Branko Glamuzina au, Luka Glamuzina au, Jesica Goldsmit av,aw, Stephan Gollasch ax, Philippe Goulletquer ay, Joanna Grabowska a, Rogan Harmer b, Phillip J. Haubrock az,ba,bb, Dekui He bc, Jeffrey W. Hean bd,be, Gábor Herczeg bf, Kimberly L. Howland aw, Ali İlhan bg, Elena Interesova bh,bi,bj, Katarína Jakubčinová bk, Anders Jelmert bl, Stein I. Johnsen bm, Tomasz Kakareko bn, Kamalaporn Kanongdate bo, Nurçin Killi p, Jeong-Eun Kim bp, Şerife Gülsün Kırankaya bq, Dominika Kňazovická br, Oldřich Kopecký br, Vasil Kostov bs, Nicholas Koutsikos bt, Sebastian Kozic a, Tatia Kuljanishvili br, Biju Kumar bu, Lohith Kumar bv, Yoshihisa Kurita bw, Irmak Kurtul bg, Lorenzo Lazzaro bb, Laura Lee bx, Maiju Lehtiniemi by, Giovanni Leonardi bz, Rob S.E.W. Leuven ca, Shan Li cb, Tatsiana Lipinskaya cc, Fei Liu cd, Lance Lloyd ce,cf, Massimo Lorenzoni cg, Sergio Alberto Luna ch, Timothy J. Lyons e,ci, Kit Magellan cj,ck, Martin Malmstrøm cl, Agnese Marchini cm, Sean M. Marr bd, Gérard Masson cn, Laurence Masson co, Cynthia H. McKenzie cp, Daniyar Memedemin cq, Roberto Mendoza ch, Dan Minchin cr,cs, Laurence Miossec ay, Seyed Daryoush Moghaddas ct, Moleseng C. Moshobane cu,cv,cw, Levan Mumladze cx, Rahmat Naddaﬁcy, Elnaz Najaﬁ-Majd cz, Aurel Năstase da, Ion Năvodaru da, J. Wesley Neal o, Sarah Nienhuis db, Matura Nimtim x, Emma T. p g ( ) 1 Prof. Weyl passed away suddenly during the ﬁnal stages of this manuscript's preparation. E-mail address: mpiria@agr.hr (M. Piria). ⁎ Corresponding author. Science of the Total Environment 788 (2021) 147868 Science of the Total Environment 788 (2021) 147868 https://doi.org/10.1016/j.scitotenv.2021.147868 0048-9697/© 2021 The Author(s). Published by Elsevier B.V. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/). 016/j.scitotenv.2021.147868 The Author(s). Published by Elsevier B.V. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/ Leonidas Vardakas bt, Gaute Velle ei,ej, Hugo Verreycken ek, Lizaveta Vintsek el, Hui Wei em,en, András Weiperth ao, Olaf L.F. Weyl eo,bd,1, Emily R. Winter dc, Radosław Włodarczyk ep, Louisa E. Wood ak, Ruibin Yang eq, Sercan Yapıcı p, Shayne S.B. Yeo ab, Baran Yoğurtçuoğlu aj, Anna L.E. Yunnie er, Yunjie Zhu es, Grzegorz Zięba a, Kristína Žitňanová bk, Stacey Clarke b Yunnie er, Yunjie Zhu es, Grzegorz Zięba a, Kristína Žitňanová bk, Stacey Clarke b a Department of Ecology and Vertebrate Zoology, Faculty of Biology and Environmental Protection, University of Lodz, 90-237 Lodz, Poland b Centre for Environment, Fisheries and Aquaculture Science, Lowestoft, Suffolk NR33 0HT, UK b Centre for Environment, Fisheries and Aquaculture Science, Lowestoft, Suffolk NR33 0HT, UK f gy, y, y, , d School of the Environment, Trent University, Peterborough, Ontario K9L 0G2, Canada f , y, g , , e Tropical Aquaculture Laboratory, Program in Fisheries and Aquatic Sciences, School of Forest Resources and Conservation, University of Florid f e Tropical Aquaculture Laboratory, Program in Fisheries and Aquatic Sciences, School of Forest Resources a ure Laboratory, Program in Fisheries and Aquatic Sciences, School of Forest Resources and Conservation, University of Florida, Ruskin, FL 33570, USA Belarusian State University, 220030 Minsk, Belarus Faculty of Biology, Belarusian State University, 220030 Minsk, Belarus g College of Science and Mathematics, Auburn University, Auburn, AL 36849, USA g College of Science and Mathematics, Auburn University, Auburn, AL 3 h Marine Science Centre, University of Basrah, PO Box 49, Basrah, Iraq h Marine Science Centre, University of Basrah, PO Box 49, Basrah, Iraq m Department of Ecological and Biological Sciences, University of Tuscia, 01100 Viterbo, Italy n Fisheries and Oceans Canada, Gulf Fisheries Centre, Moncton, New Brunswick E1C 5K4, Canada p Department of Basic Sciences, Faculty of Fisheries, Muğla Sıtkı Koçman University, 48000 Menteşe t Smithsonian Environmental Research Center, Edgewater, MD 21037, US t Smithsonian Environmental Research Center, Edgewater, MD 21037, USA t Smithsonian Environmental Research Center, Edgewater, MD 21037, USA , g , , u Coordination of Sustainable Use of Fisheries Resources, Department of Species Conservation, Ministry of Environment, 70068-900 Brasilia, Brazil g u Coordination of Sustainable Use of Fisheries Resources, Department of Species Conservation, M u Coordination of Sustainable Use of Fisheries Resources, Department of S v Leibniz Centre for Tropical Marine Research (ZMT), 28359 Bremen, Germany v Leibniz Centre for Tropical Marine Research (ZMT), 28359 Bremen, Germany w Smithsonian Tropical Research Institute, Apartado 2072, Balboa, Panam p p x Department of Environmental Technology and Management, Faculty of Environment, Kasetsart University, Bangkok 10900, Thailand x Department of Environmental Technology and Management, Faculty of Environment, Kasetsart Uni y MARE – Marine and Environmental Sciences Centre, Faculty of Sciences, University of Lisbon, 1749-016 Lisboa, Portugal y MARE – Marine and Environmental Sciences Centre, Faculty of Sciences, University of Lisbon, 1749-0 y MARE – Marine and Environmental Sciences Centre, Faculty of Sciences z Department of Animal Biology, Faculty of Sciences, University of Lisbon, 1749-016 Lisboa, Portugal z Department of Animal Biology, Faculty of Sciences, University aa Polytechnic Institute of Setúbal, 2910-761 Setúbal, Portugal aa Polytechnic Institute of Setúbal, 2910-761 Setúbal, Portugal y f g ab Department of Biological Sciences, National University of Singapore, 117558, Singapore y f g ab Department of Biological Sciences, National University of Singapore, 117558, Singapore ab Department of Biological Sciences, National University of Sing p f g y f g p g p ac Laboratory of Ecology, Fisheries and Ichthyology, Biodiversity Department - Palotina Sector, Federal University of Paraná (UFPR), Curitiba 8 p f g y f g p ac Laboratory of Ecology, Fisheries and Ichthyology, Biodiversity Departm ac Laboratory of Ecology, Fisheries and Ichthyology, Biodiversity Department - Palotina Sector, Federal University of Paraná (UFPR), Curitiba 80060-000, Brazil d y f gy y gy y p y f ( ) ad Laboratory of Coastal Benthic Ecology, French Research Institute for Exploitation of the Sea (IFREMER), 29280 Plouzané, France ad Laboratory of Coastal Benthic Ecology, French Research Institute for Exploitation of the Sea (IFREMER), 2 ae Division of Research Policy and Plan, National Research Council of Thailand, Bangkok 10900, Thailand ae Division of Research Policy and Plan, National Research Council of Thailand, Bangkok 10900, Thailand af Department of General and Applied Hydrobiology, Faculty of Biology, Soﬁa University, 1164 g.k. A global-scale screening of non-native aquatic organisms to identify potentially invasive species under current and future climate conditions Nolan dc, Anna Occhipinti-Ambrogi m, Henn Ojaveer dd,de, Sergej Olenin cs, Karin Olsson b,df,dg, Norio Onikura bw, Kathryn O'Shaughnessy dh, Daniele Paganelli di, Paola Parretti s,dj, Jiří Patoka br, Richard Thomas B. Pavia Jr at, Daniele Pellitteri-Rosa cm, Michèle Pelletier-Rousseau av, Elfritzson M. Peralta at, Costas Perdikaris dk, Dariusz Pietraszewski a, Marina Piria a,dl,⁎, Sophie Pitois b, Laura Pompei cg, Nicolas Poulet dm, Cristina Preda cq, Riikka Puntila-Dodd by, Ali T. Qashqaei dn, Tena Radočaj dl, Hossein Rahmani do, Smrithy Raj bu,dp, David Reeves dq, Milica Ristovska dr, Viktor Rizevsky ar, D. Ross Robertson w, Peter Robertson ds, Laura Ruykys dt,du, Abdulwakil O. Saba j,dv, José M. Santos q, Hasan M. Sarı bg, Pedro Segurado q, Vitaliy Semenchenko cc, Wansuk Senanan dw, Nathalie Simard av, Predrag Simonović dx, Michał E. Skóra dy, Kristína Slovák Švolíková bk, Evangelia Smeti bt, Tereza Šmídová br, Ivan Špelić dl, Greta Srėbalienė cs, Gianluca Stasolla dz, Paul Stebbing ak,ea, Barbora Števove bk, Vettath R. Suresh eb, Bettina Szajbert bf, Kieu Anh T. Ta dt, Ali Serhan Tarkan a,p, Jonathan Tempesti ec, Thomas W. Therriault ed, Hannah J. Tidbury ak, Nildeniz Top-Karakuş p, Elena Tricarico bb, Débora F.A. Troca ee, Konstantinos Tsiamis ef, Quenton M. Tuckett e, Pero Tutman eg, Umut Uyan eh, Eliza Uzunova af, Science of the Total Environment 788 (2021) 147868 L. Vilizzi, G.H. Copp, J.E. Hill et al. Leonidas Vardakas bt, Gaute Velle ei,ej, Hugo Verreycken ek, Lizaveta Vintsek el, Hui Wei em,en, András Weiperth ao, Olaf L.F. Weyl eo,bd,1, Emily R. Winter dc, Radosław Włodarczyk ep, Louisa E. Wood ak, Ruibin Yang eq, Sercan Yapıcı p, Shayne S.B. Yeo ab, Baran Yoğurtçuoğlu aj, Anna L.E. Yunnie er, Yunjie Zhu es, Grzegorz Zięba a, Kristína Žitňanová bk, Stacey Clarke b Lozenets, Department of General and Applied Hydrobiology, Faculty of Biology, Soﬁ af Department of General and Applied Hydrobiology, Faculty of Biology, Soﬁa University, 1164 g.k. Leonidas Vardakas bt, Gaute Velle ei,ej, Hugo Verreycken ek, Lizaveta Vintsek el, Hui Wei em,en, András Weiperth ao, Olaf L.F. Weyl eo,bd,1, Emily R. Winter dc, Radosław Włodarczyk ep, Louisa E. Wood ak, Ruibin Yang eq, Sercan Yapıcı p, Shayne S.B. Yeo ab, Baran Yoğurtçuoğlu aj, Anna L.E. Yunnie er, Yunjie Zhu es, Grzegorz Zięba a, Kristína Žitňanová bk, Stacey Clarke b Leonidas Vardakas bt, Gaute Velle ei,ej, Hugo Verreycken ek, Lizaveta Vintsek el, Hui Wei em,en, András Weiperth ao, Olaf L.F. Weyl eo,bd,1, Emily R. Winter dc, Radosław Włodarczyk ep, Louisa E. Wood ak, Ruibin Yang eq, Sercan Yapıcı p, Shayne S.B. Yeo ab, Baran Yoğurtçuoğlu aj, Anna L.E. Yunnie er, Yunjie Zhu es, Grzegorz Zięba a, Kristína Žitňanová bk, Stacey Clarke b Leonidas Vardakas bt, Gaute Velle ei,ej, Hugo Verreycken ek, Lizaveta Vintsek el, Hui Wei em,en, András Weiperth ao, Olaf L.F. Weyl eo,bd,1, Emily R. Winter dc, Radosław Włodarczyk ep, Louisa E. Wood ak, Ruibin Yang eq, Sercan Yapıcı p, Shayne S.B. Yeo ab, Baran Yoğurtçuoğlu aj, Anna L.E. Leonidas Vardakas bt, Gaute Velle ei,ej, Hugo Verreycken ek, Lizaveta Vintsek el, Hui Wei em,en, András Weiperth ao, Olaf L.F. Weyl eo,bd,1, Emily R. Winter dc, Radosław Włodarczyk ep, Louisa E. Wood ak, Ruibin Yang eq, Sercan Yapıcı p, Shayne S.B. Yeo ab, Baran Yoğurtçuoğlu aj, Anna L.E. Yunnie er, Yunjie Zhu es, Grzegorz Zięba a, Kristína Žitňanová bk, Stacey Clarke b Lozenets, Soﬁa, Bulgaria p f pp y gy y f gy ﬁ y g ag Animal and Plant Sciences Group, Edinburgh Napier University, Sighthill, Edinburgh EH11 4BN, UK h ag Animal and Plant Sciences Group, Edinburgh Napier University, Sighthill, Edinburgh EH11 4BN, UK ag Animal and Plant Sciences Group, Edinburgh Napier University, Sighth ah Florida Southern College, Lakeland, FL 33801, USA Florida Southern College, Lakeland, FL 33801, USA ai Department of Aquatic Resources, Institute of Freshwater Research, Swedish University of Agricultural Sciences, SE-750 07 Drottningholm, Sweden ai Department of Aquatic Resources, Institute of Freshwater Research, aj Hydrobiology section, Department of Biology, Faculty of Science, Hacettepe University, Çankaya-Ankara 06800, Turkey k aj Hydrobiology section, Department of Biology, Faculty of Science, Hacettepe University, Çankaya-An ak Centre for Environment, Fisheries and Aquaculture Science, Weymouth, Dorset DT4 8UB, UK l al Centre for Ecological Research, Balaton Limnological Institute, Tihany 8237, Hungary f g g y g y am Centre for Ecological and Evolutionary Synthesis, University of Oslo, NO-0316 Oslo, Norway am Centre for Ecological and Evolutionary Synthesis, University of Oslo, NO-0316 Oslo, Norway an Lancaster Environment Centre, Lancaster University, Lancaster, Lancas y ao Institute for Natural Resources Conservation, Department of Aquaculture, Faculty of Agriculture and Environmental Sciences, Szent István U ao Institute for Natural Resources Conservation, Department of Aquaculture, Faculty of Agricultur ao Institute for Natural Resources Conservation, Department of Aquaculture, Faculty of Agriculture and Environmental Sciences, Szent István University, Gödöllő 2100, Hungary ap D f N l R E i d L d S h l f A i l U i i f Li b 1349 017 Li b P l Institute for Natural Resources Conservation, Department of Aquaculture, Faculty of Agriculture and Environmental Sciences, Szent István Un ap Department of Natural Resources, Environment and Landscape, School of Agriculture, University of Lisbon, 1349-017 Lisbon, Portugal f p f q y f g ap Department of Natural Resources, Environment and Landscape, School of Agriculture, University of Lisbon, 1349-017 Lisbon, Portug ap Department of Natural Resources, Environment and Landscape, School of Agriculture, University of Lisbon ap Department of Natural Resources, Environment and Landscape, School q Muséum National d Histoire Naturelle, Laboratoire Biologie des Organismes Guadeloupe Antilles), Station Marine de Dinard, CRESCO, 35800 Dinard, France , g g Guadeloupe Antilles), Station Marine de Dinard, CRESCO, 35800 Dinard, France ar Laboratory of Ichthyology, Scientiﬁc and Practical Center for Bioresources, National Academy of Sciences of Belarus, Minsk 220072, Belarus ar Laboratory of Ichthyology, Scientiﬁc and Practical Center for Bioresources, National Academy of Sciences o aboratory of Ichthyology, Scientiﬁc and Practical Center for Bioresources, N epartment of Biology, Faculty of Science, Muğla Sıtkı Koçman University, 4 ğ ç y ş ğ y of Science, Research Center for the Natural and Applied Sciences, Graduate School, University of Santo Tomas, Manila, 1008, Metro Manila, Philippines f D b ik 20000 D b ik C ti p f gy y f ğ ç y ş ğ y at Department of Biological Sciences, College of Science, Research Center for the Natural and Applied Sciences, Graduate School, University of San at Department of Biological Sciences, College of Science, Research Center for the Natural and Applied Sciences au Department of Applied Ecology, University of Dubrovnik, 20000 Dubrovnik, Croatia au Department of Applied Ecology, University of Dubrovnik, 20000 Dubrovnik, Croatia Fisheries and Oceans Canada, Maurice Lamontagne Institute, Mont-Joli, Q , g , J , Q , aw Arctic and Aquatic Research Division, Freshwater Institute, Fisheries and Oceans Canada, Winnipeg, Quebec MB R3T 2N6, Canada aw Arctic and Aquatic Research Division, Freshwater Institute, Fisheries and Oceans Canada, Winnipeg esearch Division, Freshwater Institute, Fisheries and Oceans Canada, Winn aw Arctic and Aquatic Research Division, Freshwater Institute, Fisheries ax GoConsult, 22763 Hamburg, Germany ction, French Research Institute for Exploitation of the Sea (IFREMER), 449 az Senckenberg Research Institute and Natural History Museum Frankfurt, Department of River Ecology and Conservation, 63571 Gelnhausen, b az Senckenberg Research Institute and Natural History Museum Frankfurt, Department of River Ecology and b ckenberg Research Institute and Natural History Museum Frankfurt, Depa g y f p f gy y ba Nature and Environment Management Operators s.r.l., 50121 Florence, Italy ba Nature and Environment Management Operators s.r.l., 50121 Florence, Italy bb ba Nature and Environment Management Operators s.r.l., 50121 Florence bb Department of Biology, University of Florence, 50121 Florence, Italy p f gy y f y bc The Key Laboratory of Aquatic Biodiversity and Conservation, Institute of Hydrobiology, Chinese Academy of Sciences, Wuhan 430071, China bc The Key Laboratory of Aquatic Biodiversity and Conservation, Institut bd DST/NRF Research Chair in Inland Fisheries and Freshwater Ecology, South African Institute for Aquatic Biodiversity, Grahamstown 6140, South A bd DST/NRF Research Chair in Inland Fisheries and Freshwater Ecology, S gy f f q y be GroundTruth, Water, Wetlands and Environmental Engineering, Hilton, KwaZulu-Natal 3245, South Africa gy f f q be GroundTruth, Water, Wetlands and Environmental Engineering, Hilton, KwaZulu-Natal 3245, South Africa bf Behavioural Ecology Group, Department of Systematic Zoology and Ecology, ELTE Eötvös Loránd University, H-1117 Budapest, Hungary b bf Behavioural Ecology Group, Department of Systematic Zoology and Ecology, ELTE Eötvös Loránd University bf Behavioural Ecology Group, Department of Systematic Zoology and Ec bg Faculty of Fisheries, Ege University, 35100 Bornov bh Tomsk State University, Tomsk 634050, Russia y bi Institute of Systematics and Ecology of Animals, Siberian Branch of the Russian Academy of Sciences, Novosibirsk 630090, Russia bi Institute of Systematics and Ecology of Animals, Siberian Branch of the Russian Academy of Sciences, Nov bi Institute of Systematics and Ecology of Animals, Siberian Branch of bi Institute of Systematics and Ecology of Animals, S Institute of Systematics and Ecology of Animals, Siberian Branch of the Russian Academy of Sciences, Novosibirsk 630090, Russia bj Novosibirsk branch of Russian Federal Research Institute of Fisheries and Oceanography, Novosibirsk 630090, Russia f y gy f f y f bj Novosibirsk branch of Russian Federal Research Institute of Fisheries and Oceanography, Novosibirsk 630090, Russia f y gy f f y f bj Novosibirsk branch of Russian Federal Research Institute of Fisheries and Oceanography, Novosibirsk 630090, Russia f f g p y bk Department of Ecology, Faculty of Natural Sciences, Comenius University, 841 04 Bratislava, Slovakia bk Department of Ecology, Faculty of Natural Sciences, Comenius University, 841 04 Bratislava, Slovakia bl Institute of Marine Research, Flødevigen Research Station, NO-7485 His, Norway bl Institute of Marine Research, Flødevigen Research Station, NO-7485 His, Norway f , g , , y bm Norwegian Institute for Nature Research, NO-7485 Trondheim, Norway f g y bm Norwegian Institute for Nature Research, NO-7485 Trondheim, Norway m Norwegian Institute for Nature Research, NO-7485 Trondheim, Norway bn Department of Ecology and Biogeography, Faculty of Biological and Veterinary Sciences, Nicolaus Copernicus University, 87-100 Toruń, Poland b bn Department of Ecology and Biogeography, Faculty of Biological and Ve bn Department of Ecology and Biogeography, Faculty of Biological and Veterinary Sciences, Nicolaus Co bo Faculty of Environment and Resource Studies, Mahidol University, Salaya 73170, Thailand bo Faculty of Environment and Resource Studies, Mahidol University, Salaya 73170, Thailand bo Faculty of Environment and Resource Studies, Mahidol University, Sala y f y y bp College of Biological Sciences and Biotechnology, Chungnam National University, Daejeon 305-764, Republic of Korea bp College of Biological Sciences and Biotechnology, Chungnam National University, Daejeon 305-764, bp College of Biological Sciences and Biotechnology, Chungnam National University, Daejeon 305-764, Republic ollege of Biological Sciences and Biotechnology, Chungnam National Univ bq Department of Biology, Faculty of Arts and Science, Düzce University, 81620 Düzce, Turkey bq Department of Biology, Faculty of Arts and Science, Düzce University, 81620 Düzce, Turkey br Department of Zoology and Fisheries, Faculty of Agrobiology, Food and Natural Resources, Czech University of Life Sciences bs f i h i i f A i l i il d h di i i k j d i bs Department of Fisheries, Institute of Animal Science, Ss Cyril and Methodius University, Skopje bs Department of Fisheries, Institute of Animal Science, Ss Cyril and Methodius University, Skopje 10 Department of Fisheries, Institute of Animal Science, Ss Cyril and Method p f f y y pj bt Institute of Marine Biological Resources & Inland Waters, Hellenic Centre for Marine Research, Anavissos, 19013, Attica, Greece bt Institute of Marine Biological Resources & Inland Waters, Hellenic Centre for Marine Research, Anavissos, 19013, 2 Science of the Total Environment 788 (2021) 147868 L. Leonidas Vardakas bt, Gaute Velle ei,ej, Hugo Verreycken ek, Lizaveta Vintsek el, Hui Wei em,en, András Weiperth ao, Olaf L.F. Weyl eo,bd,1, Emily R. Winter dc, Radosław Włodarczyk ep, Louisa E. Wood ak, Ruibin Yang eq, Sercan Yapıcı p, Shayne S.B. Yeo ab, Baran Yoğurtçuoğlu aj, Anna L.E. Yunnie er, Yunjie Zhu es, Grzegorz Zięba a, Kristína Žitňanová bk, Stacey Clarke b Vilizzi, G.H. Copp, J.E. Hill et al. Leonidas Vardakas bt, Gaute Velle ei,ej, Hugo Verreycken ek, Lizaveta Vintsek el, Hui Wei em,en, András Weiperth ao, Olaf L.F. Weyl eo,bd,1, Emily R. Winter dc, Radosław Włodarczyk ep, Louisa E. Wood ak, Ruibin Yang eq, Sercan Yapıcı p, Shayne S.B. Yeo ab, Baran Yoğurtçuoğlu aj, Anna L.E. Yunnie er, Yunjie Zhu es, Grzegorz Zięba a, Kristína Žitňanová bk, Stacey Clarke b John's, Newfoundland a cp Northwest Atlantic Fisheries Centre, Fisheries and Oceans Canada, St. Leonidas Vardakas bt, Gaute Velle ei,ej, Hugo Verreycken ek, Lizaveta Vintsek el, Hui Wei em,en, András Weiperth ao, Olaf L.F. Weyl eo,bd,1, Emily R. Winter dc, Radosław Włodarczyk ep, Louisa E. Wood ak, Ruibin Yang eq, Sercan Yapıcı p, Shayne S.B. Yeo ab, Baran Yoğurtçuoğlu aj, Anna L.E. Yunnie er, Yunjie Zhu es, Grzegorz Zięba a, Kristína Žitňanová bk, Stacey Clarke b bu Department of Aquatic Biology & Fisheries, University of Kerala, Thiruvananthapuram, Kerala 695034, India b bv REF Division, ICAR-Central Inland Fisheries Research Institute, Kolkata, West Bengal 70012 g bw Fishery Research Laboratory, Kyushu University, Fukutsu, Fukuoka 811-3304, Japan y y y J p ment of Evolution, Ecology and Behaviour, Faculty of Health and Life Sciences, University of Liverpool, Liverpool L69 7TX, England, United Kingdom R h C Fi i h E i I i 00790 H l i ki Fi l d y y y y J p bx Department of Evolution, Ecology and Behaviour, Faculty of Health and Life Sciences, University of Liverpool, Liverpool L6 bx Department of Evolution, Ecology and Behaviour, Faculty of Health and Life Sciences, by Marine Research Centre, Finnish Environment Institute, 00790 Helsinki, Finland b by Marine Research Centre, Finnish Environment Institute, 00790 Hels by Marine Research Centre, Finnish Environment Inst bz Hierofalcon Research Group, 95124 Catania, Italy bz Hierofalcon Research Group, 95124 Catania, Italy f p, , y ca Department of Animal Ecology and Physiology, Institute for Water and Wetland Research, Radboud University & Netherlands Centre of Expertise on Exotic Species, 6500 GL Nijmegen, the Netherlands f p, , y ca Department of Animal Ecology and Physiology, Institute for Water and Wetland Research, Radboud University & Netherlands Centre of Expertise on Exotic Species, 6500 GL Nijmegen, the Netherlands ca Department of Animal Ecology and Physiology, Institute for Water and Wetland Research, Radboud University & Netherlands Centre of Ex Netherlands cb Natural History Research Center, Shanghai Natural History Museum, Branch of Shanghai Science & Technology Museum, Shanghai 200041, y g y f g gy g cc Laboratory of Hydrobiology, Scientiﬁc and Practical Center for Bioresources, National Academy of Sciences of Belarus, Minsk 220072, Belarus cd Aquatic Science Institute, Tibet Academy of Agriculture and Anim cd Aquatic Science Institute, Tibet Academy of Agriculture and Animal Husbandry Science, Lhasa 850009, China cd Aquatic Science Institute, Tibet Academy of Agriculture and Animal Husbandry Science, Lhasa ce Lloyd Environmental Pty Ltd, Somers, Victoria 3927, Australia ce Lloyd Environmental Pty Ltd, Somers, Victoria 3927, Australia y y cf School of Health and Life Sciences, Federation University Australia, Ballarat, Victoria 3350, Australia cf School of Health and Life Sciences, Federation University Australia, Ballarat, Victoria 3350, Australia cg Department of Chemistry, Biology and Biotechnologies, University of Perugia, 06123 Perugia, Italy p f y gy g y f g g y ch Facultad de Ciencias Biológicas, Universidad Autónoma de Nuevo León, Nuevo León 66455, Mexico ch Facultad de Ciencias Biológicas, Universidad Autónoma de Nuevo León, Nuevo León 66455, Mexico ci New Mexico Biopark Society, Albuquerque, NM 87102, USA ci New Mexico Biopark Society, Albuquerque, NM 87102, USA cj South African Institute for Aquatic Biodiversity, Grahamstown 6140, South Africa cj South African Institute for Aquatic Biodiversity, Grahamstow ck University of Battambang, 02360 Battambang, Cambodia ck University of Battambang, 02360 Battambang, Cambodia cl Norwegian Scientiﬁc Committee for Food and Environme g ﬁ f ( ) y cm Department of Earth and Environmental Sciences, University of Pavia, 27100 Pavia, Italy g ﬁ f ( ) y cm Department of Earth and Environmental Sciences, University of Pavia, 27100 Pavia, Italy p f , y f , , y cn Laboratoire interdisciplinaire des environnements continentaux, Centre national de la recherche scientiﬁque, Université de Lorraine, 57 cn Laboratoire interdisciplinaire des environnements continentaux, Centre national de la reche gy y y g g y cp Northwest Atlantic Fisheries Centre, Fisheries and Oceans Canada, St. Science of the Total Environment 788 (2021) 147868 1. Introduction hence posing a threat to native species, usually possess life-history traits including frequent reproduction with a high incidence of ovoviviparity, large body size and a long life span, the ability to exploit opportunisti- cally available food resources, a history of invasion success, and a close climate matching with the RA area (Statzner et al., 2008; Chan et al., 2021). In addition, these species often tolerate higher salinity, broader environmental temperatures and higher levels of organic pollution than native species (Leuven et al., 2009). Importantly, the identiﬁcation of species posing a high risk of being (or becoming) invasive in a certain RA area provides a basis for advice to policy, decision-makers and other stakeholders regarding management options for existing and potential future invasive NNS (Copp et al., 2005b, 2016b; Mumford et al., 2010; David et al., 2013; David and Gollasch, 2018, 2019). The threat posed by invasive non-native species (NNS) worldwide to native species and ecosystems requires a global approach to identify species that pose a high risk of becoming invasive across varied geogra- phies and climates (Gordon et al., 2008). Use of the same risk screening protocols worldwide can provide the larger-scale information with which to inform the decisions of policy makers and environmental man- agers in dealing with species invasions (Early et al., 2016; Shackleton et al., 2019). Indeed, reliable, evidence-based risk assessment (RA) methods are vital to decision-making in biosecurity management at na- tional and international levels in order to prevent or mitigate unwanted biological invasions (Kaiser and Burnett, 2010). Ideally, the RA out- comes should compare and prioritise NNS according to their potential invasiveness under current and future climate conditions (Barney and DiTomaso, 2010), which aids in identifying the underlying factors asso- ciated with invasion hotspots (O'Donnell et al., 2012; Chapman et al., 2020). The objectives of the present study were to: (i) construct a global da- tabase of risk screenings that span the broadest range of aquatic organ- isms possible, given available resources, across the widest possible geographical spread; (ii) subject the global database of screenings to calibration and accuracy analysis; and (iii) generate global-scale risk thresholds at the organism group and climate class/marine ecoregion levels under both current and future climate conditions. Leonidas Vardakas bt, Gaute Velle ei,ej, Hugo Verreycken ek, Lizaveta Vintsek el, Hui Wei em,en, András Weiperth ao, Olaf L.F. Weyl eo,bd,1, Emily R. Winter dc, Radosław Włodarczyk ep, Louisa E. Wood ak, Ruibin Yang eq, Sercan Yapıcı p, Shayne S.B. Yeo ab, Baran Yoğurtçuoğlu aj, Anna L.E. Yunnie er, Yunjie Zhu es, Grzegorz Zięba a, Kristína Žitňanová bk, Stacey Clarke b Vilizzi, G.H. Copp, J.E. Hill et al. G R A P H I C A L A B S T R A C T G R A P H I C A L A B S T R A C T a b s t r a c t Article history: Received 13 April 2021 Received in revised form 13 May 2021 Accepted 14 May 2021 Available online 20 May 2021 Editor: Damia Barcelo Keywords: Decision support tools AS-ISK Hazard identiﬁcation Non-native species Risk analysis Climate change Article history: Received 13 April 2021 Received in revised form 13 May 2021 Accepted 14 May 2021 Available online 20 May 2021 The threat posed by invasive non-native species worldwide requires a global approach to identify which intro- duced species are likely to pose an elevated risk of impact to native species and ecosystems. To inform policy, stakeholders and management decisions on global threats to aquatic ecosystems, 195 assessors representing 120 risk assessment areas across all six inhabited continents screened 819 non-native species from 15 groups of aquatic organisms (freshwater, brackish, marine plants and animals) using the Aquatic Species Invasiveness Screening Kit. This multi-lingual decision-support tool for the risk screening of aquatic organisms provides assessors with risk scores for a species under current and future climate change conditions that, following a statistically based calibration, permits the accurate classiﬁcation of species into high-, medium- and low-risk cat- egories under current and predicted climate conditions. The 1730 screenings undertaken encompassed wide geographical areas (regions, political entities, parts thereof, water bodies, river basins, lake drainage basins, and marine regions), which permitted thresholds to be identiﬁed for almost all aquatic organismal groups screened as well as for tropical, temperate and continental climate classes, and for tropical and temperate marine ecoregions. In total, 33 species were identiﬁed as posing a ‘very high risk’ of being or becoming invasive, and the scores of several of these species under current climate increased under future climate conditions, primarily due to their wide thermal tolerances. The risk thresholds determined for taxonomic groups and climate zones provide a basis against which area-speciﬁc or climate-based calibrated thresholds may be interpreted. In turn, the risk rankings help decision-makers identify which species require an immediate ‘rapid’ management action (e.g. eradication, control) to avoid or mitigate adverse impacts, which require a full risk assessment, and which are to be restricted or banned with regard to importation and/or sale as ornamental or aquarium/ﬁshery enhancement. Keywords: Decision support tools AS-ISK Hazard identiﬁcation Non-native species Risk analysis Climate change © 2021 The Author(s). Published by Elsevier B.V. This is an open access article under the CC BY license (http:// creativecommons.org/licenses/by/4.0/). H I G H L I G H T S a b s t r a c t • A global approach is required to identify invasive species posing high risk impact. • A global approach is required to identify invasive species posing high risk impact. p p g g p • 195 assessors screened 819 non-native species from 15 groups of aquatic or- ganisms. • Risk thresholds were identiﬁed for 14 aquatic organism groups. • The resulting risk thresholds and rankings will help management and conservation. Leonidas Vardakas bt, Gaute Velle ei,ej, Hugo Verreycken ek, Lizaveta Vintsek el, Hui Wei em,en, András Weiperth ao, Olaf L.F. Weyl eo,bd,1, Emily R. Winter dc, Radosław Włodarczyk ep, Louisa E. Wood ak, Ruibin Yang eq, Sercan Yapıcı p, Shayne S.B. Yeo ab, Baran Yoğurtçuoğlu aj, Anna L.E. Yunnie er, Yunjie Zhu es, Grzegorz Zięba a, Kristína Žitňanová bk, Stacey Clarke b John's, Newfoundland and Labrador A1A 5J7, Canada cq Faculty of Natural and Agricultural Sciences, Ovidius University of Constanta, Constanta 900527, Romania cq Faculty of Natural and Agricultural Sciences, Ovidius University of Constanta, Constanta 900527, R Marine Organism Investigations, Marina Village, Ballina, Killaloe, Clare V9 cr Marine Organism Investigations, Marina Village, Ballina, Killaloe, Clare cs Marine Research Institute, Klaipėda University, 92294 Klaipėda, Lithuania cs Marine Research Institute, Klaipėda University, 92294 Klaipėda, Lithuania , p y, p , ct Department of Biodiversity and Ecosystems Management, Environmental Sciences Research Institute, Shahid Beheshti University, 1983963113 Tehran, Iran p y p of Biodiversity and Ecosystems Management, Environmental Sciences Research Institute, Shahid Beheshti University, 1983963113 Tehran, Iran ct Department of Biodiversity and Ecosystems Management, Environmental Sciences Researc ct Department of Biodiversity and Ecosystems Management, Environmental Sciences Research Institute, Shahid Beheshti University, 198396311 cu South African National Biodiversity Institute, Biological Invasions Direc f y g cv Department of Biology, Sefako Makgatho Health Sciences University, Gauteng 0208, South Africa cv Department of Biology, Sefako Makgatho Health Sciences University, Gauteng 0208, South Africa cv Department of Biology, Sefako Makgatho Health Sciences University cw Young Water Professionals, South African Chapter, Limpopo 1685, South Africa cw Young Water Professionals, South African Chapter, Limpopo 1685, South Africa cx Institute of Zoology, Ilia State University, Tbilisi 0162, Georgia f gy y g cy Swedish University of Agricultural Sciences, Department of Aquatic Resources, Division of Coastal Research, SE-453 30 Oregrund, Sweden cy Swedish University of Agricultural Sciences, Department of Aquatic Resources, Division of cy Swedish University of Agricultural Sciences, Department of Aqu cz Department of Biology, Faculty of Sciences, Ege University, 35040 Izmir, Turkey cz Department of Biology, Faculty of Sciences, Ege University, 35040 Izmir, Turkey cz Department of Biology, Faculty of Sciences, Ege University, 35040 Izmir, Turkey a Department of Biodiversity Conservation and Sustainable Use of Natural db Ontario Ministry of Natural Resources and Forestry, Peterborough, Ontario K9J 8M5, Canada db Ontario Ministry of Natural Resources and Forestry, Peterborough, Ontario K9J 8M5, Canada dc Department of Life and Environmental Sciences, Faculty of Science a dc Department of Life and Environmental Sciences, Faculty of Science and Technology, Bournemouth University, Talbot Campus, Poole BH12 5BB, UK dd U i i f T 80012 P E i dd University of Tartu, 80012 Pärnu, Estonia de National Institute of Aquatic Resources, Technical University of Denmark, 2800 Lyngby, Denmark de National Institute of Aquatic Resources, Technical University of Den de National Institute of Aquatic Resources, Technical University of Denmark, 2800 Lyngby, Denmark Aquatic Resources, Technical University of Denmark, 2800 Lyngby, Denma df School of Zoology, Tel Aviv University, Tel Aviv 6997801, Israel df School of Zoology, Tel Aviv University, Tel Aviv 6997801, Israel dg The Inter-University Institute for Marine Sciences in Eilat, Coral Beach, Eilat 8810302, Israel dg The Inter-University Institute for Marine Sciences in Eilat, Coral Beach, Eilat 8810302, Israel dg The Inter-University Institute for Marine Sciences in Eilat, Coral B y f dh Texas Parks and Wildlife Department, Coastal Fisheries, 4200 Smith School Rd., Austin, TX 78744, USA dh Texas Parks and Wildlife Department, Coastal Fisheries, 4200 Smith School Rd., Austin, TX 78744, USA dh Texas Parks and Wildlife Department, Coastal Fisheries, 4200 Smit di CNR-IRSA Water Research Institute, 28922 Verbania, Italy Research Institute, 28922 Verbania, Italy enter in Biodiversity and Genetic Resources, InBIO Associate Laboratory and Faculty of Sciences and Technologies, University of the Azores, 9500-321 , , y dj CIBIO, Research Center in Biodiversity and Genetic Resources, InBIO Associate Laboratory and Faculty of Sciences and Technologies, University dk dj CIBIO, Research Center in Biodiversity and Genetic Resources, InBIO Associate Laboratory and F dj CIBIO, Research Center in Biodiversity and Genetic Resourc dk Department of Fisheries, Regional Unit of Thesprotia, Epirus, 46 100, Igoumenitsa, Greece dk Department of Fisheries, Regional Unit of Thesprotia, Epirus, 46 100, Igoumenitsa, Greece p f , g f p , p , , g , dl Department of Fisheries, Apiculture, Wildlife Management and Special Zoology, University of Zagreb Faculty of Agriculture, 10000 Zagreb, Cr p f g f p p g dl Department of Fisheries, Apiculture, Wildlife Management and Special dl Department of Fisheries, Apiculture, Wildlife Management and Special Zoology, University of Zagreb Faculty of Agriculture, 10000 Zagreb, Croatia d p f p f g p gy y dm Pôle écohydraulique OFB-IMFT-P, French Agency for Biodiversity, 31400 Toulouse, France dm Pôle écohydraulique OFB-IMFT-P, French Agency for Biodiversity, 31400 Toulouse, France dm Pôle écohydraulique OFB-IMFT-P, French Agency for Biodiversity, 3 dn Sahel Square, Tehran 1493889881, Iran Sahel Square, Tehran 1493889881, Iran do Sari Agricultural Sciences and Natural Resources University, Sari, 4816118771, Mazandaran, Iran q , , do Sari Agricultural Sciences and Natural Resources University, Sari, 4816118771, Mazandaran, Iran do Sari Agricultural Sciences and Natural Resources University, Sari, 4816118771, Mazandaran, Iran do Sari Agricultural Sciences and Natural Resources University, Sari, 48 nces and Natural Resources University, Sari, 4816118771, Mazandaran, Ir dp National Centre for Biological Sciences, Bangalore 560065, India dp National Centre for Biological Sciences, Bangalore 560065, India dq National Fish and Wildlife Foundation, Baton Rouge, LA 70808, USA f g dr Institute of Biology, Faculty of Natural Sciences and Mathematics, Ss Cyril and Methodius University, 1000 Skopje, Macedonia dr Institute of Biology, Faculty of Natural Sciences and Mathematics, Ss C dr Institute of Biology, Faculty of Natural Sciences and Mathematics, Ss Cyril and Methodius University, 1000 Skopje, Macedonia dr Institute of Biology, Faculty of Natural Sciences and Mathematics, Ss Cyril and Methodius University, 1000 Skopje, Macedonia f gy y f y y pj ds Modelling, Evidence and Policy Group, School of Natural and Environmental Resources, Newcastle University, Newcastle NE1 7RU, UK ds Modelling, Evidence and Policy Group, School of Natural and Environm g y p f y dt Nature and Biodiversity Conservation Agency, Vietnam Environment Administration, Ministry of Natural Resources and Environment, 10 Ton That Thuyet, Nam Tu Liem District, Hanoi, Viet Nam du Flora and Fauna Division, Department of Environment and Natural Resources, Palmerston, Northern Territory 0828, Australia d y g y, , y f , du Flora and Fauna Division, Department of Environment and Natural Resources, Palmerston, Northern Territory 0828, Australia dv School of Agriculture, Lagos State University, Epe Campus, 106101 Epe, Lagos State, Nigeria dv School of Agriculture, Lagos State University, Epe Campus, 106101 Epe, Lagos State, Nigeria dw Department of Aquatic Science, Faculty of Science, Burapha University p f q y f p y dx Faculty of Biology & Institute for Biological Research “Siniša Stanković”, University of Belgrade, Belgrade 11000, Serbia dx Faculty of Biology & Institute for Biological Research “Siniša Stanković”, University of Belgrade, Bel dx Faculty of Biology & Institute for Biological Research “Siniša Stanković gy f g y f g g f Gdańsk, Faculty of Oceanography and Geography, Institute of Oceanography, Professor Krzysztof Skóra Hel Marine Station, 84-150 Hel, Poland s 70029 Santeramo in colle Italy dy University of Gdańsk, Faculty of Oceanography and Geography, Institute of Oceanography, Professor Krzysztof Skóra Hel Marine S dy University of Gdańsk, Faculty of Oceanography and Geography, Institute of Oceanography, Professo dy University of Gdańsk, Faculty of Oceanography and Geography, Inst dz Dadoda S.r.l.s, 70029 Santeramo in colle, Italy ea APEM Ltd, A17 Embankment, Business Park, Heaton Mersey, Manchester, Cheshire SK4 3GN, UK ea APEM Ltd, A17 Embankment, Business Park, Heaton Mersey, Manchester, Cheshire SK4 3GN, UK b ea APEM Ltd, A17 Embankment, Business Park, Heaton Mersey, Manch eb Division of Mariculture, Central Marine Fisheries Research Institute, Cochin, Kerala 682018, India eb Division of Mariculture, Central Marine Fisheries Research Institute, Cochin, Kerala 682018, India eb Division of Mariculture, Central Marine Fisheries Research Institute, C ec Department of Biology, University of Pisa, 56126 Pisa, Italy p f gy y f y ed Paciﬁc Biological Station, Fisheries and Oceans Canada, Nanaimo, British Columbia V9T 6N7, Canada ed Paciﬁc Biological Station, Fisheries and Oceans Canada, Nanaimo, British Columbia V9T 6N7, Cana ed Paciﬁc Biological Station, Fisheries and Oceans Canada, Nan ee Institute of Oceanography, Federal University of Rio Grande, 96203-900 Rio Grande, Brazil ee Institute of Oceanography, Federal University of Rio Grande, 96203-900 Rio Grande, Brazil ef Institute of Oceanography, Hellenic Centre for Marine Research, Attica, f g p y f y eg Laboratory for Ichthyology and Coastal Fisheries, Institute of Oceanography and Fisheries, 21000 Split, Croatia h hyology and Coastal Fisheries, Institute of Oceanography and Fisheries, 21 eg Laboratory for Ichthyology and Coastal Fisheries, Institute of Oceanography and Fisheries, 21000 S eg Laboratory for Ichthyology and Coastal Fisheries, Institute of Oce eh Skretting Turkey, Güllük Milas, 48670, Muğla, Turkey ei Norwegian Research Centre, 5007 Bergen, Norway ej Department of Biological Sciences, University of Bergen, 5007 Bergen, Norway k ej Department of Biological Sciences, University of Berg ej Department of Biological Sciences, University of Bergen, 5007 Bergen, Norway ek Research Institute for Nature and Forest (INBO), B-1630 Linkebeek, Belgium ek Research Institute for Nature and Forest (INBO), B-1630 Linkebeek, Be f ( ) g el Institute of Botany, Faculty of Biology, Jagiellonian University, 30-387 Kraków, Poland f ( ) g el Institute of Botany, Faculty of Biology, Jagiellonian University, 30-387 Kraków, Poland f y y f gy J g y em Pearl River Fisheries Research Institute, Chinese Academy of Fishery Science, Guangzhou 510380, China em Pearl River Fisheries Research Institute, Chinese Academy of Fishery Science, Guangzhou 510380, China y f y g en Key Laboratory of Recreational ﬁsheries Research, Ministry of Agriculture and Rural Affai en Key Laboratory of Recreational ﬁsheries Research, Ministry of Agriculture and Rural Affairs, Guangzhou 510380, China y y f ﬁ y f g ff g eo Centre for Invasion Biology, South African Institute for Aquatic Biodiversity, Makhanda 6139, South Africa eo Centre for Invasion Biology, South African Institute for Aquatic Biodiversity, Makhanda 6139, South Africa f gy f f q y f partment of Biodiversity Studies and Bioeducation, Faculty of Biology and Environmental Protection, University of Lodz, 90-237 Lodz, Poland ll f i h i h i l l i i h hi ep Department of Biodiversity Studies and Bioeducation, Faculty of Biology and Environmental Protection, University of Lodz, 90- ep Department of Biodiversity Studies and Bioeducation, Faculty of Biology and Environmenta Department of Biodiversity Studies and Bioeducation, Faculty of Biology and Environmental Protection, University of Lodz, 90 237 Lodz, Poland eq College of Fisheries, Huazhong Agricultural University, Wuhan 430070, China eq College of Fisheries, Huazhong Agricultural University, Wuhan 430070, China eq College of Fisheries, Huazhong Agricultural University, W q College of Fisheries, Huazhong Agricultural University, Wuhan 430070, C er PML Applications Ltd, Plymouth, Devon PL1 3DH, UK er PML Applications Ltd, Plymouth, Devon PL1 3DH, UK es Aquaculture Technology Promotion Station of Nantong, Nantong, China es Aquaculture Technology Promotion Station of Nantong, Nantong, China 3 Science of the Total Environment 788 (2021) 147868 L. a r t i c l e i n f o a b s t r a c t © 2021 The Author(s). Published by Elsevier B.V. This is an open access article under the CC BY license (http:// creativecommons.org/licenses/by/4.0/). 1. Introduction The global- scale thresholds identiﬁed will provide a basis against which thresholds calibrated for speciﬁc RA areas may be interpreted and will also allow the ‘rapid risk screening’ of individual species for a certain RA area The ﬁrst step in the NNS risk analysis process is risk screening (i.e. hazard identiﬁcation), which aims to identify the NNS that are likely to pose an elevated threat to native species and ecosystems and, there- fore, warrant more comprehensive (i.e. full) RA (Copp et al., 2005b). Aquatic species that are likely to carry a high risk of becoming invasive, 4 Science of the Total Environment 788 (2021) 147868 L. Vilizzi, G.H. Copp, J.E. Hill et al. Change Assessment (CCA), which require the assessor to evaluate how future predicted climate conditions are likely to affect the BRA with re- spect to risks of introduction, establishment, dispersal and impact. To achieve a valid screening, for each question the assessor must provide a response, a level of conﬁdence in the response, and a justiﬁcation. In all cases, the assessor is a specialist in the biology/ecology of the aquatic organism under screening for the RA area under study. Upon comple- tion of the screening, the species receives both a BRA score and a BRA + CCA (composite) score (ranging from −20 to 68 and from −32 to 80, respectively). Scores < 1 suggest that the species is unlikely to become invasive and is therefore classiﬁed as ‘low risk’ (Pheloung et al., 1999). Higher scores classify the species as posing either a ‘me- dium risk’ or a ‘high risk’ of becoming invasive. Distinction between medium-risk and high-risk levels depends upon setting a ‘threshold’ value (see Section 2.2 Data processing and analysis). whenever speciﬁc calibration is not possible. In addition, these global- scale thresholds will place RA area-speciﬁc risk screening calibrations within a broader, global context, also accounting for climate change pre- dictions. In turn, this will enhance the value and scope of more localised calibrations to inform environmental policy and decision-makers of the relative risk rankings of aquatic NNS so as to facilitate the cost-effective allocation of management resources. 2.1. Risk screening procedure For the purposes of this study, ‘invasive species’ are deﬁned, as per Copp et al. (2005c, p. 244), as those NNS “that spread, with or without the aid of humans, in natural or semi-natural habitats, producing a sig- niﬁcant change in composition, structure, or ecosystem processes, or cause severe economic losses to human activities”. Species were evalu- ated for their potential to become invasive in the assessor(s)-deﬁned RA area using the Aquatic Species Invasiveness Screening Kit (AS-ISK: free download at www.cefas.co.uk/nns/tools/). This is a decision-support tool (Copp et al., 2016b, 2021) adapted from the Fish Invasiveness Screening Kit (Copp et al., 2009, 2005a), which itself was derived from the globally-applied Weed Risk Assessment of Pheloung et al. (1999). The AS-ISK comprises questions from the generic screening module of the European Non-native Species in Aquaculture Risk Analysis Scheme (Copp et al., 2016a) and incorporates the ‘minimum requirements’ (Roy et al., 2018) for the assessment of invasive NNS with regard to the 2014 EU Regulation 1143/2014 (European Union, 2014). As a taxon-generic toolkit, the AS-ISK is applicable to any aquatic species (other than parasites and pathogens) in virtually any climatic/marine ecoregion zone (Copp et al., 2016b; Table 1), and allows the screening of 27 groups of aquatic organisms in total (taxonomy after Ruggiero et al., 2015): mammals, birds, reptiles, amphibians, ﬁshes (freshwater, brackish, marine), tunicates, lancelets, invertebrates (freshwater, brack- ish, marine), ‘other’ animals (freshwater, brackish, marine), plants (freshwater, brackish, marine), protists (freshwater, brackish, marine), fungi (freshwater, brackish, marine), and bacteria (freshwater, brackish, marine). The ranked levels of conﬁdence (1 = low; 2 = medium; 3 = high; 4 = very high) associated with each question-related response mirror the conﬁdence rankings recommended by the International Programme on Climate Change (IPCC, 2005; see also Copp et al., 2016b). Based on the conﬁdence level (CL) allocated to each response, a conﬁdence factor (CF) is computed as: CF ¼ ∑CLQi = 4 55 ð Þ i ¼ 1, . . . , 55 ð Þ where CLQi is the conﬁdence level for the ith Question (Qi), 4 is the max- imum achievable value for conﬁdence (i.e. very high: see above) and 55 is the total number of questions. Based on the 49 Qs comprising the BRA and the six Qs comprising the CCA, the CLBRA and CLCCA are also com- puted (out of the CLTotal for all 55 Qs). 2.2. Data processing and analysis Data consisted of: (i) individual contributions to the present study by assessors invited to screen one (or more) NNS belonging to one or more aquatic organismal groups of choice (i.e. falling within their ex- pertise) for a certain RA area; and (ii) datasets from more comprehen- sive screening studies of NNS for a certain RA area, both published (see Table 1) and unpublished. For each species screened, the scien- tiﬁc name used in the original contribution or study was updated to the most recent taxonomy after the World Register of Marine Species (www.marinespecies.org), else after the Integrated Taxonomic Information System (www.itis.gov/) or FishBase (www.ﬁshbase.org). This was followed by ‘cross-checking’ for the existence of at least one peer-reviewed publication that used the updated scientiﬁc name in case of a change in taxonomy. A notable exception was the retention of the original name Crassostrea gigas instead of the recently proposed Magallana gigas for the Paciﬁc oyster (see Bayne et al., 2017). Taxonomic details of the corresponding Order and Family were also re- trieved for each species screened. The screening protocol consists of 55 questions (Copp et al., 2016b). The ﬁrst 49 questions comprise the Basic Risk Assessment (BRA), which are concerned with the biogeographical and biological aspects of the species being screened. The last six questions address the Climate Table 1 Published initial applications of the Aquatic Species Invasiveness Screening Kit (AS-ISK) by aquatic organismal group. Aquatic organismal group Reference(s) Reptiles Ruykys et al. (2021) Amphibians Ruykys et al. (2021) Freshwater ﬁshes Glamuzina et al. (2017), Li et al. (2017), Tarkan et al. (2017a), Tarkan et al. (2017b), Dodd et al. (2019), Suresh et al. (2019), Interesova et al. (2020), Moghaddas et al. (2020), Zięba et al. (2020), Glamuzina et al. (2021), Haubrock et al. (2021), Kumar et al. (2021), Moghaddas et al. (2021), Radočaj et al. (2021), Ruykys et al. (2021), Wei et al. (2021) Brackish ﬁshes Castellanos-Galindo et al. (2018), Clarke et al. (2020) Marine ﬁshes Filiz et al. (2017a), Filiz et al. (2017b), Bilge et al. (2019), Clarke et al. (2020), Lyons et al. (2020), Uyan et al. (2020) Tunicates Clarke et al. (2020) Freshwater invertebrates Paganelli et al. (2018), Ruykys et al. (2021), Semenchenko et al. (2018) Brackish invertebrates Clarke et al. (2020), Ruykys et al. (2021) Marine invertebrates Clarke et al. (2020), Killi et al. (2020), Stasolla et al. (2020), Ruykys et al. 2.2. Data processing and analysis (2021) Freshwater plants Ruykys et al. (2021) Marine plants Clarke et al. (2020), Ruykys et al. (2021) Marine protists Clarke et al. (2020) Table 1 Published initial applications of the Aquatic Species Invasiveness Screening Kit (AS-ISK) by aquatic organismal group. Table 1 conﬁdence intervals of speciﬁcities, which were computed along the en- tire range of sensitivity points (i.e. 0 to 1, at 0.1 intervals). 1 – speciﬁcity for each threshold value, where in the present context sensitivity and speciﬁcity will be the proportion of a priori invasive and non-invasive species, respectively, correctly identiﬁed as such. For ROC curve analysis to be implemented, the species selected for screen- ing must be categorised a priori as non-invasive or invasive using inde- pendent literature sources. 1 – speciﬁcity for each threshold value, where in the present context sensitivity and speciﬁcity will be the proportion of a priori invasive and non-invasive species, respectively, correctly identiﬁed as such. For ROC curve analysis to be implemented, the species selected for screen- ing must be categorised a priori as non-invasive or invasive using inde- pendent literature sources. Following Smith et al. (1999), three measures of accuracy were deﬁned: 1) for a priori invasive species: Ai = (Ir/It) × 100, where Ir is the number of a priori invasive species rejected, and It the total number of a priori invasive species screened; The a priori categorisation was as follows (see also Clarke et al., 2020): (i) a ﬁrst search was made of FishBase for any reference to the species' threat, with the species categorised as non-invasive if listed as ‘harmless’, categorised as invasive if listed as ‘potential pest’, or scored as absent if either not evaluated or not listed in the above database; (ii) a second search was made of the Centre for Agriculture and Biosci- ence International Invasive Species Compendium (CABI ISC: www. cabi.org/ISC) and the Global Invasive Species Database (GISD: www. iucngisd.org), with the species categorised as invasive if it appeared in any of such lists or scored as absent if not listed; (iii) a third search was made of the Invasive and Exotic Species of North America list (www.invasive.org), with the species categorised as invasive if it ap- peared in any of such lists or scored as absent if not listed; (iv) except for those species categorised as invasive in any (or all) of the previous three steps, a Google Scholar (literature) search was performed to check whether at least one peer-reviewed reference is found that ‘dem- onstrates’ (hence, not ‘assumes’) invasiveness/impact. 3. Results Their distribution was not nor- mal (JB = 16.378, P < 0.001), not skewed (skewness = −0.039, z = −0.675, P = 0.499), but platykurtic (kurtosis = 2.531, z = −5.210, P < 0.001) (Fig. 3b). The majority of delta values (i.e. differences be- tween BRA + CCA and BRA scores, hence accounting for climate change predictions) were equal to 0, 4, 6 and 10 (>10% of the total in all cases), and overall the proportion of the positive differences was much larger than that of the negative differences (68.8% vs 16.7%) (Fig. 3c). Across all species, the mean CL values were: CLTotal = 2.73 ± 0.01, CLBRA = 2.78 ± 0.01, and CLCCA = 2.25 ± 0.02, indicating in all cases medium to high conﬁdence (Supplementary data Table S5). Because of sample size constraints (see Vilizzi et al., 2019), group- speciﬁc thresholds for both the BRA and BRA + CCA were ﬁtted to those groups of aquatic organisms for which >10 species were screened. Consequently, for mammals and birds for which there were low numbers of taxa a combined threshold was computed by pooling together the screened species for these groups with those screened for reptiles and amphibians. This rendered the respective thresholds statistically signiﬁcant, permitting their use for distinguishing between high-risk and low-to-medium risk species until such time that RA- area-speciﬁc calibrations can be undertaken for those taxonomic groups. To highlight ‘very high risk’ species for the aquatic organismal groups with large enough sample sizes, ad hoc thresholds for the BRA and BRA + CCA were set weighted according to the range of scores for the high-risk species (see Clarke et al., 2020) and with the constraint that the species was screened for a ‘representative’ number of RA areas (i.e. weighted according to the corresponding organismal group). Additionally, climate class-speciﬁc thresholds were computed for fresh- water ﬁshes, and marine ecoregion-speciﬁc thresholds for marine ﬁshes and invertebrates – the aquatic organismal groups with large enough sample size for successful computation of such thresholds. In all cases, ROC curve analysis was carried out with the package 'pROC' (Robin et al., 2011) for R x64 v3.6.3 using 2000 bootstrap replicates for the Thresholds were computed for all screened groups of aquatic organ- isms except those represented by ≤10 species (Table 2). 3. Results In total, 1730 screenings were conducted by 195 assessors (the co- authors of this study) on 819 taxa comprising 798 species, nine sub- species, three hybrids, and nine genera (Supplementary data Table S1). Of these taxa (hereafter, loosely termed ‘species’), 562 (68.6%) were categorised a priori as non-invasive and 257 (31.4%) as invasive (Supplementary data Table S1) and were screened relative to 120 RA areas (Supplementary data Tables S2 and S3) across all six inhabited con- tinents (Fig. 1). The RA areas consisted of extensive geographical areas, regions, countries, parts of countries, states, other political entities, water bodies, river basins, lake drainage basins, and marine regions (Supplementary data Table S2). Screenings encompassed 15 groups of aquatic organisms (Fig. 2) in 104 Orders (Supplementary data Table S2), with 24 species assigned to two different groups depending on the RA area's aquatic habitat (Supplementary data Table S4). A measure of the accuracy of the calibration analysis is the Area Under the Curve (AUC), which ranges from 0 to 1: a model whose pre- dictions are 100% correct has an AUC of 1, one whose predictions are 100% wrong has an AUC of 0. In the former case there are neither ‘false positives’ (a priori non-invasive species classiﬁed as high risk, hence false invasive) nor ‘false negatives’ (a priori invasive species clas- siﬁed as low or medium risk, hence false non-invasive); in the latter case, the test cannot discriminate between ‘true positives’ (a priori inva- sive species classiﬁed as high risk, hence true invasive) and ‘true nega- tives’ (a priori non-invasive species classiﬁed as low or medium risk, hence true non-invasive). Following ROC curve analysis, the best threshold value that maximises the true positive rate and minimises the false positive rate was determined using Youden's J statistic. The BRA scores ranged from −15.0 to 55.0, with a mean = 18.6, a median = 18.0, and 5th and 95th percentiles = −4.0 and 42.5. Their distribution was not normal (JB = 36.664, P < 0.001), not skewed (skewness = 0.076, z = 1.230, P = 0.195), but platykurtic (kurtosis = 2.305, z = −9.359, P < 0.001) (Fig. 3a). The BRA + CCA scores ranged from −27.0 to 67.0, with a mean = 22.3, a median = 22.0, and 5th and 95th percentiles = −5.6 and 51.1. Table 1 Table 1 Published initial applications of the Aquatic Species Invasiveness Screening Kit (AS-ISK) by aquatic organismal group. Aquatic organismal group Reference(s) Reptiles Ruykys et al. (2021) Amphibians Ruykys et al. (2021) Freshwater ﬁshes Glamuzina et al. (2017), Li et al. (2017), Tarkan et al. (2017a), Tarkan et al. (2017b), Dodd et al. (2019), Suresh et al. (2019), Interesova et al. (2020), Moghaddas et al. (2020), Zięba et al. (2020), Glamuzina et al. (2021), Haubrock et al. (2021), Kumar et al. (2021), Moghaddas et al. (2021), Radočaj et al. (2021), Ruykys et al. (2021), Wei et al. (2021) Brackish ﬁshes Castellanos-Galindo et al. (2018), Clarke et al. (2020) Marine ﬁshes Filiz et al. (2017a), Filiz et al. (2017b), Bilge et al. (2019), Clarke et al. (2020), Lyons et al. (2020), Uyan et al. (2020) Tunicates Clarke et al. (2020) Freshwater invertebrates Paganelli et al. (2018), Ruykys et al. (2021), Semenchenko et al. (2018) Brackish invertebrates Clarke et al. (2020), Ruykys et al. (2021) Marine invertebrates Clarke et al. (2020), Killi et al. (2020), Stasolla et al. (2020), Ruykys et al. (2021) Freshwater plants Ruykys et al. (2021) Marine plants Clarke et al. (2020), Ruykys et al. (2021) Marine protists Clarke et al. (2020) Except for marine regions, for each RA area the corresponding Köppen-Geiger climate class (i.e. Tropical, Dry, Temperate, Continental, Polar: Peel et al., 2007) was identiﬁed, noting that in several cases more than one climate class applied to the same RA area. For marine regions, the classiﬁcation by Spalding et al. (2007) was used including: (i) Arctic, (ii) Temperate Northern Atlantic and Temperate Northern Paciﬁc (grouped in the present study into ‘Temperate marine’), and (iii) Central Indo-Paciﬁc, Tropical Atlantic, Tropical Eastern Paciﬁc and Western Indo-Paciﬁc (grouped in the present study into ‘Tropical marine’). The shape of the global distribution of the BRA and BRA + CCA scores was tested in R x64 v3.6.3 (R Development Core Team, 2020) using the package 'moments' v0.14 (Komsta and Novomestky, 2015), with normality, skewness and kurtosis evaluated by the Jarque-Bera (JB), D'Agostino and Anscombe tests, respectively. Computation of risk outcomes was based on receiver operating characteristic (ROC) curve analysis (Bewick et al., 2004). An ROC curve is a graph of sensitivity vs 5 Science of the Total Environment 788 (2021) 147868 L. Vilizzi, G.H. Copp, J.E. Hill et al. Table 1 The latter was then taken as ‘sufﬁcient evidence’ for categorising the species as invasive; whereas, if no evidence was found, then the species was categorised as non-invasive. Overall, the advantage of this method is that, by virtue of its meta-analytical foundation, it draws upon and com- bines previous approaches into a multi-tiered protocol. This maximises the amount of information collectable about the NNS under screening, thereby increasing the accuracy of the screening outcomes (Vilizzi, Copp and Hill, unpublished). 2) for a priori non-invasive species: An = (Na/Nt) × 100, where Na is the number of a priori non-invasive species accepted and Nt the total number of a priori non-invasive species screened; 2) for a priori non-invasive species: An = (Na/Nt) × 100, where Na is the number of a priori non-invasive species accepted and Nt the total number of a priori non-invasive species screened; 3) overall: Ao = (Na + Ir)/(Nt + It). 3) overall: Ao = (Na + Ir)/(Nt + It). In all cases, values above 50% are indicators of the accuracy of the screening tool. 3. Results For reptiles, amphibians, freshwater and marine ﬁshes, tunicates, freshwater and brackish invertebrates and marine protists, the BRA threshold was lower than the BRA + CCA one, whereas the opposite was true for brackish ﬁshes, marine invertebrates, and freshwater and marine plants. Except for marine protists (BRA), the mean AUC values (in Table 2) were in all cases > 0.5 – this conﬁrmed the ability of the toolkit to differentiate between a priori invasive and non-invasive species. After pooling, BRA and BRA + CCA thresholds could be computed for mammals and birds, and in both cases the BRA threshold was lower than the BRA + CCA one (Table 2). Based on the aquatic organismal group-speciﬁc thresholds (excluding the pooled ones), all three measures of accuracy had a mean value ≥50% for all groups except tunicates (BRA + CCA only), marine plants and ma- rine protists (both BRA and BRA + CCA) – a result of the relatively small sample sizes (Table 3). The number (and proportion) of true positives 6 Science of the Total Environment 788 (2021) 147868 L. Vilizzi, G.H. Copp, J.E. Hill et al. Fig. 1. Map of the risk assessment areas for which species were screened with the Aquatic Species Invasiveness Screening Kit (AS-ISK; see also Supplementary data Table S2). by the Dry climate was removed from further analysis (noting that the Polar climate was found only in combination with the Temperate and/ or Continental climates, hence could not be analysed separately regard- less of sample size). In total, 59 non-marine region RA areas were there- fore considered. For freshwater ﬁshes in tropical, temperate and continental climates, both the BRA and BRA + CCA thresholds were higher for the tropical climate, lower for the temperate and even lower for the continental climate, whereas the BRA + CCA was similar to the BRA in all cases (Table 5). Of the 38 marine ecoregion RA areas, four fell within the Arctic ecoregion, 24 in the Temperate grouping (including 23 RA areas in the Temperate Northern Atlantic and one in the Temperate Northern Paciﬁc ecoregions), and ten in the Tropical was consistently larger than that of the false negatives, which in all cases accounted for only 0–5.6% of the screened species for each group (Table 4). Table 2 Aquatic organismal group n BRA BRA + CCA Thr Mean LCI UCI VH Thr RAAs Thr Mean LCI UCI VH Thr RAAs Mammals 2 25.5 0.7180 0.5834 0.8525 – – 31.6 0.6639 0.5223 0.8056 – – Birds 4 25.5 0.7180 0.5834 0.8525 – – 31.6 0.6639 0.5223 0.8056 – – Reptiles 30 28.5 0.6550 0.4176 0.8924 – 2 36.5 0.5975 0.3489 0.8461 – 2 Amphibians 24 15.5 0.7983 0.6144 0.9822 – 3 19.5 0.7815 0.5934 0.9696 – 3 Freshwater ﬁshes 259 14.7 0.8446 0.7957 0.8936 30 10 17.7 0.8213 0.7691 0.8735 36 10 Brackish ﬁshes 17 38 0.7917 0.5542 1.0000 – – 29.5 0.6875 0.3969 0.9781 – – Marine ﬁshes 127 12.75 0.8254 0.7089 0.9420 – 3 19 0.7819 0.6541 0.9096 – 3 Tunicates 22 22.5 0.6417 0.3943 0.8890 – 2 36.75 0.5792 0.3119 0.8464 – 2 Freshwater invertebrates 144 13.25 0.8243 0.7185 0.9301 30 4 25.75 0.8199 0.7076 0.9322 36 4 Brackish invertebrates 11 15 0.6111 0.0247 1.0000 – – 26.9 0.7222 0.4142 1.0000 – – Marine invertebrates 151 15.1 0.8842 0.8333 0.9351 30 3 14.25 0.8483 0.7859 0.9107 36 3 Freshwater plants 15 24.5 0.8611 0.6697 1.0000 – – 12.5 0.9028 0.7522 1.0000 – – Marine plants 15 32 0.6161 0.3098 0.9224 – – 27.25 0.6250 0.3197 0.9303 – – Marine protists 18 34 0.4545 0.1639 0.7452 – – 42.75 0.5779 0.2899 0.8659 – – Marine bacteria 1 – – – – – – – – – – – – grouping (including three RA areas in the Central Indo-Paciﬁc, four in the Tropical Atlantic, one in the Tropical Eastern Paciﬁc, and two in the Western Indo-Paciﬁc ecoregions) (Supplementary data Table S2). However, owing to low sample sizes, thresholds could not be computed for any aquatic organismal group in the Arctic ecoregion. For marine ﬁshes, the BRA and BRA + CCA thresholds were higher for the temper- ate relative to the tropical ecoregion grouping, and the BRA + CCA threshold was higher than the BRA threshold in both cases (Table 5). For marine invertebrates, the BRA and BRA + CCA thresholds were higher for the tropical relative to the temperate ecoregion grouping, and the BRA threshold was similar to the BRA + CCA for the temperate ecoregion grouping but lower for the tropical ecoregion grouping (Table 5). In all cases, the mean AUC values were well above 0.5. Table 2 Table 2 Basic Risk Assessment (BRA) and BRA + Climate Change Assessment (CCA) AS-ISK thresholds from receiver operating characteristic curve analysis for the groups of aquatic organisms screened in the present study (Marine bacteria added for completeness). For each aquatic organismal group, the following is provided: number of screened species (n), threshold (Thr) values for the BRA and BRA + CCA [the Area Under the Curve (AUC) values are >0.5, and therefore are statistically valid, including for Mammals, Birds, Reptiles and Amphibians – with the latter two groups pooled together with Mammals and Birds for computation of the thresholds], with corresponding mean, lower conﬁdence interval (LCI) and upper conﬁdence interval (UCI) for the AUC, very high risk threshold (VH Thr) and minimum number of risk assessment areas (RAAs) for selection of the very high risk species. Note that threshold values are given in increments of 0.05 and rounded to the ﬁrst or second decimal (as applicable, following AS-ISK threshold notation). Table 2 In this study, ﬁshes and invertebrates represented the largest pro- portion of screened aquatic species, thus reﬂecting the composition of introduced animal species recorded for e.g. European waters (Alcaraz et al., 2005; Gherardi et al., 2009; Katsanevakis et al., 2013) but also the relative number of experts (cf. assessors) in the various aquatic or- ganismal groups. After freshwater ﬁshes, freshwater and marine inver- tebrates comprised the second most-widely screened group of aquatic organisms, with the marine invertebrates including a large proportion of Decapoda – an Order that comprises several of the world's worst in- vasive species (Lowe et al., 2000; Souty-Grosset et al., 2006). The large number of screenings for freshwater ﬁshes in this study can be at- tributed to the importance of inland waters as providers of ecosys- tem services for human societies (e.g. Wilson and Carpenter, 1999) and to the fact that these habitats are under high human-induced pressure, including NNS introductions (e.g. Hughes et al., 1998; Rahel, 2000). The broad geographical spread of most of the screened freshwater ﬁsh species reﬂects the increasing homogenisation of aquatic fauna and ﬂora as a result of worldwide introductions (e.g. McKinney, 1998; Rahel, 2000). Further, the taxonomic Orders that were more frequently screened are those usually ecologically ﬂexi- ble, able to withstand adverse ecological conditions, generally wide- spread over large spatial scales, and often of economic importance (e.g. Hulme, 2009). Based on the climate/marine ecoregion-speciﬁc thresholds, all three measures of accuracy had a mean value well above 50% in all cases and for both the BRA and BRA + CCA (Table 6). The number (and proportion) of true positives was consistently larger than that of the false negatives, which always accounted for only 0–1.6% of the screened species for each combination (Table 7). Similarly, the proportion of false positives was in most cases smaller than that of the true positives, and the proportion of medium-risk species was al- ways relatively high. 3. Results Similarly, the proportion of false positives was in most cases smaller than that of the true positives, and the proportion of the medium-risk species was always relatively high. In total, 33 species were identiﬁed as carrying a very high risk of invasiveness: 26 species based on both the BRA and BRA + CCA, four on the BRA only, and three on the BRA + CCA only (Fig. 4a, b). Of the 82 non-marine region RA areas, 56 included one climate class. To these RA areas, an additional four were added for which the second climate class (namely, Continental) was only marginally represented (Supplementary data Table S2), whereas the only RA area represented Fig. 2. Number (and corresponding percentage) of species screened according to aquatic organismal group. 7 Fig. 2. Number (and corresponding percentage) of species screened according to aquatic organismal group. Fig. 2. Number (and corresponding percentage) of species screened according to aquatic organismal group. Vilizzi, G.H. Copp, J.E. Hill et al. Science of the Total Environment 788 (202 , J.E. Hill et al. Science of the Total Envir Science of the Total Environment 788 (2021) 147868 L. Vilizzi, G.H. Copp, J.E. Hill et al. distribution of the Basic Risk Assessment (BRA) scores for the species screened with AS-ISK; (b) same for the BRA + CCA (Climate Change Co values (i.e. differences between BRA + CCA and BRA scores for each of the species screened) with corresponding percentage. 8 Fig. 3. (a) Frequency distribution of the Basic Risk Assessment (BRA) scores for the species screened with AS-ISK; (b) same for the BRA + CCA (Climate Change Component) scores; (c) same for delta CCA values (i.e. differences between BRA + CCA and BRA scores for each of the species screened) with corresponding percentage. 8 Science of the Total Environment 788 (2021) 147868 L. Vilizzi, G.H. Copp, J.E. Hill et al. L. Vilizzi, G.H. Copp, J.E. Hill et al. Table 4 Risk outcomes under current and future climate conditions Aquatic organismal group BRA BRA + CCA Non-invasive Invasive Non-invasive Invasive n % n % n % n % Reptiles Low 0 0.0 1 3.3 0 0.0 1 3.3 Medium 15 50.0 2 6.7 15 50.0 2 6.7 High 5 16.7 7 23.3 5 16.7 7 23.3 Amphibians Low 1 4.2 0 0.0 12 50.0 0 0.0 Medium 11 45.8 0 0.0 5 20.8 7 29.2 High 5 20.8 7 29.2 17 70.8 7 29.2 Freshwater ﬁshes Low 43 16.6 2 0.8 50 19.3 3 1.2 Medium 81 31.3 13 5.0 69 26.6 18 6.9 High 42 16.2 78 30.1 47 18.1 72 27.8 Brackish ﬁshes Low 0 0.0 0 0.0 0 0.0 0 0.0 Medium 9 52.9 3 17.6 8 47.1 3 17.6 High 0 0.0 5 29.4 1 5.9 5 29.4 Marine ﬁshes Low 36 28.3 1 0.8 33 26.0 1 0.8 Medium 46 36.2 1 0.8 54 42.5 4 3.1 High 27 21.3 16 12.6 22 17.3 13 10.2 Tunicates Low 0 0.0 0 0.0 0 0.0 0 0.0 Medium 7 31.8 2 9.1 11 50.0 6 27.3 High 5 22.7 8 36.4 1 4.5 4 18.2 Freshwater invertebrates Low 14 9.7 0 0.0 1 0.7 0 0.0 Medium 85 59.0 6 4.2 112 77.8 9 6.3 High 19 13.2 20 13.9 5 3.5 17 11.8 Brackish invertebrates Low 0 0.0 0 0.0 0 0.0 0 0.0 Medium 1 9.1 1 9.1 2 18.2 3 27.3 High 1 9.1 8 72.7 0 0.0 6 54.5 Marine invertebrates Low 15 9.9 0 0.0 16 10.6 0 0.0 Medium 57 37.7 8 5.3 53 35.1 10 6.6 High 20 13.2 51 33.8 23 15.2 49 32.5 Freshwater plants Low 1 5.6% 0 0.0% 4 22.2% 1 5.6% Medium 5 27.8% 2 11.1% 2 11.1% 4 22.2% High 0 0.0% 10 55.6% 0 0.0% 7 38.9% Marine plants Low 0 0.0 0 0.0 0 0.0 0 0.0 Medium 7 46.7 4 26.7 6 40.0 5 33.3 High 1 6.7 3 20.0 2 13.3 2 13.3 Marine protists Low 0 0.0 0 0.0 0 0.0 0 0.0 Medium 4 22.2 4 22.2 3 16.7 2 11.1 High 7 38.9 3 16.7 8 44.4 5 27.8 Several of the top (i.e. ‘very high risk’) species under current climate conditions achieved an even higher score under conditions of climate change (Fig. 4). Table 4 Of these species, most are either primarily warm- water/tropical or have wide thermal tolerances. For example, the red- eared slider Trachemys scripta scripta is a very common semi-aquatic turtle native to the south-eastern USA (Florida to south-eastern Virginia: Powel et al., 1991) and in its introduced range occurs in a wide variety of habitats, including slow-ﬂowing rivers, ﬂoodplain swamps, marshes, seasonal wetlands and permanent ponds (Scriber et al., 1986). Both goldﬁsh Carassius auratus and gibel carp Carassius gibelio are known to establish across a wide geographical and climatic range. These species are widespread and locally invasive both in Europe's more northerly parts, such as Finland and Poland (e.g. Grabowska et al., 2010; Puntila et al., 2013), across the Mediterranean region (e.g. Crivelli, 1995; Tarkan et al., 2012), and further aﬁeld in Australia (e.g. Beatty et al., 2017) and the Americas (Magalhães and Jacobi, 2013; Halas et al., 2018). The common lionﬁsh Pterois miles is an- other highly invasive species, especially since its invasion of the western Atlantic and Mediterranean Sea, which has been unprecedentedly rapid (Bariche et al., 2017; Schoﬁeld, 2010). The channelled applesnail Pomacea canaliculata is native to South America and has been intro- duced as an ornamental species in Europe and the Mediterranean area, but also elsewhere in the world through aquaculture (https:// www.cabi.org/isc/datasheet/68490). Phyllorhiza punctata is native to the tropical Western Paciﬁc, i.e. from Australia to Japan (Rippingale and Kelly, 1995) and has been reported across the Mediterranean Region (Abed-Navandi and Kikinger, 2007; Boero et al., 2009; Çevik et al., 2011; Gueroun et al., 2014; Deidun et al., 2017). An in-depth, species-speciﬁc description of all the ‘very high risk’ species identiﬁed in this study (Fig. 4) is provided in the Supplementary data, Appendix A1. Overall, the responses to climate-change questions tended to in- crease scores as well as thresholds for most taxonomic groups in most climatic regions. Species with broad distributions tended to possess broad thermal tolerances, suggesting that those species are likely to be able to expand their ranges, and thus impacts, poleward under future climate conditions. For example, six freshwater ﬁsh species not native to Great Britain were predicted to beneﬁt from the forecasted future cli- mate conditions, thus offering the potential to expand their ranges (Britton et al., 2010), which in Great Britain would be northward. Table 4 Risk outcomes (given as number of screened species and corresponding percentage) for the BRA and BRA + CCA for the groups of aquatic organisms for which risk thresholds were directly computed. Species are categorised a priori as either Non-invasive or Invasive (see Supplementary data Table S2). Medium-risk and high-risk outcomes for each group are based on the thresholds given in Table 2, whereas low-risk outcomes are based on a ‘default’ threshold of 1. 4.2. Table 3 Hill et al. L. Vilizzi, G.H. Copp, J.E. Hill et al. saltwater intrusion, and increased environmental disturbances, climate change is likely to favour the spread of NNS along their invasion path- ways as these conditions present opportunities for enhanced survival and lower invasion resistance of the invaded habitats (Rahel and Olden, 2008). Further, climate change may result in altered transport and/or introduction mechanisms or reduced effectiveness of control strategies (Hellmann et al., 2008). Table 3 Table 3 Accuracy measures for screenings on the groups of aquatic organisms for which BRA and BRA + CCA thresholds were directly computed (cf. Table 2). Ai = accuracy for a priori in- vasive species; An = accuracy for a priori non-invasive species; Ao = overall accuracy (see text for details). In italics, values < 50%. Table 3 Accuracy measures for screenings on the groups of aquatic organisms for which BRA and BRA + CCA thresholds were directly computed (cf. Table 2). Ai = accuracy for a priori in- vasive species; An = accuracy for a priori non-invasive species; Ao = overall accuracy (see text for details). In italics, values < 50%. Despite the large number of aquatic species screened in this study, bacteria were represented by only one species and no screenings for fungi were contributed. Risk screenings of these groups of aquatic or- ganisms would require the participation of experts in the ﬁelds of mi- crobiology and mycology, respectively. This points to the need for greater multi-disciplinarity in future risk identiﬁcation/assessment stud- ies, which is particularly important as both aquatic bacteria and fungi are known to exert in some cases severe ecological impacts once established and spread in their invasive range (Litchman, 2010), similar to their ter- restrial counterparts (Alderman, 1996; Loo, 2008). Regardless, it must be noted that pathogenic and parasitic organisms are normally evaluated separately from other NNS using risk assessment protocols speciﬁc to in- fectious agents (e.g. Peeler et al., 2007; D'hondt et al., 2015; Copp et al., 2016a). In addition, the diminutive size of these taxa could cause their presence to go un-noticed, thereby limiting knowledge of their spread and extent of invasiveness. Aquatic organismal group BRA BRA + CCA Ai An A0 Ai An A0 Reptiles 70.0 75.0 73.3 70.0 75.0 73.3 Amphibians 100.0 70.6 79.2 50.0 50.0 50.0 Freshwater ﬁshes 83.9 74.7 78.0 77.4 71.7 73.7 Brackish ﬁshes 62.5 100.0 82.4 62.5 88.9 76.5 Marine ﬁshes 88.9 75.2 77.2 72.2 79.8 78.7 Tunicates 80.0 58.3 68.2 40.0 91.7 68.2 Freshwater invertebrates 76.9 83.9 82.6 65.4 95.8 90.3 Brackish invertebrates 88.9 50.0 81.8 66.7 100.0 72.7 Marine invertebrates 86.4 78.3 81.5 83.1 75.0 78.1 Freshwater plants 83.3 100.0 88.9 58.3 100.0 72.2 Marine plants 42.9 87.5 66.7 28.6 75.0 53.3 Marine protists 42.9 36.4 38.9 71.4 27.3 44.4 9 Science of the Total Environment 788 (2021) 147868 L. Vilizzi, G.H. Copp, J.E. Table 4 Such poleward shifts in marine species distributions were initially deemed likely (VanDerWal et al., 2013), based in part on shifts observed in pre- vious warm periods (Drinkwater, 2006), and they have been predicted for some freshwater ﬁshes, e.g. channel catﬁsh Ictalurus punctatus (McCauley and Beitinger, 1992). Indeed, poleward shifts have been doc- umented in a recent meta-analysis of marine species distributions (Chaudhary et al., 2021). As such, tropical species are likely to expand poleward into temperate regions (e.g. Quero, 1998; Scavia et al., 2002), and temperate species to expand poleward into continental re- gions (Root et al., 2003; Hickling et al., 2006). Whereas, species with more limited thermal tolerances are likely to undergo shifts in their dis- tributions (and thus impacts), which are characterised by range reces- sions and range expansions in a poleward direction (Roessig et al., 2004; Rahel and Olden, 2008; Eissa and Zaki, 2011; Renaud et al., 2012). Wide thermal tolerances may be enhanced by local adaptation, such as is apparent in the cold-adapted population of eastern mosquitoﬁsh Gambusia holbrooki, which has established a self-sustaining population in Normandy, France (Beaudouin et al., 2008). This reﬂects the wide geo- graphical distribution of Gambusia holbrooki in its native range, which 4.2. Risk outcomes under current and future climate conditions The high proportion of positive ‘delta’ values (i.e. after accounting for climate change predictions: Fig. 3c) is in line with ﬁndings that climate change is likely to exacerbate the risk of introduction, establishment, dispersal and impact of several NNS, though some species might re- spond negatively to changes in climate conditions (Kernan, 2015). With predicted warmer temperatures, reduced lake ice cover, altered ﬂow regimes, increased salinity due to changes in precipitation and 10 Science of the Total Environment 788 (2021) 147868 L. Vilizzi, G.H. Copp, J.E. Hill et al. g the Mississippi River basin from the Gulf of Mexico north- way up the states of Indiana and Illinois (Aislabie et al., 2019). many cases, increased scores for BRA + CCA for freshwater For example, red lionﬁsh Pterois volitans scores in Florid creased slightly from the BRA to the BRA + CCA, with mode mainly due to a greater potential for the species to survive in igh risk species based on the BRA score (±SE) and screened for a representative number of risk assessment areas; (b) Same for the BRA + CCA. In dark gray he BRA and BRA + CCA. Table 4 Black circle (•): listing in the Centre for Agriculture and Bioscience International Invasive Species Compendium (CABI ISC); black sq asive Species Database (GISD). R = Reptiles; A = Amphibians; FF = Freshwater ﬁshes; MF = Marine ﬁshes; T = Tunicates; FI = Freshwater invertebrate ithin each aquatic organismal group, species are ordered according to decreasing score. See Table 2 for very high risk thresholds and number of risk as Fig. 4. (a) Very high risk species based on the BRA score (±SE) and screened for a representative number of risk assessment areas; (b) Same for the BRA + CCA. In dark gray, very high risk species for both the BRA and BRA + CCA. Black circle (•): listing in the Centre for Agriculture and Bioscience International Invasive Species Compendium (CABI ISC); black square (▪): listing in the Global Invasive Species Database (GISD). R = Reptiles; A = Amphibians; FF = Freshwater ﬁshes; MF = Marine ﬁshes; T = Tunicates; FI = Freshwater invertebrates; MI = Marine invertebrates. Within each aquatic organismal group, species are ordered according to decreasing score. See Table 2 for very high risk thresholds and number of risk assessment areas criteria. extends along the Mississippi River basin from the Gulf of Mexico north- ward to midway up the states of Indiana and Illinois (Aislabie et al., 2019). extends along the Mississippi River basin from the Gulf of Mexico north- ward to midway up the states of Indiana and Illinois (Aislabie et al., 2019). Thus, in many cases, increased scores for BRA + CCA for freshwater species are warranted. Further, this phenomenon, which is based on the interaction of climate and physiology, should pertain also to species in different aquatic environments such as brackish and marine systems. For example, red lionﬁsh Pterois volitans scores in Florida (USA) in- creased slightly from the BRA to the BRA + CCA, with modest increases mainly due to a greater potential for the species to survive inshore in the northern Gulf of Mexico during winter, thereby extending annual im- pacts (Lyons et al., 2020). Table 4 Conversely, within tropical climate zones, warmer conditions under climate-change scenarios are not likely to For example, red lionﬁsh Pterois volitans scores in Florida (USA) in- creased slightly from the BRA to the BRA + CCA, with modest increases mainly due to a greater potential for the species to survive inshore in the northern Gulf of Mexico during winter, thereby extending annual im- pacts (Lyons et al., 2020). Conversely, within tropical climate zones, warmer conditions under climate-change scenarios are not likely to Thus, in many cases, increased scores for BRA + CCA for freshwater species are warranted. Further, this phenomenon, which is based on the interaction of climate and physiology, should pertain also to species in different aquatic environments such as brackish and marine systems. 11 Science of the Total Environment 788 (2021) 147868 L. Vilizzi, G.H. Copp, J.E. Hill et al. Table 5 BRA and BRA + CCA thresholds from receiver characteristic curve analysis for freshwater ﬁshes relative to three climate classes (after Peel et al., 2007) and for marine ﬁshes and inver- tebrates relative to two marine ecoregion groupings (modiﬁed after Spalding et al., 2007). Abbreviations and notation as in Table 2. Table 5 BRA and BRA + CCA thresholds from receiver characteristic curve analysis for freshwater ﬁshes relative to three climate classes (after Peel et al., 2007) and for marine ﬁshes and inver- tebrates relative to two marine ecoregion groupings (modiﬁed after Spalding et al., 2007). Abbreviations and notation as in Table 2. interactions, and evolutio reater spread or heighten NN in most cases, near or total entative groups of aquatic an indicator of the accuracy Richardson, 2013). The ma curacy could be that of a larg el et al., 2007) and for marine ﬁsh Mean LCI 0.8986 0.8088 0.8493 0.7974 0.7387 0.5992 0.8208 0.6589 0.8016 0.6390 0.8871 0.8153 0.7279 0.5895 tebrates relative to two marine ecoregion groupings (modiﬁed after Spalding et al., 2007). Abbreviations and notation as in Table 2. dispersal, demography, species interactions, and evolution, not all changes in climate may result in greater spread or heighten NNS impacts (Urban et al., 2016). incite large alterations to the spread, abundance, or impacts of tropical species. The same can be said for continental climate zones with regard to cold-water species due to their existing adaptations to that climate types. Whereas range recession of a species' distribution could occur should a future temperature regime exceed the species' thermal toler- ances (Scavia et al., 2002). Table 7 Table 7 Risk outcomes (given as number of screened species and corresponding percentage) for the BRA and BRA + CCA for freshwater ﬁshes relative to three climate classes and for ma- rine ﬁshes and invertebrates relative to two marine ecoregion groupings (see Table 5). See Table 4 for details. Table 7 Risk outcomes (given as number of screened species and corresponding percentage) for the BRA and BRA + CCA for freshwater ﬁshes relative to three climate classes and for ma- rine ﬁshes and invertebrates relative to two marine ecoregion groupings (see Table 5). See Table 4 for details. Table 4 for details. Climate/marine ecoregion BRA BRA + CCA Non-invasive Invasive Non-invasive Invasive n % n % n % n % Freshwater ﬁshes Tropical Low 10 15.9 1 1.6 8 12.7 1 1.6 Medium 25 39.7 3 4.8 25 39.7 3 4.8 High 2 3.2 22 34.9 4 6.3 22 34.9 Temperate Low 31 15.5 0 0.0 38 19.0 1 0.5 Medium 59 29.5 14 7.0 39 19.5 9 4.5 High 30 15.0 66 33.0 43 21.5 70 35.0 Continental Low 8 13.8 0 0.0 8 13.8 0 0.0 Medium 6 10.3 7 12.1 4 6.9 6 10.3 High 7 12.1 30 51.7 9 15.5 31 53.4 Marine ﬁshes Temperate Low 1 2.2 0 0.0 1 2.2 0 0.0 Medium 27 58.7 2 4.3 29 63.0 2 4.3 High 8 17.4 8 17.4 6 13.0 8 17.4 Tropical Low 33 39.8 1 1.2 31 37.3 1 1.2 Medium 25 30.1 0 0.0 32 38.6 3 3.6 High 13 15.7 11 13.3 8 9.6 8 9.6 Marine invertebrates Temperate Low 15 15.5 0 0.0 12 12.4 0 0.0 Medium 49 50.5 3 3.1 49 50.5 4 4.1 High 7 7.2 23 23.7 10 10.3 22 22.7 Tropical Low 0 0.0 0 0.0 4 6.3 1 1.6 Medium 19 30.2 27 42.9 11 17.5 14 22.2 High 0 0.0 17 27.0 4 6.3 29 46.0 4.3. Management implications ( ) Despite the general emphasis on range expansion and greater im- pacts of invasive species due to warmer conditions (Rahel and Olden, 2008; Bradley et al., 2010), climate change is a complex issue for NNS risk assessment. When providing responses to CCA questions and rank- ing their conﬁdence in those responses, the assessor must consider a great breadth of information, including climate-match model predic- tions (e.g. Britton et al., 2010), if available for their RA area, as well as emissions scenarios, climate-model outputs, and time frames (Kennedy, 1990). The IPCC (2014) presents a variety of scenarios based on future emissions levels, with the extremes represented by RCP8.5 and RCP2.6. Further, there are numerous climate models that may be used for guidance, though these may differ in profound ways from one another in terms of predicted future temperature and precip- itation regimes (e.g. Kirtman et al., 2017). Other anomalies such as Artic warming, which are expected to lead to harsh, cold winters in mid- latitude areas of North America and Asia (Cohen et al., 2014; Kug et al., 2015), or non-analogue climates (Fitzpatrick and Hargrove, 2009) occur. Extreme events can set back or cancel species range expan- sion (Canning-Clode et al., 2011; Rehage et al., 2016) and thus may in- ﬂuence future risk estimates. Multi-directional range shifts are not only possible, but likely (VanDerWal et al., 2013). The time frame for such predictions is an important variable as well, given that the potential outcomes of range expansion, contraction, or oscillation in size are rela- tive to current NNS ranges. In view of the complexity of climate change interactions with biologically important factors such as physiology, The very low proportion and, in most cases, near or total absence of false negatives across the representative groups of aquatic organisms screened in the present study is an indicator of the accuracy of the risk screenings (cf. Kumschick and Richardson, 2013). The management consequences of this elevated accuracy could be that of a large number Table 4 Climate/marine ecoregion n BRA BRA + CCA Thr Mean LCI UCI Thr Mean LCI UCI Freshwater ﬁshes Tropical 63 18.4 0.9153 0.8370 0.9936 19.6 0.8986 0.8088 0.9885 Temperate 200 15.9 0.8685 0.8197 0.9174 16.0 0.8493 0.7974 0.9012 Continental 58 12.9 0.7844 0.6579 0.9110 13 0.7387 0.5992 0.8782 Marine ﬁshes Temperate 46 19.5 0.8083 0.6422 0.9745 31.5 0.8208 0.6589 0.9828 Tropical 83 12.5 0.8521 0.6994 1.0000 23.4 0.8016 0.6390 0.9643 Marine invertebrates Temperate 97 15.1 0.9236 0.8643 0.9829 15.6 0.8871 0.8153 0.9588 Tropical 63 35.75 0.7386 0.6085 0.8688 23.25 0.7279 0.5895 0.8662 dispersal, demography, species interactions, and evolution, not all changes in climate may result in greater spread or heighten NNS impacts (Urban et al., 2016). References Abed-Navandi, D., Kikinger, R., 2007. First record of the tropical scyphomedusa Phyllorhiza punctata von Lendenfeld, 1884 (Cnidaria: Rhizostomeae) in the Central Mediterra- nean Sea. Aquat. Invasions 2, 391–394. https://doi.org/10.3391/ai.2007.2.4.7. Aislabie, L.R., Verreycken, H., Chapman, D.S., Copp, G.H., 2019. Study on invasive alien spe- cies – development of risk assessments to tackle priority species and enhance pre- vention. Contract No 07.0202/2016/740982/ETU/ENV.D2. Final Report, Annex 3: Risk Assessment for Gambusia holbrooki (Girard, 1853) https://doi.org/10.13140/ RG.2.2.34828.33924. Alcaraz, C., Vila-Gispert, A., García-Berthou, E., 2005. Proﬁling invasive ﬁsh species: the importance of phylogeny and human use. Divers. Distrib. 11, 289–298. https://doi. org/10.1111/j.1366-9516.2005.00170.x. As is common in NNS risk analysis (e.g. Caley et al., 2006; Barry et al., 2008), the available scientiﬁc information (both peer-reviewed and gray literature) about the species being screened was reﬂected in the conﬁ- dence rankings assessors attributed to their responses. As such, given the robust conﬁdence levels, the present study provides a means for existing risk rankings to be adjusted and a stronger evidence base to identify: (i) which species require an immediate ‘rapid’ management ac- tion (e.g. eradication, control) to avoid or mitigate adverse impacts; (ii) which to subject to a full RA; and (iii) which to restrict or ban with regard to importation and/or sale as ornamental or aquarium/ﬁshery enhance- ment. To this end, biological monitoring programmes may explicitly search for high-scoring species because the sooner these species are found in new areas the more likely are eradication programmes to be successful. Such monitoring could be supported by developing tech- niques such as DNA metabarcoding (Brown et al., 2016) and eDNA sur- veys (Holman et al., 2019). Risk identiﬁcation therefore plays an important role in the provision of advice to policy makers, for the de- velopment of appropriate legislation, and associated regulation and management pertaining to NNS. In this perspective, the present study has also provided a means of ﬁne-tuning NNS risk analysis pro- cedures in countries that encompass more than one climatic class by the computation of generalised thresholds. In conclusion, the pres- ent study provides a comprehensive baseline to help identify (through risk screening using AS-ISK) for management priority high risk species across a range of taxonomic groups and geographi- cal/climatic regions, even where existing information on such spe- cies invasiveness/impact is limited. Alderman, D.J., 1996. Geographical spread of bacterial and fungal diseases of crustaceans. Rev. Sci. Tech. OIE 15, 603–632. CRediT authorship contribution statement of species ultimately warranting comprehensive (full) RA. In this study, accuracy was measured explicitly and represented the ‘pragmatic’ ap- proach, given that a full (comprehensive) RA, which might follow for species identiﬁed as potentially posing a high risk of invasiveness, would normally involve a major economic commitment. This is con- trary to the ‘idealistic’ approach, which would involve assessing both medium- and high-risk species, given the even higher economic com- mitment of accounting for both. An even more pragmatic approach could be therefore to base management decisions (or species' rankings) on the risk screening outcomes until such time that a full RA can be undertaken. This approach has been employed by the UK's Alien Species Group in its ‘impact’ ranking of aquatic NNS with regard to waterbody classiﬁcation under the EU Water Framework Directive (https://ec.europa.eu/environment/water/water-framework/index_ en.html). Within this context, aquatic NNS are listed provisionally as being of low, medium or high impact (UK-TAG ASG, 2021) pending the outcome of a full, and in some cases rapid, RA commissioned by the Great Britain Non-native Species Secretariat (www.nonnativespecies. org). The categorisation of the species is then subsequently conﬁrmed or changed according to the outcome of the full or rapid RA. of species ultimately warranting comprehensive (full) RA. In this study, accuracy was measured explicitly and represented the ‘pragmatic’ ap- proach, given that a full (comprehensive) RA, which might follow for species identiﬁed as potentially posing a high risk of invasiveness, would normally involve a major economic commitment. This is con- trary to the ‘idealistic’ approach, which would involve assessing both medium- and high-risk species, given the even higher economic com- mitment of accounting for both. An even more pragmatic approach could be therefore to base management decisions (or species' rankings) on the risk screening outcomes until such time that a full RA can be undertaken. This approach has been employed by the UK's Alien Species Group in its ‘impact’ ranking of aquatic NNS with regard to waterbody classiﬁcation under the EU Water Framework Directive (https://ec.europa.eu/environment/water/water-framework/index_ en.html). Within this context, aquatic NNS are listed provisionally as being of low, medium or high impact (UK-TAG ASG, 2021) pending the outcome of a full, and in some cases rapid, RA commissioned by the Great Britain Non-native Species Secretariat (www.nonnativespecies. org). The categorisation of the species is then subsequently conﬁrmed or changed according to the outcome of the full or rapid RA. CRediT authorship contribution statement LV and GHC designed the concept, co-wrote and edited the manu- script; LV analysed the data, and all other authors contributed data to and participated in the composition and editing of the manuscript. Acknowledgements This study is dedicated to our co-author, the late Prof. Olaf Weyl, whose sudden passing during ﬁeld work in November 2020 was re- ceived by us with great sadness. This study represents a contribution to the term of reference ‘d’ (“Advance knowledge base to further de- velop indicators to evaluate the status and impact of non-indigenous species in marine environments”) of the ICES working group on Intro- ductions and Transfers of Marine Organisms. The participation of GHC was supported by the Cefas Science Excellence fund. AW and ÁF were supported by TKP2020-NKA-16 project. As shown in the present study, invasive and non-invasive species could be distinguished accurately across aquatic organisms to a greater degree than would be expected by chance alone. Calibrated thresholds could be computed for several taxonomic groups, and for freshwater and marine ﬁshes and for marine invertebrates also based on climate/ecoregion. In RA areas for which no calibration is possible, e.g. due to a statistically insufﬁcient number of screenings and/or to the requirement of screening only one target species (or small group thereof), these generalised thresholds (i.e. at the organ- ism group or climate class level) can be reliably used in future risk screening applications for distinguishing between species that pose a high risk of being invasive from those posing a low-to-medium risk. Use of these thresholds may be therefore of particular relevance in cases of individual species being risk screened (e.g. Castellanos- Galindo et al., 2018; Zięba et al., 2020), including ‘rapid risk assess- ment’ studies, or for RA areas where the number of NNS is too limited for a valid calibration to be undertaken (e.g. Filiz et al., 2017b; Paganelli et al., 2018; Semenchenko et al., 2018; Dodd et al., 2019; Lyons et al., 2020). Declaration of competing interest The authors declare that they have no known competing ﬁnancial interests or personal relationships that could have appeared to inﬂu- ence the work reported in this paper. Supplementary data to this article can be found online at https://doi. org/10.1016/j.scitotenv.2021.147868. Supplementary data to this article can be found online at https://doi. org/10.1016/j.scitotenv.2021.147868. Table 6 Table 6 Accuracy measures for screenings on freshwater ﬁshes relative to three climate classes and for marine ﬁshes and invertebrates relative to two marine ecoregion groupings (see Table 5). Abbreviations and notation as in Table 3. Table 6 Accuracy measures for screenings on freshwater ﬁshes relative to three climate classes and for marine ﬁshes and invertebrates relative to two marine ecoregion groupings (see Table 5). Abbreviations and notation as in Table 3. Climate/marine ecoregion BRA BRA + CCA Ai An A0 Ai An A0 Freshwater ﬁshes Tropical 84.6 94.6 90.5 84.6 89.2 87.3 Temperate 82.5 75.0 78.0 87.5 64.2 73.5 Continental 81.1 66.7 75.9 83.8 57.1 74.1 Marine ﬁshes Temperate 80.0 77.8 78.3 80.0 83.3 82.6 Tropical 91.7 81.7 83.1 66.7 88.7 85.5 Marine invertebrates Temperate 88.5 90.1 89.7 84.6 85.9 85.6 Tropical 38.6 100.0 57.1 65.9 78.9 69.8 12 Science of the Total Environment 788 (2021) 147868 L. Vilizzi, G.H. Copp, J.E. Hill et al. References Global threats from invasive alien species in the twenty-ﬁrst century and national response capac- ities. Nat. Commun. 7, 12485. https://doi.org/10.1038/ncomms12485. Chan, J., Zeng, Y., Yeo, D.C.J., 2021. Invasive species trait-based risk assessment for non- native freshwater ﬁshes in a tropical city basin in Southeast Asia. PLoS ONE 16, e0248480. https://doi.org/10.1371/journal.pone.0248480. Eissa, A.E., Zaki, M.M., 2011. The impact of global climatic changes on the aquatic environ- ment. Procedia Environ. Sci. 4, 251–259. https://doi.org/10.1016/j.proenv.2011.03.030. Chapman, D.S., Gunn, I.D., Pringle, H.E., Siriwardena, G.M., Taylor, P., Thackeray, S.J., Willby, N.J., Carvalho, L., 2020. Invasion of freshwater ecosystems is promoted by net- work connectivity to hotspots of human activity. Glob. Ecol. Biogeogr. 29, 645–655. https://doi.org/10.1111/geb.13051. European Union, 2014. Regulation (EU) No 1143/2014 of the European Parliament and of the Council of 22 October 2014 on the prevention and management of the introduc- tion and spread of invasive alien species. Off. J. Eur. Union L 317, 35. https://eur-lex. europa.eu/legal-content/EN/TXT/?uri=CELEX%3A32014R1143. Chaudhary, C., Richardson, A.J., Schoeman, D.S., Costello, M.J., 2021. Global warming is causing a more pronounced dip in marine species richness around the equator. Proc. Nat. Acad. Sci. 118, e2015094118. https://doi.org/10.1073/pnas.2015094118. Filiz, H., Tarkan, A.S., Bilge, G., Yapıcı, S., 2017a. Assessment of invasiveness potential of Pterois miles by the aquatic species invasiveness screening kit. J. Black Sea/Mediterra- nean Environ. 23, 17–37. http://blackmeditjournal.org/wp-content/uploads/ 2.2312017-H-Filiz.pdf. Clarke, S.A., Vilizzi, L., Lee, L., Wood, L.E., Cowie, W.J., Burt, J.A., Mamiit, R.J., Hassina, A., Davison, P.I., Fenwick, G.V., Harmer, R., Skóra, M.E., Kozic, S., Aislabie, L.R., Kennerley, A., Le Quesne, W.J.F., Copp, G.H., Stebbing, P.D., 2020. Identifying poten- tially invasive non-native marine and brackish water species for the Arabian Gulf and Sea of Oman. Glob. Change Biol. 26, 2081–2092. https://doi.org/10.1111/ GCB.14964. Filiz, H., Yapıcı, S., Bilge, G., 2017b. The factors increasing of invasiveness potential of ﬁve pufferﬁshes in the eastern Mediterranean, Turkey. Nat. Eng. Sci. 2 (Suppl), 22–30. www.nesciences.com/download.php?id=238. Fitzpatrick, M.C., Hargrove, W.W., 2009. The projection of species distribution models and the problem of non-analog climate. Biodivers. Conserv. 18, 2255. https://doi.org/ 10.1007/s10531-009-9584-8. Cohen, J., Screen, J.A., Furtado, J.C., Barlow, M., Whittleston, D., Coumou, D., Francis, J., Dethloff, K., Entekhabi, D., Overland, J., Jones, J., 2014. Recent Arctic ampliﬁcation and extreme mid-latitude weather. Nat. Geosci. 7, 627–637. https://doi.org/10.1038/ ngeo2234. Gherardi, F., Gollasch, S., Minchin, D., Olenin, S., Panov, V.E., 2009. Alien invertebrates and ﬁsh in European inland waters. Handbook of Alien Species in Europe. Springer Series in Invasion Ecology vol 3. References Springer, Dordrecht, pp. 81–92. https://doi.org/10.1007/ 978-1-4020-8280-1_6. Copp, G.H., Garthwaite, R., Gozlan, R.E., 2005a. Risk identiﬁcation and assessment of non- native freshwater ﬁshes: concepts and perspectives on protocols for the UK. Cefas Science Technical Report. Cefas, Lowestoft, UK www.cefas.co.uk/publications/ techrep/tech129.pdf. Glamuzina, B., Tutman, P., Nikolić, V., Vidović, Z., Pavličević, J., Vilizzi, L., Copp, G.H., Simonović, P., 2017. Comparison of taxon-speciﬁc and taxon-generic risk screening tools to identify potentially invasive non-native ﬁshes in the River Neretva catchment (Bosnia and Herzegovina and Croatia). River Res. Appl. 33, 670–679. https://doi.org/ 10.1002/rra.3124. Copp, G.H., Garthwaite, R., Gozlan, R.E., 2005b. Risk identiﬁcation and assessment of non- native freshwater ﬁshes: a summary of concepts and perspectives on protocols for the UK. J. Appl. Ichthyol. 21, 371–373. https://doi.org/10.1111/j.1439-0426.2005.00692.x. Copp, G.H., Bianco, P.G., Bogutskaya, N., Erős, T., Falka, I., Ferreira, M.T., Fox, M.G., Freyhof, J., Gozlan, R.E., Grabowska, J., Kováč, V., Moreno-Amich, R., Naseka, A.M., Peňáz, M., Povž, M., Przybylski, M., Robillard, M., Russell, I.C., Stakėnas, S., Šumer, S., Vila- Gispert, A., Wiesner, C., 2005c. To be, or not to be, a non-native freshwater ﬁsh? J. Appl. Ichthyol. 21, 242–262. https://doi.org/10.1111/j.1439-0426.2005.00690.x. Glamuzina, B., Tutman, P., Glamuzina, L., Vidović, Z., Simonović, P., Vilizzi, L., 2021. Quantifying current and future risks of invasiveness of non-native aquatic species in highly urbanised estuarine ecosystems–a case study of the River Neretva Estuary (Eastern Adriatic Sea: Croatia and Bosnia–Herzegovina). Fish. Manag. Ecol. https:// doi.org/10.1111/FME.12463. Copp, G.H., Vilizzi, L., Mumford, J., Fenwick, G.V., Godard, M.J., Gozlan, R.E., 2009. Calibration of FISK, an invasiveness screening tool for nonnative freshwater ﬁshes. Risk Anal. 29, 457–467. https://doi.org/10.1111/j.1539-6924.2008.01159.x. Gordon, D.R., Onderdonk, D.A., Fox, A.M., Stocker, R.K., 2008. Consistent accuracy of the Australian weed risk assessment system across varied geographies. Divers. Distrib. 14, 234–242. https://doi.org/10.1111/j.1472-4642.2007.00460.x. Copp, G.H., Russell, I.C., Peeler, E.J., Gherardi, F., Tricarico, E., MacLeod, A., Cowx, I.G., Nunn, A.N., Occhipinti-Ambrogi, A., Savini, D., Mumford, J., Britton, J.R., 2016a. European non-native species in aquaculture risk analysis scheme – a summary of assessment protocols and decision making tools for use of alien species in aquaculture. Fish. Manag. Ecol. 23, 1–11. https://doi.org/10.1111/fme.12074. Grabowska, J., Kotusz, J., Witkowski, A., 2010. Alien invasive ﬁsh species in Polish waters: an overview. J. Vertebr. Biol. 59, 73–85. https://doi.org/10.25225/fozo.v59.i1.a1.2010. Gueroun, S.K.M., Kéﬁ-Daly Yahia, O., Deidun, A., Fuentes, V., Piraino, S., Daly Yahia, M.N., 2014. First record and potential trophic impact of Phyllorhiza punctata (Cnidaria: Scyphozoa) along the north Tunisian coast (South Western Mediterranean Sea). Ital. J. Zool. References David, M., Gollasch, S., 2019. Risk assessment for ballast water management – learning from the Adriatic Sea case study. Mar. Pollut. Bull. 147, 36–46. https://doi.org/10.1016/j. marpolbul.2018.02.003. Britton, J.R., Cucherousset, J., Davies, G.D., Godard, M.J., Copp, G.H., 2010. Non-native ﬁshes and climate change: predicting species responses to warming tempera- tures in a temperate region. Freshw. Biol. 55, 1130–1141. https://doi.org/ 10.1111/j.1365-2427.2010.02396.x. David, M., Gollasch, S., Leppäkoski, E., 2013. Risk assessment for exemptions from ballast water management – the Baltic Sea case study. Mar. Pollut. Bull. 75, 205–217. https:// doi.org/10.1016/j.marpolbul.2013.07.031. Brown, E.A., Chain, F.J.J., Zhan, A., MacIsaac, H.J., Cristescu, M.E., 2016. Early detection of aquatic invaders using metabarcoding reveals a high number of non-indigenous spe- cies in Canadian ports. Divers. Distrib. 22, 1045–1059. https://doi.org/10.1111/ ddi.12465. Deidun, A., Sciberras, J., Sciberras, A., Gauci, A., Balistreri, P., Salvatore, A., Piraino, S., 2017. The ﬁrst record of the white-spotted Australian jellyﬁsh Phyllorhiza punctata von Lendenfeld, 1884 from Maltese waters (western Mediterranean) and from the Ionian coast of Italy. BioInvasions Rec. 6, 119–124. https://doi.org/10.3391/bir.2017.6.2.05. Caley, P., Lonsdale, W.M., Pheloung, P.C., 2006. Quantifying uncertainty in predictions of invasiveness, with emphasis on weed risk assessment. Biol. Invasions 8, 1595–1604. https://doi.org/10.1007/s10530-004-0010-6. D'hondt, B., Vanderhoeven, S., Roelandt, S., Mayer, F., Versteirt, V., Adriaens, T., Ducheyne, E., San Martin, G., Grégoire, J.C., Stiers, I., Quoilin, S., 2015. Harmonia+ and Pandora+: risk screening tools for potentially invasive plants, animals and their pathogens. Biol. Invasions 17, 1869–1883. https://doi.org/10.1007/s10530-015-0843-1. Canning-Clode, J., Fowler, A.E., Byers, J.E., Carlton, J.T., Ruiz, G.M., 2011. ‘Caribbean creep’ chills out: climate change and marine invasive species. PloS One 6, e296547. https://doi.org/10.1371/journal.pone.0029657. Dodd, J.A., Vilizzi, L., Bean, C.W., Davison, P.I., Copp, G.H., 2019. At what spatial scale should risk screenings of translocated freshwater ﬁshes be undertaken – river basin district or climo-geographic designation? Biol. Conserv. 230, 122–130. https://doi. org/10.1016/j.biocon.2018.12.002. Castellanos-Galindo, G., Moreno, X., Robertson, R., 2018. Risks to eastern Paciﬁc marine ecosystems from sea-cage mariculture of alien Cobia. Manag. Biol. Invasions 9, 323–327. https://doi.org/10.3391/mbi.2018.9.3.14. Drinkwater, K.F., 2006. The regime shift of the 1920s and 1930s in the North Atlantic. Progr. Oceanogr. 68, 134–151. https://doi.org/10.1016/j.pocean.2006.02.011. Çevik, C., Derici, O.B., Çevik, F., Cavaş, L., 2011. First record of Phyllorhiza punctata von Lendenfeld, 1884 (Scyphozoa: Rhizostomeae: Mastigiidae) from Turkey. Aquat. Invasions 6 (Suppl. 1), S27–S28. https://doi.org/10.3391/ai.2011.6.S1.006. Early, R., Bradley, B.A., Dukes, J.S., Lawler, J.J., Olden, J.D., Blumenthal, D.M., Gonzalez, P., Grosholz, E.D., Ibañez, I., Miller, L.P., Sorte, C.J.B., Tatem, A.J., 2016. References Bariche, M., Kleitou, P., Kalogirou, S., Bernardi, G., 2017. Genetics reveal the identity and origin of the lionﬁsh invasion in the Mediterranean Sea. Sci. Rep. 7, 6782. https:// doi.org/10.1038/s41598-017-07326-1. Barney, J.N., DiTomaso, J.M., 2010. Bioclimatic predictions of habitat suitability for the bio- fuel switchgrass in North America under current and future climate scenarios. Biomass Bioenergy 34, 124–133. https://doi.org/10.1016/j.biombioe.2009.10.009. Barry, S.C., Hayes, K.R., Hewitt, C.L., Behrens, H.L., Dragsund, E., Bakke, S.M., 2008. Ballast water risk assessment: principles, processes, and methods. ICES J. Mar. Sci. 65, 121–131. https://doi.org/10.1093/icesjms/fsn004. Bayne, B.L., Ahrens, M., Allen, S.K., Anglès D'auriac, M., Backeljau, T., Beninger, P., Bohn, R., Boudry, P., Davis, J., Green, T., Guo, X., Hedgecock, D., Ibarra, A., Kingsley-Smith, P., Krause, M., Langdon, C., Lapègue, S., Li, C., Manahan, D., Mann, R., Perez-Paralle, L., Powell, E.N., Rawson, P.D., Speiser, D., Sanchez, J.-L., Shumway, S., Wang, H., 2017. The proposed dropping of the genus Crassostrea for all Paciﬁc cupped oysters and its replacement by a new genus Magallana: a dissenting view. J. Shellﬁsh Res. 36, 545–547. https://doi.org/10.2983/035.036.0301. Beatty, S.J., Allen, M.G., Whitty, J.M., Lymbery, A.J., Keleher, J.J., Tweedley, J.R., Ebner, B.C., Morgan, D.L., 2017. First evidence of spawning migration by goldﬁsh (Carassius auratus); implications for control of a globally invasive species. Ecol. Freshwat. Fish 26, 444–455. https://doi.org/10.1111/eff.12288. Beaudouin, R., Ginot, V., Monod, G., 2008. Growth characteristics of eastern mosquitoﬁsh Gambusia holbrooki in a northern habitat (Brittany, France). J. Fish Biol. 73, 2468–2484. https://doi.org/10.1111/j.1095-8649.2008.02101.x. Bewick, V., Cheek, L., Ball, J., 2004. Statistics review 13: receiver operating characteristic curves. Crit. Care 8, 508–512. https://doi.org/10.1186/cc3000. Bilge, G., Filiz, H., Yapici, S., Tarkan, A.S., Vilizzi, L., 2019. A risk screening study on the po- tential invasiveness of Lessepsian ﬁshes in the south-western coasts of Anatolia. Acta Ichthyol. Piscat. 49, 23–31. https://doi.org/10.3750/AIEP/02422. Boero, F., Putti, M., Trainito, E., Prontera, E., Piraino, S., Shiganova, T.A., 2009. First records of Mnemiopsis leidyi (Ctenophora) from the Ligurian, Tyrrhenian and Ionian seas (Western Mediterranean) and ﬁrst record of Phyllorhiza punctata (Cnidaria) from the Western Mediterranean. Aquat. Invasions 4, 675–680. https://doi.org/10.3391/ ai.2009.4.4.13. 13 Science of the Total Environment 788 (2021) 147868 L. Vilizzi, G.H. Copp, J.E. Hill et al. Bradley, B.A., Wilcove, D.S., Oppenheimer, M., 2010. Climate change increases risk of plant invasion in the Eastern United States. Biol. Invasions 12, 1855–1872. https://doi.org/ 10.1007/s10530-009-9597-y. David, M., Gollasch, S., 2018. How to approach ballast water management in European seas. Estuar. Coast. Shelf Sci. 201, 248–255. https://doi.org/10.1016/j.ecss.2016.10.018. References 82, 95–100. https://doi.org/10.1080/11250003. 2014.981306. Copp, G.H., Vilizzi, L., Tidbury, H., Stebbing, P.D., Tarkan, A.S., Miossec, L., Goulletquer, P., 2016b. Development of a generic decision-support tool for identifying potentially in- vasive aquatic taxa: AS-ISK. Manag. Biol. Invasions 7, 343–350. https://doi.org/ 10.3391/mbi.2016.7.4.04. Halas, D., Lovejoy, N., Mandrak, N.E., 2018. Undetected diversity of goldﬁsh (Carassius spp.) in North America. Aquat. Invasions 13, 211–219. https://doi.org/10.3391/ ai.2018.13.2.03. Copp, G.H., Vilizzi, L., Wei, H., Li, S., Piria, M., Al-Faisal, A.J., Almeida, D., Atique, U., Al- Wazzan, Z., Bakiu, R., Bašić, T., Bui, T.D., Canning-Clode, J., Castro, N., Chaichana, R., Çoker, T., Dashinov, D., Ekmekçi, F.G., Erős, T., Ferincz, Á., Ferreira, T., Giannetto, D., Gilles Jr., A.S., Głowacki, Ł., Goulletquer, P., Interesova, E., Iqbal, S., Jakubčinová, K., Kanongdate, K., Kim, J.-E., Kopecký, O., Kostov, V., Koutsikos, N., Kozic, S., Kristan, P., Kurita, Y., Lee, H.-G., Leuven, R.S.E.W., Lipinskaya, T., Lukas, J., Marchini, A., González Martínez, A.I., Masson, L., Memedemin, D., Moghaddas, S.D., Monteiro, J., Mumladze, L., Naddaﬁ, R., Năvodaru, I., Olsson, K.H., Onikura, N., Paganelli, P., Pavia Jr., R.T., Perdikaris, C., Pickholz, R., Pietraszewski, D., Povž, M., Preda, C., Ristovska, M., Rosíková, K., Santos, J.M., Semenchenko, V., Senanan, W., Simonović, P., Smeti, E., Števove, B., Švolíková, K., Ta, K.A.T., Tarkan, A.S., Top, N., Tricarico, E., Uzunova, E., Vardakas, L., Verreyken, H., Zięba, G., Mendoza, R., 2021. Speaking their language – development of a multilingual decision-support tool for communicating invasive spe- cies risks to decision makers and stakeholders. Environ. Modell. Softw. 135, 104900. https://doi.org/10.1016/j.envsoft.2020.104900. Haubrock, P.J., Copp, G.H., Johović, I., Balzani, P., Inghilesi, A.F., Nocita, A., Tricarico, E., 2021. North American channel catﬁsh, Ictalurus punctatus: a neglected but potentially invasive freshwater ﬁsh species? Biol. Inv. https://doi.org/10.1007/s10530-021- 02459-x. Hellmann, J.J., Byers, J.E., Bierwagen, B.G., Dukes, J.S., 2008. Five potential consequences of climate change for invasive species. Conserv. Biol. 22, 534–543. https://doi.org/ 10.1111/j.1523-1739.2008.00951.x. Hickling, R., Roy, D.B., Hill, J.K., Fox, R., Thomas, C.D., 2006. The distributions of a wide range of taxonomic groups are expanding polewards. Glob. Change Biol. 12, 450–455. https://doi.org/10.1111/j.1365-2486.2006.01116.x. Holman, L.E., de Bruyn, M., Creer, S., Carvalho, G., Robidart, J., Rius, M., 2019. Detection of introduced and resident marine species using environmental DNA metabarcoding of sediment and water. Sci. Rep. 9, 11559. https://doi.org/10.1038/s41598-019-47899- 7. Hughes, R.M., Kaufmann, P.R., Herlihy, A.T., Kincaid, T.M., Reynolds, L., Larsen, D.P., 1998. A process for developing and evaluating indices of ﬁsh assemblage integrity. Can. J. Fish. Aquat. Sci. 55, 1618–1631. https://doi.org/10.1139/f98-060. References https://doi.org/10.1371/journal.pone.0119248. Lyons, T.J., Tuckett, Q.M., Donahou, A.D., Hill, J.E., 2020. Risk screen of lionﬁshes, Pterois, Dendrochirus, and Parapterois, for southeastern United States coastal waters of the Gulf of Mexico and Atlantic Ocean. Biol. Inv. 22, 1573–1583. https://doi.org/ 10.1007/s10530-020-02203-x. Ruykys, L., Ta, K.A.T., Bui, T.D., Vilizzi, L., Copp, G.H., 2021. Risk screening of the potential invasiveness of non-native aquatic species in Vietnam. Biol. Invasions https://doi.org/ 10.1007/s10530-020-02430-2 in press. Magalhães, A.L.B.D., Jacobi, C.M., 2013. Invasion risks posed by ornamental freshwater ﬁsh trade to southeastern Brazilian rivers. Neotrop. Ichthyol. 11, 433–441. https://doi.org/ 10.1590/S1679-62252013005000003. Scavia, D., Field, J.C., Boesch, D.F., Buddemeier, R.W., Burkett, V., Cayan, D.R., Fogarty, M., Harwell, M.A., Howarth, R.W., Mason, C., Reed, D.J., 2002. Climate change impacts on US coastal and marine ecosystems. Estuaries 25, 149–164. https://doi.org/ 10.1007/BF02691304. McCauley, R., Beitinger, T., 1992. Predicted effects of climate warming on the commercial culture of the channel catﬁsh, Ictalurus punctatus. GeoJournal 28, 61–66. https://doi. org/10.1007/BF00216407. Schoﬁeld, P.J., 2010. Update on geographic spread of invasive lionﬁshes (Pterois volitans [Linnaeus, 1758] and P. miles [Bennett, 1828]) in the Western North Atlantic Ocean, Caribbean Sea and Gulf of Mexico. Aquat. Invasions 5, 117–122. https://doi.org/ 10.3391/ai.2010.5.S1.024. g McKinney, M.L., 1998. On predicting biotic homogenization: species-area patterns in ma- rine biota. Glob. Ecol. Biogeogr. Lett. 7, 297–301. https://doi.org/10.2307/2997604. Moghaddas, S.D., Abdoli, A., Kiabi, B.H., Rahmani, H., 2020. Risk assessment of the poten- tial invasiveness of Coptodon zillii (Gervais, 1848) in Anzali Wetland using AS-ISK. Model. Environ. Sci. 18, 255–270 In Persian. 10.29252/envs.18.2.255. Scriber, K.T., Evans, J.E., Morreale, S.J., Smith, M.H., Gibbons, J.W., 1986. Genetic divergence among populations of the yellow-bellied slider turtle (Pseudemys scripta) separated by aquatic and terrestrial habitats. Copeia 1986, 691–700. https://doi.org/10.2307/ 1444951. Moghaddas, S.D., Abdoli, A., Kiabi, B.H., Rahmani, H., Vilizzi, L., Copp, G.H., 2021. Identifying invasive ﬁsh species threats to RAMSAR wetland sites in the Caspian Sea region – a case study of the Anzali Wetland Complex (Iran). Fish. Manag. Ecol. 28, 28–39. https://doi.org/10.1111/fme.12453. Semenchenko, V., Lipinskaya, T., Vilizzi, L., 2018. Risk screening of non-native macroinver- tebrates in the major rivers and associated basins of Belarus using the Aquatic Species Invasiveness Screening Kit. Manag. Biol. Inv. 9, 127–136. https://doi.org/10.3391/ mbi.2018.9.2.06. Mumford, J.D., Booy, O., Baker, R.H.A., Rees, M., Copp, G.H., Black, K., Holt, J., Leach, A.W., Hartley, M., 2010. Non-native species risk assessment in Great Britain. In: Evans, A. (Ed.), What Makes an Alien Invasive? Risk and Policy Responses. Aspects of Applied Biology 104. References https://doi.org/10.1071/MF9951145. Kumschick, S., Richardson, D.M., 2013. Species-based risk assessments for biological inva- sions: advances and challenges. Divers. Distrib. 19, 1095–1105. https://doi.org/ 10.1111/ddi.12110. Robin, X., Turck, N., Hainard, A., Tiberti, N., Lisacek, F., Sanchez, J.-C., Müller, M., 2011. pROC: an open-source package for R and S+ to analyze and compare ROC curves. BMC Bioinformatics 12, 77. https://doi.org/10.1186/1471-2105-12-77. Leuven, R.S.E.W., van der Velde, G., Baijens, I., Snijders, J., van der Zwart, C., Lenders, H.J.R., bij de Vaate, A., 2009. The river Rhine: a global highway for dispersal of aquatic inva- sive species. Biol. Invasions 11, 1989–2008. https://doi.org/10.1007/s10530-009- 9491-7. Roessig, J.M., Woodley, C.M., Cech, J.J., Hansen, L.J., 2004. Effects of global climate change on marine and estuarine ﬁshes and ﬁsheries. Rev. Fish Biol. Fish. 14, 251–275. https:// doi.org/10.1007/s11160-004-6749-0. Li, S., Chen, J., Wang, X., Copp, G.H., 2017. Invasiveness screening of non-native ﬁshes for the middle reach of the Yarlung Zangbo River, Tibetan Plateau, China. River Res. Appl. 33, 1439–1444. https://doi.org/10.1002/rra.3196. Root, T.L., Price, J.T., Hall, K.R., Schneider, S.H., Rosenzweig, C., Pounds, J.A., 2003. Fingerprints of global warming on wild animals and plants. Nature 421, 57–60. https://doi.org/ 10.1038/nature01333. p g Litchman, E., 2010. Invisible invaders: non-pathogenic invasive microbes in aquatic and terrestrial ecosystems. Ecol. Lett. 13, 1560–1572. https://doi.org/10.1111/j.1461- 0248.2010.01544.x. Roy, H.E., Rabitsch, W., Scalera, R., Stewart, A., Gallardo, B., Genovesi, P., Essl, F., Adriaens, T., Bacher, S., Booy, O., Branquart, E., Brunel, S., Copp, G.H., Dean, H., D’hondt, B., Josefsson, M., Kenis, M., Kettunen, M., Linnamagi, M., Lucy, F., Martinou, A., Moore, N., Nentwig, W., Nieto, A., Pergl, J., Peyton, J., Roques, A., Schindler, S., Schönrogge, K., Solarz, W., Stebbing, P.D., Trichkova, T., Vanderhoeven, S., van Valkenburg, J., Zenetos, A., 2018. Developing a framework of minimum standards for the risk assess- ment of alien species. J. Appl. Ecol. 55, 526–538. https://doi.org/10.1111/1365- 2664.13025. Loo, J.A., 2008. Ecological impacts of non-indigenous invasive fungi as forest pathogens. In: Langor, D.W., Sweeney, J. (Eds.), Ecological Impacts of Non-Native Invertebrates and Fungi on Terrestrial Ecosystems. Springer, Dordrecht, pp. 81–96. Lowe, S., Browne, M., Boudjelas, S., De Poorter, M., 2000. 100 of the World’s Worst Invasive Alien Species: A Selection From the Global Invasive Species Database. Vol. 12. Invasive Species Specialist Group, Auckland, New Zealand. www.issg.org/pdf/ publications/worst_100/english_100_worst.pdf. Ruggiero, M.A., Gordon, D.P., Orrell, T.M., Bailly, N., Bourgoin, T., Brusca, R.C., Cavalier- Smith, T., Guiry, M.D., Kirk, P.M., 2015. A higher level classiﬁcation of all living organ- isms. PloS ONE 10, e0119248. References Changes in the Euro-Atlantic ﬁsh species composition resulting from ﬁshing and ocean warming. Ital. J. Zool. 65 (S1), 493–499. https://doi.org/10.1080/ 11250009809386873. Kernan, M., 2015. Climate change and the impact of invasive species on aquatic ecosystems. Aquat. Ecosyst. Health 18, 321–333. https://doi.org/10.1080/14634988.2015.1027636. R Development Core Team, 2020. R: A Language and Environment for Statistical Computing. R Foundation for Statistical Computing, Vienna, Austria www.r-project.org/. Š Killi, N., Tarkan, A.S., Kozic, S., Copp, G.H., Davison, P.I., Vilizzi, L., 2020. Risk screening of the potential invasiveness of non native jellyﬁshes in the Mediterranean Sea. Mar. Pollut. Bull. 150, 110728. https://doi.org/10.1016/j.marpolbul.2019.110728. Radočaj, T., Špelić, I., Piria, M., Vilizzi, L., 2021. Identifying threats from introduced and translocated non-native freshwater ﬁshes in Croatia and Slovenia under current and future climatic conditions. Global Ecol. Conserv. 27, e01520. https://doi.org/ 10.1016/j.gecco.2021.e01520. Kirtman, B.P., Misra, V., Anandhi, A., Palko, D., Infanti, J., 2017. Future climate change scenar- ios for Florida. In: Chassignet, E.P., Jones, J.W., Misra, V., Obeysekera, J. (Eds.), Florida’s Climate: Changes, Variations, & Impacts. Florida Climate Institute, Gainesville, FL, pp. 533–555. Rahel, F.J., 2000. Homogenization of ﬁsh faunas across the United States. Science 288 (5467), 854–856. https://doi.org/10.1126/science.288.5467.854. Rahel, F.J., Olden, J.D., 2008. Assessing the effects of climate change on aquatic invasive spe- cies. Conserv. Biol. 22, 521–533. https://doi.org/10.1111/j.1523-1739.2008.00950.x. pp Komsta, L., Novomestky, F., 2015. Moments – moments, cumulants, skewness, kurtosis and related tests. R package version 0.14. http://CRAN.R-project.org/package= moments. Rehage, J.S., Blanchard, J.R., Boucek, R.E., Lorenz, J.J., Robinson, M., 2016. Knocking back in- vasions: variable resistance and resilience to multiple cold spells in native vs. nonna- tive ﬁshes. Ecosphere 7, e01268. https://doi.org/10.1002/ecs2.1268. Kug, J.-S., Jeong, J.-H., Jang, Y.-S., Kim, B.-M., Folland, C.K., Min, S.-K., Son, S.-W., 2015. Two distinct inﬂuences of Arctic warming on cold winters over North America and East Asia. Nat. Geosci. 8, 759–762. https://doi.org/10.1038/ngeo2517. Renaud, P.E., Berge, J., Varpe, Ø., Lønne, O.J., Nahrgang, J., Ottesen, C., Hallanger, I., 2012. Is the poleward expansion by Atlantic cod and haddock threatening native polar cod, Boreogadus saida? Polar Biol. 35, 401–412. https://doi.org/10.1007/s00300-011- 1085-z. Kumar, L., Kumari, K., Gogoi, P., Manna, R.K., Madayi, R.C., Salim, S.M., Muttanahalli Eregowda, V., Raghavan, S.V., Das, B.K., 2021. Risk analysis of non-native three-spot cichlid, Amphilophus trimaculatus, in the River Cauvery (India). Fisher. Manag. Ecol. https://doi.org/10.1111/fme.12467 in press. Rippingale, R.J., Kelly, S.J., 1995. Reproduction and survival of Phyllorhiza punctata (Cnidaria: Rhizostomeae) in a seasonally ﬂuctuating salinity regime in Western Australia. Mar. Freshw. Res. 46, 1145–1151. References Crivelli, A.J., 1995. Are ﬁsh introductions a threat to endemic freshwater ﬁshes in the northern Mediterranean region? Biol. Conserv. 72, 311–319. https://doi.org/10.1016/0006-3207 (94)00092-5. 14 Science of the Total Environment 788 (2021) 147868 L. Vilizzi, G.H. Copp, J.E. Hill et al. Hulme, P.E., 2009. Trade, transport and trouble: managing invasive species pathways in an era of globalization. J. Appl. Ecol. 46, 10–18. https://doi.org/10.1111/j.1365- 2664.2008.01600.x. Paganelli, D., Pandolﬁ, A., Sconﬁetti, R., Marchini, A., Vilizzi, L., 2018. Potential invasiveness by non-indigenous macrozoobenthos in the secondary hydrographic system of a temperate-climate river catchment. Ecol. Indic. 88, 274–281. https://doi.org/10.1016/ j.ecolind.2018.01.037. Interesova, E., Vilizzi, L., Copp, G.H., 2020. Risk screening of the potential invasiveness of non-native freshwater ﬁshes in the River Ob basin (West Siberian Plain, Russia). Reg. Environ. Change 20, 64. https://doi.org/10.1007/s10113-020-01644-3. Peel, M.C., Finlayson, B.L., McMahon, T.A., 2007. Updated world map of the Köppen-Geiger climate classiﬁcation. Hydrol. Earth Sys. Sci. 4, 439–473. www.hydrol-earth-syst-sci- discuss.net/4/439/2007/. IPCC, 2005. Guidance Notes for Lead Authors of the IPCC Fourth Assessment Report on Addressing Uncertainties. Intergovernmental Panel on Climate Change, WMO & UNEP www.ipcc.ch/site/assets/uploads/2018/02/ar4-uncertaintyguidancenote-1.pdf. Peeler, E.J., Murray, A.G., Thebault, A., Brun, E., Giovaninni, A., Thrush, M.A., 2007. The ap- plication of risk analysis in aquatic animal health management. Prev. Vet. Med. 81, 3–20. https://doi.org/10.1016/j.prevetmed.2007.04.012. IPCC, 2014. Climate change 2014: synthesis report. In: Core Writing Team, Pachauri, R.K., Meyer, L.A. (Eds.), Contribution of Working Groups I, II and III to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change. IPCC, Geneva, Switzerland https://www.ipcc.ch/site/assets/uploads/2018/05/SYR_AR5_FINAL_full_wcover.pdf. Pheloung, P.C., Williams, P.A., Halloy, S.R., 1999. A weed risk assessment model for use as a biosecurity tool evaluating plant introductions. J. Environ. Manag. 57, 239–251. https://doi.org/10.1006/jema.1999.0297. Kaiser, B.A., Burnett, K.M., 2010. Spatial economic analysis of early detection and rapid re- sponse strategies for an invasive species. Resour. Energy Econ. 32, 566–585. https:// doi.org/10.1016/j.reseneeco.2010.04.007. Powel, R., Conant, R., Collins, J., 1991. Peterson Field Guides: Reptiles and Amphibians of Eastern/Central North America. Houghton Mifﬂin Co., Boston, U.S.A. Puntila, R., Vilizzi, L., Lehtiniemi, M., Copp, G.H., 2013. First application of FISK, the fresh- water ﬁsh invasiveness screening kit, in Northern Europe: example of Southern Finland. Risk Anal. 33, 1397–1403. https://doi.org/10.1111/risa.12069. Katsanevakis, S., Gatto, F., Zenetos, A., Cardoso, A.C., 2013. How many marine aliens in Europe. Manag. Biol. Invasions 4, 37–42. https://doi.org/10.3391/mbi.2013.4.1.05. Kennedy, V.S., 1990. Anticipated effects of climate change on estuarine and coastal ﬁsher- ies. Fisheries 15, 16–24. https://doi.org/10.1577/1548-8446(1990)015<0016: AEOCCO>2.0.CO;2. Quero, J.C., 1998. References Association of Applied Biologists, pp. 49–54. Shackleton, R.T., Shackleton, C.M., Kull, C.A., 2019. The role of invasive alien species in shaping local livelihoods and human well-being: a review. J. Environ. Manag. 229, 145–157. https://doi.org/10.1016/j.jenvman.2018.05.007. O'Donnell, J., Gallagher, R.V., Wilson, P.D., Downey, P.O., Hughes, L., Leishman, M.R., 2012. In- vasion hotspots for non-native plants in Australia under current and future climates. Glob. Change Biol. 18, 617–629. https://doi.org/10.1111/j.1365-2486.2011.02537.x. Smith, C.S., Lonsdale, W.M., Fortune, J., 1999. When to ignore advice: invasion predictions and decision theory. Biol. Invasions 1, 89–96. https://doi.org/10.1023/A:1010091918466. 15 L. Vilizzi, G.H. Copp, J.E. Hill et al. Science of the Total Environment 788 (2021) 147868 Souty-Grosset, C., Holdich, D.M., Noël, P.Y., Reynolds, J.D., Haffner, P., 2006. Atlas of cray- ﬁsh in Europe. Collection Patrimoines Naturels. Vol. 64. Museum National d’Histoire Naturelle, Paris. Alien Species Group Available at:. www.wfduk.org/sites/default/ﬁles/UKTAG% 20classiﬁcation%20of%20alien%20species%20working%20paper%20v8.pdf. Alien Species Group Available at:. www.wfduk.org/sites/default/ﬁles/UKTAG% 20classiﬁcation%20of%20alien%20species%20working%20paper%20v8.pdf. Urban, M.C., Bocedi, G., Hendry, A.P., Mihoub, J.-B., Pe’er, G., Singer, A., Bridle, J.R., Crozier, L.G., De Meester, L., Godsoe, W., Gonzalez, A., Hellmann, J.J., Holt, R.D., Huth, A., Johst, K., Krug, C.B., Leadley, P.W., Palmer, S.C.F., Pantel, J.H., Schmitz, A., Zollner, P.A., Tarvis, J.M.J., 2016. Improving the forecast for biodiversity under climate change. Science 353, 6304. https://doi.org/10.1126/science.aad8466. Spalding, M.D., Fox, H.E., Allen, G.R., Davidson, N., Ferdaña, Z.A., Finlayson, M.A.X., Halpern, B.S., Jorge, M.A., Lombana, A.L., Lourie, S.A., Martin, K.D., 2007. Marine ecoregions of the world: a bioregionalization of coastal and shelf areas. BioScience 57, 573–583. https://doi.org/10.1641/B570707. p g Stasolla, G., Tricarico, E., Vilizzi, L., 2020. Risk screening of the potential invasiveness of non-native marine crustacean decapods and barnacles in the Mediterranean Sea. Hydrobiologia https://doi.org/10.1007/s10750-020-04432-6 in press. Uyan, U., Oh, C.-W., Tarkan, A.S., Top, N., Copp, G.H., Vilizzi, L., 2020. Risk screening of the potential invasiveness of non-native marine ﬁshes in South Korea. Mar. Pollut. Bull. 153, 111018. https://doi.org/10.1016/j.marpolbul.2020.111018. 153, 111018. https://doi.org/10.1016/j.marpolbul.2020.111018. Statzner, B., Bonada, N., Dolédec, S., 2008. Biological attributes discriminating invasive from native European stream macroinvertebrates. Biol. Invasions 10, 517–530. https://doi.org/10.1007/s10530-007-9148-3. VanDerWal, J., Murphy, H.T., Kutt, A.S., Perkins, G.C., Bateman, B.L., Perry, J.J., Reside, A.E., 2013. Focus on poleward shifts in species’ distribution underestimates the ﬁngerprint of cli- mate change. Nat. Clim. Chang. 3, 239–243. https://doi.org/10.1038/nclimate1688. Suresh, V.R., Ekka, A., Biswas, D.K., Sahu, S.K., Yousuf, A., Das, S., 2019. Vermiculated sailﬁn catﬁsh, Pterygoplichthys disjunctivus (Actinopterygii: Siluriformes: Loricariidae): inva- sion, biology, and initial impacts in east Kolkata Wetlands, India. Acta Ichthyol. Piscat. 49, 221–233. References https://doi.org/10.3750/AIEP/02551. Vilizzi, L., Copp, G.H., Adamovich, B., Almeida, D., Chan, J., Davison, P.I., Dembski, S., Ekmekçi, F.G., Ferincz, Á., Forneck, S.C., Hill, J.E., Kim, J.-E., Koutsikos, N., Leuven, R.S.E.W., Luna, S.A., Magalhães, F., Marr, S.M., Mendoza, R., Mourão, C.F., Neal, J.W., Onikura, N., Perdikaris, C., Piria, M., Poulet, N., Puntila, R., Range, I.L., Simonović, P., Ribeiro, F., Tarkan, A.S., Troca, D.F.A., Vardakas, L., Verreycken, H., Vintsek, L., Weyl, O.L.F., Yeo, D.C.J., Zeng, Y., 2019. A global review and meta-analysis of applications of the freshwater Fish Invasiveness Screening Kit. Rev. Fish Biol. Fisher. 29, 529–568. https://doi.org/10.1007/s11160-019-09562-2. p g Tarkan, A.S., Gaygusuz, Ö., Gürsoy-Gaygusuz, Ç., Saç, G., Copp, G.H., 2012. Circumstantial evidence of gibel carp Carassius gibelio reproductive interference exerted on native ﬁsh species in a mesotrophic reservoir. Fish. Manag. Ecol. 19, 167–177. https://doi. org/10.1111/j.1365-2400.2011.00839.x. Tarkan, A.S., Vilizzi, L., Top, N., Ekmekçi, F.G., Stebbing, P.D., Copp, G.H., 2017a. Identiﬁca- tion of potentially invasive freshwater ﬁshes, including translocated species, in Turkey using the Aquatic Species Invasiveness Screening Kit (AS-ISK). Int. Rev. Hydrobiol. 102, 47–56. https://doi.org/10.1002/iroh.201601877. İ Wei, H., Chaichana, R., Vilizzi, L., Daengchana, P., Liu, F., Nimtim, M., Zhu, Y., Li, S., Hu, Y., Copp, G.H., 2021. Do non-native ornamental ﬁshes pose a similar level of invasion risk in neighbouring regions of similar current and future climate? Aquat. Conserv. https://doi.org/10.1002/aqc.3609 in press. Hydrobiol. 102, 47–56. https://doi.org/10.1002/iroh.201601877. İ Wilson, M.A., Carpenter, S.R., 1999. Economic valuation of freshwater ecosystem services in the United States: 1971–1997. Ecol. Appl. 9, 772–783. https://doi.org/10.1890/ 1051-0761(1999)009[0772:EVOFES]2.0.CO;2. Tarkan, A.S., Sarı, H.M., İlhan, A., Kurtul, I., Vilizzi, L., 2017b. Risk screening of non-native and translocated freshwater ﬁsh species in a Mediterranean-type shallow lake: Lake Marmara (West Anatolia). Zool. Middle East 63, 48–57. https://doi.org/ 10.1080/09397140.2017.1269398. Zięba, G., Vilizzi, L., Copp, G.H., 2020. How likely is Lepomis gibbosus to become invasive in Poland under conditions of climate warming? Acta Ichthyol. Piscat. 50, 37–51. https://doi.org/10.3750/AIEP/02390. UK-TAG ASG, 2021. Alien species classiﬁcation according to level of impact - revised list (29/01/2021). UK Technical Advisory Group on the Water Framework Directive. 16 16
https://openalex.org/W3128405472	http://jurnal.abulyatama.ac.id/index.php/tilapia/article/download/1577/pdf	Indonesian	null	Pengaruh Penggunaan Berbagai Jenis Shelter Terhadap Kelangsungan Hidup Dan Pertumbuhan Benih Ikan Kakap Putih (Lates Calcalifer)	JURNAL TILAPIA	2,021	cc-by-sa	3,208	▪ Jurnal TILAPIA, Vol. 2, No. 1, Januari 2021 : 29-35 ▪ Jurnal TILAPIA, Vol. 2, No. 1, Januari 2021 : 29-35 ISSN 2721-592X (Online) Available online at www.jurnal.abulyatama.ac.id/tilapia ISSN 2721-592X (Online) Universitas Abulyatama Jurnal TILAPIA (Ilmu Perikanan dan Perairan) Available online at www.jurnal.abulyatama.ac.id/tilapia ISSN 2721-592X (Online) Universitas Abulyatama Jurnal TILAPIA (Ilmu Perikanan dan Perairan) Pengaruh Penggunaan Berbagai Jenis.... (Nazlia, Munandar, Thaib, & Ridwan, 2021) Pengaruh Penggunaan Berbagai Jenis Shelter Terhadap Kelangsungan Hidup Dan Pertumbuhan Benih Ikan Kakap Putih (Lates Calcalifer) Suraiya Nazlia1, T Haris Munandar1, Azwar Thaib1, T Ridwan2 1Budidaya Perairan Fakultas Perikanan, Universitas Abulyatama, Aceh Besar, 23372, Indonesia. 2Balai Pembenihan Budidaya Air Payau Ujung Batee, Aceh Besar 23381, Indonesia. Email korespondensi: suraiyanazlia_psp@abulyatama.ac.id Diterima 29 Desember 2020; Disetujui 20 Januari 2021; Dipublikasi 29 Januari 2021 Abstract: White snapper (Lates calcalifer) is a type of fish that prey on the same sex or known as cannibalism. Therefore, it is necessary to control these properties. One way is to install protective media in the maintenance container. The research has been conducted at BPBAP Ujung Batee, Aceh Besar District, Aceh Province. This study aims to determine the effect of using various types of shelter on the survival and growth of white snapper fry. The length of the fish at the beginning of stocking was ± 3.5 cm, and the weight of the fish at the beginning of stocking ranged from ± 0.8 grams. The total number of fish studied was 480, the number of containers used was 12 units. For feeding, it was given 3 times a day, namely in the morning, afternoon and evening. The experimental design used in this study was a completely randomized design with 4 treatments and 3 replications with the observation parameters being survival rate, absolute length growth and absolute weight growth. The results showed that the use of various types of shelter had a very significant effect on the survival rate, and had a significant effect on the absolute length growth and absolute weight growth of the white snapper fry with the best results in treatment B with successive values of 88%, 3.7 cm, and 4.5 g. The conclusion of this research is that the best shelter to reduce cannibalism and mortality is the Paralon pipe shelter, which helps increase the survival and growth of white snapper fry. Key words: snapper (Lates calcalifer) seed, survival, weight growth, length growth, Shelter. Waktu dan Tempat Penelitian ini dilaksanakan pada bulan Agustus 2018 di Balai Perikanan Budidaya Air Payau (BPBAP) Ujung Batee Jalan Laksamana malahayati Km 16 Durung, Kecamatan Mesjid Raya, Kabupaten Aceh Besar, Provinsi Aceh. Alat dan Bahan Alat dan bahan yang digunakan dalam penelitian ini adalah bak fiber berbentuk persegi panjang, pipa paralon, jaring, baskom, thermometer, pH meter, DO meter, refraktometer, timbangan digital, instalasi aerasi, alat sipon, penggaris, scoop net, kamera, benih ikan kakap putih, dan pakan komersil dengan kandungan protein 48%. Salah satu upaya untuk menghindari kanibalisme adalah dengan pemasangan alat pelindung (shelter) di dalam wadah pemeliharaan benih ikan kakap putih. Penggunaan shelter tersebut diharapkan dapat memberikan tempat yang aman juga menekan angka kematian sehingga dapat meningkatkan kelangungan hidup serta pertumbuhan ikan. Shelter tersebut dapat berupa pipa paralon, jaring, bambu, kelapa kering dan tanaman air (Sofiandi, 2002). Berdasarkan uraian di atas maka perlu dilakukan penelitian tentang pengaruh penggunaan berbagai jenis shelter terhadap kelangsungan hidup dan pertumbuhan benih ikan kakap putih. METODE PENELITIAN Ikan kakap putih (Lates calcarifer) merupakan jenis ikan karnivora (ICRF, 2004), dan bersifat kanibal yang merupakan sifat alami ikan tersebut (Prihaningrum et al., 2015). Umumnya sifat kanibal ditemukan pada umur atau ukuran benih, dan mulai berkurang setelah benih mencapai ukuran 7-10 cm (Putro et al., 2004). Menurut Nybakken (2001), ikan kakap putih cenderung menangkap mangsa yang aktif bergerak di kolom air, dan menyergap mangsa tersebut dari tempat persembunyiannya (Maryati, 2004). Hal tersebut di atas mengungkapkan bahwa dalam kondisi lapar dan tanpa adanya media pelindung (shelter) mengakibatkan terjadinya kanibalisme (Rouse, 1997). Kata kunci : benih ikan kakap (Lates calcalifer), kelangsungan hidup, pertumbuhan berat, pertumbuhan panjang. Shelter. Kata kunci : benih ikan kakap (Lates calcalifer), kelangsungan hidup, pertumbuhan berat, pertumbuhan panjang. Shelter. Kata kunci : benih ikan kakap (Lates calcalifer), kelangsungan hidup, pertumbuhan berat, pertumbuhan panjang. Shelter. Key words: snapper (Lates calcalifer) seed, survival, weight growth, length growth, Shelter. Abstrak: Ikan kakap putih (Lates calcalifer) termasuk ikan yang memangsa antar sesama jenis atau disebut kanibalisme. Oleh karena itu, perlu dilakukan pengendalian terhadap sifat tersebut. Salah satu cara yaitu pemasangan media pelindung di dalam wadah pemeliharaan. Penelitian telah dilaksanakan di BPBAP Ujung Batee, Kabupaten Aceh Besar, Provinsi Aceh. Penelitian ini bertujuan untuk mengetahui pengaruh penggunaan berbagai jenis shelter terhadap kelangsungan hidup dan pertumbuhan benih ikan kakap putih. Ukuran panjang ikan pada awal penebaran adalah ± 3,5 cm, dan berat ikan pada awal penebaran berkisar antara ±0,8 gram. Jumlah keseluruhan ikan yang diteliti berjumlah 480 ekor, jumlah wadah yang digunakan 12 unit. Untuk pemberian pakan diberikan sebanyak 3 kali sehari, yaitu pada pagi, siang dan sore hari. Rancangan percobaan yang digunakan pada penelitian ini adalah Rancangan Acak Lengkap dengan 4 perlakuan dan 3 kali ulangan dengan parameter pengamatan adalah tingkat kelangsungan hidup, pertumbuhan panjang mutlak dan pertumbuhan berat mutlak. Hasil penelitian menunjukkan bahwa penggunaan berbagai jenis shelter berpengaruh sangat nyata terhadap tingkat kelangsungan hidup, dan berpengaruh nyata terhadap pertumbuhan panjang mutlak serta pertumbuhan berat mutlak benih ikan kakap putih dengan hasil terbaik pada perlakuan B dengan nilai berturut-turut 88%, 3,7 cm, dan 4,5 g. - 29 - Jurnal TILAPIA, Vol.2, No. 1, Januari 2021 : 29-35 http://jurnal.abulyatama.ac.id/tilapia Kesimpulan dari penelitin ini adalah shelter yang terbaik untuk menekan angka kanibalisme dan mortalitas adalah shelter pipa paralon sehingga membantu meningkatkan kelangsungan hidup dan pertumbuhan benih ikan kakap putih. Parameter Kualitas Air Pengamatan parameter kualitas air selama penelitian meliputi oksigen terlarut, pH, salinitas, dan suhu. A : Kontrol B : Pipa Paralon C : Jaring Pertumbuhan Berat Mutlak Pertumbuhan berat benih ikan kakap dapat dihitung berdasarkan rumus Effendi (1997), yaitu: Langkah selanjutnya pemeliharaan ikan uji dengan pemberian pakan, jenis pakan yang diberikan berupa pelet dengan frekuensi pemberian pakan sebanyak 3 kali sehari yaitu pada saat pagi, siang dan sore hari. Dosis pakan yang diberikan sebesar 5%. W = Wt – W0 Keterangan: Keterangan: W = Pertumbuhan berat W = Pertumbuhan berat Wt = Pertumbuhan berat pada akhir penelitian (g) Rancangan penelitian yang digunakan pada penelitian ini adalah Rancangan Acak Lengkap (RAL) non faktorial yang terdiri dari 4 perlakuan dan 3 kali ulangan. Perlakuan yang digunakan adalah sebagai berikut: W0 = Pertumbuhan berat pada awal penelitian (g) Prosedur Penelitian Langkah awal yang dilakukan adalah persiapan bak fiber berbentuk persegi panjang dengan volume 50liter air sebagai wadah penelitian. Total wadah yang digunakan adalah 12 unit. Kemudian penyiapan air laut serta pengisian kedalam media pemeliharaan tersebut. Selanjutnya penggunaan berbagai jenis shelter ke dalam wadah pemeliharaan benih ikan kakap putih dengan 4 perlakuan. Kemudian dilanjutkan dengan pemasangan kran dan selang aerasi serta pengaktifan aerasi. Dan setelah itu dilakukan penimbangan yaitu pengukuran berat dan pengukuran panjang awal benih ikan kakap putih (Lates calcalifer) sebelumnya ikan telah di - 30 - ISSN 2721-592X (Online) L0 = Panjang rata-rata pada awal penelitian (cm) adaptasikan, baru dilakukan penebaran kedalam media penelitian. Dengan penebaran sebanyak 40 ekor dalam satu wadah dan jumlah semua benih yang ditebar sebanyak 480 ekor. (Nazlia, Munandar, Thaib, & Ridwan, 2021) Hasil SR = Kelangsungan hidup benih ikan kakap putih Hasil pengamatan penggunaan berbagai jenis shelter pada benih ikan kakap putih dengan variabel yang diamati yaitu kelangsungan hidup, pertumbuhan panjang mutlak dan pertumbuhan berat mutlak selama 28 hari pemeliharaan, disajikan sebagai berikut. Nt = Jumlah ikan pada akhir penelitian Nt = Jumlah ikan pada akhir penelitian N0 = Jumlah ikan pada awal penelitian N0 = Jumlah ikan pada awal penelitian Kelangsungan Hidup Kelangsungan hidup dapat dihitung berdasarkan Effendi (1997) dengan rumus: SR = Nt N0 x 100% Analisis Data D : Pipa Paralon dan Jaring Data yang diperoleh dari hasil penelitian dianalisis secara statistik menggunakan analisis sidik ragam (ANOVA) pada selang kepercayaan 95%, untuk mengetahui pengaruh penggunaan berbagai jenis shelter. Kemudian dilakukan pengujian lanjut menggunakan Uji Beda Nyata Jujur (BNJ). Pengaruh Penggunaan Berbagai Jenis.... HASIL DAN PEMBAHASAN Keterangan: Kelangsungan Hidup Hasil perhitungan rata-rata tingkat kelangsungan hidup benih ikan kakap putih selama pemeliharaan pada berbagai jenis shelter dapat dilihat pada Tabel 1. Hasil penelitian pada Tabel di atas menunjukkan bahwa rata-rata tingkat kelangsungan hidup tertinggi selama masa pemeliharaan terdapat pada penggunaan shelter pipa paralon (perlakuan B) yaitu 88% diikuti shelter pipa paralon dan jaring (perlakuan D) yaitu 82% dan shelter jaring (perlakuan C) yaitu 70% serta yang terendah terdapat pada kontol atau tanpa penggunaan shelter (perlakuan A). Hasil analisis sidik ragam menunjukkan perbedaan yang sangat nyata (P<0,05). Hasil dilanjutkan dengan uji BNJ, menunjukkan bahwa perlakuan A berbeda nyata terhadap perlakuan B dan perlakuan D, namun tidak berbeda nyata terhadap perlakuan C. Pertumbuhan Panjang Mutlak Pertumbuhan panjang mutlak ikan uji dihitung dengan rumus Effendi (1997): L = Lt – L0 Keterangan: L = Pertumbuhan panjang mutlak (cm) L = Pertumbuhan panjang mutlak (cm) Lt = Panjang rata-rata pada akhir penelitian (cm) - 31 - Jurnal TILAPIA, Vol.2, No. 1, Januari 2021 : 29-35 http://jurnal.abulyatama.ac.id/tilapia Tabel 1. Tingkat Kelangsungan Hidup, Pertumbuhan Panjang Mutlak, dan Pertumbuhan Berat Mutlak Parameter Perlakuan A B C D SR (%) 63±2.89a 88±1.44b 70±7.50a 82±5.20b PPM (cm) 3.5±0.00a 3.7±0.06b 3.6±0.10ab 3.6±0.10ab PBM (g) 4.1±0.26a 4.5±0.09b 4.3±0.05ab 4.4±0.13ab Kelangsungan Hidup Hasil perhitungan rata-rata tingkat kelangsungan hidup benih ikan kakap putih selama pemeliharaan pada berbagai jenis shelter dapat dilihat pada Tabel 1. Hasil penelitian pada Tabel di atas menunjukkan bahwa rata-rata tingkat kelangsungan hidup tertinggi selama masa pemeliharaan terdapat pada penggunaan shelter pipa paralon (perlakuan B) yaitu 88% diikuti shelter pipa paralon dan jaring (perlakuan D) yaitu 82% dan shelter jaring (perlakuan C) yaitu 70% serta yang terendah terdapat pada kontol atau tanpa penggunaan shelter (perlakuan A). Hasil analisis sidik ragam menunjukkan perbedaan yang sangat nyata (P<0,05). Hasil dilanjutkan dengan uji BNJ, menunjukkan bahwa perlakuan A berbeda nyata terhadap perlakuan B dan perlakuan D, namun tidak berbeda nyata terhadap perlakuan C. Pertumbuhan Panjang Mutlak Hasil perhitungan pertumbuhan panjang mutlak benih ikan kakap putih selama pemeliharaan pada penggunaan berbagai jenis shelter dapat dilihat pada Tabel 1. Hasil penelitian yang terlihat pada Tabel 1 Tabel 1. Tingkat Kelangsungan Hidup, Pertumbuhan Panjang Mutlak, dan Pertumbuhan Berat Mutlak keseluruhan dari pertumbuhan panjang mutlak adalah berbeda nyata (P<0,05). Akan tetapi, pada uji lanjut hanya pada perlakuan A dan perlakuan B yang menunjukkan perbedaan pada pertumbuhan panjang mutlak. Kualitas Air Parameter kualitas air yang diamati selama pemeliharaan meliputi: pH, salinitas, DO, suhu dan amoniak dapat dilihat pada Tabel 2. Pertumbuhan Panjang Mutlak Tabel 2. Hasil parameter kualitas air ikan kakap putih selama penelitian No. Parameter kualitas air Kisaran 1. Tingkat keasaman (pH) 7,0 - 8,5 2. Salinitas 28 - 32 ppt 3. Oksigen terlarut (DO) 5,44 - 6,54 mg/l 4. Suhu 29 - 30 ºC 5. Amoniak 0,207 - 0,282 mg/l P b h selama penelitian No. Parameter kualitas air Kisaran 1. Tingkat keasaman (pH) 7,0 - 8,5 2. Salinitas 28 - 32 ppt 3. Oksigen terlarut (DO) 5,44 - 6,54 mg/l 4. Suhu 29 - 30 ºC 5. Amoniak 0,207 - 0,282 mg/l Pembahasan Hasil perhitungan pertumbuhan panjang mutlak benih ikan kakap putih selama pemeliharaan pada penggunaan berbagai jenis shelter dapat dilihat pada Tabel 1. Hasil penelitian yang terlihat pada Tabel 1 menunjukkan bahwa rata-rata pertumbuhan panjang mutlak tertinggi selama pemeliharaan terdapat pada perlakuan B (pipa paralon) yaitu 3,7 cm diikuti oleh perlakuan C (jaring) dan D (pipa paralon dan jaring) yaitu keduanya bernilai 3,6 cm, serta hasil terendah terdapat pada perlakuan A (kontrol) yaitu 3,5 cm. Hasil analisis sidik ragam menunjukkan bahwa Pertumbuhan Berat Mutlak Pertumbuhan berat mutlak rata-rata benih ikan kakap putih selama pemeliharaan dapat dilihat pada Tabel 1. Tabel di atas menunjukkan bahwa rata-rata pertumbuhan berat mutlak tertinggi selama masa pemeliharaan terdapat pada perlakuan B (pipa paralon) yaitu 4,5 g diikuti oleh perlakuan D (pipa paralon dan jaring) dan perlakuan C (jaring) yaitu masing-masing 4,4 g dan 4,3 g, serta hasil terendah terdapat pada perlakuan A (kontro) yaitu 4,1 g. Hasil analisis sidik ragam menunjukkan adanya perbedaan (P<0,05), namun pada uji lanjut hanya pada perlakuan A dan perlakuan B yang menunjukkan perbedaan dari keseluruhan unit percobaan. Pembahasan Menurut Boer (2000), tingkat kelangsungan hidup memiliki keterkaitan erat dengan mortalitas yang merupakan kematian pada suatu populasi - 32 - ISSN 2721-592X (Online) organisme yang mengakibatkan jumlahnya berkurang. Diantara penyebabnya adalah kanibalisme. Menurut Satwika (2014), untuk menekan tingkat kanibalisme atau saling serang, maka shelter adalah tempat perlindungan atau persembunyian biota pada masa pemeliharaan ikan (Effendi, 2002). Hasil uji lanjut menunjukkan bahwa perlakuan B (pipa paralon) dan perlakuan A (kontrol) memberikan perbedaan nyata terhadap pertumbuhan panjang mutlak dan pertumbuhan berat mutlak benih ikan kakap putih dalam keseluruhan unit percobaan. Ikan kakap putih merupakan ikan predator yang buas (kordi, 2011). Menurut (Rayes, 2013), pertumbuhan ikan dapat terjadi apabila energi yang disimpan lebih besar dibandingkan dengan energi yang digunakan untuk aktivitas tubuh. Sehingga diduga penggunaan shelter pipa paralon (perlakuan B) mampu mengontrol persaingan selama masa pemeliharaan seperti mudah menghindari kejaran ikan yang predator dan mampu mendapatkan makanan dengan baik. Rahmawati et al., (2015) mengemukakan bahwa penggunaan potongan pipa paralon pada ikan nokturnal dapat berfungsi sebagai tempat berlindung dan manipulasi penetrasi cahaya yang secara tidak langsung mempengaruhi aktifitas makan sehingga berdampak pada meningkatnya pertumbuhan. Berdasarkan hasil penelitian bahwa penggunaan berbagai jenis shelter menghasilkan tingkat kelangsungan hidup benih ikan kakap putih berada diatas 60%. Perlakuan berbagai jenis shelter tersebut memberikan pengaruh yang sangat nyata (P<.0,05) terhadap tingkat kelangsungan hidup ikan kakap putih. Hasil uji lanjut menunjukkan bahwa perlakuan B (pipa paralon) memberikan pengaruh yang sangat nyata dari perlakuan A (kontrol). Hal ini memperlihatkan bahwa penggunaan berbagai jenis shelter dapat mengurangi tingkat kanibalisme pada benih ikan kakap putih. Menurut (Sahputra, Khalil, & Zulfikar, 2017), tingginya tingkat kelangsungan hidup disebabkan oleh faktor lingkungan hidup baik internal maupun eksternal, serta tidak terjadinya kompetisi antara benih ikan kakap putih. Selain itu, untuk memperoleh tingkat kelangsungan hidup ikan yang baik dan tinggi maka perlu diperhatikan ketersediaan makanan yang cukup, dan penanganan yang baik serta didukung oleh kesesuaian padat tebar dan jumlah ikan yang ditebar (Royce, 1972 dan Krebs, 1972). Menurut Effendi (1979) selain dari faktor-faktor tersebut nilai kualitas air seperti oksigen terlarut, karbodioksida, nitrat dan lain-lain juga mempengaruhi kelangsungan hidup ikan. (Nazlia, Munandar, Thaib, & Ridwan, 2021) Kesimpulan Dari hasil penelitian dapat disimpulkan bahwa penggunaan berbagai jenis shelter sebagai media pelindung menunjukkan perbedaan yang sangat nyata terhadap kelangsungan hidup. Dan perbedaan yang nyata terhadap pertumbuhan panjang mutlak serta pertumbuhan bobot mutlak. Shelter yang terbaik untuk kelangsungan hidup dan pertumbuhan terdapat pada perlakuan B (pipa paralon) sehingga dapat menekan angka kanibalisme dan mortalitas benih ikan kakap putih (Lates calcarifer). Nyabakken. (2001). Seasona variation in meat and liver histopathology of white snapper (Lates calcalifer, Bloch) di Karamba Jaring Apung. Balai Besar Perikanan Budidaya Laut, Lampung: 66 hlm. Nyabakken. (2001). Seasona variation in meat and liver histopathology of white snapper (Lates calcalifer, Bloch) di Karamba Jaring Apung. Balai Besar Perikanan Budidaya Laut, Lampung: 66 hlm. Prihaningrum, A., Aditya, T. W., Saputra, Y. (2015). Petunjuk Teknis Budidaya Ikan Kakap Putih (Lates calcalifer, Bloch) Di Keramba Jaring Apung. Balai Besar Perikanan Budidaya Laut, Lampung: 66 hlm. Prihaningrum, A., Aditya, T. W., Saputra, Y. (2015). Petunjuk Teknis Budidaya Ikan Kakap Putih (Lates calcalifer, Bloch) Di Keramba Jaring Apung. Balai Besar Perikanan Budidaya Laut, Lampung: 66 hlm. KESIMPULAN DAN SARAN Maryati. (2004). Kajian respons kekebalan tubuh dan performa pertumbuhan ikan kakap putih Lates calcarifer Bloch melalui suplementasi protein hidrolisis pada pakan Immune response and growth performance of Asian sea bass Lates calcarifer. DAFTAR PUSTAKA Boyd, C. E. (1982). Water quality management for pond fish culture. Elsevier Scientific Publishing Co. Boyd, C.E. 1982. Water Quality Management in Aquaculture and Fisheries Science. Else- vier Scientific Publishing Company. Amsterdam. 3125p. Boyd, C. 1998. Water Quality for Pond Aquaculture. Anburn University. Alabama. Effendie, M. I. 1979. Biologi Perikanan. Yayasan Pustaka Nusantara. Yogyakarta. 163 hal. Effendie, M. I. 2002. Biologi Perikanan. Yogyakarta. Yayasan Pustaka Nusantara. Halaman:5. Kordi, 2011. Budidaya Ikan Laut. Rineka Cipta, Jakarta Kualitas Air Menurut Boyd (1998), kualitas air terdiri dari keseluruhan faktor fisika, kimai, dan biologi yang akan mempengaruhi ketahanan hidup, reproduksi, pertumbuhan, dan lain sebainya. Oleh karena itu penting untuk memperhatikan aspek kualitas air. Parameter yang diamati pada penelitian ini adalah pH, salinitas, DO, suhu dan amoniak. Hasil pengukuran pH air selama penelitian berkisar 7,0-8,5. Nilai ini masih dalam batas toleransi untuk budidaya ikan kakap putih. Menurut SNI (1999), ikan kakap putih hidup pada pH 7,0-13. Suhu selama penelitian berkisar 29oC- 30oC. Kisaran salinitas selama pemeliharaan 28-32%o yang menurut Pengaruh Penggunaan Berbagai Jenis.... Pertumbuhan merupakan penambahan ukuran, Panjang, atau bobot ikan dalam suatu periode waktu tertentu yang dipengaruhi oleh pakan, jumlah ikan yang mengkonsumsi pakan, suhu, umur dan ukuran - 33 - (Nazlia, Munandar, Thaib, & Ridwan, 2021) Jurnal TILAPIA, Vol.2, No. 1, Januari 2021 : 29-35 http://jurnal.abulyatama.ac.id/tilapia SNI (2014) berkisar 28-33 ppt. Kisaran suhu optimal bagi kehidupan dan pertumbuhan ikan kakap putih adalah 23 o-32 oC. Pada kisaran tersebut konsumsi oksigen berkisar 5,44-6,54 mg/l dengan kebutuhan oksigen terlarut ikan kkap putih dewasa adalah ≥4 ppm (Boyd,1982). Selanjutnya, pengukuran amoniak selama penelitian berkisar antara 0,207 - 0,282 mg/l. Menurut Svodova et al., (1993), toleransi amoniak pada benih ikan kakap putih pada konsentrasi 0,05- 0,13 mg/l. Menurut Boyd (1982), kenaikan amoniak dapat meningkatkan konsumsi oksigen di jaringan, merusak insang dan mengurangi kemampuan darah mengangkut oksigen. Amoniak berasal dari sisa pakan dan feses yang dihasilkan oleh ikan itu sendiri dan bahan organic lainnya. Selain itu, kondisi pH dan suhu sangat berpengaruh terhadap amoniak. Saran Perlu dilakukan penelitian lebih lanjut tentang tata letak dan tingkatan shelter yang baik untuk menekan kanibalisme pada benih ikan kakap putih (Lates calcalifer). Putro, R., Pratiwi, R., & Astuti, N. W. W. (2004). Pengaruh Perbedaan Waktu Awal Putro, R., Pratiwi, R., & Astuti, N. W. W. (2004). Pengaruh Perbedaan Waktu Awal - 34 - ISSN 2721-592X (Online) Pemberiaan Pakan Buatan terhadap Pertumbuhan dan Aktivitas Enzim Pencernaan Larva Ikan Kerapu Bebek Cromileptes altivelis. Jurnal Biologii Indonesia, 8(2). Ikan Kakap Putih (Lates calcarifer, Bloch). SNI : 01 - 6147 – 1999. Badan Standar Nasional. Jakarta. Sofiandi, A. 2002. Pengaruh Perbedaan Shelter Terhadap Pertumbuhan dan Kelangsungan Hidup Udang Galah (Macrobachium rosenbergii de Man). Program Studi Budidaya Perairan. Fakultas Perikanan dan Ilmu Kelautan. Institut Pertanian Bogor. Rayes, R. D. (2013). Pengaruh Perubahan Salinitas Terhadap Pertumbuhan Dan Sintasan Ikan Kakap Putih (Lates Calcarifer Bloch). Jurnal KELAUTAN. 6 (1): 47 - 56 Svobodova., Richarf L., Jana M., and Blanka v. 1993. Water qualitu and fish health. EIFAC Technical paper 54. Rouse, D. B.,1997. Production of Australian Red Claw Crayfish. Auburn University. Alabama. USA. Royce, W.F., 1972. Introduction to the Practice of Fishery Science. XI. Academic press Inc. New York San Fransisco. London 428.pp Sahputra, I., Khalil, M., & Zulfikar, Z. (2017). Pemberian jenis pakan yang berbeda terhadap pertumbuhan dan kelangsungan hidup benih ikan kakap putih (Lates calcalifer, Bloch). Acta Aquatica: Aquatic Sciences Journal, 4(2), 65. https://doi.org/10.29103/aa.v4i2.305 Satwika, H. 2014. Pengaruh Penggunaan Shelter Berbeda Terhadap Tingkat Kelangsungan Hidup Dan Pertumbuhan Lobster Pasir (Panulirus homarus) Pada Kegiatan Pendederan Secara Indoor. (Skripsi). Depaetemen Budidaya Perairan Fakultas Perikanan dan Ilmu Kelautan. Institut Pertanian Bogor. Bogor. SNI. 2014. Standar Nasional Indonesia. Benih Ikan Kakap Putih (Lates calcarifer, Bloch). SNI : 6145.3 : 2014). Badan Standar Nasional. Jakarta. Pengaruh Penggunaan Berbagai Jenis.... (Nazlia, Munandar, Thaib, & Ridwan, 2021) - SNI. 1999. Standar Nasional Indonesia. Benih - 35 -
https://openalex.org/W3194197920	https://journals.plos.org/plosone/article/file?id=10.1371/journal.pone.0256916&type=printable	English	null	Gender effects on outcomes of psychosomatic rehabilitation are reduced	PloS one	2,021	cc-by	7,307	PLOS ONE RESEARCH ARTICLE Conclusion Overall, the results show a gender-fair effectiveness of the rehabilitation. The new findings could be explained by changes in living conditions, gender roles, or better treatment methods. Copyright: © 2021 Burghardt et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited. Results Received: March 17, 2021 Accepted: August 19, 2021 Published: August 27, 2021 Gender-specific differences in the course of treatment were all too small to be clinically rele- vant. The differences were smallest in the middle-aged cohort. However, at the time of admission, women reported a slightly higher symptom burden. Peer Review History: PLOS recognizes the benefits of transparency in the peer review process; therefore, we enable the publication of all of the content of peer review and author responses alongside final, published articles. The editorial history of this article is available here: https://doi.org/10.1371/journal.pone.0256916 Editor: Stephan Doering, Medical University of Vienna, AUSTRIA Editor: Stephan Doering, Medical University of Vienna, AUSTRIA Editor: Stephan Doering, Medical University of Vienna, AUSTRIA OPEN ACCESS Citation: Burghardt J, Riffer F, Sprung M (2021) Gender effects on outcomes of psychosomatic rehabilitation are reduced. PLoS ONE 16(8): e0256916. https://doi.org/10.1371/journal. pone.0256916 Juliane BurghardtID1, Friedrich Riffer2,3, Manuel Sprung1,2,3 1 Division of Clinical Psychology, Department Psychology and Psychodynamics, Karl Landsteiner University of Health Science, Krems an der Donau, Austria, 2 University Hospital for Psychosomatic Medicine Eggenburg, Psychosomatisches Zentrum Waldviertel, Eggenburg, Austria, 3 Psychiatric Rehabilitation Clinic Gars am Kamp, Psychosomatisches Zentrum Waldviertel, Gars am Kamp, Austria Juliane.Burghardt@kl.ac.at * Juliane.Burghardt@kl.ac.at * Juliane.Burghardt@kl.ac.at a1111111111 a1111111111 a1111111111 a1111111111 a1111111111 Objective The study examined whether psychiatric/psychosomatic rehabilitation continues to have a better course of treatment for women than men. Gender effects on outcomes of psychosomatic rehabilitation are reduced Juliane BurghardtID1, Friedrich Riffer2,3, Manuel Sprung1,2,3 Methods We compared the course of global symptom severity, health-related quality of life and func- tioning between admission and discharge in patients (848 men, 1412 women) at an Austrian psychiatric/psychosomatic rehabilitation clinic. PLOS ONE PLOS ONE Introduction In line with the male gender role, men prefer a higher level of con- trol over the therapeutic process and report a need for action-oriented problem-solving strate- gies [12]. Accordingly, therapy and rehabilitation should be implemented in a gender-sensitive manner [12,14]. Notwithstanding these recommendations, a study of psychiatric/psychothera- peutic treatments in 2007 concluded that interventions are largely gender-neutral [15]. G d diff t l id t i th f t t t d diff Forschungs- und Bildungsges.m.b.H.; NFB) supported this research by financing the endowed professorship in clinical psychology held by MS. The Open Access Publishing Fund of Karl Landsteiner University of Health Sciences, Krems, Austria supports the publication of this manuscript. The VAMED Institute for Gender Medicine financed the position held by JB. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript. Competing interests: Prof. Dr. Manuel Sprung is the scientific director of the University Clinic for Psychosomatic Medicine Eggenburg, PSZW (Eggenburg and Gars). Chief physician Dr. Riffer is the medical director of the PSZW. This does not alter our adherence to PLOS ONE policies on sharing data and materials. Gender differences are not only evident in the course of treatment; women and men differ in the severity of symptoms [16], as well as in the frequency [9,17], and in the course of mental disorders [18]. The greater treatment effects in women could therefore be explained by the greater burden of symptoms upon admission [4]. In addition, the living conditions and thus the resources of men and women also differ, e.g. characterized by lower employment of women and a higher burden of childcare [19,20]. Gender roles have been changing signifi- cantly for years. As a result, women increasingly describe themselves as more masculine and less feminine than before [21,22]. The living conditions of men and women have also become more and more equal in the last few decades due to the increase in the employment of women and rising incomes [23]. As a consequence of the changed gender roles and the change in liv- ing conditions, the question must be asked whether the gender differences can still be repli- cated. Do women continue to benefit more from psychosomatic rehabilitation programs than men? To answer this question, the present study analyzes gender-specific differences in the effects of inpatient psychosomatic/psychiatric rehabilitation in an Austrian rehabilitation clinic. Introduction Psychosomatic rehabilitation focusses on the psychotherapeutic treatment of mental disorders in a stationary context with the goal to reduce symptoms and increase functioning and quality of life. Many studies have shown that psychosomatic rehabilitation shows better treatment effects for women than for men [1–3]. This treatment advantage for women is not limited to psychosomatic rehabilitation interventions, but is reflected in a number of rehabilitation set- tings (see [4]) e.g. for chronic back pain [5] or the treatment of cardiovascular diseases [6]. The Data Availability Statement: All relevant data are within the manuscript and its Supporting information files. Funding: The Lower Austria research and education company (NFB) (Niedero¨sterreichische 1 / 12 PLOS ONE \| https://doi.org/10.1371/journal.pone.0256916 August 27, 2021 PLOS ONE Gender effects on psychosomatic rehabilitation outcomes are reduced meta-analytical MESTA study confirmed the better treatment effects for women in psychoso- matic rehabilitation interventions [3]: An increase in the proportion of women in the patient sample was accompanied by an improvement in the treatment outcome between admission and discharge (β = .22). Later studies also replicated this effect of psychosomatic rehabilitation. For example de Vries and colleagues [1] showed a greater reduction in symptoms (depression, psychological stress, and psychosocial health) for female patients than for male patients with data from 2008 to 2010. A more recent study with data from 2013 could not find any signifi- cant differences in the effectiveness of rehabilitation in terms of symptom burden or resilience [7]. Only self-regulatory ability and work motivation showed the effect of gender on treatment outcomes. Multiple reasons for the gender differences were suggested. It was argued that that rehabilitation interventions are better tailored to the needs of women than to the needs of men [2]. These different needs are based on gender roles [8,9]. Among other things, seeking help and support contradicts the male gender role, which is why men often show less help seeking behavior than women [10] and communicate less effectively with health care providers. For instance, some men may understate pain or hide emotions [8] or are less accurate about their medical records [11]. The loss of autonomy during treatment is possibly more problematic for men than for women [12]. The patient role is incongruent with the male gender role [13], which is active and agentic. Competing interests: Prof. Dr. Manuel Sprung is the scientific director of the University Clinic for Psychosomatic Medicine Eggenburg, PSZW (Eggenburg and Gars). Chief physician Dr. Riffer is the medical director of the PSZW. This does not alter our adherence to PLOS ONE policies on sharing data and materials. Introduction Since older cohorts often have more traditional living conditions with larger gender dif- ferences [9], an additional analysis compares gender differences in different age groups. If the gender differences decrease due to social changes, younger cohorts could show smaller gender differences than older cohorts. Participants F01-F09 Mental disorders due to known physiological conditions, F10-F19 Mental and behavioral disorders due to psychoactive substance use, F20-F29 Schizophrenia, schizotypal, delusional, and other non-mood psychotic disorders, F30-F39 Mood disorders, F40-F48 Anxiety, dissociative, stress-related, somatoform and other nonpsychotic mental disorders, F50-F59 Behavioral syndromes associated with physiological disturbances and physical factors, F60-F69 Disorders of adult personality and behavior, F70-F79 Intellectual disabilities, F80-F89 Pervasive and specific developmental disorders, F90-F98 Behavioral and emotional disorders with onset usually occurring in childhood and adolescence, F99-F99 Unspecified mental disorder; E00-E90 Endocrine, nutritional and metabolic diseases; I00-I99 Diseases of the Circulatory System; M00-M99 diseases of the musculoskeletal system and connective tissue; Z73 burnout (the information on the main diagnoses of 29 patients is missing in the database). F01-F09 Mental disorders due to known physiological conditions, F10-F19 Mental and behavioral disorders due to psychoactive substance use, F20-F29 Schizophrenia, schizotypal, delusional, and other non-mood psychotic disorders, F30-F39 Mood disorders, F40-F48 Anxiety, dissociative, stress-related, somatoform and other nonpsychotic mental disorders, F50-F59 Behavioral syndromes associated with physiological disturbances and physical factors, F60-F69 Disorders of adult personality and behavior, F70-F79 Intellectual disabilities, F80-F89 Pervasive and specific developmental disorders, F90-F98 Behavioral and emotional disorders with onset usually occurring in childhood and adolescence, F99-F99 Unspecified mental disorder; E00-E90 Endocrine, nutritional and metabolic diseases; I00-I99 Diseases of the Circulatory System; M00-M99 diseases of the musculoskeletal system and connective tissue; Z73 burnout (the information on the main diagnoses of 29 patients is missing in the database). + Information from 29 patients (1.3% of the sample) was missing. + Information from 29 patients (1.3% of the sample) was missing. # Pension, has applied for pension payment, rehabilitation pay or sick pay. # Pension, has applied for pension payment, rehabilitation pay or sick pay. clinic’s internal processes. The sample comprised 848 (37.5%) men and 1412 (62.5%) women. The patients were between 18 and 74 years old at the time of admission. Their mean age was 46.1 years (SD = 8.8; median = 47.7). Details are given in Table 1. Male patients were on aver- age somewhat older than female patients, t(2258) = 2.251; p < 0.05, 95% CI [0.11; 1.61], Cohen’s ds = 0.10. The most common diagnoses (main diagnosis) of patients were F30-F39 mood disorders (63.6%) and F40-F48 neurotic, stress and somatoform disorders (28.3%). Participants The present study is a retrospective analysis of data collected as part of the routine examina- tion. The study analyses data from 2260 rehabilitation patients, with complete examination results (i.e. all primary outcomes in the admission and discharge surveys) who were treated between July 2011 and January 2015. Depending on the outcome, measurements from 71 to 66% of the total patient sample are available. Missing measurements are mainly due to the PLOS ONE \| https://doi.org/10.1371/journal.pone.0256916 August 27, 2021 2 / 12 Gender effects on psychosomatic rehabilitation outcomes are reduced PLOS ONE Table 1. Sample characteristics by age cohort and sex. Sex WHO age cohorts male female <35 35–55 >55 Frequency 848 1412 270 1628 362 Age M (SD) 46.7 (8.9) 45.8 (8.8) 28.8 (4.0) 46.5 (5.18) 57.3 (2.5) Sex 37.5% 62.5% - - - male - - 10.7% 70.6% 18.6% female - - 12.7% 72.9% 14.4% Diagnosis F01-09 4 (0.5%) 1 (0.1%) 0 (0.0%) 1 (0.1%) 4 (1.1%) F10-19 8 (1.0%) 11 (0.8%) 5 (1.9%) 13 (0.8%) 1 (0.3%) F20-29 16 (1.9%) 21 (1.5%) 10 (3.7%) 27 (1.7%) 0 (0.0%) F30-39 528 (63.2%) 892 (63.9%) 140 (52.4%) 1035 (64.5%) 245 (68.1%) F40-48 228 (27.3%) 403 (28.9%) 91 (34.1%) 447 (27.9%) 93 (25.8%) F50-59 4 (0.5%) 6 (0.4%) 1 (0.4%) 4 (0.2%) 5 (1.4%) F60-69 12 (1.4%) 23 (1.6%) 15 (5.6%) 19 (1.2%) 1 (0.3%) F70-79 0 (0.0%) 0 (0.0%) 0 (0.0%) 0 (0.0%) 0 (0.0%) F80-89 0 (0.0%) 0 (0.0%) 0 (0.0%) 0 (0.0%) 0 (0.0%) F90-98 2 (0.2%) 0 (0.0%) 2 (0.7%) 0 (0.0%) 0 (0.0%) E00-90 3 (0.4%) 0 (0.0%) 0 (0.0%) 2 (0.1%) 1 (0.3%) I00-99 1 (0.1%) 0 (0.0%) 0 (0.0%) 1 (0.1%) 0 (0.0%) M00-99 0 (0.0%) 1 (0.1%) 0 (0.0%) 1 (0.1%) 0 (0.0%) Z73 29 (3.5%) 38 (2.8%) 3 (1.1%) 54 (3.4%) 10 (2.8%) Employment+ unemployed 190 (28.9%) 294 (28.1%) 75 (36.2%) 343 (28.1%) 66 (23.9%) retired# 33 (5.0%) 69 (6.6%) 12 (5.8%) 63 (5.2%) 27 (9.8%) employed 434 (66.1%) 682 (65.3%) 120 (58.0%) 813 (66.7%) 183 (66.3%) Table 1. Sample characteristics by age cohort and sex. https://doi.org/10.1371/journal.pone.0256916.t001 PLOS ONE \| https://doi.org/10.1371/journal.pone.0256916 August 27, 2021 Participants At the time of admission, 1,116 (65.6%) of the patients were employed, 484 (28.4%) were unemployed and 102 (6.0%) were retired or had applied for a pension or were receiving reha- bilitation or sick pay. Gender and age-specific differences with regard to diagnoses and occu- pational status (employment) are described in Table 1. Male and female patients did not differ PLOS ONE \| https://doi.org/10.1371/journal.pone.0256916 August 27, 2021 3 / 12 PLOS ONE Gender effects on psychosomatic rehabilitation outcomes are reduced significantly with regard to diagnosis or occupational status. Further information on the total sample can be found in Riffer and colleagues [24–26]. Materials Basic socio-demographic and clinical data. Gender, age, diagnoses (i.e. the main diagno- ses upon discharge) and information on the professional status (employment) were deter- mined using the basic documentation of the hospital information system. Gender (male, female) was defined in terms of biological sex. Age referred to the chronological age (in years) at the time of admission. Diagnoses were classified using the ICD-10 [27]. Symptom burden, quality of life and functional ability. The routine examination survey assessed the general symptom burden, quality of life and functional ability (primary out- comes). General symptom severity was determined using the Symptom Checklist (SCL-90) [28] or the Brief Symptom Checklist (BSCL) [29]. The checklists measure the subjectively per- ceived impairments caused by physical and psychological symptoms in a total of nine symp- tom areas with 90, respectively 53 items on a 5-point response scale. The BSCL had been derived from the SCL-90 by selecting the items with the highest item loading. Previous data showed that the BSCL and SCL-90 are highly correlated [r = .92 to.99], [30]. Both instruments provide a Global Severity Index (GSI), which is a commonly used measure of general psycho- logical distress. Items were averaged to form the index. Higher values imply higher symptom severity. The quality of life was determined with the WHO Quality of Life questionnaire (WHOQOL-BREF) [31]. It assesses the subjective health-related quality of life in four sub- areas as well as globally using 26 items (with a 5-point response scale). The quality of life was determined by adding values for each area. This provided measures for physical, psychological, social, and environmental quality of life as well as a global value. Higher scores denote better quality of life. Functional ability was determined using the Global Assessment of Functioning (GAF) scale of the Diagnostic and Statistical Manual of Mental Disorders (DSM-IV) [32]. Based on external assessments by the attending physician, the GAF is used to determine a sin- gle value for the current general functional level of the patient (graded into 10 levels). Higher scores imply better functioning. Relevant comparative values of other Austrian psychiatric rehabilitation patients can be found in a current meta-analysis [33]. Procedure The study used a naturalistic one-group pre-post design without a control group. The analysis used data of patients who were treated at the psychiatric rehabilitation clinic Gars am Kamp in Austria. Patients answered the questionnaires in a computer assessment room in the presence of a trained professional. The self-report measures were assessed using the Hogrefe test system. This system provides a platform that administers standardized questionnaires licensed by Hogrefe. It provides a user-friendly surface and assures data integrity. The GAF was assessed during individual medical examinations. The analyses were approved by the ethics commission of the Karl Landsteiner University of Health Sciences (Nr: 1006/2021). We complied with the requirements of the current version of the Declaration of Helsinki [34]. The patients agreed to the data collection and usage when they started treatment, all analyses were conducted on pseudonymized data. Results Results at pre- (T1) and posttest (T2), separately for each gender To test the effectiveness of the treatment, we compared the outcomes of T1 (pretest) and T2 (posttest) separately by gender. The analyzes in Table 2 show that all outcomes at T2 improved compared to T1, for both men and women. The changes show small to large effects. The big- gest effects were found for functional ability (GAF), the smallest in the areas of social and envi- ronmental quality of life (QOLsocial and QOLenvironment). Treatment All patients in the study sample took part in a standard multidisciplinary therapy program of 22½ hours per week during a planned 6-week inpatient stay at a psychiatric/psychosomatic rehabilitation clinic in Austria. The therapy, which largely takes place in open groups that are 4 / 12 PLOS ONE \| https://doi.org/10.1371/journal.pone.0256916 August 27, 2021 PLOS ONE Gender effects on psychosomatic rehabilitation outcomes are reduced not specific to the disorder, was carried out by a multidisciplinary treatment team in accor- dance with the treatment plan of the pension insurance institution (Pensionsversicherungsan- stalt; www.pv.at) responsible for psychiatric/psychosomatic rehabilitation in Austria. Psychosomatic rehabilitation in Germany and psychiatric rehabilitation in Austria are very similar approaches to treat mental disorders. Both emphasize psychotherapeutic interventions, but also include psychopharmacological and various other complementary interventions, for instance excise and physical therapy. Both mainly treat patients with depressive or anxiety- related disorders. Further details on the treatment program can be found elsewhere [25]. Due to the legal requirements applied to all psychiatric rehabilitation clinics in Austria the treat- ment program is comparable to that of other Austrian rehabilitation clinics [33]. The length of stay of the patients in the study sample was between 39 and 62 days (M = 42.1; SD = 3.5; Modus = 41.0). There were no significant gender- or age-specific differences in the length of stay. Analyses The data was extracted from the Hogrefe test system and the clinic information system. To simplify analysis, only complete data sets were evaluated. The analyzes tested for gender-spe- cific differences at the beginning of treatment (admission survey) and for gender-specific dif- ferences in the changes in the examination results (primary outcomes) between admission and discharge. The final sample had a power of over .99 to find small effects for this interaction between gender and measurement time (GPower Version 3.1.9.2., [35]). The gender-specific pre-post and baseline values at admission were compared with t-tests and repeated-measure ANOVAs using SPSS 26. These treatment effects were tested both in the overall sample and separately for three age cohorts. The level of significance was set at α <0.05 (two-sided). Cohen’s d values above 0.8 are interpreted as large effects, between 0.5 and 0.2 as medium effects and effects below 0.2 are interpreted as small effects [36]. Corresponding values for Z2 p recommend that values from 0.14 should be interpreted as large, from 0.06 as medium and from 0.01 as small effects [37]. PLOS ONE Effect size M SD M SD t df p r dav ꜜGSI men 1.10 0.67 0.80 0.68 13.78 847 < 0.001 0.55 0.44 women 1.31 0.66 0.87 0.70 24.0 1411 < 0.001 0.52 0.65 ꜛQOLphysical men 55.52 18.31 64.40 20.95 -17.073 847 < 0.001 0.71 0.45 women 51.80 18.77 62.62 19.91 -26.51 1411 < 0.001 0.69 0.56 ꜛQOLpsychological men 48.91 19.18 58.48 21.01 -17.589 847 < 0.001 0.69 0.48 women 41.18 18.83 53.36 20.32 -28.125 1411 < 0.001 0.66 0.63 ꜛQOLsocial men 55.33 22.83 61.32 23.09 -9.55 847 < 0.001 0.68 0.26 women 55.99 23.12 61.77 22.79 -11.14 1411 < 0.001 0.64 0.25 ꜛQOLenvironment men 69.44 15.93 71.32 17.14 -4.54 847 < 0.001 0.74 0.11 women 65.42 16.15 67.75 16.72 -7.10 1411 < 0.001 0.72 0.14 ꜛQOLglobal men 45.71 21.11 58.13 22.66 -17.68 847 < 0.001 0.57 0.57 women 42.69 21.46 56.70 22.31 -24.48 1411 < 0.001 0.52 0.64 ꜛGAF men 59.68 7.12 66.51 8.46 -38.461 847 < 0.001 0.79 0.88 women 59.45 6.67 65.81 7.82 -45.72 1411 < 0.001 0.75 0.88 ꜜlower value = positive; ꜛhigher value = positive. Corr = correlation between T1 and T2. https://doi.org/10.1371/journal.pone.0256916.t002 Table 3. Group comparison of outcomes at admission (time T1). Outcomes sex (male vs. female) t df p ds GSI -7.059 2258 < .001 0.32 QOLphysical 4.605 2258 < .001 0.20 QOLpsychological 9.377 2258 < .001 0.41 QOLsocial -0.665 2258 .506 < 0.001 QOLenvironment 5.748 2258 < .001 0.25 QOLglobal 3.260 2258 < .01 0.14 GAF 0.781 2258 .435 < 0.005 https://doi.org/10.1371/journal.pone.0256916.t003 PLOS ONE \| https://doi.org/10.1371/journal.pone.0256916 August 27, 2021 6 / 12 Table 2. Treatment outcomes for men and women. https://doi.org/10.1371/journal.pone.0256916.t002 Outcome comparisons at admission (T1) The level of distress at the time of admission (T1) was then compared between men and women. Table 3 contains the corresponding t-tests for independent samples at admission (T1). In line with earlier findings, women were more distressed than men at admission. This difference was most pronounced in the psychological area of quality of life (QOLmental) and general symptom burden (GSI). Significant differences were also found in the areas of environ- ment (QOLenvironment), physical (QOLphysical), and global quality of life (QOLglobal). 5 / 12 PLOS ONE \| https://doi.org/10.1371/journal.pone.0256916 August 27, 2021 Gender effects on psychosomatic rehabilitation outcomes are reduced PLOS ONE PLOS ONE Table 2. Treatment outcomes for men and women. Outcomes T1 T2 t-test Corr. Effect size M SD M SD t df p r dav ꜜGSI men 1.10 0.67 0.80 0.68 13.78 847 < 0.001 0.55 0.44 women 1.31 0.66 0.87 0.70 24.0 1411 < 0.001 0.52 0.65 ꜛQOLphysical men 55.52 18.31 64.40 20.95 -17.073 847 < 0.001 0.71 0.45 women 51.80 18.77 62.62 19.91 -26.51 1411 < 0.001 0.69 0.56 ꜛQOLpsychological men 48.91 19.18 58.48 21.01 -17.589 847 < 0.001 0.69 0.48 women 41.18 18.83 53.36 20.32 -28.125 1411 < 0.001 0.66 0.63 ꜛQOLsocial men 55.33 22.83 61.32 23.09 -9.55 847 < 0.001 0.68 0.26 women 55.99 23.12 61.77 22.79 -11.14 1411 < 0.001 0.64 0.25 ꜛQOLenvironment men 69.44 15.93 71.32 17.14 -4.54 847 < 0.001 0.74 0.11 women 65.42 16.15 67.75 16.72 -7.10 1411 < 0.001 0.72 0.14 ꜛQOLglobal men 45.71 21.11 58.13 22.66 -17.68 847 < 0.001 0.57 0.57 women 42.69 21.46 56.70 22.31 -24.48 1411 < 0.001 0.52 0.64 ꜛGAF men 59.68 7.12 66.51 8.46 -38.461 847 < 0.001 0.79 0.88 women 59.45 6.67 65.81 7.82 -45.72 1411 < 0.001 0.75 0.88 ꜜlower value = positive; ꜛhigher value = positive. Corr = correlation between T1 and T2. Gender differences in changes between T1—T2 The course of treatment for men and women was compared with the help of a repeated mea- sures ANOVAs, controlling for the differences in symptom burden at the time of admission (T1). Table 4 reports the results of the 2 (time) × 2 (gender) ANOVAs per outcome variable. Treatment effects were gender-specific (= significant interaction between the factors time × gender) for general symptom burden (GSI) and quality of life in the physical and psy- chological domain (QOLphysical, QOLmental), as well as functional ability (GAF). Functional ability showed a more pronounced improvement in men; the other two measures showed Table 2. Treatment outcomes for men and women. Outcomes T1 T2 t-test Corr. https://doi.org/10.1371/journal.pone.0256916.t002 Gender differences in changes between T1—T2 The course of treatment for men and women was compared with the help of a repeated mea- sures ANOVAs, controlling for the differences in symptom burden at the time of admission (T1). Table 4 reports the results of the 2 (time) × 2 (gender) ANOVAs per outcome variable. Treatment effects were gender-specific (= significant interaction between the factors time × gender) for general symptom burden (GSI) and quality of life in the physical and psy- chological domain (QOLphysical, QOLmental), as well as functional ability (GAF). Functional ability showed a more pronounced improvement in men; the other two measures showed Table 3. Group comparison of outcomes at admission (time T1). Outcomes sex (male vs. female) t df p ds GSI -7.059 2258 < .001 0.32 QOLphysical 4.605 2258 < .001 0.20 QOLpsychological 9.377 2258 < .001 0.41 QOLsocial -0.665 2258 .506 < 0.001 QOLenvironment 5.748 2258 < .001 0.25 QOLglobal 3.260 2258 < .01 0.14 GAF 0.781 2258 .435 < 0.005 Table 3. Group comparison of outcomes at admission (time T1). PLOS ONE \| https://doi.org/10.1371/journal.pone.0256916 August 27, 2021 6 / 12 Gender effects on psychosomatic rehabilitation outcomes are reduced PLOS ONE Table 4. Group comparison in the improvement of the outcomes of admission and discharge surveys (T1—T2). Outcomes 2 (time) × 2 (sex) ANOVA 2 (time) × 2 (sex) ANOVA by age cohort Factor F df p η2 Age cohorts Fꜜ df p η2 GSI time 650.28 1 < .001 0.22 <35 4.17 1 .042 0.015 sex 30.44 1 < .001 0.013 35–55 9.34 1 .002 0.006 time × sex 20.76 1 < .001 0.009 >55 10.63 1 .001 0.029 QOLphysical time 882.19 1 < .001 0.28 <35 1.82 1 .179 0.007 sex 12.48 1 < .001 0.005 35–55 2.88 1 .090 0.002 time × sex 8.59 1 .003 0.004 >55 6.22 1 .013 0.017 QOLpsychological time 965.40 1 < .001 0.30 <35 10.70 1 .001 0.038 sex 66.93 1 < .001 0.03 35–55 2.02 1 .156 0.001 time × sex 13.81 1 < .001 0.006 >55 10.86 1 .001 0.029 QOLsocial time 202.51 1 < .001 0.08 <35 3.48 1 .063 0.013 sex 0.37 1 n.s. - 35–55 1.50 1 .221 0.001 time × sex 0.07 1 n.s. - >55 0.36 1 .548 0.001 QOLenvir. Gender differences in changes between T1—T2 time 62.82 1 < .001 0.03 <35 7.65 1 .006 0.028 sex 32.47 1 < .001 0.01 35–55 0.18 1 .675 < 0.001 time × sex 0.72 1 n.s. - >55 0.69 1 .407 0.002 QOLglobal time 829.75 1 < .001 0.27 <35 3.64 1 .058 0.013 sex 7.14 1 .008 0.003 35–55 < 0.01 1 .959 < 0.001 time × sex 2.99 1 n.s. - >55 8.26 1 .004 0.022 GAF time 3400.07 1 < .001 0.60 <35 5.16 1 .024 0.019 sex 2.34 1 n.s. - 35–55 4.13 1 .042 0.003 time × sex 4.25 1 .039 0.002 >55 1.04 1 .309 0.003 ꜜF-values for interaction time by sex. greater effects in women. At the same time, all effect sizes were so small that the differences should be viewed as clinically insignificant. PLOS ONE \| https://doi.org/10.1371/journal.pone.0256916 August 27, 2021 Discussion Both men and women showed substantial improvements in all examined outcome areas (gen- eral symptom burden, quality of life and functional ability). Although some of these improve- ments showed significant gender-specific differences, the effect sizes of these differences were so small that they can be regarded as clinically negligible. Older findings that showed moderate gender differences in the course of treatment could therefore not be replicated. Substantial gender differences were neither evident in externally rated functional ability (GAF) nor in self- report measures. At the time of admission, women showed greater symptom burden than men. The gender differences in treatment outcomes were most pronounced in the oldest and youngest age group, but remained small in these groups too. Accordingly, all age groups, both men and women, showed broad treatment success. The age group patterns did not match a linear decrease in the gender difference in the sense of a cohort effect. If the equalization of liv- ing conditions between men and women were the reason for the decrease in gender differ- ences, it should be least pronounced in the youngest age group. The interpretation is complicated, however, by the fact that the analysis confounded age and cohort effects. The data pattern fits an explanation of gender differences through gender roles. The relatively low gender effect in the middle age group could be based on the gender differences being driven by masculine gender role orientations. These show the smallest differences between men and women in middle adulthood [38,39]. However, to test this explanation a direct measure of gen- der role orientation would be necessary. Since most patients in this study fall within the middle age group the very small overall gender effect is driven by this age group. However, previous meta-analyses of psychosomatic rehabilitation interventions show that this age distribution is representative of psychiatric rehabilitation clinics in Austria and psychosomatic rehabilitation clinics Germany [3,33]. Earlier studies faced the question of whether women only benefited more from rehabilita- tion interventions because they were more distressed than men when they were admitted [4]. In lieu of the current data, this interpretation appears to be less likely, since there are no longer any gender differences in treatment outcomes, but women are still more distressed at admis- sion. Gender-specific differences in the change between T1—T2 per age cohort For further exploration, we tested whether the gender effects differed between age cohorts. As the living conditions of men and women become increasingly similar, it would be possible that older cohorts show greater gender differences than younger ones. To test this, we repeated the repeated measures ANOVAs for the three age groups. The age groups were formed in accor- dance with the WHO age limits and corresponded to the ranges 18–35 years = young adults, 35–55 years = middle-aged adults, over 55 years = older adults. If the low gender effects were based on cohort effects, the oldest cohort should show the largest gender-specific treatment effects. The results did not match these expectations. Gender differences in treatment effects were slightly larger in both the youngest and oldest age groups than for the entire sample. The results are shown in Table 4. The youngest age group showed gender differences in the course of treatment, with at least a small effect size (η2 > 0.01) in the general symptom burden, the quality of life in the psychological, social and environmental domains as well as in the global evaluation (QOLmental, QOLsocial, QOLenvironment, QOLglobal). These measurements showed a better course of treatment in women. Functional ability (GAF) also showed a small gender dif- ference, however, men showed the better course in this measure. The oldest cohort also reported a more positive course of treatment for women of at least small size in terms of gen- eral symptom burden and quality of life in the physical, mental and global domains PLOS ONE \| https://doi.org/10.1371/journal.pone.0256916 August 27, 2021 7 / 12 PLOS ONE Gender effects on psychosomatic rehabilitation outcomes are reduced (QOLphysical, QOLmental, QOLglobal). The middle age group showed no (clinically relevant) gen- der differences over the course of treatment. PLOS ONE \| https://doi.org/10.1371/journal.pone.0256916 August 27, 2021 Discussion The greatest effect was found in work motivation, although this effect also remained small. Work motivation and related procedures could show greater effects than, for example, symptom burden, because there are gender-typical differ- ences in working environments. Accordingly, it would make sense to investigate further treat- ment outcomes in future studies. It should also be examined whether other rehabilitation measures continue to show gender effects. The strengths of the present study are the sample size and the inclusion of heterogeneous patient groups. An important extension of the present study would be the examination of post-treatment follow-up data. Previous findings showed that gender differences at follow-up assessment compared to the discharge assessment disappeared [3]. Although men benefit more from rehabilitation at discharge assessment, men may show poorer results at the follow- up assessment. Since men show more health risk behavior such as substance abuse or sleep deprivation [45], therapeutic effects could decrease especially strongly after rehabilitation. One of the limitations of the present study is that gender role orientation was not studied. The divi- sion into age groups was done post hoc and confounded age and cohort effects. Additionally, the data was collected between 2011 and 2015. It is thus, plausible that new changes have occurred. This and the fact that the data was only gathered in one clinic makes it necessary to replicate our findings. Another limitation is that no control group was included in the study. Therefore, it cannot be ruled out that without treatment the effects would be lower and that the observed gender-specific differences in the change in T1—T2 could also be found in a con- trol group without treatment. However, a controlled study on psychiatric/psychosomatic reha- bilitation in Austria confirmed that the improvements in psychiatric/psychosomatic rehabilitation are systematically greater than in an untreated control group [46]. Unfortu- nately, gender differences were not investigated in this study. In the future, more controlled studies should be carried out that also take gender-specific differences into account. In contrast to earlier studies, gender-specific differences in the outcomes of psychiatric/psy- chosomatic rehabilitation were scarcely detectable. The present findings confirm that rehabili- tation interventions can be equally effective among men and women, even though women remain more distressed upon admission than men. Thus, the higher symptom burden of women at admission does not necessarily lead to a difference in treatment effectiveness. Discussion This suggests that the earlier gender difference in treatment outcomes was more likely due to an earlier lack of fit between needs and services; as previously suggested [2,4]. The most positive interpretation of the results would be that treatment methods are now more optimally aimed at both men and women. However, this interpretation is less in line with the larger gen- der differences in the youngest and oldest cohort. Alternatively, the gender differences that influenced the course of therapy differently, such as differences in gender roles or living condi- tions, could have decreased sufficiently in the population in certain age groups. The results may be surprising in view of the amount of evidence that has highlighted the extent of the dif- ferences in disease rates between men and women. Gender-specific differences in psychopa- thology are favored by a number of biological, cultural, cognitive, and affective factors and manifest themselves in a variety of ways [40]. For example, depression in men is more often characterized by a mixture of internalization (anxiety, depression) and externalization disor- ders (substance abuse, aggression) [41] while women report more somatic symptoms [42]. The gender-specific differences in the symptoms of depression have been neurobiologically [43] and neuroendocrinologically confirmed [44]. The depression pattern typical for men was more strongly characterized by alcohol abuse and suicide than that of women. These gender differences could be missing in the present study, since patients represent a selective sample. 8 / 12 PLOS ONE \| https://doi.org/10.1371/journal.pone.0256916 August 27, 2021 PLOS ONE Gender effects on psychosomatic rehabilitation outcomes are reduced Previous work has noted that the gender differences could be partly attributed to biases within measures [18]. Since men and women often report different symptoms questionnaires can be biased towards capturing traditionally ‘feminine’ symptoms of a disorder. Therefore, it is nota- ble that our study used the SCL-90 lists, which is the most common instrument used in studies within the MESTA-meta-analysis, which did find a gender effect β = .22 [3]. It is therefore unlikely that our results hinge on measurement biases. Even though, the absence of pronounced gender effects on efficacy may seem surprising, the present results only superficially contradict more recent findings. Although earlier studies continued to confirm the existence of gender differences in psychosomatic rehabilitation inter- ventions, the corresponding effects were small for data from 2008 to 2010 [1] or limited to a few variables for data from 2013 [7]. PLOS ONE \| https://doi.org/10.1371/journal.pone.0256916 August 27, 2021 Acknowledgments The results of the present study were in part presented during a poster presentation at the 18th annual conference of the Austrian Society for Psychiatry, Psychotherapy and Psychosomatics. This article is distributed under the terms of the Creative Commons Attribution 4.0 Inter- national License https://creativecommons.org/licenses/by/4.0/deed.en), which permits unre- stricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. Author Contributions Conceptualization: Juliane Burghardt, Friedrich Riffer, Manuel Sprung. Data curation: Manuel Sprung. Formal analysis: Juliane Burghardt, Manuel Sprung. Resources: Friedrich Riffer, Manuel Sprung. Resources: Friedrich Riffer, Manuel Sprung. Supervision: Manuel Sprung. Validation: Juliane Burghardt. Writing – original draft: Manuel Sprung. Writing – original draft: Manuel Sprung. Writing – review & editing: Juliane Burghardt, Friedrich Riffer, Manuel Sprung. Writing – review & editing: Juliane Burghardt, Friedrich Riffer, Manuel Sprung. Discussion It remains unclear whether the decline in gender differences on treatment effectiveness was caused by changes in society as a whole or by specific characteristics of the present treatment. The extent of the gender differences in rehabilitation outcomes was particularly small in the middle age group, which is consistent with the typical course of changes in gender roles over the lifespan. This would correspond to an explanation by changes in society as a whole. If the finding of comparatively bigger gender effects among older or younger patients replicates, this might imply a need for gender and age-sensitive treatments instead of a gender-sensitive treat- ment that ignores age. 9 / 12 PLOS ONE \| https://doi.org/10.1371/journal.pone.0256916 August 27, 2021 PLOS ONE Gender effects on psychosomatic rehabilitation outcomes are reduced References 1. de Vries U, Lange M, Franke W, Petermann F (2011) Differenzielle Effekte stationa¨rer psychosoma- tischer Rehabilitation. Phys Med Rehabil Kurortmed 21:290–295. https://doi.org/10.1055/s-0031- 1291369 2. Dinger-Broda A, Broda M (1997) Geschlechtsspezifische Unterschiede in der psychosomatischen Rehabilitationv. Prax Klin Verhal Rehabil 7–12. 3. Steffanowski A, Lo¨schmann C, Schmidt J, Wittmann WW, Nu¨bling N (2005) Meta-Analyse der Effekte stationa¨rer psychosomatischer Rehabilitation—MESTA-Studie. http://forschung.deutsche- rentenversicherung.de/ForschPortalWeb/rehaDoc.pdf?rehaid= C8009F9A95868E3EC1256E99004424E7. Accessed 25 Jun 2020. 4. Schro¨der A (2000) Ma¨nner und Frauen in der Rehabilitation. Z Fu¨r Med Psychol 9:37–42. 5. Mohr B, Gra¨f T, Forster M, Krohn-Grimberghe B, Kurzeja R, Mantel F, et al (2008) Der Einfluss von Depressivita¨t und Geschlecht auf den Rehabilitationserfolg bei chronischem Ru¨ckenschmerz: Eine Pilotstudie. Rehabil 47:284–298. https://doi.org/10.1055/s-2008-1076708 PMID: 18937161 6. Adams KF, Sueta CA, Gheorghiade M, O’Connor CM, Schwartz TA, Koch GG, et al (1999) Gender dif- ferences in survival in advanced heart failure. Insights from the FIRST study. Circulation 99:1816– 1821. https://doi.org/10.1161/01.cir.99.14.1816 PMID: 10199877 7. von Ho¨rsten N, Schulz W, Gissendanner SS, Schmid-Ott G (2019) Geschlechterunterschiede im Ver- lauf und Erfolg psychosomatischer Rehabilitation. Phys Med Rehab Kuror 29:190–198. https://doi.org/ 10.1055/a-0852-3471 10 / 12 PLOS ONE \| https://doi.org/10.1371/journal.pone.0256916 August 27, 2021 PLOS ONE Gender effects on psychosomatic rehabilitation outcomes are reduced 8. O’Neil JM (2008) Summarizing 25 Years of Research on Men’s Gender Role Conflict Using the Gender Role Conflict Scale: New Research Paradigms and Clinical Implications. Couns Psychol 36:358–445. https://doi.org/10.1177/0011000008317057 9. Seedat S, Scott K, Angermeyer MC, et al (2009) Cross-national associations between gender and men- tal disorders in the world health organization world mental health surveys. Arch Gen Psychiatry 66:785–795. https://doi.org/10.1001/archgenpsychiatry.2009.36 PMID: 19581570 10. Yousaf O, Grunfeld EA, Hunter MS (2015) A systematic review of the factors associated with delays in medical and psychological help-seeking among men. Health Psychol Rev 9:264–76. https://doi.org/10. 1080/17437199.2013.840954 PMID: 26209212 11. Sieverding M, Arbogast AL, Zintel S, von Wagner C (2020) Gender differences in self-reported family history of cancer: A review and secondary data analysis. Cancer Med n/a:. https://doi.org/10.1002/ cam4.3405 PMID: 32835456 12. Seidler ZE, Rice SM, Oliffe JL, Fogarty AS, Dhillon HM (2018) Men In and Out of Treatment for Depres- sion: Strategies for Improved Engagement. Aust Psychol 53:405–415. https://doi.org/10.1111/ap. 12331 13. Courtenay WH (2000) Constructions of masculinity and their influence on men’s well-being: a theory of gender and health. Soc Sci Med 1982 50:1385–1401. https://doi.org/10.1016/s0277-9536(99)00390-1 PMID: 10741575 14. References Stengler K, Glaesmer H, Dietrich S (2011) Gender- und geschlechtsspezifische Aspekte in der psychia- trischen und psychotherapeutischen Forschung: eine bibliometrische Analyse. Z Fu¨r Psychiatr Psychol Psychother 59:305–310. https://doi.org/10.1024/1661-4747/a000086 15. Bekker MHJ, van Mens-Verhulst J (2007) Anxiety Disorders: Sex Differences in Prevalence, Degree, and Background, But Gender-Neutral Treatment. Gend Med 4:S178–S193. https://doi.org/10.1016/ s1550-8579(07)80057-x PMID: 18156102 16. Cavanagh A, Wilson CJ, Kavanagh DJ, Caputi P (2017) Differences in the Expression of Symptoms in Men Versus Women with Depression: A Systematic Review and Meta-analysis. Harv Rev Psychiatry 25:29–38. https://doi.org/10.1097/HRP.0000000000000128 PMID: 28059934 17. Boyd A, Van de Velde S, Vilagut G, de Graaf R, O׳Neill S, Florescu S, et al (2015) Gender differences in mental disorders and suicidality in Europe: Results from a large cross-sectional population-based study. J Affect Disord 173:245–254. https://doi.org/10.1016/j.jad.2014.11.002 PMID: 25462424 18. Piccinelli M, Wilkinson G (2000) Gender differences in depression. Critical review. Br J Psychiatry 177:486–92. https://doi.org/10.1192/bjp.177.6.486 PMID: 11102321 19. World Economic Forum (2019) Global Gender Gap Report 2020. 9. World Economic Forum (2019) Global Gender Gap R 20. Worringen U, Zwingmann C (2001) Zur Einfu¨hrung: Geschlechtsspezifische Rehabilitationsforschung in Deutschland. In: Worringen U, Zwingmann C (eds) Rehabilitation weiblich- ma¨nnlich Geschlechts- spezifische Rehabilitationsforschung. Juventa. 21. Donnelly K, Twenge JM (2017) Masculine and feminine traits on the Bem Sex-Role Inventory, 1993– 2012: A cross-temporal meta-analysis. Sex Roles J Res 76:556–565. https://doi.org/10.1007/s11199- 016-0625-y 22. Twenge JM (1997) Changes in masculine and feminine traits over time: A meta-analysis. Sex Roles 36:305–325. https://doi.org/10.1007/BF02766650 23. Statistik Austria (2020) Internationaler Frauentag 2020: Frauen holen bezu¨glich Bildungsniveau und Erwerbsta¨tigkeit auf; Teilzeit und niedrigere Erwerbseinkommen fu¨hren zu gro¨ßeren sozialen Risiken. In: Stat. Austria. http://www.statistik.at/web_de/statistiken/menschen_und_gesellschaft/soziales/ gender-statistik/erwerbstaetigkeit/122737.html. Accessed 2 Jul 2020. 24. Riffer F, Sprung M, Kaiser E (2017) Evaluationsergebnisse der Rehabilitationsklinik Gars am Kamp. Spectr Psychiatr 3:34–37. 25. Riffer F, Sprung M, Streibl L, Kaiser E (2018) Stationa¨re medizinische Rehabilitation von Patienten mit psychiatrischen oder psychosomatischen Erkrankungen: erste Evaluationsergebnisse der Rehabilita- tionsklinik Gars am Kamp. In: Riffer F, Kaiser E, Sprung M, Streibl L (eds) Das Fremde: Flucht— Trauma—Resilienz: Aktuelle traumaspezifische Konzepte in der Psychosomatik. Springer Berlin Hei- delberg, Berlin, Heidelberg, pp 227–241. 26. Riffer F, Sprung M, Streibl L, Kaiser E (2018) Relevanz von Erkrankungsart und beruflichem Status fu¨r die Ergebnisse psychiatrischer Rehabilitation. neuropsychiatrie 32:33–43. https://doi.org/10.1007/ s40211-017-0254-x PMID: 29236245 27. Dilling H, Mombour W, Schmidt MH, World-Health-Organization (2011) Internationale Klassifikation psychischer Sto¨rungen: ICD-10, Kapitel V (F, klinisch-diagnostische Leitlinien). Hans Huber, Bern. 28. References Franke GH, Derogatis LR (1995) Die Symptom-Checkliste von Derogatis-Deutsche Version. Beltz, Go¨ttingen. 11 / 12 PLOS ONE \| https://doi.org/10.1371/journal.pone.0256916 August 27, 2021 PLOS ONE Gender effects on psychosomatic rehabilitation outcomes are reduced 29. Franke GH (2000) BSI brief symptom inventory von LR Derogatis (Kurzform der SCL-90-R). Dtsch Ver- sion-Man Go¨tt Ger Beltz Test GmbH. 30. Derogatis LR (1993) BSI Brief Symptom Inventory. Administration, Scoring, and Procedures Manual, 4th ed. National Computer Systems, Minneapolis, MN, US. 31. Angermeyer MC, Kilian R, Matschinger H (2000) WHOQOL-100 und WHOQOL-BREF (WHO-QOL): deutschsprachigen Version der WHO Instrumente zur Erfassung von Lebensqualita¨t. Hogrefe, Go¨ttingen. 32. American-Psychiatric-Association (1996) Diagnostisches und statistisches Manual psychischer Sto¨run- gen DSM-IV. Hogrefe, Go¨ttingen. 33. Sprung M, Mu¨nch HM, Kaiser E, Streibl L, Riffer F (2018) Meta-Analyse der Evaluationsergebnisse psy- chiatrischer-psychosomatischer Rehabilitation in O¨ sterreich. neuropsychiatrie. https://doi.org/10.1007/ s40211-018-0290-1 PMID: 30328582 34. World-Medical-Association (2013) World medical association declaration of helsinki: Ethical principles for medical research involving human subjects. JAMA 310:2191–2194. https://doi.org/10.1001/jama. 2013.281053 PMID: 24141714 35. Faul F, Erdfelder E, Lang A-G, Buchner A (2007) GPower 3: A flexible statistical power analysis pro- gram for the social, behavioral, and biomedical sciences. Behav Res Methods 39:175–191. https://doi. org/10.3758/bf03193146 PMID: 17695343 36. Cohen J (1988) Statistical power analysis for the behavioral sciences, 2. ed. Erlbaum, Hillsdale, NJ. 37. Lakens D (2013) Calculating and reporting effect sizes to facilitate cumulative science: a practical primer for t-tests and ANOVAs. Front Psychol 4:863. https://doi.org/10.3389/fpsyg.2013.00863 PMID: 24324449 38. Pinquart M, So¨rensen S (2001) Gender Differences in Self-Concept and Psychological Well-Being in Old Age A Meta-Analysis. J Gerontol Ser B 56:P195–P213. https://doi.org/10.1093/geronb/56.4.p195 PMID: 11445606 39. Puglisi JT, Jackson DW (1981) Sex Role Identity and Self Esteem in Adulthood. Int J Aging Hum Dev 12:129–138. https://doi.org/10.2190/4EPU-UK3E-X7KA-QBDK PMID: 7203674 40. Hyde JS, Mezulis AH (2020) Gender Differences in Depression: Biological, Affective, Cognitive, and Sociocultural Factors. Harv Rev Psychiatry 28:4–13. https://doi.org/10.1097/HRP.0000000000000230 PMID: 31913978 41. Rice SM, Fallon BJ, Aucote HM, Mo¨ller-Leimku¨hler AM (2013) Development and preliminary validation of the male depression risk scale: Furthering the assessment of depression in men. J Affect Disord 151:950–958. https://doi.org/10.1016/j.jad.2013.08.013 PMID: 24051100 42. Silverstein B, Ajdacic-Gross V, Rossler W, Angst J (2017) The gender difference in depressive preva- lence is due to high prevalence of somatic depression among women who do not have depressed rela- tives. J Affect Disord 210:269–272. https://doi.org/10.1016/j.jad.2017.01.006 PMID: 28068614 43. PLOS ONE \| https://doi.org/10.1371/journal.pone.0256916 August 27, 2021 References Knott V, Mahoney C, Kennedy S, Evans K (2001) EEG power, frequency, asymmetry and coherence in male depression. Psychiatry Res Neuroimaging 106:123–140. https://doi.org/10.1016/s0925-4927(00) 00080-9 PMID: 11306251 44. Walther A, Rice T, Kufert Y, Ehlert U (2017) Neuroendocrinology of a Male-Specific Pattern for Depres- sion Linked to Alcohol Use Disorder and Suicidal Behavior. Front Psychiatry 7:206. https://doi.org/10. 3389/fpsyt.2016.00206 PMID: 28096796 45. Courtenay WH (2000) Behavioral Factors Associated with Disease, Injury, and Death among Men: Evi- dence and Implications for Prevention. J Men’s Stud 9:81–142. https://doi.org/10.3149/jms.0901.81 46. Rabenstein R, Pintzinger N, Knogler V, Kirnbauer V, Lenz G, Schosser A (2015) Wirksamkeit eines ambulanten, verhaltenstherapeutisch orientierten Rehabilitationsprogramms—eine Wartelistenkontroll- gruppenstudie. Verhaltenstherapie 25:192–200. https://doi.org/10.1159/000437449 12 / 12
https://openalex.org/W1964349500	https://hal.science/hal-01197552/file/50496_20120531084510639_1.pdf	English	null	Transcriptome Analysis Describing New Immunity and Defense Genes in Peripheral Blood Mononuclear Cells of Rheumatoid Arthritis Patients	PloS one	2,009	cc-by	9,055	Transcriptome Analysis Describing New Immunity and Defense Genes in Peripheral Blood Mononuclear Cells of Rheumatoid Arthritis Patients Vitor Hugo Teixeira, Robert Olaso, Marie-Laure Magniette, Sandra Lasbleiz, Laurent Jacq, Catarina Resende Oliveira, Pascal Hilliquin, Ivo Gut, François Cornelis, Elisabeth Petit-Teixeira To cite this version: Vitor Hugo Teixeira, Robert Olaso, Marie-Laure Magniette, Sandra Lasbleiz, Laurent Jacq, et al.. Transcriptome Analysis Describing New Immunity and Defense Genes in Peripheral Blood Mononu- clear Cells of Rheumatoid Arthritis Patients. PLoS ONE, 2009, 4 (8), pp.e6803. 10.1371/jour- nal.pone.0006803. hal-01197552 Abstract Background: Large-scale gene expression profiling of peripheral blood mononuclear cells from Rheumatoid Arthritis (RA) patients could provide a molecular description that reflects the contribution of diverse cellular responses associated with this disease. The aim of our study was to identify peripheral blood gene expression profiles for RA patients, using Illumina technology, to gain insights into RA molecular mechanisms. Methodology/Principal Findings: The Illumina Human-6v2 Expression BeadChips were used for a complete genome-wide transcript profiling of peripheral blood mononuclear cells (PBMCs) from 18 RA patients and 15 controls. Differential analysis per gene was performed with one-way analysis of variance (ANOVA) and P values were adjusted to control the False Discovery Rate (FDR,5%). Genes differentially expressed at significant level between patients and controls were analyzed using Gene Ontology (GO) in the PANTHER database to identify biological processes. A differentially expression of 339 Reference Sequence genes (238 down-regulated and 101 up-regulated) between the two groups was observed. We identified a remarkably elevated expression of a spectrum of genes involved in Immunity and Defense in PBMCs of RA patients compared to controls. This result is confirmed by GO analysis, suggesting that these genes could be activated systemically in RA. No significant down-regulated ontology groups were found. Microarray data were validated by real time PCR in a set of nine genes showing a high degree of correlation. Conclusions/Significance: Our study highlighted several new genes that could contribute in the identification of innovative clinical biomarkers for diagnostic procedures and therapeutic interventions. Citation: Teixeira VH, Olaso R, Martin-Magniette M-L, Lasbleiz S, Jacq L, et al. (2009) Transcriptome Analysis Describing New Immunity and Defense Genes in Peripheral Blood Mononuclear Cells of Rheumatoid Arthritis Patients. PLoS ONE 4(8): e6803. doi:10.1371/journal.pone.0006803 Editor: Benjamin Rich, Harvard Institute of Medicine, United States of America Editor: Benjamin Rich, Harvard Institute of Medicine, United States of America Received April 7, 2009; Accepted August 3, 2009; Published August 27, 2009 Received April 7, 2009; Accepted August 3, 2009; Published August 27, 2009 Copyright: 2009 Teixeira et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited. Transcriptome Analysis Describing New Immunity and Defense Genes in Peripheral Blood Mononuclear Cells of Rheumatoid Arthritis Patients Vitor Hugo Teixeira1,2., Robert Olaso3., Marie-Laure Martin-Magniette4,5, Sandra Lasbleiz6, Laurent Jacq1, Catarina Resende Oliveira7, Pascal Hilliquin8, Ivo Gut9, Franc¸ois Cornelis1,6,8,10, Elisabeth Petit- Teixeira1 1 GenHotel-EA3886, Evry University - Paris 7 University Medical School, AutoCure European Consortium member, Evry, France, 2 Faculty of Medicine, University of Coimbra, Coimbra, Portugal, 3 CEA - IG – CNG, Functional genomic platform, Department of Translational Research, Evry, France, 4 UMR - AgroParisTech/INRA, MIA 518, Paris, France, 5 URGV- UMR - INRA 1165, CNRS 8114, Evry University, Evry, France, 6 Lariboisie`re Hospital, APHP, Paris, France, 7 Center for Neurosciences and Cell Biology, Faculty of Medicine, University of Coimbra, Coimbra, Portugal, 8 Sud Francilien Hospital Center, Corbeil-Essonnes, France, 9 CEA - IG - CNG - Department of Translational Research, Evry, France, 10 Clinical Genetics Unit, Lariboisie`re Hospital, APHP, Paris, France Abstract Funding: This work was supported by Association Franc¸aise des Polyarthritiques, Socie´te´ Franc¸aise de Rhumatologie, Association Rhumatisme et Travail, European Union for AutoCure, Association Polyarctique, Agence Franc¸aise de Se´curite´ Sanitaire des Produits de Sante´, Groupe Taitbout, Genopole, Federation Franc¸aise de Cardiologie, Societe Francaise de Cardiologie, Boehringer Ingelheim and the French Ministry of Research and Education. Vitor Hugo Teixeira’s work was supported by Foundation for Science and Technology, Portugal (grant SFRH/BD/23304/2005). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript. Competing Interests: The authors have declared that no competing interests exist. E-mail: vitor@polyarthrite.net * E-mail: vitor@polyarthrite.net . These authors contributed equally to this work. . These authors contributed equally to this work. HAL Id: hal-01197552 L’archive ouverte pluridisciplinaire HAL, est destinée au dépôt et à la diffusion de documents scientifiques de niveau recherche, publiés ou non, émanant des établissements d’enseignement et de recherche français ou étrangers, des laboratoires publics ou privés. HAL is a multi-disciplinary open access archive for the deposit and dissemination of sci- entific research documents, whether they are pub- lished or not. The documents may come from teaching and research institutions in France or abroad, or from public or private research centers. Citation: Teixeira VH, Olaso R, Martin-Magniette M-L, Lasbleiz S, Jacq L, et al. (2009) Transcriptome Analysis Describing New Immu Peripheral Blood Mononuclear Cells of Rheumatoid Arthritis Patients. PLoS ONE 4(8): e6803. doi:10.1371/journal.pone.0006803 Probe synthesis, hybridization and detection cRNA was synthesized, amplified and purified using the Illumina TotalPrep RNA Amplification Kit (Ambion Inc.) following manufacturer recommendations. Briefly, 200 ng of RNA was reverse transcribed. After second strand synthesis, the cDNA was transcribed in vitro and cRNA labelled with biotin-16- UTP. Labelled probe hybridization to Illumina BeadChips human-6v2 was carried out using Illumina’s BeadChip 6v2 protocol. These beadchips contain 48,701 unique 50-mer oligonucleotides in total, with hybridization to each probe assessed at ,30 different beads on average. 22,403 probes (46%) are targeted at Reference Sequence (RefSeq) [20] transcripts and the remaining 26,298 probes (54%) are for other transcripts, generally less well characterized (including predicted transcripts). Considering several comparative studies, Illumina and other microrray technologies have similar performances [16–18]. However, these studies showed that each approach was able to detect specific genes, meaning an increase in knowledge by each platform. To complete previous studies on RA with other microarrays [7], we applied Illumina large-scale gene expression profiling in PBMCs of RA patients to potentially gain insights into molecular mechanism of this disease. We identified new genes involved in different functional Immunity and Defense related mechanisms as pro-inflammation, anti-microbial activity, cellular stress and immunomodulatory functions in Rheumatoid Arthritis. Beadchips were scanned on the Illumina BeadArray 500GX Reader using Illumina BeadScan image data acquisition software (version 2.3.0.13). Illumina BeadStudio software (version 1.5.0.34) was used for preliminary data analysis. To assess quality metrics of each run, several quality control procedures were implemented. Total RNA control samples were analyzed with each run. The Illumina BeadStudio software was used to view control summary reports, scatter plots of the total RNA control results from different days and scatter plots of daily run samples. The scatter plots compared control against control or sample against sample and calculated a correlation coefficient (Figure 1). Viewing the scatter plots determined whether controls across different days varied in quality, indicating a reduction in assay performance, and highlighted those samples that were of lower quality. The control summary report is generated by the BeadStudio software, which evaluates the performance of the built-in controls of the BeadChips across particular runs. This allows the user to look for variations in signal intensity, hybridization signal, background signal and the background to noise ratio for all samples analyzed in that run. Data are expressed as log2 ratios of fluorescence intensities of the experimental and the common reference sample. Study population The study and all protocols presented here were approved by the Ethics Committees of Biceˆtre and Saint Louis Hospitals (Paris, France) and all study participants provided written informed consent. All RA patients satisfied the revised criteria of the American College of Rheumatology [19] according to the rheumatologist in charge of the patient. A rheumatologist university fellow reviewed all clinical data. Characteristics of the 18 RA French Caucasian Patients are reported in Table 1. Among the control group consisted of 15 RA French Caucasian healthy individuals, 11 were females (mean 6 Standard Deviation Table 1. Clinical and demographic characteristics of the RA patients. Table 1. Clinical and demographic characteristics of the RA patients. Clinical features patients RA Patients (n = 18) Mean age (years) 60 Women (%) 72.3 Caucasian (%) 100 RF-positive (%) 88.9 ACPA-positive (%) 90.9 (out of 11 RA patients) Mean disease duration (years) 8.6 Erosions (%) 63.2 Disease Activity Score 28 (DAS28) mean 5.22 Disease-Modifying Anti-Rheumatic Drugs use (%) 100 Anti-TNF therapy 0 doi:10.1371/journal.pone.0006803.t001 All microarray data reported in this study is described in accordance with MIAME guidelines and have been deposited in the National Center for Biotechnology Information Gene Expression Omnibus (GEO, http://www.ncbi.nlm.nih.gov/geo/) public repository, and they are accessible through GEO accession (GSE15573). Introduction antibodies (ACPA) in the serum [3,4], the differential responsive- ness to treatment [5,6], and the variability in clinical presentation. In addition, several gene expression profiling studies of synovial tissues and peripheral blood mononuclear cells (PBMCs) from RA patients showed marked variation in gene expression profiles that allowed to identify distinct molecular disease mechanisms involved in RA pathology [7,8]. The relative contribution of the different mechanisms may vary among patients and in different stages of disease. Thus, the broad goals of expression profiling in RA are to (i) improve our understanding of the pathogenic mechanisms underlying RA, (ii) identify new drugs targets, (iii) assess activity of Rheumatoid Arthritis (RA) is an autoimmune disease charac- terized by chronic and persistent joint synovial tissue inflammation associated with the destruction of affected joints [1]. The multifactorial nature of the disease provides a high RA heterogeneity with specific combinations of a genetic background and environmental factors that influence the susceptibility, severity and outcome of the disease [2]. The RA heterogeneity is demonstrated by the presence of distinct autoantibody specificities, like rheumatoid factor (RF) and anti-cyclic citrullinated peptide PLoS ONE \| www.plosone.org August 2009 \| Volume 4 \| Issue 8 \| e6803 August 2009 \| Volume 4 \| Issue 8 \| e6803 1 RA Gene Expression Profiling (SD) age at enrolment 56.9 6 6.6). In all comparisons mentioned, the groups were age and sex-matched. the disease, (iv) predict future outcomes, such as responsiveness therapy, overall disease severity, and organ specific risk and (v) develop new diagnostic tests [9]. (SD) age at enrolment 56.9 6 6.6). In all comparisons mentioned, the groups were age and sex-matched. Probe synthesis, hybridization and detection The Illumina data were then normalized using the ‘normalize quantiles’ function in the BeadStudio Software. Isolation of total RNA PBMCs gene expression profiling allows both pathogenetic and pathophysiological processes identification as demonstrated in several types of diseases: cancer [10], asthma [11], systemic lupus erythematosus (SLE) [12], cardiovascular diseases [13] and psychiatric disorders [14]. Pathogenetic processes are primarily associated with the cause of a disease. Then, microarrays could lead to the identification of abnormal genes and gene activities that may not be only limited to PBMCs, but could occur in cells of pathological tissue as well. In contrast, pathophysiological changes in lymphocytic gene expression are considered an essentially normal reaction of the immune system to a pathological stimulus. Therefore, pathophysiological gene profiles may be shared in a variety of diseases, whereas pathogenetic gene expression is expected to be disease specific [15]. The differences in expression profiles provide opportunities to stratify RA patients based on molecular criteria that may require different treatment strategies. Peripheral blood (PB) was drawn in PAXgene RNA isolation tubes (PreAnalytix) from 18 patients and 15 controls. Total RNA was isolated from PBMCs using the PAXgene RNA isolation kit (PreAnalytix). Total RNA yield (ng) was determined spectropho- tometrically using the NanoDrop ND-1000 (Wilmington). Total RNA profiles were recorded using a Bioanalyzer 2100 (Agilent). RNA integrity numbers were determined and the mean value was 8.07+/20.51 SD and a Coefficient of Variation (CV) of 6.4%. Real-Time PCR Total RNA was reverse transcribed using Superscript III and oligo(dT) primers (Invitrogen) according to the manufacturer’s instructions. Real-time quantitative PCR was carried out using the SYBR-green master mix (Applied BioSystems) in an Mx 3005P thermocycler (Agilent). PCR conditions were 95uC for 10 min, followed by 40 cycles of 95uC for 15 sec and, 60uC for 1 min. At August 2009 \| Volume 4 \| Issue 8 \| e6803 2 RA Gene Expression Profiling Figure 1. Scatter plots representation of signal intensities. Scatter plot of technical replicates: (A) same reference RNA undergoing two different hybridization or (B) same reference RNA from two different labeling runs. Typical scatter plots of data obtained from two patients (C) or a patient and a control (D). Pearson correlation coefficient is indicated in each scatter plot. doi:10.1371/journal.pone.0006803.g001 Figure 1. Scatter plots representation of signal intensities. Scatter plot of technical replicates: (A) same reference RNA undergoing two different hybridization or (B) same reference RNA from two different labeling runs. Typical scatter plots of data obtained from two patients (C) or a patient and a control (D). Pearson correlation coefficient is indicated in each scatter plot. doi:10.1371/journal.pone.0006803.g001 signals are explained by the patient status. One contrast was built to determine an expression difference between controls and RA patients [22]. Since the number of individuals is large, the residual variance was used to calculate the statistic test [23]. The raw P values were adjusted by the Benjamini-Hochberg procedure, which controls the False Discovery Rate (FDR) [24]. For the contrast, a gene is considered differentially expressed if the Benjamini-Hochberg-corrected P value is less than 0.05. the end of the amplification reaction, melting curve analyses were performed to confirm the specificity as well as the integrity of the PCR products by the presence of a single peak. Gene-specific primers were designed inside or nearby the microarray sequence targeted, using Primer Express Software (PE Applied Biosystems). Primers sequences of all genes analyzed are provided in Table S1. Absence of cross contamination and primer dimer was checked on genomic DNA and water. From a list of 8 housekeeping genes (HKG), we chose HMBS and ALDOA, which meet the criteria of less variation between samples and compatible expression level with the studied genes. The geometric mean of HKG expression was used to normalize the expression of genes of interest [21]. Real-Time PCR Standard curves were generated from assays made with serial dilutions of reference cDNA to calculate PCR efficiencies (100%+/215%, with r2 . = 0.997). Ct samples were transformed into quantity values using the formula (1+Efficiency)Ct. Only means of triplicate with a CV of less than 10% were analyzed. Inter-plate variation was below 8%. the end of the amplification reaction, melting curve analyses were performed to confirm the specificity as well as the integrity of the PCR products by the presence of a single peak. Gene-specific primers were designed inside or nearby the microarray sequence targeted, using Primer Express Software (PE Applied Biosystems). Primers sequences of all genes analyzed are provided in Table S1. Absence of cross contamination and primer dimer was checked on genomic DNA and water. From a list of 8 housekeeping genes (HKG), we chose HMBS and ALDOA, which meet the criteria of less variation between samples and compatible expression level with the studied genes. The geometric mean of HKG expression was used to normalize the expression of genes of interest [21]. Standard curves were generated from assays made with serial dilutions of reference cDNA to calculate PCR efficiencies (100%+/215%, with r2 . = 0.997). Ct samples were transformed into quantity values using the formula (1+Efficiency)Ct. Only means of triplicate with a CV of less than 10% were analyzed. Inter-plate variation was below 8%. Genes that were expressed at significantly different levels between patients and controls were analyzed by supervised hierarchical clustering (uncentered correlation, complete linkage) [25] to visualize the correlation of co-expressed genes in Treeview (available at http://rana.lbl.gov/EisenSoftware.htm). For an interpretation of the biological processes that are represented by the genes that show a significantly different level of expression in RA patients compared to the controls, we applied Gene Ontology analysis in the PANTHER database at http:// www.pantherdb.org (Applied Biosystems) [26]. PANTHER uses the binomial statistics tool to compare our gene list to a reference list (NCBI: Homo sapiens genes) determining the statistically significant over- or under- representation of PANTHER biological process [27]. After, for each biological process in PANTHER, the Gene expression profiling in PMBCs of RA patients p p g p Genome-wide transcriptional profiles of PBMCs from 18 RA patients and 15 age and sex-matched controls were measured on microarrays that contain 48,701 unique 50-mer oligonucleotides in total, with a mean ,30 hybridizations per sequence. Data were analyzed using ANOVA. Using this test with a FDR of 5% we identified 380 transcripts with significant expression. The proportion of detected transcripts was substantially higher among RefSeq genes (91%) than non-RefSeq genes (9%), reflecting the greater degree of knowledge and certainty about the existence of RefSeq transcripts. Four genes, represented more than once in this list, were averaged from sequences with the same Unigene identifier. Significant difference in expression level between the two groups was observed for 339 RefSeq genes. Among them, 238 were downregulated (Table S2) and 101 were upregulated (Table S3). The significant gene expression differences between RA patients and controls were visualized in a cluster diagram (Figure 2). Genes upregulated in RA To categorize the up-regulated 101 genes into functional biological groups we used the PANTHER database (described in Methods section). We observed an elevated expression of a spectrum of genes involved in Immunity and Defense, nucleoside, nucleotide and nucleic acid metabolism, signal transduction, protein metabolism and modification, mRNA transcription, transport and developmental processes in the peripheral blood of RA patients compared to controls. Compared to a NCBI Homo sapiens reference list, the differentially up-regulated genes list revealed three biological processes significantly over represented (P,0.05) (Figure 3A). Then, the fold change of the genes associated to each biological process was compared to the overall distribution of fold changes. Immunity and Defense was the only significant functional biological process after Bonferroni correction (P = 0.03) (Figure 3B). This cluster of genes involved in Immunity and Defense process contains the S100 family proteins S100 calcium-binding protein A8 (S100A8), S1000A9 and S100A12, the orosomucoid family proteins ORM1 and ORM2, as well as other inflammatory mediators like lymphocyte antigen 96 (LY96), cathelicidin antimicrobial peptide (CAMP), thioredoxin (TXN), allograft inflammatory factor 1 (AIF1), nuclear factor of activated T-cells 5 (NFAT5), F5 (coagulation factor V), SLC11A1 (solute carrier family 11, member 1) and PGLYRP1 (peptidoglycan recognition protein 1). These genes are characterized in Table 2. Figure 2. Cluster diagram of the expression of 339 significantly expressed genes in 18 RA patients and 15 controls. Genes are organized by hierarchical clustering based on overall similarity in expression patterns. Red represents relative expression greater than the median expression level across all samples, and green represents an expression level lower than the median. Black indicates intermediate expression. doi:10.1371/journal.pone.0006803.g002 Statistical analysis Statistical analysis on microarray data was performed using one- way analysis of variance (ANOVA) per gene where the normalized PLoS ONE \| www.plosone.org August 2009 \| Volume 4 \| Issue 8 \| e6803 3 RA Gene Expression Profiling genes associated with that term are evaluated according to the likelihood that their fold changes were drawn randomly from the overall distribution of fold changes. The Mann-Whitney U Test (Wilcoxon Rank-Sum Test) is used to determine the P value that, say, if specific biological process genes have random fold changes relative to the overall list of values that was input. A significant P value indicates that the distribution (fold change) for this category is non-random and different from the overall distribution [28]. In both statistical tests, processes with a P value ,0.05 were considered significant after Bonferroni correction which was applied to adjust for multiple comparisons. Correlations between two set of data were measured using Pearson coefficient. Figure 2. Cluster diagram of the expression o expressed genes in 18 RA patients and 15 organized by hierarchical clustering based on overall s Figure 2. Cluster diagram of the expression of 339 significantly expressed genes in 18 RA patients and 15 controls. Genes are organized by hierarchical clustering based on overall similarity in expression patterns. Red represents relative expression greater than the median expression level across all samples, and green represents an expression level lower than the median. Black indicates intermediate expression. doi:10.1371/journal.pone.0006803.g002 Results Gene expression profiling in PMBCs of RA patients Genes downregulated in RA Compared to the same NCBI Homo sapiens reference list, the mainstream of the genes that showed a lower expression in RA PLoS ONE \| www.plosone.org August 2009 \| Volume 4 \| Issue 8 \| e6803 4 Figure 3. Gene ontology analysis of the most representative biological processes represented by the 101 up-regulated genes in RA patients. Gene Ontology (GO) analysis in the PANTHER database was applied for an interpretation of the biological processes that are represented by the genes showing a higher significantly different expression level in RA patients compared to the controls. A and B - Biological processes with a P value ,0.05 (red bars) were considered significant after Bonferroni correction. The GO analysis P-value was plotted on the y axis versus biological processes on the x axis. A - The binomial statistics tool was used to compare our gene list to a reference list (NCBI: Homo sapiens genes) determining the statistically significant (P value) over- or under- representation of PANTHER biological process. B - The Mann-Whitney U Test was used to determine significant P values which indicate that the distribution (fold change) for each biological process is non-random and different from the overall distribution. OP - Oxidative Phosphorylation; ET - Electron Transport; ID - Immunity and Defense; NNNAM - Nucleoside, Nucleotide and Nucleic Acid Metabolism; T – Transport. doi:10.1371/journal.pone.0006803.g003 RA Gene Expression Profiling RA Gene Expression Profiling Figure 3. Gene ontology analysis of the most representative biological processes represented by the 1 Figure 3. Gene ontology analysis of the most representative biological processes represented by the 101 up-regulated genes in RA patients. Gene Ontology (GO) analysis in the PANTHER database was applied for an interpretation of the biological processes that are represented by the genes showing a higher significantly different expression level in RA patients compared to the controls. A and B - Biological processes with a P value ,0.05 (red bars) were considered significant after Bonferroni correction. The GO analysis P-value was plotted on the y axis versus biological processes on the x axis. A - The binomial statistics tool was used to compare our gene list to a reference list (NCBI: Homo sapiens genes) determining the statistically significant (P value) over- or under- representation of PANTHER biological process. Genes downregulated in RA B - The Mann-Whitney U Test was used to determine significant P values which indicate that the distribution (fold change) for each biological process is non-random and different from the overall distribution. OP - Oxidative Phosphorylation; ET - Electron Transport; ID - Immunity and Defense; NNNAM - Nucleoside, Nucleotide and Nucleic Acid Metabolism; T – Transport. doi:10.1371/journal.pone.0006803.g003 p doi:10.1371/journal.pone.0006803.g003 regulated genes list. However, no significant biological process was found (after Bonferroni correction) when we evaluate all the down-regulated genes associated to biological processes, according with their fold changes drawn randomly from the overall distribution of fold changes (data not shown). patients are linked to different biological processes such as nucleoside, nucleotide and nucleic acid metabolism, mRNA transcription and regulation, cell cycle, intracellular protein traffic. A small number of down-regulated genes were also involved in oncogenesis like runt-related transcription factor 3 (RUNX3), SMAD family member 7 (SMAD7), PHD finger protein 17 (PHF17) as well as interleukin-1 receptor-associated kinase 2 (IRAK2), interleukin 2 receptor, beta (IL2RB), CD96 and SH2B adaptor protein 3 (SH2B3) related to immune functions. Therefore, in PANTHER database classification we found four significant functional biological processes in our 238 down- Discussion Microarray technology has been used to discriminate differenc- es in gene expression profiles in tissues and PBMCs. Both synovial tissue and PBMCs have been used to evaluate differences in the gene expression profiles in RA [7,8]. If expression-based profiling is to be of practical importance, sample accessibility becomes crucial. In this context, peripheral mononuclear cells are key sentinels of host defence, being used to identify novel disease mediators, disease variants and treatment responses [7,29,30]. Transcriptome studies using Illumina and other technologies showed that each approach was able to detect specific genes, y ( g ) The S100 calcium-binding proteins (S100A8, S100A9, and S100A12) are characterized by strong prevalence in cells of myeloid origin. Activated phagocytes expressing S100A8 and S100A9 proteins are among the first cells infiltrating inflammatory lesions in the synovium [31]. The expression of S100A8 and S100A9 was found to be strongest at the cartilage-pannus junction, which is the prime site of cartilage destruction and bone erosion in arthritis [32]. S100A12 is strongly expressed in inflamed synovial tissue, whereas it is nearly undetectable in synovia of control subjects or patients after successful treatment [33]. Further studies showed increased levels of S100A8/S100A9 and S100A12 concentrations in synovial fluid (SF) and serum in RA patients [34–36]. In addition, S100A9, S100A8 and S100A12 levels were associated with body mass index, presence of ACPA and RF and presence of ACPA, respectively [37]. Furthermore, several expression profiling studies in PBMCs showed a highly expression of S100A8, S100A9 and S100A12 in RA patients compared to controls [8,38,39]. Therefore, pro-inflammatory S100 proteins are attractive therapeutic targets for immune interventions in the treatment of RA. Figure 4. Correlation between microarray and real time PCR data. The scatter plot compares mean expression of RA patients/ controls ratio for nine genes. Each point represents the RA patients/ controls ratio from the microarray (y axis) and real time PCR (x axis). Pearson correlation coefficient is indicated in the scatter plot. doi:10.1371/journal.pone.0006803.g004 Figure 4. Correlation between microarray and real time PCR data. The scatter plot compares mean expression of RA patients/ controls ratio for nine genes. Each point represents the RA patients/ controls ratio from the microarray (y axis) and real time PCR (x axis). Pearson correlation coefficient is indicated in the scatter plot. doi:10.1371/journal.pone.0006803.g004 Human allograft inflammatory factor-1 (AIF-1) is a Ca2+- binding EF-hand protein encoded within the major histocompat- ibility complex class (MHC) III region of chromosome 6. Real time PCR validation In all samples, we confirmed the expression of five up-regulated genes (LY96, S100A12, ORM2, ORM1, RPL31) and four down regulated genes (IL2RB, DNMT1, RUNX3 and IRF1) in RA patients, by real-time PCR. From microarray and real time PCR PLoS ONE \| www.plosone.org August 2009 \| Volume 4 \| Issue 8 \| e6803 August 2009 \| Volume 4 \| Issue 8 \| e6803 August 2009 \| Volume 4 \| Issue 8 \| e6803 5 RA Gene Expression Profiling Table 2. Differentially up-regulated transcripts linked to Immunity and Defense biological process in RA patients. Gene Bank Name Genome Location GeneID (NCBI) Transcript Identifier Fold Change RA vs Controls S100A8 S100 calcium binding protein A8 1q21 6279 NM_002964.3 2.9 S100A9 S100 calcium binding protein A9 1q21 6280 NM_002965.2 2.0 S100A12 S100 calcium binding protein A12 1q21 6283 NM_005621.1 1.8 AIF1 Allograft inflammatory factor 1 6p21.3 199 NM_001623.3 1.6 TXN Thioredoxin 9q31 7295 NM_003329.1 2.1 NFAT5 Nuclear factor of activated T-cells 5, tonicity-responsive 16q22.1 10725 NM_173215.1 1.3 CAMP (LL37) Cathelicidin antimicrobial peptide 3p21.3 820 NM_004345.3 2.5 LY96 (MD-2) Lymphocyte antigen 96 8q21.11 23643 NM_015364.2 2.6 ORM1 Orosomucoid 1 9q31–q32 5004 NM_000607.1 2.7 ORM2 Orosomucoid 2 9q32 5005 NM_000608.2 1.5 SLC11A1 (NRAMP1) Solute carrier family 11, member 1 2q35 6556 NM_000578.3 1.5 PGLYRP1 Peptidoglycan recognition protein 1 19q13.2–q13.3 8993 NM_005091.1 2.0 F5 Coagulation factor V 1q23 2153 NM_000130.4 1.7 doi:10.1371/journal.pone.0006803.t002 meaning an increase in knowledge by each platform [16–18]. To complete previous studies on RA with Affymetrix or double colour microarrays, we decided to use Illumina technology. Our study did not confirm a specific expression for the genes regulated by interferon type I, as described in a RA large-scale expression profiling [8] (two colours analysis protocol using Stanford University microarrays) (data not shown). Our analysis revealed only one significantly increased biological mechanism: Immunity and Defense. This process was already highlighted by other studies in RA, as several genes that we described (S100A8, S1000A9, S100A12 and AIF1). Additionally, our study identified new genes like LY96/MD-2, NFAT5, TXN, CAMP/LL37, ORM1, ORM2, SLC11A1, PGLYRP1 and F5. These genes are involved in different functional Immunity and Defense related mechanisms as pro- inflammation, anti-microbial activity, oxidative and osmotic cellular stress and immunomodulatory functions (Figure 5). data, we calculated the RA patients/controls ratio for each genes expression. The qPCR expression data of the nine genes showed a high correlation with the microarray expression data (r = 0.937) (Figure 4). Discussion AIF-1 is produced by macrophages and lymphocytes, and its synthesis is PLoS ONE \| www.plosone.org August 2009 \| Volume 4 \| Issue 8 \| e6803 August 2009 \| Volume 4 \| Issue 8 \| e6803 6 RA Gene Expression Profiling Figure 5. Biological functions of Immunity and Defense genes highlighted in our Rheumatoid Arthritis study. The genes differentially expressed in Immunity and Defense process are stratified in four functional related mechanisms: pro-inflammation, anti-microbial activity, oxidative and osmotic cellular stress and immunomodulatory functions. doi:10.1371/journal.pone.0006803.g005 Figure 5. Biological functions of Immunity and Defense genes highlighted in our Rheumatoid Arthritis study. The genes differentially expressed in Immunity and Defense process are stratified in four functional related mechanisms: pro-inflammation, anti-microbial activity, oxidative and osmotic cellular stress and immunomodulatory functions. doi:10.1371/journal.pone.0006803.g005 NFAT5, the primordial member of the NFAT family, is expressed by almost all cells and is activated in response to osmotic stress. In lymphocytes, NFAT5 controls the osmotic stress-induced expression of several cytokines, including tumour-necrosis factor (TNF) and lymphotoxin-b. NFAT5-deficient mice have impaired T-cell function under hyperosmotic conditions and decreased cellularity of the thymus and spleen [48]. In RA, NFAT5 mRNA is expressed in proliferating RA-synovial fibroblasts (RA-SF) but not in nonproliferating RA-SF. Furthermore, NFAT5 mRNA is expressed in RA synovium - but not in normal individuals - as well as at sites of bone destruction. NFAT5 could be then related not only with proliferation but also with the activation and invasion of RA-SF in vivo [49]. mediated by several cytokines, such as interferon-c (IFN-c) [40]. Kimura et al. have demonstrated that AIF-1 is expressed in synovial and mononuclear cells in RA synovial tissue and increases proliferation of cultured synoviocytes and IL-6 production by these synoviocytes and PBMCs [41]. A recent study, showed an increased expression of AIF1 mRNA in RA patients PMBCs compared with controls as well as in synovial macrophages in the lining layer of all the inflamed RA synovial membranes compared with non-inflamed OA controls [42]. AIF-1 plays therefore an important role in the pathogenesis of RA by affecting key processes such as the activation of synovial cell proliferation and the inflammatory cytokine cascade including IL-6 in joints and may represent a new molecular target in RA therapy. The LY96 (MD-2) acts as an extracellular adaptor protein in the activation of TLR4 by the lipopolysaccharide (LPS) of Gram negative bacteria, a well-known inducer of the innate immune response [43]. Discussion In RA, recent studies have described pDCs as perpetuators of synovial inflammation and modulators of B cell responses in the synovial tissue [57]. The mRNA up-regulation of the LL37 gene found in RA patients comparing to controls in our study could be due to an immunological response to infectious agents such as bacteria and viruses. Moreover the LL37 over-expression could be implicated in a pDCs-dependent mechanism involved in the perpetuation of RA inflammation through the abolition of self-tolerance and subsequent emergence of self-reactive lymphocytes. Human alpha-1-acid glycoprotein (AGP) - also called oroso- mucoid - is a 37-kDa molecule consisting of a heavily glycosylated single polypeptide chain. Alpha-1-acid glycoprotein 1 (AGP1) and alpha-1-acid glycoprotein 2 (AGP2), coded by ORM1 and ORM2 genes respectively, are positive acute phase proteins [58,59]. AGPs plasma concentration may increase several fold during acute phase reactions such as inflammation or chronic disease. AGPs have a strong immunomodulatory function [60]. It was shown that under pathological conditions not only the total concentration of AGPs but its glycosylation pattern may be altered [59]. Smith and colleagues have demonstrated that the AGPs populations in the serum and synovial fluid of RA patients are distinct in terms of glycosylation pattern. This discovery has direct functional significance since only the serum AGPs population is capable of blocking leucocyte adhesion [61]. Furthermore, Haston et al. have shown that AGPs can influence MMP-13 activity. It is hypothesized that AGPs may form part of a negative feedback mechanism which is inadequate to prevent disease progression in rheumatoid arthritis. These processes may exacerbate the increased turnover of collagen characteristic of the disease [62]. These results suggest an interesting role for AGP in RA pathogenesis. The multi-step coagulation complex system is activated by tissue factor (TF), which is exposed to blood. In this process, factor 5 (FV) is cleaved and activates factor Va (FVa). After several proteins interactions, the prothrombinase complex (FXa–FVa) converts prothrombin to thrombin which is generated in a large amount. Thrombin activates FV and FVIII and platelets converting fibrinogen to a fibrin clot [70]. Accumulation of fibrin in the RA synovium exceeds that in control tissue by a wide margin and represents one of the most striking pathologic features of rheumatoid synovitis. For some time, this fibrin deposition has been considered to be a serious contributor to permanent damage by maintaining a vicious circle of inflammation [71,72]. Discussion have demonstrated that human TXN is a novel target gene induced by IFN-c [54]. RA is a Th1-driven disease which has IFN-c as characteristic Th1 cytokine and subsequently the TXN mRNA up-regulation in PBMCs could suggest an involvement in RA IFN-c pathway. Moreover, Kim et al. have demonstrated that human TXN is a novel target gene induced by IFN-c [54]. RA is a Th1-driven disease which has IFN-c as characteristic Th1 cytokine and subsequently the TXN mRNA up-regulation in PBMCs could suggest an involvement in RA IFN-c pathway. in inflammation and tissue destruction of RA [66,67]. Bacterial or viral infection may play a role in triggering the development of RA and TNFa and iNOS are key players in enhanced antimicrobial activity of activated macrophages [65]. PGRPs are innate immunity proteins, recognizing bacterial peptidoglycan, and acting in antibacterial immunity. PGRP-1 seems to belong to the innate immune arm of effectors molecules, such as antimicrobial peptides and C-type lectins, among others. This protein was shown to be almost exclusively present as a soluble protein in the granules of polymorphonuclear leucocytes (PMN) [68]. Saha et al. examined the immunomodulating activities of the PGRPs in a peptidoglycan-induced arthritis mice model. They showed that a systemic injection of peptidoglycan (PG) or muramyldipeptide (MDP) induces an acute arthritis of the joints of the feet in BalbC mice. Peptidoglycan-induced arthritis PGLYRP-12/2 mice, had a MDP-induced activation of proin- flammatory genes than WT mice. Moreover, PGLYRP-12/2 mice have longer-lasting MDP-induced arthritis than WT mice. The anti-inflammatory function of PGLYRP-1 manifests itself only in the later stages of MDP-induced arthritis, which is consistent with the local release of PGLYRP-1 from PMN granules after PMNs’ arrival into the mice foot [69]. These data point to that PGRP-1 could have a specialized but nevertheless significant role in signalling events like arthritis in mammals. CAMP (LL37) is an antimicrobial peptide which has a broad of antimicrobial activity. LL-37 has the potential to participate in the innate immune response both by killing bacteria and by recruiting a cellular immune response [55]. Gilliet and Lande recently found that LL37, overexpressed in psoriatic skin, is the key mediator of plasmacytoid dendritic cells (pDCs) activation in psoriasis. LL37 converts nonstimulatory self-DNA into a potent trigger of pDCs to produce IFN. pDCs respond to self-DNA if coupled with an antimicrobial peptide, suggesting that modified self-DNA drives autoimmunity in psoriasis by activating TLR9 [56]. Discussion In the extravascular coagulation at the arthritic synovial joint sequential activity of factor Xa (in the presence of cofactor Va) and of thrombin leads to fibrin deposition in the joint [73]. Thus, mRNA F5 up-regulation in RA patients could enhance the production of factor V that under cleavage produces factor Va. This Va increase could lead to the augmentation of thrombin and consequently fibrin in RA joints. In conclusion, our study highlighted several new genes (LY96/ MD-2, NFAT5, TXN, CAMP/LL37, ORM1, ORM2, SLC11A1, PGLYRP1 and F5) in PBMCs of RA patients that could contribute in the identification of innovative clinical biomarkers for diagnostic procedures and therapeutic interventions. Nevertheless, compar- ative analysis with another disease involving an inflammatory process could clarify the relation between the expression profiling and the pathophysiological processes specifically involved in Rheumatoid Arthritis.z p g SLC11A1 (formerly called NRAMP1) is a gene that is important in macrophage-mediated natural resistance to a variety of intracellular pathogens. Exogenous and endogenous agents that mediate inflammation by activating the macrophage can cause NRAMP1 translocation to the membrane of the phagolysosome, where it serves as a cation transporter. The significant increase in iron deposition observed in the synovial membrane of rheumatoid arthritis patients, and foam cells in atherosclerotic lesions, could be attributable to NRAMP1 [63]. In human RA synovium NRAMP1 was detected in macrophages and neutrophils in the linning and subinitimal zone, as well as in inflammatory infiltrates, but was absent in fibroblasts [64]. NRAMP1 has also pleiotropic effects on macrophage function, including upregulation of chemokine/ cytokine gene, TNFa, IL-1b, inducible nitric oxide synthase (iNOS), MHC expression as well as tumoricidal and antimicrobial activity. These effects are involved in resistance to infection and may also be involved in induction and maintenance of autoimmune disease [65]. TNFa and IL-1b play important roles Discussion Knockout studies in mice have demonstrated that MD-2 is indispensable for LPS responses [44]. Multiple evidences points to the potential importance of TLR signalling in RA pathogenesis by means of the presence of different TLR ligands and functional TLR receptors in inflamed joints from patients with RA [45]. TLR4 mRNA is highly expressed in the synovium at early stages of RA as well as at later stages of the disease. In vitro stimulation of the RA synovial fibroblasts with TLR4 ligand (LPS) produces a wide range of proinflammatory cytokines, chemokines, and tissue destructive enzymes [46]. Recently, the potential value of TLR4 or signals derived from this receptor as therapeutic targets has become clearer - mainly as a result of studies in animal models of joint inflammation but also in human RA [47]. Interestingly, in our study, mRNA TLR4 was up-regulated in RA patients compared to controls but not significantly. Thus, the TLR4-MD2-LPS complex could be involved in the activation of synovial fibroblasts and contribute to the development of synovial inflammation and joint destruction. Oxidative stress to essential cell components caused by oxygen free radicals is generally considered as a serious mechanism in RA pathogenicity [50]. Thioredoxin (TXN), a cellular reducing catalyst induced by oxidative stress, is involved in the stimulation of the DNA-binding activity of NF-kb transcription factor [51]. Increased cytokine production driven by NF-kb can enhance expression of vascular adhesion molecules that attract leucocytes into the joint, as well as matrix metalloproteinases that help to degrade the extracellular matrix [50]. TXN concentrations were found significantly elevated in SF and serum of RA patients. The positive correlation between the SF TXN and the serum C Reactive Protein in the absence of a high concentration of SF TNF-a may indicate that TXN is involved in the prolongation and persistence of the RA inflammation, because high concentrations of TXN could stimulate NF-kB activation in the presence of the otherwise insufficient concentration of TNF-a [52]. Thus, TXN monitoring in RA patients could provide useful information regarding the extent of oxidative stress. Furthermore, truncated thioredoxin (Txn80) stimulates monocytes/macrophages to induce IL-12, implying that it is involved in immune inflammatory reactions directing Th1 immunity and IFN-c production [53]. PLoS ONE \| www.plosone.org August 2009 \| Volume 4 \| Issue 8 \| e6803 7 PLoS ONE \| www.plosone.org RA Gene Expression Profiling Moreover, Kim et al. References 25. Eisen MB, Spellman PT, Brown PO, Botstein D (1998) Cluster analysis and display of genome-wide expression patterns. Proc Natl Acad Sci U S A 95: 14863–14868. 1. Seldin MF, Amos CI, Ward R, Gregersen PK (1999) The genetics revolution and the assault on rheumatoid arthritis (review). Arthritis Rheum 42: 1071–1079. 26. Mi H, Lazareva-Ulitsky B, Loo R, Kejariwal A, Vandergriff J, et al. (2005) The PANTHER database of protein families, subfamilies, functions and pathways. Nucleic Acids Res 33: D284–288. 2. Firestein GS (2003) Evolving concepts of rheumatoid arthritis [review]. Nature 423: 356–361. 3. Zendman AJ, van Venrooij WJ, Pruijn GJ (2006) Use and significance of anti- CCP autoantibodies in rheumatoid arthritis. Rheumatology (Oxford) 45: 20–25. 27. Cho RJ, Campbell MJ (2000) Transcription, genomes, function. Trends in Genetics 16: 409–415. 4. Hueber W, Kidd BA, Tomooka BH, Lee BJ, Bruce B, et al. (2005) Antigen microarray profiling of autoantibodies in rheumatoid arthritis. Arthritis Rheum 52: 2645–2655. 28. Clark AG, Glanowski S, Nielsen R, Thomas PD, Kejariwal A, et al. (2003) Inferring nonneutral evolution from human-chimp-mouse orthologous gene trios. Science 302(5652): 1960–1963. 5. Lipsky PE, van der Heijde DM, St Clair EW, Furst DE, Breedveld FC, et al. (2000) Infliximab and methotrexate in the treatment of rheumatoid arthritis. Anti-Tumor Necrosis Factor Trial in Rheumatoid Arthritis with Concomitant Therapy Study Group. N Engl J Med 343: 1594–1602. 29. Shaffer AL, Wright G, Yang L, Powell J, Ngo V, et al. (2006) A library of gene expression signatures to illuminate normal and pathological lymphoid biology. Immunol Rev 210: 67–85. 30. Yamagata T, Benoist C, Mathis DA (2006) Shared gene-expression signature in innate-like lymphocytes. Immunol Rev 210: 52–66. 6. Edwards JC, Szczepanski L, Szechinski J, Filipowicz-Sosnowska A, Emery P, et al. (2004) Efficacy of B-cell-targeted therapy with rituximab in patients with rheumatoid arthritis. N Engl J Med 350: 2572–2581. 31. Foell D, Roth J (2004) Proinflammatory S100 proteins in arthritis and autoimmune disease. Arthritis Rheum 50(12): 3762–3771. 7. Toonen EJ, Barrera P, Radstake TR, van Riel PL, Scheffer H, et al. (2008) Gene expression profiling in rheumatoid arthritis: current concepts and future directions. Ann Rheum Dis 67(12): 1663–1669. 32. Youssef P, Roth J, Frosch M, Costello P, Fitzgerald O, et al. (1999) Expression of myeloid related proteins (MRP) 8 and 14 and the MRP8/14 heterodimer in rheumatoid arthritis synovial membrane. J Rheumatol 26: 2523–2528. 8. References van der Pouw Kraan TC, Wijbrandts CA, van Baarsen LG, Voskuyl AE, Rustenburg F, et al. (2007) Rheumatoid arthritis subtypes identified by genomic profiling of peripheral blood cells: assignment of a type I interferon signature in a subpopulation of patients. Ann Rheum Dis 66(8): 1008–1014. 33. Foell D, Kane D, Bresnihan B, Vogl T, Nacken W, et al. (2003) Expression of the pro-inflammatory protein S100A12 (EN-RAGE) in rheumatoid and psoriatic arthritis. Rheumatology (Oxford) 42: 1383–1389. 9. Baechler EC, Batliwalla FM, Reed AM, Peterson EJ, Gaffney PM, et al. (2006) Gene expression profiling in human autoimmunity. Immunol Rev 210: 120–137. 34. Frosch M, Strey A, Vogl T, Wulffraat NM, Kuis W, et al. (2000) Myeloid-related proteins 8 and 14 are specifically secreted during interaction of phagocytes and activated endothelium and are useful markers for monitoring disease activity in pauciarticular-onset juvenile rheumatoid arthritis. Arthritis Rheum 43: 628–637. 10. Alizadeh AA, Eisen MB, Davis RE, Ma C, Lossos IS, et al. (2000) Distinct types of diffuse large B-cell lymphoma identified by gene expression profiling. Nature 403(6769): 503–511. 35. Kane D, Roth J, Frosch M, Vogl T, Bresnihan B, et al. (2003) Increased perivascular synovial membrane expression of myeloid related proteins in psoriatic arthritis. Arthritis Rheum 48: 1676–1685. 11. Brutsche MH, Brutsche IC, Wood P, Brass A, Morrison N, et al. (2001) Apoptosis signals in atopy and asthma measured with cDNA arrays. Clin Exp Immunol 123(2): 181–187. 36. Rouleau P, Vandal K, Ryckman C, Poubelle PE, Boivin A, et al. (2003) The calcium-binding protein S100A12 induces neutrophil adhesion, migration, and release from bone marrow in mouse at concentrations similar to those found in human inflammatory arthritis. Clin Immunol 107: 46–54. 12. Mandel M, Achiron A (2006) Gene expression studies in systemic lupus erythematosus. Lupus 15(7): 451–456. y p ( ) 13. Henriksen PA, Kotelevtsev Y (2002) Application of gene expression profiling to cardiovascular disease. Cardiovasc Res 54(1): 16–24. 37. Chen YS, Yan W, Geczy CL, Brown MA, Thomas R (2009) Serum levels of soluble receptor for advanced glycation end products and of S100 proteins are associated with inflammatory, autoantibody, and classical risk markers of joint and vascular damage in rheumatoid arthritis. Arthritis Res Ther 11: R39. 14. Colangelo V, Schurr J, Ball MJ, Pelaez RP, Bazan NG, et al. (2002) Gene expression profiling of 12633 genes in Alzheimer hippocampal CA1: transcription and neurotrophic factor down-regulation and up-regulation of apoptotic and pro-inflammatory signaling. References J Neurosci Res 70(3): 462–473. 38. Bovin LF, Rieneck K, Workman C, Nielsen H, Sørensen SF, et al. (2004) Blood cell gene expression profiling in rheumatoid arthritis. Discriminative genes and effect of rheumatoid factor. Immunol Lett 93: 217–226. apoptotic and pro-inflammatory signaling. J Neurosci Res 70( 15. Gladkevich A, Nelemans SA, Kauffman HF, Korf J (2005) Microarray profiling of lymphocytes in internal diseases with an altered immune response: potential and methodology. Mediators Inflamm 14: 317–330. 39. Batliwalla FM, Baechler EC, Xiao X, Li W, Balasubramanian S, et al. (2005) Peripheral blood gene expression profiling in rheumatoid arthritis. Genes Immun 6: 388–397. gy 16. MAQC Consortium (2006) The MicroArray Quality Control (MAQC) project shows inter- and intraplatform reproducibility of gene expression measurements. Nat Biotechnol 24(9): 1151–161. 40. Utans U, Arceci RJ, Yamashita Y, Russell ME (1995) Cloning and characterization of allograft inflammatory factor-1: a novel macrophage factor identified in rat cardiac allografts with chronic rejection. J Clin Invest 95: 2954–2962. 17. Barnes M, Freudenberg J, Thompson S, Aronow B, Pavlidis P (2005) Experimental comparison and cross-validation of the Affymetrix and Illumina gene expression analysis platforms. Nucleic Acids Res 33(18): 5914–5923. 41. Kimura M, Kawahito Y, Obayashi H, Ohta M, Hara H, et al. (2007) A critical role for allograft inflammatory factor-1 in the pathogenesis of rheumatoid arthritis. J Immunol 178(5): 3316–3322. 18. Maouche S, Poirier O, Godefroy T, Olaso R, Gut I, et al. (2008) Performance comparison of two microarray platforms to assess differential gene expression in human monocyte and macrophage cells. BMC Genomics 9: 302. 42. Harney SM, Vilarin˜o-Gu¨ell C, Adamopoulos IE, Sims AM, Lawrence RW, et al. (2008) Fine mapping of the MHC Class III region demonstrates association of AIF1 and rheumatoid arthritis. Rheumatology (Oxford) 47(12): 1761–1767. 19. Arnett FC, Edworthy SM, Bloch DA, McShane DJ, Fries JF, et al. (1988) The American Rheumatism Association 1987 revised criteria for the classification of rheumatoid arthritis. Arthritis Rheum 31: 315–324. 43. Viriyakosol S, Tobias PS, Kitchens RL, Kirkland TN (2001) MD-2 binds to bacterial lipopolysaccharide. J Biol Chem 276: 38044–38051. 20. Pruitt KD, Tatusova T, Maglott DR (2007) NCBI reference sequences (RefSeq): a curated non-redundant sequence database of genomes, transcripts and proteins. Nucleic Acids Res 35: D61–65. 44. Nagai Y, Akashi S, Nagafuku M, Ogata M, Iwakura Y, et al. (2002) Essential role of MD-2 in LPS responsiveness and TLR4 distribution. Nat Immunol 3(7): 667–672. 21. Supporting Information Table S1 Primers sequences of the nine genes analyzed by real- time PCR Found at: doi:10.1371/journal.pone.0006803.s001 (0.04 MB DOC) Table S2 List of 238 downregulated genes differentially expressed between RA patients and controls Found at: doi:10.1371/journal.pone.0006803.s002 (0.28 MB DOC) PLoS August 2009 \| Volume 4 \| Issue 8 \| e6803 August 2009 \| Volume 4 \| Issue 8 \| e6803 PLoS ONE \| www.plosone.org 8 RA Gene Expression Profiling Table S3 List of 101 upregulated genes differentially expressed between RA patients and controls Table S3 List of 101 upregulated genes differentially expressed between RA patients and controls Author Contributions Conceived and designed the experiments: VHT EPT. Performed the experiments: VHT RO. Analyzed the data: VHT RO MLMM EPT. Contributed reagents/materials/analysis tools: VHT RO SL LJ PH IG FC EPT. Wrote the paper: VHT RO MLMM CRO IG EPT. p Found at: doi:10.1371/journal.pone.0006803.s003 (0.14 MB DOC) Found at: doi:10.1371/journal.pone.0006803.s003 (0.14 MB DOC) Acknowledgments We thank Jeanne-Antide Perrier and Magali de Surville for microarray and real time PCR technical assistance. We are grateful to the RA patients and individual controls for their participation in this study. We thank Jeanne-Antide Perrier and Magali de Surville for microarray and real time PCR technical assistance. We are grateful to the RA patients and We thank Jeanne-Antide Perrier and Magali de Surville for microarray and real time PCR technical assistance. We are grateful to the RA patients and individual controls for their participation in this study g p individual controls for their participation in this study. RA Gene Expression Profiling RA Gene Expression Profiling 48. Macian F (2005) NFAT proteins: key regulators of T-cell development and function. Nat Rev Immunol 5(6): 472–484. 61. Smith KD, Pollqcchi A, Field M, Watson J (2002) The heterogeneity of the glycosylation of the alpha-1-acid glycoprotein between the sera and synovial fluid in Rheumatoid Arthritis. Biomed Chromatogr 16: 261–266. 49. Masuda K, Masuda R, Neidhart M, Simmen BR, Michel BA, et al. (2002) Molecular profile of synovial fibroblasts in rheumatoid arthritis depends on the stage of proliferation. Arthritis Res Ther 4(5): R8. 62. Haston JL, FitzGerald O, Kane D, Smith KD (2003) The influence of alpha1- acid glycoprotein on collagenase-3 activity in early rheumatoid arthritis. Biomed Chromatogr 17(6): 361–364. g p ( ) 50. Filippin LI, Vercelino R, Marroni NP, Xavier RM (2008) Redox signalling and the inflammatory response in rheumatoid arthritis. Clin Exp Immunol 152(3): 415–422. 63. Awomoyi AA (2007) The human solute carrier family 11 member 1 protein (SLC11A1): linking infections, autoimmunity and cancer? FEMS Immunol Med Microbiol 49(3): 324–329. 51. Matthews JR, Wakasugi N, Virelizier JL, Yodoi J, Hay RT (1992) Thioredoxin regulates the DNA binding activity of NF-kB by reduction of a disulfide bond involving cysteine 62. Nucleic Acids Res 20(15): 3821–3830. 64. Telfer JF, Brock JH (2002) Expression of ferritin, transferrin receptor, and non- specific resistance associated macrophage proteins 1 and 2 (Nramp1 and Nramp2) in the human rheumatoid synovium. Ann Rheum Dis 61(8): 741–744. 52. Yoshida S, Katoh T, Tetsuka T, Uno K, Matsui N, et al. (1999) Involvement of thioredoxin in rheumatoid arthritis: its costimulatory roles in the TNF-alpha- induced production of IL-6 and IL-8 from cultured synovial fibroblasts. J Immunol 163(1): 351–358. 65. Blackwell JM (1996) Structure and function of the natural-resistance associated macrophage protein (Nramp1), a candidate protein for infectious and autoimmune disease susceptibility. Mol Med Today 2: 205–211. 53. Pekkari K, Avila-Carin˜o J, Bengtsson A, Gurunath R, Scheynius A, et al. (2001) Truncated thioredoxin(Trx80) induces production of interleukin-12 and enhances CD14 expression in human monocytes. Blood 97: 3184–3190. 66. Chu CQ, Field M, Feldmann M, Maini RN (1991) Localization of tumor necrosis factor alpha in synovial tissues and at the cartilage–pannus junction in patients with rheumatoid arthritis. Arthritis Rheum 34: 1125–1132. p y 54. Kim SH, Oh J, Choi JY, Jang JY, Kang MW, et al. (2008) Identification of human thioredoxin as a novel IFN-gamma-induced factor: mechanism of induction and its role in cytokine production. References Vandesompele J, De Preter K, Pattyn F, Poppe B, Van Roy N, et al. (2002) Accurate normalization of real-time quantitative RT-PCR data by geometric averaging of multiple internal control genes. Genome Biol 3(7). 45. Brentano F, Kyburz D, Schorr O, Gay R, Gay S (2005) The role of Toll-like receptor signalling in the pathogenesis of arthritis. Cell Immunol 233: 90–96. 22. Gentleman R, Carey V (2002) Bioconductor. RNews, 2, 1116. 46. Ospelt C, Brentano F, Rengel Y, Stanczyk J, Kolling C, et al. (2008) Overexpression of toll-like receptors 3 and 4 in synovial tissue from patients with early rheumatoid arthritis: toll-like receptor expression in early and longstanding arthritis. Arthritis Rheum 58(12): 3684–3692. 23. Irizarry RA, Hobbs B, Collin F, Beazer-Barclay YD, Antonellis KJ, et al. (2003) Exploration, Normalization, and Summaries of High Density Oligonucleotide Array Probe Level Data. Biostatistics 4(2): 249–264. 47. O’Neill LA (2008) Primer: Toll-like receptor signaling pathways—what do rheumatologists need to know? Nat Clin Pract Rheumatol 4(6): 319–327. 24. Benjamini Y, Hochberg Y (1995) Controlling the false discovery rate: a practical and powerful approach to multiple testing, J R Statistics Soc 57: 289–300. PLoS ONE \| www.plosone.org August 2009 \| Volume 4 \| Issue 8 \| e6803 August 2009 \| Volume 4 \| Issue 8 \| e6803 9 RA Gene Expression Profiling BMC Immunol 9: 64. p 67. Miossec P, Dinarello CA, Ziff M (1986) Interleukin-1 lymphocyte chemotactic activity in rheumatoid arthritis synovial fluid. Arthritis Rheum 29: 461–470. 67. Miossec P, Dinarello CA, Ziff M (1986) Interleukin-1 lymphocyt induction and its role in cytokine production. BMC Immunol 9: 68. Boneca IG (2009) Mammalian PGRPs in the spotlight. Cell Host Microbe 5(2): 109–111. 55. Gallo RL, Murakami M, Ohtake T, Zaiou M (2002) Biology and clinical relevance of naturally occurring antimicrobial peptides. J Allergy Clin Immunol 110(6): 823–831. 69. Saha S, Qi J, Wang S, Wang M, Li X, et al. (2009) PGLYRP-2 and Nod2 are both required for peptidoglycan-induced arthritis and local inflammation. Cell Host Microbe 5: 137–150. ( ) 56. Gilliet M, Lande R (2008) Antimicrobial peptides and self-DNA in autoimmune skin inflammation. Curr Opin Immunol 20(4): 401–407. 70. Dahlba¨ck B, Villoutreix BO (2005) The anticoagulant protein C pathway. FEBS Lett 579(15): 3310–3316. 57. Lebre MC, Tak PP (2009) Dendritic cells in rheumatoid arthritis: Which subset should be used as a tool to induce tolerance? Hum Immunol. In press. should be used as a tool to induce tolerance? Hum Immunol. In 71. Barnhart MI, Riddle JM, Bluhm GB, Quintana C (1967) Fibrin promotion and lysis in arthritic joints. Ann Rheum Dis 26(3): 206–18. 58. Yuasa I, Umetsu K, Vogt U, Nakamura H, Nanba E, et al. (1997) Human orosomucoid polymorphism: molecular basis of the three common ORM1 alleles, ORM1F1, ORM1F2, and ORM1S. Hum Genet 99: -393–398. p y p alleles, ORM1F1, ORM1F2, and ORM1S. Hum Genet 99: -39 72. Weinberg JB, Pippen AMM, Greenberg CS (1991) Extravascular fibrin formation and dissolution in synovial tissue of patients with osteoarthritis and rheumatoid arthritis. Arthritis Rheum 34: 996–1005. 59. Ceciliani F, Pocacqua V (2007) The acute phase protein alpha1-acid glycoprotein: a model for altered glycosylation during diseases. Curr Protein Pept Sci 8: 91–108. rheumatoid arthritis. Arthritis Rheum 34: 996–1005. 73. Busso N, Hamilton JA (2002) Extravascular coagulation and the plasminogen activator/plasmin system in rheumatoid arthritis. Arthritis Rheum 46(9): 2268–2279. p 60. Hochepied T, Berger FG, Baumann H, Libert C (2003) Alpha(1)-acid glycoprotein: an acute phase protein with inflammatory and immunomodulating properties. Cytokine Growth Factor Rev 14(1): 25–34. PLoS ONE \| www.plosone.org August 2009 \| Volume 4 \| Issue 8 \| e6803 10
https://openalex.org/W3085451882	https://boa.unimib.it/bitstream/10281/284449/1/10.1007_s11205-020-02485-7.pdf	English	null	Targeting Policies for Multidimensional Poverty and Social Fragility Relief Among Migrants in Italy, Using F-FOD Analysis	Social indicators research	2,020	cc-by	10,053	* Stefania M. L. Rimoldi stefania.rimoldi@unimib.it 1 23 Social Indicators Research An International and Interdisciplinary Journal for Quality-of-Life Measurement ISSN 0303-8300 Soc Indic Res DOI 10.1007/s11205-020-02485-7 Targeting Policies for Multidimensional Poverty and Social Fragility Relief Among Migrants in Italy, Using F-FOD Analysis Stefania M. L. Rimoldi, Alberto Arcagni, Marco Fattore & Elisa Barbiano di Belgiojoso Targeting Policies for Multidimensional Poverty and Social Fragility Relief Among Migrants in Italy, Using F-FOD Analysis Stefania M. L. Rimoldi, Alberto Arcagni, Marco Fattore & Elisa Barbiano di Belgiojoso Social Indicators Research An International and Interdisciplinary Journal for Quality-of-Life Measurement ISSN 0303-8300 Soc Indic Res DOI 10.1007/s11205-020-02485-7 1 23 Your article is published under the Creative Commons Attribution license which allows users to read, copy, distribute and make derivative works, as long as the author of the original work is cited. You may self- archive this article on your own website, an institutional repository or funder’s repository and make it publicly available immediately. 1 23 Social Indicators Research https://doi.org/10.1007/s11205-020-02485-7 ORIGINAL RESEARCH Targeting Policies for Multidimensional Poverty and Social Fragility Relief Among Migrants in Italy, Using F‑FOD Analysis Stefania M. L. Rimoldi1 · Alberto Arcagni2 · Marco Fattore1 · Elisa Barbiano di Belgiojoso1 Accepted: 28 August 2020 © The Author(s) 2020 Accepted: 28 August 2020 © The Author(s) 2020 Accepted: 28 August 2020 © The Author(s) 2020 Alberto Arcagni alberto.arcagni@unimib.it; alberto.arcagni@uniroma1.it Keywords Multidimensional poverty · F-FOD · Partially ordered sets · Ranking Marco Fattore marco.fattore@unimib.it 1 Introduction In a recent paper by Arcagni et al. (2019), multidimensional poverty and social fragility of migrants’ families in Lombardy (Italy) have been thoroughly explored, revealing a quite com- plex pattern in the levels and shapes of their social conditions. By using concepts from par- tial order theory, the study provides realistic estimates of poverty diffusion and sheds light on the social differences between and within migrants’ groups, indirectly implying that no poverty relief policy can be effective, without accounting for the nuances of such inhomo- geneous scenario. Drawing upon the same datasets, in the present paper we take a step fur- ther towards policy targeting and propose a new statistical procedure for ranking migrants’ subpopulations, with the aim of providing policy-makers with actionable criteria supporting the definition of poverty relief interventions. To this goal, the critical issue is how to “con- dense” information on poverty and fragility into the respective rankings, still reflecting the complexity of these social traits and their nuanced patterns among migrants’ groups. Indeed, any ranking built out of multidimensional indicator systems is essentially a dimensionality reduction process, which unavoidably loses information, in favour of simplicity and action- ability. While some information loss must be accepted, the ranking method should never- theless be designed in such a way to preserve, as much as possible, that part of informa- tion considered as more valuable. This leads to the use of partially ordered set theory, as the reference formal framework of our work, and to the adoption of so-called Fuzzy First- Order Dominance (F-FOD) analysis, as the cornerstone of our ranking procedure. We differ to a later section the description of such a statistical tool; here it suffices to say that F-FOD allows for multidimensional statistical distributions defined on partially ordered sets to be compared, in terms of relative dominance, without preliminarily “collapsing” them into some kind of synthetic indicators. In its essence, F-FOD performs a “synthesis of multidimensional comparisons between distributions”, rather than performing a “comparison of synthesized multidimensional distributions”, as one would do by using more classical aggregative proce- dures, like the much-revered Counting Approach of Alkire and Foster (2011a, b), or even the non-aggregative posetic approach to poverty evaluation, employed in Arcagni et al. (2019). Abstract In this paper, we apply the novel Fuzzy First-Order Dominance (F-FOD) methodology to rank migrant subpopulations in Lombardy (Italy), in terms of multidimensional poverty and social fragility, for the year 2014, with the purpose to possibly provide useful support to policy-makers, in targeting relief interventions from poverty and discomfort. The F-FOD methodology allows for the direct comparison of different distributions of poverty and fra- gility, assessed by means of suitable ordinal multi-indicator systems, so extending to this more complex setting, the usual univariate first-order dominance criterion. It also provides complimentary “incomparability” scores, to assess to what extent the final rankings are reliable or instead forcing. It turns out that the levels of poverty and fragility of migrant subpopulations are quite different and, in particular, that the time since migrations has a key impact, on the identification of most critical cases, which typically involve recently migrated people. Evidence also emerges that the temporal poverty/fragility trajectories of migrants, distinguished by country of origin, follow different paths, suggesting how policy interventions must be properly, and differently, tuned to be effective. Alberto Arcagni alberto.arcagni@unimib.it; alberto.arcagni@uniroma1.it Marco Fattore marco.fattore@unimib.it Elisa Barbiano di Belgiojoso elisa.barbiano@unimib.it 1 Department of Statistics and Quantitative Methods, University of Milano-Bicocca, Milan, Italy 2 Department of MEMOTEF, Sapienza University of Rome, Rome, Italy 1 Department of Statistics and Quantitative Methods, University of Milano-Bicocca, Milan, Ital 2 Department of MEMOTEF, Sapienza University of Rome, Rome, Italy 6789) 3 S. M. L. Rimoldi et al. 1 Introduction This way, the dimensionality reduction step is “postponed” to the end the statistical process, making F-FOD inherently more “information and complexity preserving” than procedures passing through the evaluation of poverty or fragility at individual level. In this respect, the F-FOD procedure extends the classical univariate stochastic first-order dominance criterion, to the multidimensional ordinal setting, typical of social evaluation studies, and meets the need for more suitable poverty measurement tools; in particular, it helps overcoming the clas- sical approaches based on monetary poverty lines, in particular when focusing on foreigners (Arcagni et al. 2019; di Belgiojoso et al. 2009; Busetta 2016; Lemmi et al. 2013). Th i i f h d i l h d l i l h h d The motivation of the present study is not only methodological, however; the paper does aim also at providing new insights on migrants’ poverty and fragility, given the relevance of the migration phenomenon in the Mediterranean area and its impact on Italian soci- ety, and in Lombardy in particular. Reducing poverty is indeed the first goal of the UN 2030 Agenda for Sustainable Development, adopted by all United Nations Member States in 2015. According to it, countries recognize that ending poverty and other forms of dep- rivation is a key factor to promote peace and prosperity for people and the planet, now and into the future since, as stated by Pradhan et al. (2017, p. 1172) “SDG 1 (No poverty) has synergetic relationship with most of the other goals” (e.g. improvement of global health and wellbeing). Additionally, as many scholars acknowledge, poverty among migrants is generally more acute and persistent than among the host population (Bárcena-Martìn 1 3 Targeting Policies for Multidimensional Poverty and Social… and Pérez-Moreno 2016; Berti et al. 2014; Kazemipur and Halli 2001; Lelkes 2007; de Bustillo and Anton 2011; Obućina 2014; Pastor 2014) and so reducing migrants’ poverty is a key step to reduce the overall level of poverty in the whole society, particularly in countries with large immigration flows, like Italy. This motivates the focus of the present paper, which is specifically devoted to the ranking of migrants’ families, for policy tar- geting. More concretely, we cluster migrant families by country of origin and years since migration, two features that have been observed to be linked to migrants’ social conditions (Arcagni et al. 1 Introduction 2019), and compare the corresponding frequency distributions on a multi- dimensional set of poverty and fragility indicators, by using the aforementioned F-FOD procedure. From the resulting set of pairwise comparisons, a final ranking is eventually obtained, identifying the most critical subpopulations, to be considered by policy-makers. Besides, a complementary measure of the degree of “incomparability” among clusters is provided, to assess the reliability of the rankings and to preserve a feeling of the intrinsic and irreducible complexity of migrant’s social conditions. The paper is structured as follows: Sect. 2 provides some background about research on poverty; Sect. 3 describes the input data used in the subsequent analysis. Section 4 outlines the F-FOD procedure; Sect. 5 illustrates the results and Sect. 6 draws the conclusions. 2 Background Research on Migrants’ Poverty overcome the traditional approach, based on the comparison of family income to a prede- termined threshold, or poverty line, which proves problematic and not so effective in the case of migrants (Arcagni et al. 2019; di Belgiojoso and Rimoldi 2005; di Belgiojoso et al. 2009; Busetta 2016; Lemmi et al. 2013). A common thread among most of these proposals is the acknowledgement that poverty is a complex multidimensional trait, which cannot be assessed just in monetary terms.f This paper focuses on two main suffering domains, namely poverty and social fragil- ity; the first refers to actual deprivation status and the second, to the exposure to critical conditions that might evolve into poverty, or worsen already deprived situations (Arcagni et al. 2019). To capture the complexity of deprivation (Koutsampelas 2015), beyond purely monetary aspects, poverty status is assessed by adding to income the dimensions of liv- ing arrangements, economic aid, and medical treatment renunciation. As for living arrange- ments, an extensive literature considers homeownership as a factor of economic success and a marker of well-being (Bárcena-Martìn and Pérez-Moreno 2016; Constant et al. 2009; Davidov and Weick 2011; Gobillon and Solignac 2015; Myers and Lee 1998). For migrants, in particular, homeownership is an important step toward a definitive settlement in the host country, reflecting the will and the commitment to stay (Constant et al. 2009; Davidov and Weick 2011; Rimoldi and di Belgiojoso 2016); on the contrary, precarious accommoda- tion and shared dwelling, often overcrowded, correspond to certain poverty status (Diana and Strozza 2014; Robinson et al. 2007). Recourse to economic aid is a frequent practice in conditions of poverty: we assume that recourse to institutional aid corresponds to non- transitory poverty, while recourse to one’s own kinship or friendship network support cor- responds to transitory poverty, based on research evidence showing how recourse to institu- tional aid decreases as the length of stay increases (Hansen and Lofstrom 2003; Lie 2002).1 Unmet medical needs are strictly linked to poverty (Busetta 2016; Fernades and Pereira 2009; Koolman 2007), moreover, empirical evidence shows that migrants in Italy experi- ment knowledge barriers to health care such as scarce information about national regula- tions (Busetta 2016; Giannoni 2010). Unsatisfied education needs and socialization needs, although important factors in explaining migrants’ poverty and integration (Kazemipur and Halli 2001), could not be included because not available in the survey. 1 This hypothesis is valid when dealing with the stable resident foreign population, as it is the case of the sample used in this paper; otherwise, migrants in the first period after immigration have difficulties to access institutional opportunities and are forced to lean on their social network. 2 Background Research on Migrants’ Poverty Measuring poverty among migrants’ families is the first step in the process of contrast- ing poverty in the Italian society, where migration has introduced an additional feature of complexity, particularly in the issue of deprivation (Istat 2018). Some specific traits of cur- rent poverty in Italy come indeed from the social transformation of the Italian population over the last decades. As well known, Italy has been experiencing large flows of migrants from poor countries since the end of the Nineties. Over time, these migrants settled in Italy forming a population that, at the end of 2018, amounted to 5.3 million residents (Istat 2019). Although, in some measure, the integration process has been achieved for a part of foreigners, there is still a large share of them living in poor conditions. Currently, the proportion of families living below the national (relative) poverty line among the families entirely composed of foreigners is 34.5% in 2017 (23.9% among mixed families) vs. 10.5% among Italian families (Istat 2018). Moreover, migrants are the poorest among the poor and a large number of studies in various contexts, confirm that poverty among migrants tends to be more marked and persistent than among natives in various countries (Bárcena- Martìn and Pérez-Moreno 2016; Kazemipur and Halli 2001; Lelkes 2007; de Bustillo and Anton 2011; Obućina 2014; Pastor 2014) and in Italy as well (Berti et al. 2014). Vari- ous explanations have been attempted, to better account for the link between migration and poverty. They usually refer to migrants’ legal status, to their educational attainments and poor host language proficiency, to relatively lower job skills, to age and family status and the length of stay (Hansen and Wahlberg 2009; Sullivan and Ziegert 2008). Indeed, in Arcagni et al. (2019), it is underlined how capturing migrants’ poverty is a difficult task from many points of view. Migrants are an unstable population, due to its frequent cross- border or “in transit” nature; moreover, methodological issues arise due to problems in data collection and to the specificity of migrants’ social conditions, which require measurement methodologies to be properly tuned (Aaberge et al. 1999; di Belgiojoso et al. 2009; Berti et al. 2014; Galloway and Aaberge 2005; Koutsampelas 2015; Lemmi et al. 2013; Rimoldi and di Belgiojoso 2016). Therefore, various attempts have been carried out in research to 3 3 S. M. L. Rimoldi et al. 2 Background Research on Migrants’ Poverty Social fragility here is intended as the vulnerability of families deriving from individual unstable working conditions together with work-family models, which describe the rela- tionship between family size and the number of its members in employment (Fellini and Migliavacca 2010). Therefore, based on the fact that increasingly precarious work con- ditions directly affect consumption, housing and health (Obućina 2014; Pemberton et al. 2014), we describe social fragility through the working dynamics (ranging from the condi- tion of persistent unemployment to stable employment). To catch the family dimension, we introduced both a variable describing the effects of remittances on consumption and saving behaviour due to the presence of a family in the country of origin, and a variable account- ing for the fragility of families with just a single income source, which are commonly con- sidered as having a higher risk of poverty (Curatolo and Wolleb 2010). Finally, not having a residence permit represents, for an immigrant and his family, a serious element of precar- iousness, not allowing access to any opportunities offered by social policies at the various territorial levels to support the integration. 1 3 Targeting Policies for Multidimensional Poverty and Social… We address this issue by employing procedures based on the theory of partially ordered sets—as introduced by Fattore (2016) and Fattore and Arcagni (2018)—which are prop- erly designed to work fuzzy measures out of ordinal multi-indicator systems. Results com- ing from Arcagni et al. (2019) support such “posetic” approach, showing that deprivation among migrant families has a highly fuzzy and nuanced trait, with a limited quota of com- pletely or almost completely poor units and a much larger part of population sharing, to dif- ferent extents, deprivation facets. In fact, the overall picture of migrants’ poverty hides spe- cific patterns of poverty by country (or geographical area) of origin: Sub-Saharan migrants face the hardest condition, while Chinese and Filipinos have the best scores both for pov- erty and fragility. In addition, country of origin intertwines with years since migration in explaining poverty and fragility performances (for example, East Europeans migrants, although with a long mean duration of presence—usually negatively correlated to poverty -, suffer from partial deprivation, due to their specific migration model, i.e. being mostly alone with family left behind and often one-income family). The posetic approach adopted in Arcagni et al. 2 Background Research on Migrants’ Poverty (2019) aimed primarily at assessing the poverty and fragility degrees of individuals, comparing them to some multidimensional benchmarks, in a “threshold-like” spirit. From individual degrees, poverty and fragility patterns were outlined and synthetic indicators were computed, to compare migrants’ sub- groups. Here, the aim is somehow different, in that we want to get rankings of migrants’ subpopulations, without going through the computation of any individual scores, but directly quantifying the degree of dominance among subpopulations, when these are mul- tidimensionally compared against the set of poverty and fragility attributes. As mentioned in the Introduction, the F-FOD procedure provides such a quantification and finally leads to poverty and fragility rankings. 2 ORIM stands for “Osservatorio regionale per l’integrazione e la multietnicità?” (“Regional observatory on integration and multiethnicity”). 3 341 subjects have been excluded, due to the too small size of their nationality group and 47 subjects have been excluded because years since migration were not available. 3 Data This way, the sample is representative also for specific sub- groups, such as illegal migrants or ethnic groups. The data comprise a set of socio-demo- graphic variables (family composition, migratory characteristics, economic information…) and a yearly thematic in-depth section. The 2014 survey involved 4000 subjects living, legally or illegally, in Lombardy, including migrants in a mixed union, and the in-depth section was specifically devoted to migrants’ economic conditions. 2011; Sanguilinda et al. 2017). This way, the sample is representative also for specific sub- groups, such as illegal migrants or ethnic groups. The data comprise a set of socio-demo- graphic variables (family composition, migratory characteristics, economic information…) and a yearly thematic in-depth section. The 2014 survey involved 4000 subjects living, legally or illegally, in Lombardy, including migrants in a mixed union, and the in-depth section was specifically devoted to migrants’ economic conditions. i As highlighted in Arcagni et al. (2019), migrants’ poverty and fragility are mainly shaped by country (or geographical area) of origin and years since migration. In the pre- sent paper, we thus investigate on migrants’ group poverty and fragility along these two dimensions. Overall, we selected 3609 subjects.3 Table 1 shows the final composition of the population object of study.l In the following, we briefly recall the variables used to describe the two domains of poverty and social fragility for migrant families, and the complimentary covariates used to break down the set into sub-groups. 3 Data Data on the socio-economic conditions of migrants and migrants’ families in Lombardy come from the 2014 ORIM2 Survey, held as part of the regional monitoring activity on foreign population. ORIM surveys are conducted each year on a sample of migrants aged 15 and over and coming from “less developed countries” and also from Central and East- ern Europe. Interviewees are randomly selected according to the Centre Sampling Method (Baio et al. 2011), which ensures representativeness at both the regional level and provin- cial level (Blangiardo 2008). The method is based on a two-stage design: municipalities are the first level units selected according to their share of migrants and their demographic representativeness at the regional level, migrants (second level units) are then randomly selected among those who frequent a set of aggregation centres previously identified in each of the first level units. The underlying hypothesis of this method is that in everyday life migrants frequent some services or places, defined “aggregation centres” (such as insti- tutions, place of worship or entertainment, training centres, care centres, meeting points, ethnic shops, telephone centres…) and the profile of centers frequented by each migrant can be used a posteriori to estimate the weights to correct the initial probability of inclu- sion in the sample. The weights are inversely related to the inclusion probability, as this guarantees the representativeness of the sample and corrects the initial bias (Baio et al. 1 3 S. M. L. Rimoldi et al. Table 1 Population by country (or geographical area) of origin and years since migration. Source: own elaborations on ORIM data Country/geographical area of origin Years since migration Total < 2 3–5 6–9 10–14 15+ Albania (AL) 6 10 25 94 102 237 China (CHN) 13 35 26 53 66 193 India, Pakistan, Sri Lanka, Bangladesh (IPSB) 31 56 94 169 147 497 Latin American (LA) 17 29 74 163 141 424 North Africa (NA) 29 60 142 246 525 1002 Philippines (PH) 0 7 11 12 39 69 Romania (RO) 15 25 42 95 49 226 Sub-Saharan Africa (SSA) 72 97 141 186 218 714 Ukraine and Moldova (UM) 11 48 62 100 26 247 Total 194 367 617 1118 1313 3609 le 1 Population by country (or geographical area) of origin and years since migration. Source: own borations on ORIM data 2011; Sanguilinda et al. 2017). 1 3 3.1 Poverty and Social Fragility Attributes Dependent family in country of origin, coded on a 2-degrees scale: 1 = “Yes”; 2 = “N 4. One-income family, coded on a 2-degrees scale 1 = “Yes”; 2 = “No”. Among the complimentary variables, we focused on the following in order to target the groups: Among the complimentary variables, we focused on the following in order to target the groups: 1. Country or geographical area of origin (“Sub Saharan Africa (SSA)”; “NorthAfrica (NA)”; “Albania (AL)”; “Romania (RO)”; “Ukraine and Moldova (UM)”; “Latin America (LA)”; “Philippines (PH)”; “China (CHN)”; “India, Pakistan, Sri Lanka and Bangladesh (IPSB)”). 2. Years since migration of the forerunner (in years, quantitative). This variable appears to be usually negatively correlated with poverty since, as migrants stay longer in the host country, their poverty level tends to decrease. The variable is coded in the following cat- egories: “0–2”, “3–5”, “6–9”, “10–14” and “15 or more”. For the crossed analysis with country of origin, we used the following categories “0–5”, “6–9”, “10 or more”, so as to reduce the number of subpopulations to consider, without oversimplifying the analysis. 4 Legal status refers to the sojourn condition of migrants. More specifically, “legal” migrants are migrants with a sojourn permit, Italian or European citizens while “illegal” migrants are those migrants without per- mit (undocumented). 5 3.1 Poverty and Social Fragility Attributes Poverty status is based on four attributes: 1. Living arrangement, coded on a 5-degrees scale: 1 = “Living in encampment, shacks, squatting”; 2 = “Tenant with other migrants (not relatives)”; 3 = “Living at the work- place”; 4 = “Tenant alone or with family”; 5 = “Homeowner”. 1. Living arrangement, coded on a 5-degrees scale: 1 = “Living in encampment, shacks, squatting”; 2 = “Tenant with other migrants (not relatives)”; 3 = “Living at the work- place”; 4 = “Tenant alone or with family”; 5 = “Homeowner”. 2. Economic aid, coded on a 3-degrees scale: 1 = “Institutional aid”; 2 = “Kinship’s aid”; 3 = “No aid”. 2. Economic aid, coded on a 3-degrees scale: 1 = “Institutional aid”; 2 = “Kinship’s aid”; 3 = “No aid”. 3. Medical treatment renunciation, coded on a 3-degrees scale: 1 = “Renunciation to medi- cal treatment or resort to remedies”; 2 = “Return to the country of origin for medical treatment”; 3 = “No renunciation”. 1 3 Targeting Policies for Multidimensional Poverty and Social… 4. Equivalent monthly net family income (in euros), coded on a 4-degrees scale, according to the quartile the equivalent family income (I) belongs to: 1 = I ≤ 845; 2 = 846 ≤ I ≤ 1206; 3 = 1207 ≤ I ≤ 1639; 4 = I > 1639. The family income is adjusted according to the “migrants equivalence scale”, introduced and estimated by Rimoldi and di Belgiojoso on the basis of the Engel’s law (Rimoldi and di Belgiojoso 2016). Migrants’ social fragility is described and assessed based on the following four ordinal attributes: 1. Working dynamics, coded on a 6-degrees scale: 1 = “Persistent unemployment”; 2 = “Run into unemployment” (employed one year before the survey, but unemployed at the time of the survey); 3 = “Run into instability or persistent precariousness” (stably employed one year before the survey, but precariously employed at the survey or precari- ously employed at the survey and one year before); 4 = “Non-active status” (housewife, student and retired); 5 = “Improving condition” (unemployed one year before the sur- vey, but employed at the time of the survey); 6 = “Persistent and stable employment” (employed one year before the survey and employed at the time of the survey). 2. Legal status,4 coded on a 2-degrees scale: 1 = “Illegal”; 2 = “Legal”. 3. 5 The discussion for poverty status is completely analogous. 4 Partially Ordered Set and the F‑FOD Procedure In the following, we provide some basic definitions and concepts of partially ordered set (poset) theory and outline the F-FOD procedure, in an intuitive manner. Formal details can be found in cited references. We motivate the use of partially ordered set by means of a simple example. Consider the variables used to describe migrants’ social fragility5; the list of scores describing each 4 Legal status refers to the sojourn condition of migrants. More specifically, “legal” migrants are migrants with a sojourn permit, Italian or European citizens while “illegal” migrants are those migrants without per- mit (undocumented). 5 1 3 S. M. L. Rimoldi et al. individual fragility pattern is here called a (fragility) profile. Although the input variables are ordinal, profiles are multidimensional entities that in general cannot be ordered, due to the existence of possible conflicting scores between them. To clarify this point, consider three migrants A, B and C, with the following fragility profiles: A—6211 (persistent and stable employment, legal status, with a dependent family in the country of origin and one-income family), B—5222 (improving condition, legal status, without a dependent family in the coun- try and with more than one-income in the family) and C—2122 (run into unemployment, ille- gal status, without a dependent family in the country and with more than one-income in the family). The profile of migrant C is more fragile than that of B, due to unemployment and illegal status; on the contrary, the profiles of migrants A and C are not comparable (i.e. they are incomparable). In fact, while profile A is more fragile than C on one attribute, it is less fragile on another, so that the two profiles cannot be ordered. As a result, the set of all the pos- sible profiles cannot be completely ordered in terms of fragility and is thus named a partially ordered set (or a poset, for short) i.e. a set where the possibility of comparing elements is only partial. Partially ordered sets arise naturally when multidimensional systems of ordinal attrib- utes are to be dealt with and provide the formal setting for the analysis carried on in the next section. 4 Partially Ordered Set and the F‑FOD Procedure For this reason, here we provide some definitions and basic tools from partial order theory, so to make the present paper self-contained and clarify subsequent developments.ii A (finite) poset P = (X, ≤) is a set X of finite cardinality k, equipped with a partial order relation ≤, i.e. with a reflexive, antisymmetric and transitive binary relation (Davey and Priestley 2002; Neggers and Kim 1998; Schröder 2003). An easy way to specify the struc- ture of a poset, is by stating the dominance between pairs of elements of the underlying set X. This is more effectively done by introducing so-called incidence matrix Z, which is a k x k matrix defined by Zij = { 1 if xi ≤xj in poset P 0 otherwise 1 3 for xi, xj ∈X. for xi, xj ∈X. i j We already introduced in the example the concept of incomparability between two pro- files. In general, two elements xi, xj ∈X are incomparable if neither xi ≤xj and xj ≤xi , i.e. Zij = Zji = 0. j j Posets can be graphically represented by so-called Hasse diagrams, a kind of directed acyclic graphs which help visualizing the partial order structure (usually, when the number of elements of X is not too large). For exemplification purposes, the Hasse diagram of the partial order built upon the set of all possible fragility profiles is represented in Fig. 1. i In the Hasse diagram, each node corresponds to a poset element; if xi < xj in the input poset, then node corresponding to xj is placed higher than the node correspond- ing to xi and an edge is drawn between them, if and only if there is no other xh such that xi < xh < xj . Downward paths in the Hasse diagram reconstruct the dominances between poset elements; two nodes not connected by any downward path are thus incomparable. Simple as they may seem, posets may comprise a great deal of information on input data, useful to perform evaluations on multi-indicator systems or to build rankings of statistical units and of sub-groups of the population of interest. The formal development of the procedures for the statistical treatment of partially ordered structures is however not entirely trivial and can be found in cited references. Here we simply outline, by means of a simple example, the F-FOD procedure, later employed to rank target sub- populations in terms of poverty or fragility. 1 3 Targeting Policies for Multidimensional Poverty and Social… Targeting Policies for Multidimensional Poverty and Social… Fig. 1 Hasse diagram of the social fragility poset Fig. 2 Statistical distributions on the poset built upon four binary attributes. Color intensities are propor- tional to frequencies. (Color figure online) Fig. 1 Hasse diagram of the social fragility poset Fig. 2 Statistical distributions on the poset built upon four binary attributes. Color intensities are propor- tional to frequencies. (Color figure online) Fig. 2 Statistical distributions on the poset built upon four binary attributes. Color intensities are propor- tional to frequencies. (Color figure online) Consider an indicator system comprising four binary attributes, evaluated on three dif- ferent populations. for xi, xj ∈X. Different units in each population may share the same scores on the four attributes, so that to each binary profile a relative frequency is associated. Suppose that the statistical distributions associated to the three populations can be represented as in Fig. 2, where the more intense the color, the higher the corresponding frequency.i Intuitively, the first distribution is better (i.e. higher) than the other two and the third is better than the second. However, it is not clear how to make this feeling objective, given that the domain of such distributions is only partially ordered and not all profiles are com- parable. The solution to this issue is provided by the Fuzzy First-Order Dominance pro- cedure (Fattore and Arcagni 2019), which is an extension, to the partially ordered case, of the first-order dominance criterion, used to compare distributions defined over the real axis (or over subsets of it). F-FOD combines poset theory and fuzzy set theory, with the approach to first-order dominance with multiple discrete indicators, proposed by Arndt 1 3 3 3 S. M. L. Rimoldi et al. et al. (2012), overcoming the main limitations of the latter and providing a natural way to quantify dominance degrees between statistical distributions over posets. et al. (2012), overcoming the main limitations of the latter and providing a natural way to quantify dominance degrees between statistical distributions over posets. With reference to the above example, the basic idea behind the F-FOD can be sketched as follows. 1. Consider the set of 16 binary profiles of the poset underlying the distributions of Fig. 2 and form all of the 16! so-called linear orderings corresponding to their permutations. Such orderings can be subdivided into two classes: the class of linear orderings pre- serving the comparabilities of the input poset (i.e. such that xi < xj in the poset implies xi < xj in the linear ordering) and the set of linear orderings that do not (i.e. such that there is at least one comparability xi < xj in the poset such that xj < xi in the linear order- ing). The linear orderings belonging to the first class are called the linear extensions of the poset; they are all the possible orderings of poset elements compatible with the constraint imposed by the comparability structure of the original partial order relation.i 2. for xi, xj ∈X. A fundamental theorem of poset theory states that any finite poset is equivalent to the set of its linear extensions, so that the poset can be uniquely reconstructed from such a set (Schröder 2003). As a consequence, the quantification of the relative dominance of two distributions over the input poset can be addressed by quantifying their relative dominance on the corresponding linear extensions. 3. Interestingly, a linear extension can be considered as a univariate ordinal variable, so in it one can quantify the relative degree of first-order dominance of two distributions in a rather straightforward way, i.e. by computing the probability that a unit randomly sampled from the first distribution is dominated, in the considered linear extension, by a unit randomly sampled from the second. 4. Such “elementary” dominance degrees can be computed for each linear extension and finally averaged to get the final dominance score between the two distributions. 4. Such “elementary” dominance degrees can be computed for each linear extension and finally averaged to get the final dominance score between the two distributions. ii 5. Repeating steps 3 and 4 for each distribution pair, the F-FOD procedure eventually provides a matrix Δ, whose entry Δij is the degree of dominance of distribution j over distribution i. ii 5. Repeating steps 3 and 4 for each distribution pair, the F-FOD procedure eventually provides a matrix Δ, whose entry Δij is the degree of dominance of distribution j over distribution i. 6. Once matrix Δ is available, one can compute a degree of dominance of distribution i over the others by averaging its dominance degree over each other distribution, i.e. by computing the following index: 6. Once matrix Δ is available, one can compute a degree of dominance of distribution i over the others by averaging its dominance degree over each other distribution, i.e. by computing the following index: domj = 1 h −1 ( h ∑ i=1 Δij −1 ) . domj = 1 h −1 ( h ∑ i=1 Δij −1 ) . where h is the number of compared distributions. where h is the number of compared distributions. where h is the number of compared distributions. Remark In practice, the above procedure is implemented in a slightly simplified manner, i.e. by considering not all of the linear extensions of the poset, but the subset of so-called lexicographic linear extensions (Fattore and Arcagni 2019). Suppose you order poset pro- files based on their score on the first variable and, in case of ties, on their score on the second variable and so on, in an “alphabetic” fashion; the final linear order can be proved to be a linear extension of the input poset, called lexicographic (Fattore and Arcagni 2019). Permuting the order of the variables in the above process, one gets a set of k! (where k is the number of variables) lexicographic linear extensions, which can be proved to uniquely determine the input poset, as the set of all of linear extensions does. In general, the number of lexicographic linear extensions is much less than that of all linear extensions and so 1 3 1 Targeting Policies for Multidimensional Poverty and Social… Table 2 Δ-matrix and dominance scores for the toy example Pop. 1 Pop. 2 Pop. 3 Pop. 1 1.000 0.000 0.187 Pop. 2 1.000 1.000 0.670 Pop. 3 0.899 0.492 1.000 Dominance 0.950 0.246 0.428 Table 3 I-matrix and incomparability scores for the toy example Pop. 1 Pop. 2 Pop. 3 Pop. 1 0.000 0.000 0.187 Pop. 2 0.000 0.000 0.492 Pop. 3 0.187 0.492 0.000 Incomparability 0.094 0.246 0.340 emplying just them, heavily reduces the computational burden of the procedure and allows F-FOD to be applied to larger multi-indicator systems, than that used in the above toy example. The procedure is implemented in the R package parsec (Arcagni 2017) available on the R Comprehensive Archive Network (R Core Team 2019). emplying just them, heavily reduces the computational burden of the procedure and allows F-FOD to be applied to larger multi-indicator systems, than that used in the above toy example. The procedure is implemented in the R package parsec (Arcagni 2017) available on the R Comprehensive Archive Network (R Core Team 2019). Table 2 shows the Δ-matrix for the toy example and the corresponding average domi- nance scores. From it one can directly rank the distributions. Notice that the average dominance scores provide metric information, which enriches the ranking, by providing the “dominance distance” between pairs of distributions. where h is the number of compared distributions. As mentioned in the Introduction, any ranking extraction implies some unavoidable information loss, since incomparable multidimensional inputs are made comparable and forcedly placed on a single axis. The F-FOD procedure is no exception, but interestingly it provides a way to complement the average dominance scores with information on the degree of incomparability of a distribution, with respect to the others; this way, one can assess to what extent a distribution is forced into the ranking. To compute such incom- parability scores, the dominance matrix Δ is turned into the incomparability matrix I, which is symmetric and whose entries are defined by Iij = min(Δij,Δji) (so, the higher the degree of dominance, the lower the degree of incomparability between two distribu- tions), and then the following incomparability score is computed for each distribution: incj = 1 h −1 ( h ∑ i=1 Iij −1 ) . Table 3 reports the incomparability matrix, for the three populations example.i We conclude this methodological section with a final comment. The F-FOD may at first seem an aggregative procedure (like the construction of composite indicators), in that it computes the final dominance scores between two distributions, by aggregating “elemen- tary” dominance scores. But here it is the key difference with respect to classical scor- ing procedures: the aggregation is not performed on the input variables, but on the linear extensions, which is a way to exploit the information comprised in the structure of the underlying poset. This is why, in the Introduction, we stressed that the F-FOD is more “information and complexity preserving” than other tools, since it exploits the deep struc- ture of the input data. 1 3 3 S. M. L. Rimoldi et al. 0 0,1 0,2 0,3 0,4 0,5 0,6 0,7 0,8 0,9 0 0,05 0,1 0,15 0,2 0,25 0,3 0,35 0,4 0,45 0,5 Dominance Incomparability Poverty Fragility (labels in Italic) SSA_0-2 SSA_3-5 UM_0-2 AL_0-2 NA_0-2 CHN_0-2 SSA_0-2 PH_3-5 PH_10-14 CHN_10-14 CHN_15+ CHN_3-5 PH_15+ CHN_15+ SSA_6-9 IPBS_0-2; 3-5; 6-9 UM_0-2 LA_0-2 RO_0-2 IPBS_0-2;0-5 Fig. 3 Poverty and fragility scores (dominance and incomparability) by country of origin and length of stay in Italy Poverty Fig. 3 Poverty and fragility scores (dominance and incomparability) by country of origin and length of stay in Italy 1 3 5 Results The group “India-Pakistan-Sri Lanka-Bangladesh” (13.8% of the population) also shows low dominance scores up to 9 years since migration on the poverty side, and up to 5 years since migration on the fragility side. As for fragility, the duration of presence seems to be a less incisive protection factor for some groups like Ukrainians and Moldovans (6.8% of the population), and it is almost unin- fluential for Chinese (5.3%), who are in favorable position whatever the duration of pres- ence. Notice the role of the variable “Dependent family in country of origin” in fragility scores, which is especially important for Ukrainians and Moldovans. Chinese and Filipinos with long duration of presence (10 years at least) suffer from neither poverty nor fragility, although the most recently arrived Filipinos (3–5 years) report a high score for fragility. Poverty and fragility rankings, however, must be considered in the light of the incom- parability scores, that tell about the level of certainty of the ranking positions. Figure 3 illustrates the rankings skewed by the incomparability: points laying on the right side of the graph are more uncertain than the others (since they are to a larger extent incomparable with the other elements). Therefore, it is immediately apparent that fragility is more uncer- tain than poverty. Moreover, while uncertainty is low both at the top (i.e. for SSA_0–2, SSA_3–5, AL_0–2) and at the bottom (i.e. for PH_10–14, CI_15+) of the ranking of pov- erty (making these positions rather sure), it does not decrease along with the fragility score, so that fragility is almost certain only when it is high (i.e. for UM_02, SSA_02, PH_3–5). It should be noticed that the incomparability scores measure to what degree a distribution is, on average, incomparable with the others; but since there are “many ways” to be incompa- rable, similar scores do not mean similar poverty/fragility distributions. Thus, subpopulations placed nearby on the incomparability axis of Fig. 3, need not be similar, in social terms. Finally, it is of considerable interest to jointly examine the trajectories of poverty and fragility along with the duration of the presence. From Fig. 4 (detailed data can be found in the “Appendix 1” to this paper), we can derive some interesting tips (recall that low domi- nance scores mean high poverty/fragility levels, so that over the bisector groups are placed subpopulations for which the poverty level is higher than the fragility level). 5 Results We applied the F-FOD procedure to our data described for both poverty and social fragility. From Fig. 3 (detailed data can be found in the “Appendix 1” to this paper), it turns out that migrant groups can be indeed ordered by poverty and fragility dominance scores (the higher the scores, the lower the poverty and fragility levels), even if some of the subgroups are quite incomparable to the others, confirming the existence of quite complex patterns in migrants’ social conditions. Indeed, some sub-populations in highly critical conditions neatly emerge from the data and have the lowest incomparability level (namely Sub-Saharan migrants). Based on the reported poverty and fragility scores, the primary role of years since migration clearly emerges. Almost all the migrant groups most recently arrived (0–2 years since migration) are on top of the rankings of both poverty and fragility (i.e. they have low dominance scores): as for poverty, except for Latin Americans, Ukrainians and Moldovan and Romanians (whose dominance score is higher), migrants with the lowest length of stay account for 4.2% of total population; as for fragility, disregarding Chinese, they account for 5.0% of population. Symmetrically, migrants with the longest duration of presence (10 years at least) generally occupy the most favorable position in poverty and fragility rankings; notice that they account for 67.4% of the total population.f Interestingly, poverty and fragility hit the groups in different ways, both by duration of the presence and by country. For instance, as long as years since migration increases (at least up to 10 years), Filipinos tend to reduce both their poverty and fragility while Chi- nese, while reducing poverty, tend to maintain their level of fragility, although very mod- est. It should be noticed, however, that the results pertaining to Filipinos can be affected by instability, since the small dimension of their sample (see Table 1). 1 3 Targeting Policies for Multidimensional Poverty and Social… More in depth, from Fig. 3 it also emerges that Sub-Saharan Africans (SSA)—19.8% of the total population—are not only the poorest, but they are also among the most frag- ile groups: they report the lowest scores of poverty dominance (that means highest pov- erty) even for a long duration of presence (up to 9 years) and the lowest scores of fragility dominance, although only for those more recently arrived (years since migration < 6). 5 Results A general trend towards reduction over time of both poverty and fragility can be clearly noticed, even if along different trajectories, for different groups. The figure compares migrants who are at a different stage of their migration experience (length of stay) by area of origin with a cross-sectional approach. The trajectories do not describe the change in poverty and fragil- ity among a cohort of migrants during their migration experience. Thus, the results should be read with caution because some context effects could bias the trends. However, this figure allows us to have an idea of the possible evolution along the migration experience. At the beginning of the migration experience, migrants suffer from different levels of pov- erty and fragility (e.g. for Sub-Saharans and North-Africans, poverty is much higher than fragility), but as time in migration passes, poverty and fragility tend to converge towards lower and more similar levels. Migrants from Ukraine and Moldova and from Philippines start their migration with more fragility than poverty, but over time their fragility trans- forms into poverty, although on lower levels. Filipinos start with some fragility, but they rapidly escape from it. Conversely, Albanians start their migration as relatively poor, but they then turn poverty into fragility during the first years of migration; after 10 years, their 1 1 3 S. M. L. Rimoldi et al. 0 0,1 0,2 0,3 0,4 0,5 0,6 0,7 0,8 0,9 1 0 0,1 0,2 0,3 0,4 0,5 0,6 0,7 0,8 0,9 1 Fragility Poverty AL CHN IPSB LA NA PH RO SSA UM Fig. 4 Poverty and fragility scores comparison colored by geographical area/country of origin and arrows indicating the increasing years since migration Fig. 4 Poverty and fragility scores comparison colored by geographical area/country of origin and arrows indicating the increasing years since migration poverty level decreases. Chinese’s trajectory moves quite horizontally after 5 years since migration, meaning they are almost untouched by fragility along their durable migration. These results indicate that the most urgent condition refers to migrants from Sub-Saha- ran Africa especially those recently arrived. To improve their condition, policies should be activated (or if existing, improved) to provide them with support in many multiple direc- tions: first, social policies could be allowed also to undocumented migrants (e.g. 5 Results the case of children denied access to social services because their parents are irregular migrants); second, living arrangement could be improved by encouraging social housing; third, bar- riers to access to medical treatment could be removed by information campaign; finally, a stable income could derive from helping migrants’ labour market integration by matching demand and supply. For example, the high level of fragility of recently arrived Ukrainians and Moldovans strongly depend on their occupational instability due to the fact that they usually work in the domestic sector without a contract, moreover they have a dependent family in their country of origin and their income is the sole family income. Policies aimed at increasing the regular recruitment of domestic workers could reduce their fragility. 1 3 6 Conclusions In this paper, we have addressed the problem of ranking migrant subgroups in terms of pov- erty and fragility in a multidimensional ordinal setting, using novel statistical tools, based on poset theory. The aim of the paper is to possibly provide insights to policy-makers, in view 1 3 1 3 1 Targeting Policies for Multidimensional Poverty and Social… of targeting poverty relief interventions, by resolving the differences among migrant clus- ters and identifying the most critical cases. Reducing poverty is unanimously recognized as a priority for all governments, and scholars agree in identifying migrants as the poorest among the poor; however, measuring poverty among migrants is still an open issue and this paper aims to provide a contribute in this respect. Considering the limited budget available for poverty contrasting, tools capable to target and focus on most problematic subpopula- tions are indeed desirable, for effective and efficient policy-making. In this perspective, the F-FOD procedure employed in this paper proves valuable, providing decision-makers with both rankings and scores, useful to quantifying the priorities and the impact of possible poli- cies. The relevance of these features is confirmed also in the present paper, where it emerges the primary role of years since migration in ranking poverty and fragility: indeed, migrants most recently arrived (up to 5 years since migration, that is 15.5% of total population) suffer, with different intensity by country of origin, from both poverty and fragility. This suggests implementing policies to facilitate the initial settling of migrants, for example through hous- ing policies (for poverty), or legal status policies (for fragility). The paper also identifies recently arrived Sub-Saharan and Albanian subpopulations as the most critical cases: the first stay in poverty also for longed duration of presence, up to 9 years, the second only when just arrived. These groups, which account for 4.8% of total migrants, are quite certainly the poorest, suggesting policy-makers to design policies tuned to these subpopulations’ needs. As noted in background section, results from previous research (Arcagni et al. 6 Conclusions 2019) evi- denced specific patterns of poverty by country of origin: country of origin here represents the variable which conveys all the dimensions (such as pre-emigration conditions, migra- tory paths, cultural dimensions, etc.) not specifically invoked to explain poverty in the host country; therefore, this paper aims to offer to policy makers a cross cut of migrants’ poverty through this dimension, helping them to tune more conveniently their policy interventions.f On a more general ground, the paper shows how it is possible to develop effective statistical and measuring tools in a multidimensional ordinal setting and, in particular, how it is possible to build rankings in a sound way, providing a firm basis to the produc- tion of these kind of statistics, for policy-makers. Funding Open access funding provided by Università degli Studi di Milano - Bicocca within the CRUI- CARE Agreement. Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Com- mons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/. References Aaberge, R., Andersen, A. S., & Wennemo, T. (1999). Extent, level and distribution of low income in Norway 1979–1995. In B. Gustafsson & P. J. Pedersen (Eds.), Poverty and low income in the nordic countries (pp. 131–168). Vermont: Ashgate. Alkire, S., & Foster, J. (2011a). Counting and multidimensional poverty measurement. Journal of Public Economics, 95(7–8), 476–487. Alkire, S., & Foster, J. (2011b). Understandings and misunderstandings of multidimensional poverty measurement. The Journal of Economic Inequality, 9(2), 289–314. Arcagni, A. (2017). Parsec: An R package for partial orders in socio-economics. In M. Fattore & Bruggemann (Eds.), Partial order concepts in applied sciences (pp. 275–289). Berlin: Springer. Arcagni, A., di Belgiojoso, E. B. B., Fattore, M., & Rimoldi, S. M. (2019). Multidimensional analysis of deprivation and fragility patterns of migrants in Lombardy, using partially ordered sets and self- organizing maps. Social Indicators Research, 141(2), 551–579. g g p Arndt, C., Distante, R., Hussain, M. A., Østerdal, P. L., Pham, L. H., & Ibraimo, M. (2012). Ordinal wel- fare comparisons with multiple discrete indicators: A first order dominance approach and applica- tion to child poverty. In WIDER working paper, 2012/36, ISBN 978-92-9230-499-7. p y g p p Baio, G., Blangiardo, G. C., & Blangiardo, M. (2011). Centre sampling technique in foreign migration surveys: Amethodological note. Journal of Official Statistics, 27(3), 451–465. fi Bárcena-Martìn, E., & Pérez-Moreno, S. (2016). Immigrant-native gap in poverty: A cross-national European perspective. Review of Economics of the Household. https://doi.org/10.1007/s1115 0-015-9321-x. Berti, F., D’Agostino, A., Lemmi, A., & Neri, A. (2014). Poverty and deprivation of migrants versus natives in Italy. International Journal of Social Economics, 41(8), 630–649. Blangiardo, G. C. (2008). L’immigrazione straniera in Lombardia. La settima indagine regionale. Rap- porto 2017. Milano: ISMU-ORIM. Busetta, A. (2016). Foreigners in Italy: Economic living conditions and unmet medical needs. Gen 71, 2–3. Busetta, A. (2016). Foreigners in Italy: Economic living conditions and unmet medical needs. Genus, 71, 2–3. C t t A R b t R & Zi K (2009) Eth i id tit d i i t h hi 71, 2–3. Constant, A., Roberts, R., & Zimmermann, K. (2009). Ethnic identity and immigrant homeownership. 71, 2 3. Constant, A., Roberts, R., & Zimmermann, K. (2009). Ethnic identity and immigrant homeownership. Urban Studies, 46(9), 1879–1898. ( ) Curatolo, S., & Wolleb, G. (2010). Income vulnerability in Europe. In C. Ranci (Ed.), Social vulnerabil-i ( ) Curatolo, S., & Wolleb, G. (2010). Appendix See Table 4. See Table 4. 1 3 3 S. M. L. Rimoldi et al. e 4 Poverty and fragility scores (dominance and incomparability) by country of origin and length of in Italy Table 4 Poverty and fragility scores (dominance and incomparability) by country of origin and length stay in Italy Poverty Fragility Dominance Incomparability Dominance Incomparabilit AL_0–2 0.2847 0.2633 0.4461 0.411 AL_3–5 0.5094 0.4155 0.464 0.4402 AL_6–9 0.599 0.4009 0.4902 0.4587 AL_10–14 0.6388 0.369 0.6569 0.4267 AL_15+ 0.6214 0.3843 0.6614 0.4262 CI_0–2 0.3584 0.332 0.6344 0.4212 CI_3–5 0.542 0.4012 0.7387 0.3862 CI_6–9 0.4293 0.3857 0.7084 0.4104 CI_10–14 0.6851 0.3376 0.7218 0.4097 CI_15+ 0.8333 0.2146 0.7586 0.3684 IPSB_0–2 0.4239 0.3883 0.4414 0.4282 IPSB_3–5 0.4457 0.4014 0.4879 0.4452 IPSB_6–9 0.4619 0.4029 0.453 0.4324 IPSB_10–14 0.5089 0.4218 0.5103 0.4515 IPSB_15+ 0.516 0.4315 0.5355 0.4613 LA_0–2 0.5426 0.4072 0.4248 0.4057 LA_3–5 0.4414 0.4121 0.4324 0.4169 LA_6–9 0.5448 0.4167 0.6389 0.4439 LA_10–14 0.5775 0.4067 0.5939 0.455 LA_15+ 0.6133 0.3874 0.6322 0.447 NA_0–2 0.3422 0.3199 0.4158 0.4036 NA_3–5 0.5127 0.4211 0.4901 0.4641 NA_6–9 0.5074 0.422 0.512 0.4584 NA_10–14 0.4555 0.4083 0.47 0.4434 NA_15+ 0.5202 0.4267 0.5364 0.4569 PH_3–5 0.5924 0.3866 0.3578 0.3867 PH_6–9 0.4426 0.3891 0.4726 0.444 PH_10–14 0.8482 0.2012 0.6957 0.3867 PH_15+ 0.6085 0.387 0.7548 0.3873 RO_0–2 0.5251 0.4269 0.4845 0.4493 RO_3–5 0.5039 0.4289 0.5401 0.4595 RO_6–9 0.4538 0.4022 0.5627 0.4636 RO_10–14 0.6392 0.3699 0.612 0.4562 RO_15+ 0.655 0.3621 0.6822 0.4193 SSA_0–2 0.125 0.125 0.3569 0.3531 SSA_3–5 0.2649 0.2562 0.4249 0.4113 SSA_6–9 0.3888 0.3678 0.4931 0.4523 SSA_10–14 0.4673 0.4149 0.4943 0.4541 SSA_15+ 0.4806 0.4218 0.5415 0.4573 UM_0–2 0.5395 0.3805 0.2743 0.2743 UM_3–5 0.5067 0.3926 0.5705 0.4606 UM_6–9 0.559 0.4052 0.62 0.4579 Targeting Policies for Multidimensional Poverty and Social… Table 4 (continued) Poverty Fragility Dominance Incomparability Dominance Incomparability UM_10–14 0.5631 0.4077 0.5254 0.4655 UM_15+ 0.5657 0.402 0.5353 0.4563 References Assimilation effects on poverty among migrants in Norway. Journal of Population Economics, 18(4), 691–718. Giannoni, M. (2010). Migrants inequalities in unmet needs for access to health care in Italy. In Paper presented at the 15th Aies conference, Moncalieri, Italy.f Gobillon, L., & Solignac, M. (2015). Homeownership of migrants in France: Selection effects related to international migration flows, Discussion Paper 9517. Institute for the Study of Labor (IZA). l Hansen, J., & Lofstrom, M. (2003). Immigrant assimilation and welfare participation do migrants assim- ilate into or out of welfare? Journal of Human Resources, 38(1), 74–98. Hansen, J., & Wahlberg, R. (2009). Poverty and its persistence: A comparison of natives and migrants Sweden. Review of the Economics of the Household, 7, 105–132. Istat. (2018). La povertà in Italia. https://www.istat.it/it/archivio/217650. Istat. (2019). Demographic statistics Istat. http://www.demo.istat.it/str2018/index.html. Kazemipur, A., & Halli, S. S. (2001). Migrants and’New Poverty’: The case of Canada. International Migration Review, 35(4), 1129–1156. Koolman, X. (2007). Unmet need for health care in Europe, in EUROSTAT, Comparative EU statistics on income and living conditions: Issues and challenges. In Proceedings of the EU-SILC conference (pp. 181–191). (pp ) Koutsampelas, C. (2015). Blind spots of traditional poverty measurement: The case of migrants. Migra- tion Letters, 12(2), 103–112. Lelkes, O. (2007). Poverty among migrants in Europe. Policy Brief April 2007. Vienna: European Centre for Social Welfare Policy and Research. Lemmi, A., Berti, F., Betti, G., D’Agostino, A., Gagliardi, F., Gambacorta, R., et al. (2013). Povertà e deprivazione. In C. Saraceno, N. Sartor, & G. Sciortino (Eds.), Stranieri e disuguali (pp. 149–174). Bologna: Il Mulino. g Lie, B. (2002). Immigration and immigrants (pp. 1–113). Statistical Analyses 54, Statistics Norway, Norway. Myers, D., & Lee, S. W. (1998). Immigrant trajectories into home ownership: A temporal analysis of residential assimilation. International Migration Review, 32(3), 593–625.i Neggers, J., & Kim, S. H. (1998). Basic posets. Singapore: World Scientific. i Obućina, O. (2014). Paths into and out of poverty among migrants in Sweeden. Acta Sociologica, 57(1), 5–23. Pastor, M. (2014). Stepping stone or sink hole? Migrants, poverty, and the future of Metropolitan Amer- ica. In Conference paper, innovating to end urban poverty conference, March 27–28, University of Southern California, Los Angeles. Pemberton, S., Phillimore, J., & Robinson, D. (2014). Causes and experiences of poverty among eco- nomic migrants in the UK. In IRiS working paper series, 4, University of Birmingham. References Income vulnerability in Europe. In C. Ranci (Ed.), Social vulnerabil- ity in Europe. The new configuration of social risk. Chipenham: Palgrave Macmillan. i Davey, B. A., & Priestley, B. H. (2002). Introduction to lattices and order. Cambridge: CUP. D id E & W i k S (2011) T i i h hi i i d i i i Davey, B. A., & Priestley, B. H. (2002). Introduction to lattices and order. Cambridge: CUP. y y g Davidov, E., & Weick, S. (2011). Transition to homeownership among immigrant groups and nat West Germany, 1984–2008. Journal of Immigrant & Refugee Studies, 9(4), 393–415. de Bustillo, R. M., & Anton, J. I. (2011). From rags to riches? Immigration and poverty in Spain. Popu- lation Research and Policy Review, 30, 661–676. di Belgiojoso, E. B., Chelli, F. M., & Paterno, A. (2009). Povertá e standard di vita della popolazione straniera in Lombardia [Poverty and living standard of foreign population in Lombardy Region]. Rivista Italiana di Economia, Demografia e Statistica, LXIII, 3(4), 23–30. di Belgiojoso, E. B., & Rimoldi, S. (2005). Povertá e immigrazione straniera: Resoconto dellesperienza di unindagine pilota nella realtá lombarda. In G. Rovati (Ed.), Le dimensioni della povertá. Stru- menti di misura e politiche. Roma: Carocci. p Diana, P., & Strozza, S. (2014). Strategie abitative degli immigrati nel casertano: la costruzione di una tipologia. In P. Donadio, G. Gabrielli, & M. Massari (Eds.), Uno come te. Europei e nuovi europei nei percorsi di integrazione (pp. 262–279). Milano: Franco Angeli. p g (pp ) g Fattore, M. (2016). Partially ordered sets and the measurement of multidimensional ordinal deprivation. Social Indicators Research, 128(2), 835–858. 1 3 3 S. M. L. Rimoldi et al. Fattore, M., & Arcagni, A. (2018). A reduced posetic approach to the measurement of multidimensional ordinal deprivation. Social Indicators Research. https://doi.org/10.1007/s11205-016-1501-4.i Fattore, M., & Arcagni, A. (2019). F-FOD: Fuzzy first-order dominance analysis and populations rank- ing over ordinal multi-indicator systems. Social Indicators Research, 144(1), 1–29. i lti-indicator systems. Social Indicators Research, 1 Fellini, I., & Migliavacca, M. (2010). Unstable employment in Western Europe: Exploring the individual and household dimensions. In C. Ranci (Ed.), Social vulnerability in Europe. The new configura- tion of social risk. Chippenham: Palgrave Macmillan. Fernades, A., & Pereira, M. J. (2009). Health and migration in the European Union: Better health for all in an inclusive society. London: Pro-book, Publishing Limited.f Galloway, T. A., & Aaberge, R. (2005). Sullivan, D. H., & Ziegert, A. L. (2008). Hispanic immigrant poverty: Does ethnic origin matter? Popula- tion Research and Policy Review, 27(6), 667. Publisher’s Note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. Targeting Policies for Multidimensional Poverty and Social… References Pradhan, P., Costa, L., Rybski, D., Lucht, W., & Kropp, J. P. (2017). A systematic study of sustainable development goal (SDG) interactions. Earth’s Future, 5, 1169–1179. https://doi.org/10.1002/2017E F000632. Core Team. (2019). R: A language and environment for statistical computing. R Foundation for Statis- tical Computing, Vienna, Austria. https://www.R-project.org/. Rimoldi, S. M. L., & di Belgiojoso, E. B. (2016). Poor migrants! Evidence from the Italian Case. Athens Journal of Social Sciences, 3(2), 99–112. f Robinson, R., Reeve, K., & Casey, R. (2007). The housing pathways to migrants. York: Joseph Rowntree Foundation. guilinda, I. S., di Belgiojoso, E. B., González, Ferrer A., Rimoldi, S. M. L., & Blangiardo, G. C. (2017). Surveying immigrants in Southern Europe: Spanish and Italian strategies in comparative perspective. Sanguilinda, I. S., di Belgiojoso, E. B., González, Ferrer A., Rimoldi, S. M. L., & Blangiardo, G. C. (2017). Surveying immigrants in Southern Europe: Spanish and Italian strategies in comparative perspective. Comparative Migration Studies, 5, 17. https://doi.org/10.1186/s40878-017-0060-4. Schröder, B. S. (2003). Ordered sets (pp. 29–30). Berlin: Springe 1 3 Publisher’s Note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. 3 1
https://openalex.org/W2041658927	https://link.springer.com/content/pdf/10.1007/s10584-014-1084-5.pdf	English	null	The impacts of climate change on river flood risk at the global scale	Climatic change	2,014	cc-by	9,778	Climatic Change (2016) 134:387–401 DOI 10.1007/s10584-014-1084-5 Climatic Change (2016) 134:387–401 DOI 10.1007/s10584-014-1084-5 The impacts of climate change on river flood risk at the global scale Nigel W. Arnell & Simon N. Gosling Received: 25 January 2013 /Accepted: 7 February 2014 /Published online: 6 March 2014 # The Author(s) 2014. This article is published with open access at Springerlink.com Abstract This paper presents an assessment of the implications of climate change for global river flood risk. It is based on the estimation of flood frequency relationships at a grid resolution of 0.5×0.5°, using a global hydrological model with climate scenarios derived from 21 climate models, together with projections of future population. Four indicators of the flood hazard are calculated; change in the magnitude and return period of flood peaks, flood-prone population and cropland exposed to substantial change in flood frequency, and a generalised measure of regional flood risk based on combining frequency curves with generic flood damage functions. Under one climate model, emissions and socioeconomic scenario (HadCM3 and SRES A1b), in 2050 the current 100-year flood would occur at least twice as frequently across 40 % of the globe, approximately 450 million flood-prone people and 430 thousand km2 of flood-prone cropland would be exposed to a doubling of flood frequency, and global flood risk would increase by approximately 187 % over the risk in 2050 in the absence of climate change. There is strong regional variability (most adverse impacts would be in Asia), and considerable variability between climate models. In 2050, the range in increased exposure across 21 climate models under SRES A1b is 31–450 million people and 59 to 430 thousand km2 of cropland, and the change in risk varies between −9 and +376 %. The paper presents impacts by region, and also presents relationships between change in global mean surface temperature and impacts on the global flood hazard. There are a number of caveats with the analysis; it is based on one global hydrological model only, the climate scenarios are constructed using pattern-scaling, and the precise impacts are sensitive to some of the assumptions in the definition and application. N. W. Arnell (*) Walker Institute for Climate System Research, University of Reading, Reading, UK e-mail: n.w.arnell@reading.ac.uk Electronic supplementary material The online version of this article (doi:10.1007/s10584-014-1084-5) contains supplementary material, which is available to authorized users. S. N. Gosling School of Geography, University of Nottingham, Nottingham, UK 1 Introduction (2013) counted each year the number of people living in 1×1° grid cells and flood-prone areas respectively where the simulated flood peak exceeded the current 100-year flood. Feyen et al. (2009, 2012) combined simulated flood frequency curves with flood depth-damage functions to estimate current and future average annual damage. The aims of this paper are (i) to assess the implications of climate change for a number of indicators of flood hazard, across the global domain, and (ii) to assess the effect of climate model uncertainty by using scenarios constructed from a wide range of climate models. The study uses a global-scale hydrological model to simulate river flows (Gosling and Arnell 2011; Arnell and Gosling 2013). Four sets of indicators of flood hazard are used, representing changes in flood frequency, changes in risk, and the numbers of people and area of cropland exposed to changes in hazard. Climate scenarios are derived from the 21 climate models in the Coupled Model Intercomparison Project Phase 3 (CMIP3) data set (Meehl et al. 2007a). 1 Introduction One of the most frequently cited impacts of future climate change is a potential increase in the river flood hazard. There have, however, been very few assessments of changing flood hazard Electronic supplementary material The online version of this article (doi:10.1007/s10584-014-1084-5 contains supplementary material, which is available to authorized users. 388 Climatic Change (2016) 134:387–401 at any scale, and most of these have focused on just one indicator of flood hazard: changes in the frequency of occurrence of specific frequency events (e.g. Bell et al. 2007; Prudhomme et al. 2003; Milly et al. 2002; Lehner et al. 2006; Hirabayashi et al. 2008; Dankers and Feyen 2009; Dankers et al. 2013). A key conclusion of such studies is that the projected effects of climate change on the flood hazard may be very substantial, but are very dependent on climate scenario. Very few studies have considered indicators of the human impact of changes in the flood hazard. Kleinen and Petschel-Held (2007) summed the numbers of people living in river basins where the return period of the current 50-year return period event reduces due to climate change. Hirabayashi and Kanae (2009) and Hirabayashi et al. (2013) counted each year the number of people living in 1×1° grid cells and flood-prone areas respectively where the simulated flood peak exceeded the current 100-year flood. Feyen et al. (2009, 2012) combined simulated flood frequency curves with flood depth-damage functions to estimate current and future average annual damage. at any scale, and most of these have focused on just one indicator of flood hazard: changes in the frequency of occurrence of specific frequency events (e.g. Bell et al. 2007; Prudhomme et al. 2003; Milly et al. 2002; Lehner et al. 2006; Hirabayashi et al. 2008; Dankers and Feyen 2009; Dankers et al. 2013). A key conclusion of such studies is that the projected effects of climate change on the flood hazard may be very substantial, but are very dependent on climate scenario. Very few studies have considered indicators of the human impact of changes in the flood hazard. Kleinen and Petschel-Held (2007) summed the numbers of people living in river basins where the return period of the current 50-year return period event reduces due to climate change. Hirabayashi and Kanae (2009) and Hirabayashi et al. 2.1 Introduction 2.1 Introduction Four indicators of the flood hazard are considered here (Section 2.5) based on the simulation of the flood frequency curve under current and changed climatic conditions. Impacts are summarised across 20 world regions (Supplementary Table 1). 2.2 Climate scenarios 2.2 Climate scenarios Pattern-scaled scenarios are used here in preference to scenarios extracted directly from climate model output for two reasons. First, time series of climate model output incorporate not only a gradual climate change trend, but also multi- decadal variability. This complicates comparisons between time periods, emissions scenarios and climate models. Second, climate scenarios can be constructed to represent a wider range of forcings than in the original set of model simulations: not all the CMIP3 models, for example, were run with all SRES emissions scenarios. This study uses the ClimGen software (Osborn 2009) to derive change patterns and apply them to the CRU TS3.1 baseline climatology to produce scenarios for future monthly climate. The patterns define change in mean monthly temperature, vapour pressure and cloud cover, mean monthly precipitation, the inter-annual variance in monthly precipitation, and the mean monthly number of rain-days. A simple interpolation procedure is used to downscale the change fields from the native climate model resolution to the 0.5×0.5° resolution of the baseline climatology. Climate scenarios for the four SRES emissions scenarios were constructed by rescaling the climate model patterns to the change in global mean surface temperature as estimated with SRES emissions by the simple climate model MAGICC (Osborn 2009). Different MAGICC model parameters are used for each climate model in order to represent different sensitivities to climate change, so the change in global mean surface temperature for a given emissions scenario varies between models (for three of the climate models, patterns were scaled by the average global mean surface temperature from the other models). Climate scenarios for specific prescribed changes in global mean surface temperature were constructed simply by scaling to the defined changes. 2.2 Climate scenarios The effects of climate change are represented by two types of scenario. The first characterise changes in climate under the four IPCC SRES emissions scenarios, corresponding to different rates of future emissions of greenhouse gases. The second characterise changes in climate associated with specific prescribed changes in global mean surface temperature. These sce- narios allow an assessment of the relationship between rate of climate forcing and impact response, and a preliminary evaluation of the magnitude of impact at different levels of change in temperature. The CRU TS3.1 data set (Harris et al. 2013) is used here to characterise current climate, and the period 1961–1990 is taken as the climate baseline: this is approximately 0.3 °C above pre-industrial. Both sets of scenarios are constructed by pattern-scaling the output from 21 of the climate models in the CMIP3 multi-model dataset (Meehl et al. 2007a, b) (Supplementary Table 2). The 21 climate models do not represent a set of independent models, and are not of course to be interpreted as predictions. For the sake of this analysis are all assumed to be equally plausible representations of possible future climates. Pattern-scaling assumes that the spatial fields of change per degree change in global mean surface temperature in climate variables extracted from climate model output can be rescaled to 389 Climatic Change (2016) 134:387–401 match a defined change in global mean surface temperature. Whilst this has been demonstrated to be broadly reasonable for moderate amounts of temperature and precipitation change (Mitchell 2003; Tebaldi and Arblaster 2014), it may not hold for high temperature changes or where forcing stabilises or declines. Pattern-scaled scenarios are used here in preference to scenarios extracted directly from climate model output for two reasons. First, time series of climate model output incorporate not only a gradual climate change trend, but also multi- decadal variability. This complicates comparisons between time periods, emissions scenarios and climate models. Second, climate scenarios can be constructed to represent a wider range of forcings than in the original set of model simulations: not all the CMIP3 models, for example, were run with all SRES emissions scenarios. match a defined change in global mean surface temperature. Whilst this has been demonstrated to be broadly reasonable for moderate amounts of temperature and precipitation change (Mitchell 2003; Tebaldi and Arblaster 2014), it may not hold for high temperature changes or where forcing stabilises or declines. 2.3 Socio-economic scenarios Gridded population and GDP estimates through the 21st century at a spatial resolution of 0.5× 0.5° were taken from the IMAGE 2.3 projections for the SRES storylines (van Vuuren et al. 2007). The population projections for A1 and B1 are the same, with a global population total of 8.1 billion in 2050, and the totals for A2 and B2 are 10.4 and 9 billion respectively. Global per capita GDP is highest in the A1 world, and lowest in A2, but there are large regional differences between the four scenarios. The A1 world is, in most regions, the least diverse, and A2 shows the greatest difference between regions. Estimates of the numbers of people living in flood-prone areas at a spatial resolution of 0.5×0.5° were derived by combining 5′ flood-prone areas identified on the UN PREVIEW Global Risk Data Platform (preview.grid.unep.ch; Peduzzi et al. 2009), with the CIESIN GRUMP population data set for the year 2000 (CIESIN 2004). It is assumed that the proportion of grid cell population living in flood-prone areas does not change over time, and that flood extent does not change. Approximately 623 million people live in flood-prone areas in 2000, rising to 843, 1,083 and 934 million in 2050 under the A1/B1, A2 and B2 population projections respectively (Supplementary Table 3 shows the regional distribution of flood-prone populations). A similar approach was used by Jongman et al. (2012) to estimate exposure to river flooding. The extent of cropland within each 0.5×0.5° grid cell is taken from Ramankutty et al.’s (2008) 5′ data set of crop areas in 2000. It is assumed here that the crop extent does not change through the 21st century. Flood-prone cropland is estimated by overlaying the PREVIEW 390 Climatic Change (2016) 134:387–401 flood-prone areas with the cropland extent. Approximately 7 % of global cropland (around 1 million km2) is here identified as flood-prone (see Supplementary Table 3 for the regional distribution). flood-prone areas with the cropland extent. Approximately 7 % of global cropland (around 1 million km2) is here identified as flood-prone (see Supplementary Table 3 for the regional distribution). There are, of course, limitations to these estimates of flood-prone populations and crop- lands. They are based on the accuracy of the delineated floodplains, and are likely to be underestimates, because only the areas prone to flooding from “major” rivers are identified. 2.3 Socio-economic scenarios Small-scale flooding from small rivers, or indeed flash-flooding, is not included, and neither is flash-flooding within urban areas caused by intense rainfall. 2.4 The hydrological model River flows are simulated at a spatial resolution of 0.5×0.5° using Mac-PDM.09 (Gosling et al. 2010; Gosling and Arnell 2011; Arnell and Gosling 2013), a daily water balance model. The model is driven by monthly input climate data, disaggregated statistically to the daily scale, for a period of 30 years. The flood frequency distribution for a grid cell is estimated by fitting a Generalised Extreme Value (GEV) distribution by the method of L- moments to simulated annual maximum daily flows. The model is run 20 times for each grid cell, with different stochastic disaggregations of the monthly input climate data, and the average hydrological behaviour calculated across the 20 repetitions (Gosling and Arnell 2011). River flows are not routed from one grid cell to another, so the frequency curves just represent floods generated within a 0.5×0.5° (approximately 2,500 km2) catchment. Flood frequency curves tend to become less steep (smaller coefficient of variation) as catchment area increases, so the grid scale curves here are probably steeper than the actual frequency curves in cells with large upstream contributions. It is assumed that there is no change in catchment land use with time. Mac-PDM.09 simulates well the average annual water balance and the distribution of flows through the year (Gosling and Arnell 2011), and produces estimates within the range of other global hydrological models (Haddeland et al. 2011). Validation of simulated flood frequency curves is more challenging because of a lack of observed data. Avisual comparison with Meigh et al’s (1997) regional GEV distributions suggests that the simulated flood frequency curve is close to the “observed” curve in some regions (e.g. South Korea and much of wet south east Asia), is steeper than the observed curve in some regions (e.g. west Africa and south west India), and flatter in other regions (e.g. Zimbabwe and Malawi). There is no clear pattern in the differences between simulated and observed frequency distributions, but no evidence of a consistent bias. 2.5 Indicators of flood hazard 2.5.1 Flood frequency Changes in flood frequency are indexed by (i) change in the return period of the current 100- year flood and (ii) change in the magnitude of the 100-year flood. The focus on the 100-year event enables a direct comparison with other studies (Lehner et al. 2006; Hirabayashi et al. 2008; Dankers and Feyen 2009). 2.5.4 Flood risk Flood risk is indexed by the average annual flood loss, calculated by combining the proba- bility density function of flood magnitudes with a function relating flood magnitude to flood loss. The calculation of the flood risk indicator in this analysis involves two stages. The first stage estimates grid cell indicative average annual flood loss, by combining the grid cell flood frequency curve with a generalised function relating flood magnitude to notional flood loss (not expressed in monetary terms). The second stage multiplies this grid cell indicative average annual flood loss by grid cell flood-prone population to produce “grid cell flood risk”, and sums across cells to produce watershed, regional or global totals of “regional flood risk”. Scaling by grid cell GDP rather than population would introduce the spatial variability in the value of assets exposed to flooding. The approach uses generalised non-monetised loss functions which can be applied consis- tently in each grid cell because it is currently impossible to construct realistic damage functions for each grid cell across the world. The two key issues in the construction of a generalised loss function are (i) the shape of the relationship between flood magnitude and flood loss, and (ii) the return period at which damage begins. The analysis here assumes flood damage starts at the (baseline) 20-year flood level, and focuses on a linear loss function (with loss of 1 corre- sponding to the flood 50 % larger than the 20-year flood). It is assumed in this study that there is no protection against flooding. Grid cell average annual loss, and grid cell and regional risk, can therefore be interpreted as representing the sum of flood protection costs and residual impacts. Sensitivity analyses (Section 3.5) assess the effects of different assumptions about the starting frequency, the shape of the loss function, and level of protection, and of using GDP rather than population to scale grid cell risk. Feyen et al. (2009, 2012) calculated a similar index of flood risk for Europe, although in more detail. They estimated flood depths from simulated river flows, using a high-resolution digital elevation model, and used country-specific depth-damage functions to estimate average annual damage at each model grid cell. They also truncated the damage functions at defined return periods to represent the effect of flood protection, with the protection standard assumed to depend on country GDP. Ward et al. 2.5.3 Cropland exposed to change in hazard This indicator is constructed in a very similar way to the previous indicator, but this time sums the flood-prone cropland in grid cells exposed to an increase or decrease in the frequency of flooding. Again, the indicator is based on changes in the frequency of the current 20-year flood. 2.5.2 Population exposed to change in river flood hazard This indicator represents the numbers of people exposed to a “substantial” change in the frequency of flooding. It is calculated by counting the number of flood-prone people in grid 391 Climatic Change (2016) 134:387–401 cells within a region in which the return period of the current 20-year flood declines to less than 10 years or increases to more than 40 years (a doubling or halving of frequency). The 20- year flood is used as the basis of the indicator because it is assumed that this approximates the lowest level at which development in unprotected floodplains occurs. It is not appropriate to calculate the regional net effect, because the consequences of an increase and decrease in flood frequency are not equivalent. cells within a region in which the return period of the current 20-year flood declines to less than 10 years or increases to more than 40 years (a doubling or halving of frequency). The 20- year flood is used as the basis of the indicator because it is assumed that this approximates the lowest level at which development in unprotected floodplains occurs. It is not appropriate to calculate the regional net effect, because the consequences of an increase and decrease in flood frequency are not equivalent. This indicator is similar in principle to that used by Kleinen and Petschel-Held (2007), but they calculated flood frequency curves at the outlet of major river basins and assumed that everybody within that basin was affected by a change in flood frequency. The current analysis operates at a far finer spatial resolution. 2.5.4 Flood risk (2013) use a generic global stage-damage function, which is almost linear between a depth of 0 and 5 m. 392 Climatic Change (2016) 134:387–401 3.1 Change in flood frequency: a physical overview River floods are generated differently in different geographic environments. They may be generated by intense rainfall exceeding the infiltration capacity of soil, or by rain falling on saturated ground; in this case, the amount of flooding from a given amount of rainfall depends on the extent of saturation. Floods may be generated by the melting of accumulated snow. The effects of climate change on flood characteristics therefore vary across space, depending on flood generating mechanism. Where floods are largely generated by intense rainfall and antecedent conditions are not relevant, then changes in flood characteristics are strongly influenced by changes in the frequency of intense rainfall. Where the extent of saturation is important, then changes in flood characteristics are influenced not only by changes in intense rainfall, but changes in the occurrence of saturated conditions over time; this will depend on both accumulated rainfall and evaporation. Where snowmelt floods are currently important, future floods may increase if snow accumulation increases, and would occur earlier if snowmelt occurs earlier; if higher temperatures mean more winter precipitation falls as rain, then snow accumulation would reduce and snowmelt peaks reduce. At the extreme, the flood regime may change from one dominated by spring snowmelt floods to one characterised by smaller, more frequent rain-fed floods in winter. The effects of climate change on flood characteristics are therefore dependent on context, and are not necessarily a simple function of change in precipitation. 3.2 Change in flood frequency characteristics 3.2 Change in flood frequency characteristics Figure 1 shows the change in the magnitude of the estimated 100-year flood under the SRES A1b emissions scenario for 2050 for the seven illustrative climate models, plus a “consistency map” showing changes with all 21 climate model patterns. Changes across large parts of the world are greater than plus or minus 20 %, and although there are important regional differences between the different climate model patterns, the “consistency map” shows strong agreement on the direction of change across much of the world. There are consistent increases in flood magnitude across humid tropical Africa, south and east Asia, much of South America, and in high latitude Asia and North America. There are consistent decreases in flood magnitude around the Mediterranean, in south west Africa, central America, central Europe and the European parts of Russia. In other parts of the world—including western Europe and much of North America—there is less consistency in change. This is similar to the results of other regional studies (e.g. Lehner et al. (2006), Dankers and Feyen (2009) and Hirabayashi et al. (2008)) which found differences in impacts between scenarios. Increases in the magni- tude of the 100-year flood occur where precipitation increases during at least the flood- generating season(s). Decreases in the magnitude of the 100-year flood occur not only where precipitation decreases (such as around the Mediterranean), but also where precipitation in the future falls as rain rather than snow and the resulting rain-generated peaks are smaller than the current snowmelt-generated peaks. This occurs for example across parts of central Europe and north eastern North America. The percentage change in the magnitude of the 100-year flood is generally larger than the change in the 20-year flood (Supplementary Figure 1). The pattern of change in the return period of the current 100-year flood is similar to the change in the magnitude of the 100-year flood (Supplementary Figure 2). Across large parts of the world the frequency of the current 100-year flood reduces to less than once in 50 years or more than once every 200 years (Fig. 2a and b). For example, with the HadCM3 climate model 393 Climatic Change (2016) 134:387–401 Fig. 1 Percentage change in the magnitude of the estimated 100-year flood under SRES A1b emissions in 2050, for seven climate models, plus the consistency (expressed as a percentage of the total number of models) in projected change across 21 climate models. 3.2 Change in flood frequency characteristics Grid cells where the change is less than the standard deviation due to natural unforced multi-decadal variability (see Arnell and Gosling 2013) are shaded grey. For the consistency plot, grid cells where baseline average annual runoff is less than 10 mm/year are shaded grey Fig. 1 Percentage change in the magnitude of the estimated 100-year flood under SRES A1b emissions in 2050, for seven climate models, plus the consistency (expressed as a percentage of the total number of models) in projected change across 21 climate models. Grid cells where the change is less than the standard deviation due to natural unforced multi-decadal variability (see Arnell and Gosling 2013) are shaded grey. For the consistency plot, grid cells where baseline average annual runoff is less than 10 mm/year are shaded grey Fig. 1 Percentage change in the magnitude of the estimated 100-year flood under SRES A1b emissions in 2050, for seven climate models, plus the consistency (expressed as a percentage of the total number of models) in projected change across 21 climate models. Grid cells where the change is less than the standard deviation due to natural unforced multi-decadal variability (see Arnell and Gosling 2013) are shaded grey. For the consistency plot, grid cells where baseline average annual runoff is less than 10 mm/year are shaded grey pattern the current 100-year flood would occur twice as often across 40 % of the world and over 60 % of south east Asia, central Africa, eastern Europe and Canada. pattern the current 100-year flood would occur twice as often across 40 % of the world and over 60 % of south east Asia, central Africa, eastern Europe and Canada. 3.3 Populations exposed to change in flood frequency 3.3 Populations exposed to change in flood frequency Table 1 shows the global totals of people exposed to changes in the frequency of flooding in 2050, under the four emissions and associated socio-economic scenarios. 3.2 Change in flood frequency characteristics Under A1b, for example, the popula- tion exposed to a doubling of flood frequency ranges from 31 to 449 million across all 21 models. The difference in absolute impact between A1b and A2 is largely due to the higher population in A2 (the temperature change in 2050 is similar between the two), whilst the difference in absolute impact between A1b and B1 is entirely due to the difference in emis- sions/temperature (because the populations are the same). B2 has a smaller climate change than A1b but a higher population, so has only a slightly smaller absolute impact than A1b. The numbers exposed to an increase in frequency are greater than those who would see a reduction in frequency in most models, but in some more people see a reduction in flood frequency. climate model pattern, for example, the total population exposed to a doubling of flood frequency ranges from 323 million under B1 emissions and population to 570 million under A2 emissions and population; considerably more people are exposed to an increase in flood frequency than would see a decrease. The variability in impact, however, is greater between climate models than between future emissions and socio-economic scenarios. Under A1b, for example, the popula- tion exposed to a doubling of flood frequency ranges from 31 to 449 million across all 21 models. The difference in absolute impact between A1b and A2 is largely due to the higher population in A2 (the temperature change in 2050 is similar between the two), whilst the difference in absolute impact between A1b and B1 is entirely due to the difference in emis- sions/temperature (because the populations are the same). B2 has a smaller climate change than A1b but a higher population, so has only a slightly smaller absolute impact than A1b. The numbers exposed to an increase in frequency are greater than those who would see a reduction in frequency in most models, but in some more people see a reduction in flood frequency. The reasons for the variability in global impact between climate models are primarily due to regional differences in the projected precipitation, and hence flood frequency, particularly in the areas of south and east Asia where most flood-prone people live. This is illustrated further in Fig. 3.2 Change in flood frequency characteristics 2c and d, which shows the regional proportions of people exposed to substantial changes in the frequency of flooding in 2050, under the A1b emissions and socio-economic scenario. Figure 3a and b show the relationship between change in global mean surface temperature (relative to 1961–1990) and the global proportion of flood-prone people in 2050 exposed to a doubling or halving in the frequency of flooding. In general, a greater proportion of global flood-prone population is exposed to an increase in flood frequency than a decrease, and the range between the climate model patterns is large. For all but two of the models there is little impact until temperatures reach 0.5 °C above the 1961–1990 mean, and for most the rate of change in impact slows as temperature increases. This is because the indicator is based on the exceedance of a threshold.. The population exposed to a doubling of flood frequency is sensitive to the selection of return period (Supplementary Figure 3). A greater proportion of regional population is exposed to a doubling in frequency of the 100-year flood than to a doubling in frequency of the 20-year flood (62 % of global flood-prone population under HadCM3 compared with 53 %). The proportion exposed to a halving of flood frequency is less sensitive. 3.4 Cropland exposed to change in flood frequency The global flood-prone cropland extent exposed to changes in the frequency of flooding in 2050 is shown in Table 1, under the four emissions scenarios (the cropland area is the same under each scenario, so the differences are entirely due to different changes in temperature). Under the HadCM3 climate model pattern, for example, the area of cropland exposed to an increase in flood frequency ranges from 315 to 428 thousand km2. As with the population exposed to change in frequency, the difference between climate model patterns is greater than the difference between emissions scenarios, and with most models a greater proportion of flood-prone cropland sees an increase in flood frequency than sees a decrease. Again, IPSL and CSIRO-MK3.0 show a greater proportion of cropland with a decrease in flood frequency; HadGEM also shows more cropland with a decrease in flood frequency than an increase, in contrast to projected changes in exposed populations. As with populations exposed to changing flood frequency, the differences in global totals between models are due to differences in projected regional climates and hence flood frequencies (Fig. 3.2 Change in flood frequency characteristics With the HadCM3 394 Climatic Change (2016) 134:387–401 Table 1 Global population and cropland exposed to an increase or decrease in flood frequency; 2050 Millions of people hadcm3 hadgem1 echam5 cgcm3.1 ccsm30 ipsl-cm4 csiro-Mk3.0 min max Population (m) A1B Frequency doubles 447 202 302 316 320 95 41 31 449 843 Frequency halves 76 94 55 64 46 264 130 40 264 A2 Frequency doubles 570 246 388 375 369 108 49 37 588 1,083 Frequency halves 86 108 60 76 49 342 168 43 342 B1 Frequency doubles 323 116 217 190 192 68 24 15 328 843 Frequency halves 65 76 47 47 34 237 110 27 237 B2 Frequency doubles 445 175 302 286 271 84 34 22 445 934 Frequency halves 71 88 49 59 39 278 125 35 278 Millions of km2 of cropland hadcm3 hadgem1 echam5 cgcm3.1 ccsm30 ipsl-cm4 csiro-Mk3.0 min max Cropland (m km2 A1B Frequency doubles 428 199 273 303 309 120 66 59 428 1,058 Frequency halves 180 191 164 111 105 424 195 81 424 A2 Frequency doubles 410 182 257 277 286 112 61 54 413 1,058 Frequency halves 174 181 155 103 97 417 185 77 417 B1 Frequency doubles 315 131 188 196 187 89 40 29 315 1,058 Frequency halves 159 152 137 78 77 386 148 55 386 B2 Frequency doubles 374 162 231 247 238 102 53 43 374 1,058 Frequency halves 171 171 150 94 91 410 168 70 410 % change in flood risk hadcm3 hadgem1 echam5 cgcm3.1 ccsm30 ipsl-cm4 csiro-Mk3.0 min max A1b 187 68 87 90 90 30 28 −9 376 A2 176 68 85 87 85 25 25 −12 371 B1 122 44 58 58 60 14 16 −13 294 B2 155 56 72 74 74 20 21 −12 336 The percentage change in risk is relative to the situation in 2050 in the absence of climate change 395 Climatic Change (2016) 134:387–401 climate model pattern, for example, the total population exposed to a doubling of flood frequency ranges from 323 million under B1 emissions and population to 570 million under A2 emissions and population; considerably more people are exposed to an increase in flood frequency than would see a decrease. The variability in impact, however, is greater between climate models than between future emissions and socio-economic scenarios. 3.2 Change in flood frequency characteristics The differences between impacts on flood-prone people and flood-prone cropland reflect 396 Climatic Change (2016) 134:387–401 0 10 20 30 40 50 60 70 80 90 100 N.Africa W.Africa C.Africa E.Africa SnAfrica S.Asia SEAsia EAsia CentralAsia Australasia W.Europe C.Europe E.Europe W.Asia Canada US Brasil % of flood-prone population c Flood-prone population exposed to a doubling of flood frequency 0 10 20 30 40 50 60 70 80 90 100 N.Africa W.Africa C.Africa E.Africa SnAfrica S.Asia SEAsia EAsia CentralAsia Australasia W.Europe C.Europe E.Europe W.Asia Canada US Brasil % of flood-prone population d Flood-prone population exposed to a halving of flood frequency 0 10 20 30 40 50 60 70 80 90 100 N.Africa W.Africa C.Africa E.Africa SnAfrica S.Asia SEAsia EAsia CentralAsia Australasia W.Europe C.Europe E.Europe W.Asia Canada US Meso-America Brasil SouthAmerica % of flood-prone population e Flood-prone cropland exposed to a doubling of flood frequency 0 10 20 30 40 50 60 70 80 90 100 N.Africa W.Africa C.Africa E.Africa SnAfrica S.Asia SEAsia EAsia CentralAsia Australasia W.Europe C.Europe E.Europe W.Asia Canada US Meso-America Brasil SouthAmerica % of flood-prone population f Flood-prone cropland exposed to a halving of flood frequency -100 0 100 200 300 400 500 N.Africa W.Africa C.Africa E.Africa SnAfrica S.Asia SEAsia EAsia CentralAsia Australasia W.Europe C.Europe E.Europe W.Asia Canada US Brasil % change in flood risk g Change in regional flood risk 786 HadCM3 HadGEM1 ECHAM5 CGCM3.1(T47) CCSM3 IPSL-CM4 CSIRO Mk3.0 0 10 20 30 40 50 60 70 80 90 100 N.Africa W.Africa C.Africa E.Africa SnAfrica S.Asia SEAsia EAsia CentralAsia Australasia W.Europe C.Europe E.Europe W.Asia Canada US Meso-America Brasil South America % region a Proportion of region where the current 100-year flood is exceeded more frequently than once in 50 years 0 10 20 30 40 50 60 70 80 90 100 N.Africa W.Africa C.Africa E.Africa SnAfrica S.Asia SEAsia EAsia CentralAsia Australasia W.Europe C.Europe E.Europe W.Asia Canada US Meso-America Brasil South America Meso-America South America Meso-America South America Meso-America South America % region b Proportion of region where the current1 00-year flood is exceeded less frequently than once in 200 years 0 10 20 30 40 50 60 70 80 90 100 N.Africa W.Africa C.Africa E.Africa SnAfrica S.Asia SEAsia EAsia CentralAsia Australasia W.Europe C.Europe E.Europe W.Asia Canada US Meso-America Brasil South America g b Proportion of region where the current1 00-year flood is exceeded less frequently than once in 200 years Proportion of region where the current 100-year flood is exceeded more frequently than once in 50 years Proportion of region where the current1 00-year flood is exceeded less frequently than once in 200 years 0 10 20 30 40 50 60 70 80 90 100 N.Africa W.Africa C.Africa E.Africa SnAfrica S.Asia SEAsia EAsia CentralAsia Australasia W.Europe C.Europe E.Europe W.Asia Canada US Meso-America Brasil South America % region a Proportion of region where the current 100-year flood is exceeded more frequently than once in 50 years b a 0 10 20 30 40 50 60 70 80 90 100 N.Africa W.Africa C.Africa E.Africa SnAfrica S.Asia SEAsia EAsia CentralAsia Australasia W.Europe C.Europe E.Europe W.Asia Canada US Brasil % of flood-prone population c Flood-prone population exposed to a doubling of flood frequency 0 10 20 30 40 50 60 70 80 90 100 N.Africa W.Africa C.Africa E.Africa SnAfrica S.Asia SEAsia EAsia CentralAsia Australasia W.Europe C.Europe E.Europe W.Asia Canada US Brasil % of flood-prone population d Flood-prone population exposed to a halving of flood frequency 0 10 20 30 40 50 60 70 80 90 100 N.Africa W.Africa C.Africa E.Africa SnAfrica S.Asia SEAsia EAsia CentralAsia Australasia W.Europe C.Europe E.Europe W.Asia Canada US Meso-America Brasil SouthAmerica % of flood-prone population e Flood-prone cropland exposed to a doubling of flood frequency 0 10 20 30 40 50 60 70 80 90 100 N.Africa W.Africa C.Africa E.Africa SnAfrica S.Asia SEAsia EAsia CentralAsia Australasia W.Europe C.Europe E.Europe W.Asia Canada US Meso-America Brasil SouthAmerica % of flood-prone population f Flood-prone cropland exposed to a halving of flood frequency -100 0 100 200 300 400 500 N.Africa W.Africa C.Africa E.Africa SnAfrica S.Asia SEAsia EAsia CentralAsia Australasia W.Europe C.Europe E.Europe W.Asia Canada US Brasil % change in flood risk g Change in regional flood risk 786 HadCM3 HadGEM1 ECHAM5 CGCM3.1(T47) CCSM3 IPSL-CM4 CSIRO Mk3.0 N.Africa W.Africa C.Africa E.Africa SnAfrica S.Asia SEAsia EAsia CentralAsia Australasia W.Europe C.Europe E.Europe W.Asia Canada US Meso-America Brasil South America N.Africa W.Africa C.Africa E.Africa SnAfrica S.Asia SEAsia EAsia CentralAsia Australasia W.Europe C.Europe E.Europe W.Asia Canada US Meso-America Brasi South America Meso-America South America Meso-America South America Meso-America South America Fig. 3.2 Change in flood frequency characteristics 2 Regional change in indicators of flood risk under SRES A1b emissions in 2050. a and b proportion of region where the return period of the current 100-year flood changes to less than 50 years or greater than 200 years. c and d flood-prone population exposed to a doubling or halving of flood frequency. e and f flood- prone cropland exposed to a doubling or halving of flood frequency. g change in regional flood risk. All 21 climate models are shown, and seven illustrative models are highlighted. The grey bars represent the range across all models 0 10 20 30 40 50 60 70 80 90 100 N.Africa W.Africa C.Africa E.Africa SnAfrica S.Asia SEAsia EAsia CentralAsia Australasia W.Europe C.Europe E.Europe W.Asia Canada US Brasil % of flood-prone population c Flood-prone population exposed to a doubling of flood frequency Meso-America South America 0 10 20 30 40 50 60 70 80 90 100 N.Africa W.Africa C.Africa E.Africa SnAfrica S.Asia SEAsia EAsia CentralAsia Australasia W.Europe C.Europe E.Europe W.Asia Canada US Brasil % of flood-prone population d Flood-prone population exposed to a halving of flood frequency Meso-America South America d c Flood-prone population exposed to a halving of flood frequency f f 0 10 20 30 40 50 60 70 80 90 100 N.Africa W.Africa C.Africa E.Africa SnAfrica S.Asia SEAsia EAsia CentralAsia Australasia W.Europe C.Europe E.Europe W.Asia Canada US Meso-America Brasil SouthAmerica % of flood-prone population e Flood-prone cropland exposed to a doubling of flood frequency e Flood-prone cropland exposed to a halving of flood frequency 0 10 20 30 40 50 60 70 80 90 100 N.Africa W.Africa C.Africa E.Africa SnAfrica S.Asia SEAsia EAsia CentralAsia Australasia W.Europe C.Europe E.Europe W.Asia Canada US Meso-America Brasil SouthAmerica Flood-prone cropland exposed to a halving of flood frequency M S M S -100 0 100 200 300 400 500 N.Africa W.Africa C.Africa E.Africa SnAfrica S.Asia SEAsia EAsia CentralAsia Australasia W.Europe C.Europe E.Europe W.Asia Canada US Brasil % change in flood risk g Change in regional flood risk 786 HadCM3 HadGEM1 ECHAM5 CGCM3.1(T47) CCSM3 IPSL-CM4 CSIRO Mk3.0 Meso-America South America g HadCM3 HadGEM1 ECHAM5 CGCM3.1(T47) CCSM3 IPSL-CM4 CSIRO Mk3.0 Fig. 2 Regional change in indicators of flood risk under SRES A1b emissions in 2050. a and b proportion of region where the return period of the current 100-year flood changes to less than 50 years or greater than 200 years. 3.2 Change in flood frequency characteristics 2). 3.2 Change in flood frequency characteristics c and d flood-prone population exposed to a doubling or halving of flood frequency. e and f flood- prone cropland exposed to a doubling or halving of flood frequency. g change in regional flood risk. All 21 climate models are shown, and seven illustrative models are highlighted. The grey bars represent the range across all models Fig. 2 Regional change in indicators of flood risk under SRES A1b emissions in 2050. a and b proportion of region where the return period of the current 100-year flood changes to less than 50 years or greater than 200 years. c and d flood-prone population exposed to a doubling or halving of flood frequency. e and f flood- prone cropland exposed to a doubling or halving of flood frequency. g change in regional flood risk. All 21 climate models are shown, and seven illustrative models are highlighted. The grey bars represent the range across all models Climatic Change (2016) 134:387–401 397 0 10 20 30 40 50 60 70 80 3 2 1 0 oC above 1961-1990 a Flood-prone population exposed to doubling of flood frequency 0 10 20 30 40 50 60 70 80 3 2 1 0 % of flood-prone population % of flood-prone population oC above 1961-1990 b Flood-prone population exposed to halving of flood frequency 0 10 20 30 40 50 60 3 2 1 0 oC above 1961-1990 c Flood-prone cropland exposed to doubling of flood frequency 0 10 20 30 40 50 60 3 2 1 0 % of flood-prone cropland % of flood-prone cropland oC above 1961-1990 d Flood-prone cropland exposed to halving of flood frequency -100 0 100 200 300 400 500 3 2 1 0 % changein risk oC above 1961-1990 e Change in global flood risk HadCM3 HadGEM1 ECHAM5/MPI-OM CGCM3.1(T47) CCSM3 IPSL-CM4 CSIRO-MK3.0 Fig. 3 Relationship between global temperature increase (relative to 1961–1990 mean) and the global-scale impacts of flooding in 2050; a and b population exposed to change in flood frequency, c and d cropland exposed to change in flood frequency and e global flood risk. All 21 climate models are shown, and seven illustrative models are highlighted. 3.2 Change in flood frequency characteristics The SRES A1b 2050 population is assumed 0 10 20 30 40 50 60 70 80 3 2 1 0 oC above 1961-1990 a Flood-prone population exposed to doubling of flood frequency % of flood-prone population 0 10 20 30 40 50 60 70 80 3 2 1 0 % of flood prone population oC above 1961-1990 b Flood-prone population exposed to halving of flood frequency d a % of flood-prone population 0 10 20 30 40 50 60 3 2 1 0 % of flood-prone cropland oC above 1961-1990 d Flood-prone cropland exposed to halving of flood frequency d 0 10 20 30 40 50 60 3 2 1 0 oC above 1961-1990 c Flood-prone cropland exposed to doubling of flood frequency % of flood-prone cropland -100 0 100 200 300 400 500 3 2 1 0 % changein risk oC above 1961-1990 e Change in global flood risk 50 e e HadCM3 HadGEM1 ECHAM5/MPI-OM CGCM3.1(T47) CCSM3 IPSL-CM4 CSIRO-MK3.0 Fig. 3 Relationship between global temperature increase (relative to 1961–1990 mean) and the global-scale impacts of flooding in 2050; a and b population exposed to change in flood frequency, c and d cropland exposed to change in flood frequency and e global flood risk. All 21 climate models are shown, and seven illustrative models are highlighted. The SRES A1b 2050 population is assumed the different distribution of the two exposed sectors. The cropland exposed to a doubling of flood frequency is sensitive to the selection of return period (Supplementary Figure 3). The relationship between change in global mean surface temperature and the global proportion of flood-prone cropland exposed to a change in the frequency of flooding is shown in Fig. 3c and d. The shapes of the functions are similar to those for flood-prone population exposed to changes in flood frequency. 3.5 Change in flood risk The percentage change in global flood risk is shown in Table 1. As with the other indicators, the variation between climate models is greater than the difference between emissions and socio-economic scenarios. With the HadCM3 climate model pattern, for example, global flood risk increases by 122 % under B1 and 187 % under A1b, but the range across all 21 climate models under A1b is from a decrease in risk of 9 % to an increase of 376 %. Only three of the 21 models show a decrease (and, incidentally, all are models with another variant amongst the 21 which show an increase in risk). 398 Climatic Change (2016) 134:387–401 Regional change in flood risk in 2050 (under the A1b scenario) is summarised in Fig. 2g, for the seven illustrative climate models. The global total is strongly influenced by changes in south and east Asia, which varies between climate models. In all models, flood risk decreases in some regions. All seven of the models here show a reduction in regional flood risk across western, central and eastern Europe, although in each case there are strong sub-regional variations. In most cases, flood risk in the UK, France and Ireland increases, but is offset by larger decreases in risk in Germany. Change in risk at the cell-level is strongly related to change in the frequency with which flooding begins (Supplementary Figure 4). The relationship between change in global mean surface temperature and global flood risk in 2050 is shown in Fig. 3e. The wide range between climate model patterns is clear, but two other points are of interest. First, the rate of change in risk does not decrease with increasing temperature, unlike with the other indicators, and for some model patterns accelerates. This is because the indicator is not based on the exceedance of a threshold. Second, for four models global risk decreases with small increases in temperature, but increases thereafter. This is partly because in some populous flood-prone areas the effects of higher temperatures and hence evaporation initially lead to reductions in flooding before being offset by increases in precipitation, and partly because in some regions the proportion of precipitation falling as snow and hence snowmelt floods falls as temperatures rise. Table 1 and Figs. 3.5 Change in flood risk 2g and 3e show change in flood risk assuming a linear damage function, with damage starting at the baseline 20-year flood level, and weighting each cell by flood- prone population. The estimated change in regional risk varies in detail with the assumed shape of the damage function and starting threshold (Supplementary Figure 5a), but the broad patterns are similar and the differences between assumptions are small compared with the differences between the climate models. Similarly, there is generally little difference in regional risk when grid-cell risk is scaled by GDP rather than population (Supplementary Figure 5b). Change in risk is also slightly sensitive to the assumed level of protection (Supplementary Figure 5c: the percentage change is higher when it is assumed that there is some protection to the baseline 50-year flood), but the difference is again small compared with the difference between climate model patterns. The estimated change in risk is also relatively insensitive to whether risk is aggregated over just flood-prone areas or over all populated grid cells (Supplementary Figure 5d). 4 Caveats There are a number of important caveats with this analysis, primarily relating to the climate scenarios as they are applied, the hydrological model used to construct grid cell frequency curves, and the indicators of flood risk. To some extent, they are attributable to the global-scale of analysis, which requires generalisation across large spatial domains; all require further investigation. The climate scenarios define changes in mean monthly precipitation and the year-to-year variability in precipitation. However, they do not include changes in the intensity of large daily precipitation events (an increasing proportion of precipitation falling in larger events is a robust response to climate change: Held and Soden (2006)), and they do not characterise changes in the frequency or spacing of flood-producing precipitation events. They therefore potentially underestimate the effect of climate change on flooding. The hydrological model assumes globally-consistent within-cell routing parameters, so reproduces daily flow regimes of “typical” catchments, rather than catchments with either very rapid or very slow flood responses. It does not route flood flows from one grid cell to Climatic Change (2016) 134:387–401 399 another, so assumes that the grid cell frequency curve adequately characterises the flood hazard in a cell even where in practice flooding in the cell is caused by floods generated upstream. This would probably tend to overestimate the slope of the frequency curve in areas prone to flooding generated from considerable distances upstream, because flood frequency curves tend to flatten as catchment area increases. Also, the flood-prone area is likely to be underestimated, and new data bases (e.g. Pappenberger et al. 2012) may give different indications of regional exposure to flood risk. Dankers et al. (2013) show how projected changes in flood magnitudes can vary considerably between different global hydrological models, largely due to their different representation of evaporation and snowmelt processes. New attempts to estimate current global flood risk (e.g. Ward et al. 2013) use hydrological models at a much finer resolution (of the order of 1x1km2), but these have not yet been applied to estimate future risk. The indicators of changing exposure to flood hazard are based on simple measures of change (a doubling or halving of flood frequency) applied consistently across the globe. Similarly, the grid-cell risk estimates assume that the relationship between flood magnitude and flood loss follows a generic loss function, and that the return period at which flood loss begins is consistent across the globe. 4 Caveats Finally, the indicators do not incorporate the effects of existing or future adaptation; they are to be interpreted as measures of exposure to hazard, rather than actual impact. The change in risk indicator can be interpreted as incorporating both the damages caused by flooding and the costs of investment in protection against loss. Other studies (e.g. Hirabayashi et al. 2013) have used different indicators. The impacts presented here are based on CMIP3 climate models. Climate scenarios from the later generation CMIP5 models are now available, although have been run with different forcings. Arnell and Lloyd-Hughes (2014) estimated populations exposed to changes in flooding using the same approach as in this study, with CMIP5 models and different socio- economic assumptions. The most direct comparison is between the A2 scenario here which produces a range across models of 37–588 million for the population exposed to increased flooding (Table 1), and the combination of RCP8.5 and SSP3 in Arnell and Lloyd-Hughes (2014) which has a range of 118–567 million. The CMIP5 models do not therefore produce substantially different indications of the range in potential impacts. References Arnell NW, Gosling SN (2013) The impacts of climate change on hydrological regimes at the global scale. J Hydrol 486:351–364 Arnell NW, Lloyd-Hughes B (2014) The global-scale impacts of climate change on water resources and flooding under new climate and socio-economic scenarios. Clim Chang 122:127–140 Bell VA et al (2007) Use of a grid-based hydrological model and regional climate model outputs to assess changing flood risk. Int J Climatol 27:1657–1671 Center for International Earth Science Information Network (CIESIN) CU, (IFPRI); IFPRI, Bank; TW, (CIAT); CIdAT (2004) Global Rural-Urban Mapping Project (GRUMP), alpha version: population grids. Socioeconomic Data and Applications Center (SEDAC), Columbia University. Available at http://sedac. ciesin.columbia.edu/gpw. (1 April 2011). NY, Palisades CIdAT (2004) Global Rural Urban Mapping Project (GRUMP), alpha version: population grids. Socioeconomic Data and Applications Center (SEDAC), Columbia University. Available at http://sedac. ciesin.columbia.edu/gpw. (1 April 2011). NY, Palisades Dankers R, Feyen L (2009) Flood hazard in Europe in an ensemble of regional climate scenarios. J Geophys Res Atmos 114. doi:10.1029/2008jd011523 Dankers R et al (2013) A first look at changes in flood hazard in the ISI-MIP ensemble. Proc Natl Acad Sci. doi: 10.1073/pnas.1302078110 Feyen L, Barredo JI, Dankers R (2009) Implications of global warming and urban land use change on flooding in Europe. In: Feyen J, Shannon K, Neville M (eds) Water and urban development paradigms—towards an integration of engineering, design and management approaches. CRC Press, London, pp 217–225 Feyen L et al (2012) Fluvial flood risk in Europe in present and future climates. Clim Chang 1 62 Gosling SN, Arnell NW (2011) Simulating current global river runoff with a global hydrological model: model revisions, validation, and sensitivity analysis. Hydrol Process 25:1129–1145 Gosling SN et al (2010) Global hydrology modelling and uncertainty: running multiple ensembles with a campus grid. Philos Trans R Soc A Math Phys Eng Sci 368:4005–4021 Haddeland I et al (2011) Multimodel estimate of the global terrestrial water balance: setup and first results. J Hydrometeorol 12:869–884 Harris I, Jones PD, Osborn TJ, Lister DH (2013) Updated high-resolution grids of monthly climatic observations: the CRU TS3.10 data set. Int J Climatol. doi:10.1002/joc.3711 Held I, Soden BJ (2006) Robust responses of the hydrological cycle to global warming. J Clim 19:5686–56 Held I, Soden BJ (2006) Robust responses of the hydrological cycle to global warming. J Clim 19:5686–5699 Hirabayashi Y, Kanae S (2009) First estimate of the future global population at risk of flooding. 5 Conclusions The key conclusion of this paper is that climate change has the potential to substantially change human exposure to the flood hazard, but that there is considerable uncertainty in the magnitude of this impact between different projections of regional change in climate (partic- ularly precipitation). For example, under one climate model pattern (HadCM3) and future scenario (A1b), in 2050 approximately 450 million flood-prone people would be exposed to a doubling of flood frequency, as would around 430 thousand km2 of flood-prone cropland. The total global flood risk would increase by 187 %, compared to the situation in the absence of climate change. At the same time, flood frequency would be reduced for around 75 million people and 180 thousand km2 of flood-prone cropland. With the HadCM3 climate model pattern, most of these impacts would arise in south and east Asia, where precipitation is projected to increase across flood-prone areas. Other climate models project different changes in precipitation in these populous areas (some more, most less), and this is the primary reason why estimates of impact vary between climate models. The ranges in 2050 (with the A1b scenario) in estimated numbers of people and cropland exposed to a doubling of flood frequency, and change in risk are 31–449 million, 59–428 thousand and −9 % to 376 % respectively across 21 climate models. The range between climate models is considerably 400 Climatic Change (2016) 134:387–401 larger than the range (for a given climate model) between emissions and socio-economic scenarios, and is largely driven by changes in projected flood characteristics in Asia. Acknowledgments The research presented in this paper was conducted as part of the QUEST-GSI project, funded by the UK Natural Environment Research Council (NERC) under the QUEST programme (grant number NE/E001890/1). The climate scenarios were constructed using the ClimGen software package developed by Dr Tim Osborn, University of East Anglia. Summary statistics from the simulated river flow data, including flood statistics, are available at badc.nerc.ac.uk (search for QUEST-GSI). We thank the referees and editors for their comments. Open Access This article is distributed under the terms of the Creative Commons Attribution License which permits any use, distribution, and reproduction in any medium, provided the original author(s) and the source are credited. References Hydrol Res Lett 3:6–9 e d , Sode J ( 006) obust espo ses o t e yd o og ca cyc e to g oba wa g. J C 9:5686 56 Hirabayashi Y, Kanae S (2009) First estimate of the future global population at risk of flooding. Hydrol Res L 3:6–9 Hirabayashi Y et al (2008) Global projections of changing risks of floods and droughts in a changing climate. Hydrol Sci J 53:754–772 y Hirabayashi Y et al (2013) Global flood risk under climate change. Nat Clim Chang. doi:10.1038/nclimate1 Hirabayashi Y et al (2013) Global flood risk under climate change. Nat Clim Chang. doi:10.1038/nclimate1911 Jongman B, Ward PJ, Aerts JCJH (2012) Global exposure to river and coastal flooding: long term trends and challenges. Glob Environ Chang 22:823–835 Jongman B, Ward PJ, Aerts JCJH (2012) Global exposure to river and coastal flooding: long term trends and challenges. Glob Environ Chang 22:823–835 Kleinen T, Petschel-Held G (2007) Integrated assessment of changes in flooding probabilities due to climate change. Clim Chang 81:283–312 Lehner B et al (2006) Estimating the impact of global change on flood and drought risks in Europe: a continental, integrated analysis. Clim Chang 75:273–299 401 Climatic Change (2016) 134:387–401 Meehl GA et al (2007a) The WCRP CMIP3 multimodel dataset—a new era in climate change research. Bull Am Meteorol Soc 88:1383 Meehl GA et al (2007b) Global climate projections. In: Solomon S (ed) Climate change 2007: the physical science basis. Contribution of working group 1 to the fourth assessment report of the intergovernmental panel on climate change. Cambridge University Press, Cambridge g g p panel on climate change. Cambridge University Press, Cambridge Meigh JR, Farquharson FAK, Sutcliffe JV (1997) Aworldwide comparison of regional flood estimation meth and climate. Hydrol Sci J 42:225–244 Milly PCD et al (2002) Increasing risk of great floods in a changing climate. Nature 415:514–517 Mitchell TD (2003) Pattern scaling—an examination of the accuracy of the technique for describing future climates. Clim Chang 60:217–242 g Osborn TJ (2009) A user guide for ClimGen: a flexible tool for generating monthly climate data sets scenarios. Climatic Research Unit, University of East Anglia, Norwich, 17pp Pappenberger F et al (2012) Deriving global flood hazard maps of fluvial floods through a physical mo cascade. Hydrol Earth Syst Sci 16:4143–4156 y y Peduzzi P et al (2009) Assessing global exposure and vulnerability towards natural hazards: the disaster risk index. References Nat Hazards Earth Syst Sci 9:1149–1159 Peduzzi P et al (2009) Assessing global exposure and vulnerability towards natural hazards: the disaster risk index. Nat Hazards Earth Syst Sci 9:1149–1159 y Prudhomme C, Jakob D, Svensson C (2003) Uncertainty and climate change impact on the flood regime of small UK catchments. J Hydrol 277:1–23 Prudhomme C, Jakob D, Svensson C (2003) Uncertainty and climate change impact on the flood regime of small UK catchments. J Hydrol 277:1–23 Ramankutty N et al (2008) Farming the planet: 1. Geographic distribution of global agricultural lands in the year 2000. Glob Biogeochem Cycles 22. doi:10.1029/2007gb002952 Ramankutty N et al (2008) Farming the planet: 1. Geographic distribution of global agricultural lands in the year 2000. Glob Biogeochem Cycles 22. doi:10.1029/2007gb002952 Tebaldi C, Arblaster JM (2014) Pattern-scaling: its strengths and limitations, and an update on the latest model simulations. Clim Chang. doi:10.1007/s10584-013-1032-9 Tebaldi C, Arblaster JM (2014) Pattern-scaling: its strengths and limitations, and an update on the latest model simulations. Clim Chang. doi:10.1007/s10584-013-1032-9 Tebaldi C, Arblaster JM (2014) Pattern-scaling: its strengths and simulations. Clim Chang. doi:10.1007/s10584-013-1032-9 Vuuren DP et al (2007) Stabilizing greenhouse gas concentrations at low levels: an assessment of reduction strategies and costs. Clim Chang 81:119–159 van Vuuren DP et al (2007) Stabilizing greenhouse g strategies and costs. Clim Chang 81:119–159 Ward PJ et al (2013) Assessing flood risk at the global scale: model setup, results and sensitivity. Environ Res Lett 8:044019. doi:10.1088/1748-9326/8/4/044019 Ward PJ et al (2013) Assessing flood risk at the global scale: model setup, results and sensitivity. Environ Res Lett 8:044019. doi:10.1088/1748-9326/8/4/044019
https://openalex.org/W2736193154	https://europepmc.org/articles/pmc5504729?pdf=render	English	null	Successful treatment of refractory complete separation of an esophagojejunal anastomosis after laparoscopic total gastrectomy: a case report	BMC research notes	2,017	cc-by	5,214	© The Author(s) 2017. This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/ publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated. Abstract Background: Anastomotic leakage after total gastrectomy occurs despite improvements in surgical techniques and patient management. Although many cases of dehiscence can be managed non-operatively, major leakage requires a second surgery and can potentially lead to death. Therefore, accurate and immediate diagnosis and treatment are essential. Case presentation: In this report, we describe a 66-year-old Japanese man who was diagnosed with a complete separation of an esophagojejunal anastomosis after laparoscopic total gastrectomy with oral contrast radiography using Gastrografin®. The severe complication was successfully treated by re-anastomosis after two emergency drain- age surgeries. After the second surgery, the esophageal end formed a fistula with the jejunum, but balloon dilation failed to open the fistula. Therefore, oral ingestion and conservative treatment were considered unsuitable, and we performed esophagojejunal re-anastomosis 7 months after the first surgery. At a follow-up examination 2 years after re-anastomosis, the patient weighed 47 kg, and his ingestion had recovered to 80% of that before surgery. Conclusions: Complete separation of an esophagojejunal anastomosis is a rare but severe complication of total gas- trectomy. Therefore, we consider that once separation is diagnosed, aggressive and urgent re-operation and effective drainage are useful. Moreover, it is necessary to take great care to minimize the operative morbidity associated with esophagojejunal anastomosis. Keywords: Anastomotic leakage, Esophagojejunal anastomosis, Laparoscopic total gastrectomy Successful treatment of refractory complete separation of an esophagojejuna anastomosis after laparoscopic total gastrectomy: a case report Shinichi Oka* , Shinichi Sakuramoto, Motohiro Chuman, Kenichi Aratani, Mitsuo Wakata, Yutaka Miyawaki, Hisashi Gunji, Hiroshi Sato, Koujun Okamoto, Shigeki Yamaguchi and Isamu Koyama Correspondence: oka1@cameo.plala.or.jp Department of Gastroenterological Surgery, Saitama Medical University International Medical Center, 1397‑1 Yamane, Hidaka, Saitama 350‑1298, Japan Oka et al. BMC Res Notes (2017) 10:267 DOI 10.1186/s13104-017-2589-6 Oka et al. BMC Res Notes (2017) 10:267 DOI 10.1186/s13104-017-2589-6 BMC Research Notes CASE REPORT Successful treatment of refractory complete separation of an esophagojejunal anastomosis after laparoscopic total gastrectomy: a case report Shinichi Oka , Shinichi Sakuramoto, Motohiro Chuman, Kenichi Aratani, Mitsuo Wakata, Yutaka Miyawaki, Hisashi Gunji, Hiroshi Sato, Koujun Okamoto, Shigeki Yamaguchi and Isamu Koyama Open Access Open Access CASE REPORT Backgroundh death [5]. Therefore, accurate and immediate diagnosis and treatment are essential. We searched PubMed, limit- ing our search to articles published in English during the last 20 years, but could not find any literature reports of dehiscence with complete separation after total gastrec- tomy. In previous reports, major leakage often indicated that the esophageal end and the elevated jejunal end were partially separated; however, in the present case, the ends were completely separated (by approximately 5 cm). Therefore, the present case appears to be very rare and involved potentially fatal complications. Here, we report a case of a complete separation of an esophagojejunal anastomosis after laparoscopic total gastrectomy (LTG), The incidence of anastomotic leakage after total gastrec- tomy is 8.0–14.5% [1–4], and these levels persist despite improvements in surgical techniques and patient man- agement. Although many cases of anastomotic leakage are minor and can be cured by non-operative treatments, such as fasting or hyperalimentation, major leakage occa- sionally requires a second surgery and can even lead to *Correspondence: oka1@cameo.plala.or.jp Department of Gastroenterological Surgery, Saitama Medical University International Medical Center, 1397‑1 Yamane, Hidaka, Saitama 350‑1298, Japan Oka et al. BMC Res Notes (2017) 10:267 Page 2 of 9 Oka et al. BMC Res Notes (2017) 10:267 which we successfully treated by re-anastomosis after two emergency drainage surgeries. which we successfully treated by re-anastomosis after two emergency drainage surgeries. 25-mm EEA-XL stapler (Covidien, Mansfield, MA, USA). No evidence of vascular or organ injury was observed. 25-mm EEA-XL stapler (Covidien, Mansfield, MA, USA). No evidence of vascular or organ injury was observed. Moreover, an intraoperative air leak test performed by immersion of the anastomosis in saline solution was per- formed at the end of the stapling, and no bubble escaped from the anastomosis. The operative time was 4 h 23 min, and minimal bleeding occurred. 25-mm EEA-XL stapler (Covidien, Mansfield, MA, USA). No evidence of vascular or organ injury was observed. Moreover, an intraoperative air leak test performed by immersion of the anastomosis in saline solution was per- formed at the end of the stapling, and no bubble escaped from the anastomosis. The operative time was 4 h 23 min, and minimal bleeding occurred. Case presentation A 66-year-old Japanese man underwent LTG with D1 plus lymph node dissection and Roux-en-Y reconstruc- tion for gastric cancer. His treatment course is summa- rized in Fig. 1. At hospitalization, his nutritional status, glucose tolerance, liver function, and lipid metabolism were all within normal limits; therefore, elective LTG was performed. Esophageal transection was performed with an endoscopic linear stapler. Then, a transorally inserted anvil (OrVil™; Covidien, Mansfield, MA, USA) was intro- duced into the esophagus. The operator indicated that the OrVil™ tube had reached the esophageal stump and made a small hole in the center of the esophageal stump with scissors. The tube was then extracted through the hole until the anvil reached the esophageal stump. The tube was then disconnected from the anvil by cutting the connecting thread and was removed from the abdomi- nal cavity. Next, the jejunum was separated with a linear stapler approximately 25 cm from the Treitz ligament (Fig. 2). The elevated jejunum was retrocolically ele- vated. Esophagojejunostomy was then performed using a The pathological findings were as follows: L, Post, Type 0–IIb, 54 × 36 mm, por 2 > sig, pT1a (M), UL (+), ly0, v0, pN0 (0/47), pPM0, pDM0, pStage IA. Following our usual clinical pathway, the patient was mobile by day 1 after surgery, drinking water on day 2, and had started a liquid diet by day 3. Although he devel- oped a fever of approximately 38 °C on day 3, no asso- ciated abdominal pain was reported, and the drainage fluid was yellow and clear. However, on day 5, the patient complained of sudden abdominal pain while walking, and the drainage fluid had become turbid. Therefore, we stopped the diet immediately. On a subsequent abdomi- nal computed tomography (CT) scan, no effusion was found near the esophagojejunal anastomosis or under the diaphragm; therefore, we concluded that the drainage was sufficient and that the leakage was minor. Thus, we began antibiotic medication. Despite our intervention, Fig. 1 Clinical course of the patient. The patient’s clinical symptoms resolved, including his abdominal pain and fever, and his blood data improved (his white blood cell count was 8670/mm3 on day 7 after the second surgery). ABPC aminobenzylpenicillin, BT body temperature, CMZ cefmetazole, CT computed tomography, MCIPC cloxacillin, WBC white blood cell Fig. 1 Clinical course of the patient. Case presentation The patient’s clinical symptoms resolved, including his abdominal pain and fever, and his blood data improved (his white blood cell count was 8670/mm3 on day 7 after the second surgery). ABPC aminobenzylpenicillin, BT body temperature, CMZ cefmetazole, CT computed tomography, MCIPC cloxacillin, WBC white blood cell Oka et al. BMC Res Notes (2017) 10:267 Page 3 of 9 Fig. 2 Esophagojejunostomy using a transoral anvil. A transoral anvil (OrVil™ 25 mm) was used for esophagojejunostomy. No vascular or organ injury occurred during the surgery Fig. 2 Esophagojejunostomy using a transoral anvil. A transoral anvil (OrVil™ 25 mm) was used for esophagojejunostomy. No vascular or organ injury occurred during the surgery Therefore, a second emergency surgery was performed. We found an accumulation of ascites near the esopha- geal hiatus and the elevated jejunum stump. Although the esophageal end had been sutured during the previous surgery, the sutures had failed to hold, and the stump was open, with the esophagus pulled into the mediastinum, allowing saliva to leak into the abdominal cavity. How- ever, we did not re-close the esophageal end but instead opted to leave a tube in situ to drain saliva. The elevated jejunal end was also open, but we closed this end firmly using serosa-muscle-layer sutures that introverted the mucosa. We also created an intestinal fistula to permit nutrition, which was formed at 20 cm on the anal side from the Y portion. the abdominal pain continued to worsen, and the patient could not walk; therefore, we performed oral contrast radiography with Gastrografin® on day 6. This procedure showed that the contrast medium immediately leaked from the esophageal end of the anastomosis to the pel- vic space, and we diagnosed complete separation of the anastomosis (Fig. 3). We considered lavage and drainage necessary and performed emergency surgery on the same day. During emergency laparoscopy, saliva and contami- nated ascites were observed to cause a thick white collec- tion around the esophagojejunal anastomosis. Complete separation of the esophagojejunal anastomosis was con- firmed after irrigating the upper abdomen; the ends were separated by approximately 5 cm, and the remaining staples formed a B shape (Fig. 4). No evidence of color change was present in any layer of the anastomosis at the site of the elevated jejunum, suggesting that this dehis- cence was unlikely due to insufficient blood perfusion. Case presentation Re-anastomosis was considered too difficult in the pres- ence of ongoing inflammation; therefore, the esophageal and jejunal openings were closed by sutures to prevent the leakage of digestive fluid and saliva. The surgery was completed by leaving drainage tubes in the esophageal end, in both sub-diaphragmatic spaces, and in the pouch of Douglas.i After the second surgery, the patient’s clinical symp- toms resolved, including his abdominal pain and fever, and an improvement in his blood data (white blood cell count, 8670/mm3 on day 7 after the second surgery) was noted. We then started enteral nutrition on day 43 (Fig. 1), and although oral contrast radiography showed leakage of the contrast medium from the dehiscence into the drainage tube at 2 and 3 months after the sur- gery, it had stopped by 4 months postoperatively. By this time, contrast medium was observed to flow from the esophageal end to the elevated jejunum, suggesting the development of a fistula (Fig. 6). The fistula measured approximately 6 cm in length and 0.2 cm in diameter. Given the location and path of the fistula, we performed balloon dilation (Fig. 7) as a non-operative treatment every 1 or 2 weeks. Although we dilated the fistula to On day 12, 6 days after the first emergency surgery, a CT scan suggested continued peritonitis with evidence of diffuse peritoneal hypertrophy and fluid accumulation in the anterior area of the inferior duodenal flexure (Fig. 5). Oka et al. BMC Res Notes (2017) 10:267 Page 4 of 9 Fig. 3 Oral contrast radiography. Gastrografin® collections were evident, suggesting dehiscence Fig. 3 Oral contrast radiography. Gastrografin® collections were evident, suggesting dehiscence Due to the lack of a satisfactory outcome from non- operative treatment, we chose to perform esophagoje- junal re-anastomosis after 7 months. A complete fistula was observed between the oesophagus and the jeju- num (Fig. 8) at the point where complete separation of the esophagojejunal anastomosis had previously been observed (Fig. 4). During surgery, we exposed and taped the fistula and resected a 6-cm portion; the esopha- geal end showed inflammation-related hypertrophy and sclerosis, which were pulling it into the mediastinum; therefore, we dissected the esophagus beyond the dehis- cence-induced inflammation and resected the sclerotic esophageal tissue using a linear stapler. We also resected the elevated jejunum stump, including the previous anas- tomosis site, using a linear stapler. Case presentation After confirming that both the esophagus and elevated jejunum had sufficient blood perfusion and were not tense or twisted, we cre- ated a side-to-side esophagojejunal anastomosis using an Endo-GIA stapler (Covidien, Mansfield, MA, USA), over- lapping both ends on the left side of the esophagus, again using an Endo-GIA stapler. Fig. 4 The refractory total separation of the esophagojejunal anas- tomosis region. Laparoscopy revealed saliva and contaminant ascites causing a thick white collection around the esophagojejunal anasto- mosis. The esophagojejunal anastomosis was completely separated, with the ends approximately 5 cm apart, and the remaining staples formed a B shape Fig. 4 The refractory total separation of the esophagojejunal anas- tomosis region. Laparoscopy revealed saliva and contaminant ascites causing a thick white collection around the esophagojejunal anasto- mosis. The esophagojejunal anastomosis was completely separated, with the ends approximately 5 cm apart, and the remaining staples formed a B shape Oral contrast radiography performed 4 days after the re-anastomosis showed no signs of dehiscence (Fig. 9); therefore, oral ingestion was started 8 days after surgery. However, the patient developed an abdominal abscess 0.8 cm in diameter, oral intake was not achieved, and we concluded that non-operative treatment alone would not be sufficient. Page 5 of 9 Oka et al. BMC Res Notes (2017) 10:267 caused his general condition to worsen. Because of the endocarditis and cardiac tamponade, it was suggested that extremely severe inflammation had been caused by partial injury of the pericardium during esophageal resection in the re-anastomosis. We provided ultra- sound- and CT-guided abdominal, pericardial, and tho- racic drainage, together with antibiotic treatment, and the patient’s general condition improved. Later, ingestion was started on day 28 after the re-anastomosis, and the patient was discharged 8 months after the initial surgery, 47 days after the re-anastomosis. Fig. 5 Computed tomography 6 days after the first emergency surgery. Fluid accumulation could be observed in the anterior area of the inferior duodenal flexure together with peritoneal hypertrophy of the whole abdominal area Endoscopic examination at 1 year after re-anastomosis showed no stenosis at the anastomosis site, food obstruc- tion, or reflux esophagitis (Fig. 10). Approximately 2 years after discharge, the patient weighed 47 kg, and his ingestion had recovered to 80% of that before sur- gery. The patient continues to enjoy sporting activities as he previously did and attends our outpatient clinic every 3 months. Fig. Case presentation 5 Computed tomography 6 days after the first emergency surgery. Fluid accumulation could be observed in the anterior area of the inferior duodenal flexure together with peritoneal hypertrophy of the whole abdominal area Discussion Although surgical techniques have improved, dehis- cence still occurs with a certain frequency, with postop- erative esophagojejunal anastomotic leakage occurring at a rate of 0.5–11.0% [1–12]. According to a report of secondary to a wound infection on day 9 after the re- anastomosis, progressing to endocarditis, cardiac tam- ponade, and bilateral pleural effusions by day 18, which Fig. 6 Oral contrast radiography revealed a fistula. Contrast medium was observed leaking from the dehiscence into the drainage tube after 2 and 3 months, but this had stopped by 4 months postoperatively, and the contrast medium was observed to flow from the esophagus to the jejunum, suggesting that a fistula had formed between the two ends Fig. 6 Oral contrast radiography revealed a fistula. Contrast medium was observed leaking from the dehiscence into the drainage tube after 2 and 3 months, but this had stopped by 4 months postoperatively, and the contrast medium was observed to flow from the esophagus to the jejunum, suggesting that a fistula had formed between the two ends Oka et al. BMC Res Notes (2017) 10:267 Page 6 of 9 Fig. 7 Balloon dilatation was attempted to widen the fistula. The size of the fistula was approximately 6 cm in length and 0.2 cm in diameter when noted. Therefore, to avoid further invasive surgery, we first opted to repeat dilation every 1 or 2 weeks. Although we successfully increased the diameter of the fistula to 0.8 cm, the patient remained unable to ingest food i Fig. 7 Balloon dilatation was attempted to widen the fistula. The size of the fistula was approximately 6 cm in length and 0.2 cm in diameter when noted. Therefore, to avoid further invasive surgery, we first opted to repeat dilation every 1 or 2 weeks. Although we successfully increased the diameter of the fistula to 0.8 cm, the patient remained unable to ingest food listed in this report, but several studies have indicated that anastomotic leakage is observed significantly more often in laparoscopic surgery than in open total gas- trectomy or laparoscopic distal gastrectomy [12–14]. However, conflicting data exist [15], and several studies have reported that the surgeon’s experience of LTG is related to the rate of anastomosis-related complications [16–20]. According to Jeong et al., multivariate analysis showed that postoperative morbidity significantly dif- fered according to the surgeon’s experience (fewer than 45 cases) [19]. Fig. 8 Surgical findings at re-anastomosis. Discussion We performed esophago- jejunal re-anastomosis at 7 months after the last surgery. During surgery, a complete fistula was found between the ends of the esophagus and the jejunum Several devices, such as circular staplers or linear sta- plers, are used by surgeons for various methods includ- ing the OrVil™ method [21], overlap method [22], and the functional end-to-end anastomosis method [23]. We are attempting to standardize the techniques used for esophagojejunal anastomosis using the OrVil™ method for LTG [21]. In our case, the surgeon’s experience of LTG included approximately 50 cases. Furthermore, we performed esophagojejunostomy using OrVil™ as usual in this case. This level of experience may be sufficient. However, it may be necessary to take great care to mini- mize operative morbidity associated with esophagojeju- nal anastomosis. Fig. 8 Surgical findings at re-anastomosis. We performed esophago- jejunal re-anastomosis at 7 months after the last surgery. During surgery, a complete fistula was found between the ends of the esophagus and the jejunum the Japanese National Clinical Database for digestive surgery, the incidence of anastomotic leakage after total gastrectomy was 4.4% (881 of 20,011 cases) in 2011 [12]. The details (whether open or laparoscopic) were not Page 7 of 9 Oka et al. BMC Res Notes (2017) 10:267 Fig. 9 Oral contrast radiography 4 days after re-anastomosis. The outflow of contrast medium was good, and no stenosis or reflux occurred fluoroscopy and abdominal CT examinations [25]. How- ever, sometimes anastomotic leakage cannot be diagnosed clearly from gastric fluoroscopy alone, and evaluations of the distribution of an intraperitoneal abscess or the effec- tiveness of drainage can be difficult [26]. fi In the present case, a CT scan was performed to evalu- ate whether any intraperitoneal fluid accumulation had occurred. No effusion was found near the esophagojeju- nal anastomosis or under the diaphragm; however, oral contrast radiography with Gastrografin® showed that the contrast medium was leaking from the esophageal end of the anastomosis to the pelvic space, and we diagnosed a complete separation of the anastomosis. Therefore, we consider that both fluoroscopic examination and CT scans should be performed immediately when dehiscence is suspected, even when no contrast medium drains from drainage tubes.f Anastomotic leakage may simply affect quality of life but can also cause abdominal pain, fever, and difficulties with ingestion that require the patient to be hospital- ized for prolonged periods [3, 5, 6]. Discussion Minor leakage rarely becomes severe, whereas major leakage sometimes leads to peritonitis accompanied by sepsis and multiple organ failure or death. When the dehiscence is major, urgent re- operation is sometimes needed, but the reported mortal- ity rate with re-laparotomy is as high as 37.5% [1]. Fig. 9 Oral contrast radiography 4 days after re-anastomosis. The outflow of contrast medium was good, and no stenosis or reflux occurred Fig. 10 Gastrointestinal endoscopy 1 year after re-anastomosis. Endoscopy at 1 year after re-anastomosis showed no stenosis, food obstruction, or reflux esophagitis In our case, the anastomosed ends of the esophagus and elevated jejunum were completely separated. We considered lavage and drainage of the abdominal cavity to be necessary and performed an emergency surgery. Moreover, an additional emergency surgery was required to suppress inflammation. After two emergency drainage surgeries, we finally performed re-anastomosis surgery successfully despite the complete separation of the anas- tomosis. In our case, we consider that infection control and aggressive urgent re-operation were the most impor- tant and useful procedures adopted. Despite developing a complete anastomotic separation, which is an extremely severe complication, our patient was eventually treated successfully. However, it is note- worthy that treatment plans were changed regularly dur- ing the treatment course based on the patient’s general condition and nutritional status, as well as the develop- ment of wound infection and an intraperitoneal abscess. Consequently, re-anastomosis was not performed until 7 months later. We have now observed many times that the elevated jejunum can reach the esophageal stump in a natural state without any twisting. We consider that this procedure can prevent excessive tension. In addition, we now add an extra three or four non-absorbent sutures (4-0) through all layers of the anastomosis to reinforce the anastomosis. Fig. 10 Gastrointestinal endoscopy 1 year after re-anastomosis. Endoscopy at 1 year after re-anastomosis showed no stenosis, food obstruction, or reflux esophagitis Symptoms of anastomotic leakage include abdomi- nal pain, peritoneal irritation, backache, lumbago, and continuous remittent fever over 38 °C, accompanied by leukocytosis and an elevated C-reactive protein level [24]. Diagnosis is established based on these symptoms and laboratory data, together with findings on gastric Oka et al. BMC Res Notes (2017) 10:267 Page 8 of 9 Page 8 of 9 Page 8 of 9 Competing interests Competing interests The authors declare that they have no competing interests. Authors’ contributions SO and SS enrolled the patient based on the clinical and surgical aspects, performed the endoscopic operations, drafted the article, and read and approved the final manuscript. MC, KA, MW, YM, HG, HS, KO, SY, and IK are members of the Department of Gastroenterological Surgery, Saitama Medical University International Medical Center. All authors read and approved the final manuscript. 12. Deguchi Y, Fukagawa T, Morita S, Ohashi M, Saka M, Katai H, et al. Iden- tification of risk factors for esophagojejunal anastomotic leakage after gastric surgery. World J Surg. 2012;36:1617–22. 12. Deguchi Y, Fukagawa T, Morita S, Ohashi M, Saka M, Katai H, et al. Iden- tification of risk factors for esophagojejunal anastomotic leakage after gastric surgery. World J Surg. 2012;36:1617–22. 13. Watanabe M, Miyata H, Gotoh M, Baba H, Kimura W, Tomita N, et al. Total gastrectomy risk model: data from 20,011 Japanese patients in a nation- wide internet-based database. Ann Surg. 2014;260:1034–9. 14. Okabe H, Tsunoda S, Tanaka E, Hisamori S, Kawada H, Sakai Y. Is laparoscopic total gastrectomy a safe operation? A review of various anastomotic techniques and their outcomes. Surg Today. 2015;45:549–58. Abbreviations ABPC i b g 10. Meyer L, Meyer F, Dralle H, Ernst M, Lippert H, Gastinger I. Insufficiency risk of esophagojejunal anastomosis after total abdominal gastrectomy for gastric carcinoma. Langenbecks Arch Surg. 2005;390:510–6. 10. Meyer L, Meyer F, Dralle H, Ernst M, Lippert H, Gastinger I. Insufficiency risk of esophagojejunal anastomosis after total abdominal gastrectomy for gastric carcinoma. Langenbecks Arch Surg. 2005;390:510–6. ABPC: aminobenzylpenicillin; BT: body temperature; CMZ: cefmetazole; CT: computed tomography; LTG: laparoscopic total gastrectomy; MCIPC: cloxacil- lin; WBC: white blood cell. 11. Hyodo M, Hosoya Y, Hirashima Y, Haruta H, Kurashina K, Saito S, et al. Mini- mum leakage rate (0.5%) of stapled esophagojejunostomy with sacrifice of a small part of the jejunum after total gastrectomy in 390 consecutive patients. Dig Surg. 2007;24:169–72. 11. Hyodo M, Hosoya Y, Hirashima Y, Haruta H, Kurashina K, Saito S, et al. Mini- mum leakage rate (0.5%) of stapled esophagojejunostomy with sacrifice of a small part of the jejunum after total gastrectomy in 390 consecutive patients. Dig Surg. 2007;24:169–72. Conclusions 3. Isozaki H, Okajima K, Ichinona T, Hara H, Fujii K, Nomura E. Risk factors of esophagojejunal anastomotic leakage after total gastrectomy for gastric cancer. Hepatogastroenterology. 1997;44:1509–12. 3. Isozaki H, Okajima K, Ichinona T, Hara H, Fujii K, Nomura E. Risk factors of esophagojejunal anastomotic leakage after total gastrectomy for gastric cancer. Hepatogastroenterology. 1997;44:1509–12. In conclusion, we present a rare case of the complete separation of an esophagojejunal anastomosis after LTG, which we successfully treated by re-anastomosis after two emergency drainage surgeries. y 4. Nomura S, Sasako M, Katai H, Sano T, Maruyama K. Decreasing complica- tion rates with stapled esophagojejunostomy following a learning curve. Gastric Cancer. 2000;3:97–101. 4. Nomura S, Sasako M, Katai H, Sano T, Maruyama K. Decreasing complica- tion rates with stapled esophagojejunostomy following a learning curve. Gastric Cancer. 2000;3:97–101. 5. Lang H, Piso P, Stukenborg C, Raab R, Jähne J. Management and results of proximal anastomotic leaks in a series of 1114 total gastrectomies for gastric carcinoma. Eur J Surg Oncol. 2000;26:168–71. 5. Lang H, Piso P, Stukenborg C, Raab R, Jähne J. Management and results of proximal anastomotic leaks in a series of 1114 total gastrectomies for gastric carcinoma. Eur J Surg Oncol. 2000;26:168–71. The complete separation of an esophagojejunal anasto- mosis is a rare but severe complication. Therefore, once such a case is diagnosed, aggressive urgent re-operation and effective drainage are required. We believe that this is the most important and useful approach. Moreover, it may be necessary to take additional care to minimize operative morbidity associated with esophagojejunal anastomosis. 6. Sierzega M, Kolodziejczyk P, Kulig J. Impact of anastomotic leakage on long-term survival after total gastrectomy for carcinoma of the stomach. Br J Surg. 2010;97:1035–42. 7. Zilling T, Olseen P, Walther BS. Prediction of hospital stay after total gas- trectomy. Anticancer Res. 1997;17:1355–9. 8. Pol B, LeTreut YP, Hardwigsen J, Rosset E, Houvenaeghel G, Delpero JR. Mechanically stapled esophagojejunostomy. Results of a prospective series of 176 cases. Hepatogastroenterology. 1997;44:458–66. 9. Lamb PJ, Griffin SM, Chandrashekar MV, Richardson DL, Karat D, Hayes N. Prospective study of routine contrast radiology after total gastrectomy. Br J Surg. 2004;91:1015–9. 9. Lamb PJ, Griffin SM, Chandrashekar MV, Richardson DL, Karat D, Hayes N. Prospective study of routine contrast radiology after total gastrectomy. Br J Surg. 2004;91:1015–9. Availability of data and materials 16. Jeong O, Ryu SY, Zhao XF, Jung MR, Kim KY, Park YK. Short-term surgical outcomes and operative risks of laparoscopic total gastrectomy (LTG) for gastric carcinoma: experience at a largevolume center. Surg Endosc. 2012;26:3418–25. All patient data and clinical approaches adopted are present in the medical files of Saitama Medical University International Medical Center in Japan. Bibliographic data of the references are available on PubMed, and the conclu- sions are based on the opinion of the expert involved in this study. 17. Jung HS, Seong IO, Ham MY, Hse MJ, Cho HP, Kyo YS. Short-term out- comes of laparoscopic versus open total gastrectomy: a matched-cohort study. Am J Surg. 2013;206:346–51. Consent for publication Written informed consent was obtained from the patient and from the patient’s family to publish this Case Report and any accompanying images. 18. Sakuramoto S, Kikuchi S, Futawatari N, Katada N, Moriya H, Hirai K, et al. Laparoscopy-assisted pancreas- and spleen-preserving total gastrec- tomy for gastric cancer as compared with open total gastrectomy. Surg Endosc. 2009;23:2416–23. Received: 10 October 2016 Accepted: 30 June 2017 Received: 10 October 2016 Accepted: 30 June 2017 21. Sakuramoto S, Kikuchi S, Futawatari N, Moriya H, Katada N, Yamashita K, et al. Technique of esophagojejunostomy using transoral placement of the pretilted anvil head after laparoscopic gastrectomy for gastric cancer. Surgery. 2009;147:742–7. Acknowledgements Th h ld l k The authors would like to thank the surgical and surgical critical nursing teams for their helpful assistance in surgery and postoperative management. 15. Lee SE, Ryu KW, Nam BH, Lee JH, Kim YW, Yu JS, et al. Technical feasibility and safety of laparoscopy-assisted total gastrectomy in gastric cancer: a comparative study with laparoscopy-assisted distal gastrectomy. J Surg Oncol. 2009;100:392–5. 1. Schubert D, Scheidbach H, Kuhn R, Wex C, Weiss G, Eder F, et al. Endo- scopic treatment of thoracic esophageal anastomotic leaks by using silicone-covered, self-expanding polyester stents. Gastrointest Endosc. 2005;61:891–6. 2. Kataoka M, Masaoka A, Hayashi S, Honda H, Hotta T, Niwa T, et al. Prob- lems associated with the EEA stapling technique for esophagojejunos- tomy after total gastrectomy. Ann Surg. 1989;209:99–104. Publisher’s Note 19. Jeong O, Ryu SY, Choi WY, Piao Z, Park YK. Risk factors and learning curve associated with postoperative morbidity of laparoscopic total gastrec- tomy for gastric carcinoma. Ann Surg Oncol. 2014;21:2994–3001. Springer Nature remains neutral with regard to jurisdictional claims in pub- lished maps and institutional affiliations. 20. Park JM, Jin SH, Lee SR, Kim H, Jung IH, Cho YK, et al. Complications with laparoscopically assisted gastrectomy: multivariate analysis of 300 consecutive cases. Surg Endosc. 2008;22:2133–9. Received: 10 October 2016 Accepted: 30 June 2017 26. Kim TH, Kim JH, Shin CI, Kim SH, Han JK, Choi BI. CT findings suggesting anastomotic leak and predicting the recovery period following gastric surgery. Eur Radiol. 2015;25(7):1958–66. Oka et al. BMC Res Notes (2017) 10:267 Page 9 of 9 Oka et al. BMC Res Notes (2017) 10:267 24. Kiudelis M, Bernotas J, Mickevicius A, Endzinas Z, Meleckas A. Risk factors of esophagojejunal anastomosis leakage after total gastrectomy. Lith Surg. 2013;12:20–4. 25. Woodfield CA, Levine MS. The postoperative stomach. Eur J Radiol. 2005;53(3):341–52. 26. Kim TH, Kim JH, Shin CI, Kim SH, Han JK, Choi BI. CT findings suggesting anastomotic leak and predicting the recovery period following gastric surgery. Eur Radiol. 2015;25(7):1958–66. Page 9 of 9 Page 9 of 9 Page 9 of 9 • We accept pre-submission inquiries • Our selector tool helps you to ﬁnd the most relevant journal • We provide round the clock customer support • Convenient online submission • Thorough peer review • Inclusion in PubMed and all major indexing services • Maximum visibility for your research Submit your manuscript at www.biomedcentral.com/submit Submit your next manuscript to BioMed Central and we will help you at every step: g 25. Woodfield CA, Levine MS. The postoperative stomach. Eur J Radiol. 2005;53(3):341–52. 24. Kiudelis M, Bernotas J, Mickevicius A, Endzinas Z, Meleckas A. Risk factors of esophagojejunal anastomosis leakage after total gastrectomy. Lith Surg. 2013;12:20–4. 25. Woodfield CA, Levine MS. The postoperative stomach. Eur J Radiol. 2005;53(3):341–52. 24. Kiudelis M, Bernotas J, Mickevicius A, Endzinas Z, Meleckas A. Risk factors of esophagojejunal anastomosis leakage after total gastrectomy. Lith Surg. 2013;12:20–4. References g y 22. Inaba K, Satoh S, Ishida Y, Taniguchi K, Isogai J, Kanaya S, et al. Overlap method: novel intracorporeal esophagojejunostomy after laparoscopic total gastrectomy. J Am Coll Surg. 2010;211:e25–9. 1. Schubert D, Scheidbach H, Kuhn R, Wex C, Weiss G, Eder F, et al. Endo- scopic treatment of thoracic esophageal anastomotic leaks by using silicone-covered, self-expanding polyester stents. Gastrointest Endosc. 2005;61:891–6. 23. Uyama I, Sugioka A, Fujita J, Komori Y, Matsui H, Hasumi A. Lapa- roscopic total gastrectomy with distal pancreatosplenectomy and D2 lymphadenectomy for advanced gastric cancer. Gastric Cancer. 1999;2:230–4. 2. Kataoka M, Masaoka A, Hayashi S, Honda H, Hotta T, Niwa T, et al. Prob- lems associated with the EEA stapling technique for esophagojejunos- tomy after total gastrectomy. Ann Surg. 1989;209:99–104. Submit your next manuscript to BioMed Central and we will help you at every step: • We accept pre-submission inquiries • Our selector tool helps you to ﬁnd the most relevant journal • We provide round the clock customer support • Convenient online submission • Thorough peer review • Inclusion in PubMed and all major indexing services • Maximum visibility for your research Submit your manuscript at www.biomedcentral.com/submit Submit your next manuscript to BioMed Central and we will help you at every step:
https://openalex.org/W4306194761	https://zenodo.org/records/7196928/files/2Proceedings-14.12.2019-17-22.pdf	English	null	THE LINGUA FRANCA OF 21ST CENTURY SKILLS	Zenodo (CERN European Organization for Nuclear Research)	2,022	cc-by	2,538	THE LINGUA FRANCA OF 21ST CENTURY SKILLS THE LINGUA FRANCA OF 21ST CENTURY SKILLS Mrs. Lami M.C, Research Scholar, Department of Education, Kuvempu University Shankaraghatta, Shivamogga Dist. Dr. S.S Patil, Professor and Dean, Department of Education, Kuvempu University Shankaraghatta, Shivamogga Dist. THE LINGUA FRANCA OF 21ST CENTURY SKILLS Mrs. Lami M.C, Research Scholar, Department of Education, Kuvempu University Shankaraghatta, Shivamogga Dist. Dr. S.S Patil, Professor and Dean, Department of Education, Kuvempu University Shankaraghatta, Shivamogga Dist. THE LINGUA FRANCA OF 21ST CENTURY SKILLS Mrs. Lami M.C, Research Scholar, Department of Education, Kuvempu University Shankaraghatta, Shivamogga Dist. Dr. S.S Patil, Professor and Dean, Department of Education, Kuvempu University Shankaraghatta, Shivamogga Dist. ABSTRACT 21st century witnesses a tremendous agitation in every field. Education in its fullest sense, fully equipped to shape the future citizens to compete the world of turmoil with their full potential. But the education system unaware of the important threat of making students to cope with the reality. Parents, teachers and even the stakeholders are merely focusing on to the near future whereas the younger ones are to be moulded for a life time. The moment they get graduated and opens up to the world of ‘Startups’, they witness a huge gap between the reality and the theories they have learned. We educate children for a better job with technologies and lots more but the world they have to encounter with, is really a contrary version of what they have thought of. When the 21st century demands us to be critical and creative; teaching, learning and evaluation itself should transform to suit to fit the needs of today. Obsolete pedagogy and methodology in no way help the students to foster the skills needed for 21st century. Desiderata of Lingua Franca should be made mandatory so as to enhance the capabilities of the 21st century students. Keywords: 21st century skills, Lingua Franca of 21st century, desiderata of English Innovative Techniques of School Education and Teacher Education for 21st Century Innovative Techniques of School Education and Teacher Education for 21st Century Innovative Techniques of School Education and Teacher Education for 21st Century Innovative Techniques of School Education and Teacher Education for 21st Century Innovative Techniques of School Education and Teacher Education for 21st Century attention. Youths have the potentiality to move the world to build up a healthy and peaceful environment in the societies. But to bring up the potentialities in youth, the nations should take an oath to provide quality education that shape students with relevant skills to meet the challenges of the world. This transformative vision of education will be fulfilled only through right education. A right education in 21st century which include the development of 21st century skills, the evaluation pattern and lots to count. Education in the new Era "Every word is endowed with a spirit, therefore the speaker or expounder should carefully deliver his words at the appropriate time and place, for the impression which each word makeths is clearly evident and perceptible." (Bahai Writings) The advent of the internet has unleashed forces in our world that we are challenged to harness in a meaningful and constructive way. At one level there is no doubt that the internet is a tremendous instrument in bringing about greater unity and integration but “learning to utilize the Internet in a manner conducive to material and spiritual progress is an immense challenge.” Technology and financial supports alone will not be enough for today’s challenges but quality education in the proper direction should be made mandatory for it to build a civilization that is materially and spiritually prosperous. Survey reports by UNESCO discussed the needs of today’s youth as follows Survey reports by UNESCO discussed the needs of today’s youth as follows • Youth needs to move from traditional to international- from 3Rs to beyond that. • Need to be more responsible and critical. • New approaches and modalities. • New approaches and modalities. Hence, UNESCO’s members have come up with a few changes for Quality education which entails: Hence, UNESCO’s members have come up with a few changes for Quality education • Curricular changes • Changes in evaluative techniques • Proper educational plans, policies and designs • Education for sustainable development INTRODUCTION 21st century witnesses a tremendous agitation in every field. Education in its fullest sense, fully equipped to shape the future citizens to compete the world of turmoil with their full potential. But the education system lacks the important threat of making students to cope with the reality. Parents, teachers and even the stakeholders are merely focusing on to the near future whereas the younger ones are to be moulded for a life time. The moment they get graduated and opens up to the world of ‘Startups’, they witness a huge gap between the reality and the theories they have learned. We educate children for a better job with technologies and lots more but the world they have to encounter with is really a contrary version of what they have thought of. The youngsters equipped with obsolete weapon/ strategies are thrown out by labeling them as ‘good for nothing’ ‘not fit for the job’ and so on. Hence, the need of developing our education system to fit to the 21st century indeed is a necessity and being emphasized by UNESCO in 1991 itself and it is the hottest topic which needs the right Page 17 Srinivas University Srinivas University The Pathway from 3Rs to 4Cs The Pathway from 3Rs to 4Cs Srinivas University Page 18 Page 18 Innovative Techniques of School Education and Teacher Education for 21st Century Innovative Techniques of School Education and Teacher Education for 21st Century The basic skills, 3Rs, in the ancient years, were emphasized a lot due to its ever advancing need in knowledge generation. But in the science and technological era of great ‘Doers’, just mastering the 3Rs won’t fetch any success to the youngsters whereas the new era upholds the necessity of 3Rs being boldly supported by the 21st century skills- especially the 4Cs. Education hitherto given prominence to such basic skills had transferred its attention towards developing the 21st century skills to make the students cope with their future successfully. ‘James Burns’ was rephrasing the 3Rs into a different version which needs attention- 3Rs instead of ‘Reading, wRiting and aRithmatic’ into ‘Respect, Responsibility and Relationship’ which should trigger the young minds. The new age of technology is in an oblivion towards the values. Social Media outburst have made people to openly distress the feelings of others without any bit of guilt. Anyone in the world can harm anyone of you with a fraction of second. The descent of public discourse into greater enmity enlarged the gap in relations- with societies, nations and so on. UNESCO had urged the nations to be prepared for the coming of ‘the end of century’ through their report ‘Education for 21st Century’ in 1991. This report, in an earlier time, foreseen the problems of the education system and proposed an education system that will fit in to the 21st century needs and necessities. Years later, in 2015, the focus of UNESCO shifted towards the soft skills and other mandatory skills to be developed in students by terming those as 21st century skills. The (Learning Skills) 4C Skills- Critical Thinking, Creativity, Communication and Collaboration- are majorly emphasized, along with the Literary Skills and Life Skills, in recent years. Innovative Techniques of School Education and Teacher Education for 21st Century Innovative Techniques of School Education and Teacher Education for 21st Century Innovative Techniques of School Education and Teacher Education for 21st Century Communication skill in English, especially, is an invaluable skill. Mushrooming of spoken English classes, coaching etc are the clear evidence of it. In the world of ‘Startups’, multinational companies, corporate sectors etc rely mainly on English to attract clients and customers nationally and internationally and to prosper and progress in their business. Efficient English communication fetches high profile and high demand in this technological era. Competent English speaking has a great impact on the future of students. Command over English language. To be successful in the 21st century, we need skills those are necessary for this century. The practiced ability only will turn in to skill. English language mastery provides confidence in order to practice and succeed in any skill in the 21st century. Success depends mainly on updating your knowledge base in the respective area which possible through mastery over English Language. Early exposure to English helps to polish a good command of the language. Living in a world of ICT - 21st century- it is essential to converse with people of different nationality which is possible only through a lingua franca- right now it’s other than English Language. The Interconnection of 4Cs and the Lingua Franca (English) ritical thinking teaches students to pose queries and to investigate independently the trut • Critical thinking teaches students to pose queries and to investigate independently the truth. • Creativity makes students to think divergently to produce something novel. • Collaboration teaches students the importance of team work and its benefits. • Communication teaches students to commune effectively to influence people. • Communication teaches students to commune effectively to influence people. It is evident that English language plays a vital role in developing these skills. The more one reads and comprehends the more will be his/her thinking process. When youngsters abreast with current information with the help of internet, it is to be appreciated. Right way of communication and mastery over English are some of the main determinants. Srinivas University English- the Lingua Franca of the Era Effective communication is the heart of every endeavor. To reach the population, communication is the best tool. To tackle problems, to negotiate, to adopt and adapt, to befriend and so on. The Global nature of English language help youngsters to build up their confidence to meet the challenges they face. Knowledge of English- the Library language- improves and increases one’s knowledge in different area as the reservoir of knowledge can be found easily in English language than the native languages of others. The world is interconnected with the help of technology. Now the boundaries of nations are not really a boundary for communication. When the outburst of internet- chatting, video calling etc- helped youngsters to get more opportunities outside their country and could flourish their businesses into a vast interconnected world. With the advent of media such as, Facebook, Instagram and YouTube, people themselves have become international. Through online dealings and more they could commune with people of different land. Hence, the importance of a Lingua Franca truly is a blessing. We are aware that the Lingua Franca of 21st Century is English indeed. Page 19 Srinivas University Evaluation of the new Era One of the main threats of education system in improving the English language is its evaluation system. UNESCO recommended the nations to adopt a better evaluative system to suit 21st century. The advent of constructivism emphasized on learner centered curriculum, new approaches, methods, techniques and Models to attain the goal of student centric teaching-learning curricula but still when it comes to evaluation of languages- first language as well as the second and third, for that matter- no priority were given. The general objective of languages set aside for the specific objectives of the language comprehension. LSRW should be evaluated with its necessities by giving priority to each skill. “Blessed is that teacher who shall arise to instruct the children, and to guide the people into the pathways of God, the Bestower, the Well-Beloved” “Blessed is that teacher who shall arise to instruct the children, and to guide the people into the pathways of God, the Bestower, the Well-Beloved” The requisite for building a better society starts from shaping the younger generation. One of the main determinants of that indeed is the teacher- the Mentor of new Era. A Mentor, in all aspects an exemplary personae to the students. The necessity of communication and the right way of talking should be penetrated into students mind through the mentor. The efficiency of a mentor in English language surely influences the students. This in turn make the students to master over the target Srinivas University Page 20 Innovative Techniques of School Education and Teacher Education for 21st Century Innovative Techniques of School Education and Teacher Education for 21st Century language. Along with being influential by the target language, knowledge of English helps the teacher to be a mentor in all its respect. Amassing information and keeping abreast with the new knowledge, innovations, and being a tech savvy help a mentor in all ways. Your knowledge of the content will be measured by your knowledge in English- the lingua franca, and the knowledge of recent innovations and technologies. 21st century students are far better than their teachers in technology and the ‘Apps’. Hence, up-dating your knowledge is mandatory in which English is a desideratum in 21st century. MOOC courses helps these mentors in lifelong learning. Transacting mentor’s expertise and experience, and socializing with each other is an important part of the educational system. Today's teachers are closely watched because of all of the changes in curriculum and the Common Core. Desiderata to enhance English Language • Well-equipped liberated class rooms where the intellectual curiosity and desire to learn English language is maintained. • Well-equipped liberated class rooms where the intellectual curiosity and desire to learn English language is maintained. • Don’t be afraid of committing errors. • Practice speaking skill then to other skills • Practice polishes the skill • Practice polishes the skill • Use JAM (Just A Minute) technique. • Imaginary conversation in target language. • Read articles and review it in target language • Grammar Quiz application/online grammar quizzes. • Self-evaluation. Srinivas University Page 21 Page 21 Innovative Techniques of School Education and Teacher Education for 21st Century Innovative Techniques of School Education and Teacher Education for 21st Century Conclusion Technology has shaped the ways we communicate; starting from print media to internet and in the future the artificial intelligence. In parallel with this, the Global nature of English- Globish has been uplifted and given importance in 21st century and to enhance the 21st century skills. Reference: 6) https://newsroom.unl.edu/announce/csmce/5344/29195 1) Cox, Janella (2019) Characteristics of a 21st-Century Teacher retrieved from https://www.thoughtco.com/characteristics-of-a-21st-century-teacher-2081448 2) E2030: Education and Skills for the 21st Century (2017) retrieved from http://www.unesco.org/new/fileadmin/MULTIMEDIA/FIELD/Santiago/pdf/Habilidades- SXXI-Buenos-Aires-Eng.pdf 3) E2030: EDUCATION AND SKILLS FOR THE 21ST CENTURY (2017) http://www.unesco.org/new/fileadmin/MULTIMEDIA/FIELD/Santiago/pdf/Meeting- Report-Buenos-Aires-2017-E2030-LAC-ENG.pdf 4) ‘Why Learning English Is So Important In The 21st Century’ retrieved from https://www.englishexplorer.com.sg/why-learning-english-is-so-important-in-the-21st- century/ 5) Thoughful Learning: What are the 4 C's of learning skills?(2016) retrieved from 6) https://newsroom.unl.edu/announce/csmce/5344/29195 1) Cox, Janella (2019) Characteristics of a 21st-Century Teacher retrieved from https://www.thoughtco.com/characteristics-of-a-21st-century-teacher-2081448 2) E2030: Education and Skills for the 21st Century (2017) retrieved from http://www.unesco.org/new/fileadmin/MULTIMEDIA/FIELD/Santiago/pdf/Habilidades- SXXI-Buenos-Aires-Eng.pdf 3) E2030: EDUCATION AND SKILLS FOR THE 21ST CENTURY (2017) http://www.unesco.org/new/fileadmin/MULTIMEDIA/FIELD/Santiago/pdf/Meeting- Report-Buenos-Aires-2017-E2030-LAC-ENG.pdf 3) E2030: EDUCATION AND SKILLS FOR THE 21ST CENTURY (2017) http://www.unesco.org/new/fileadmin/MULTIMEDIA/FIELD/Santiago/pdf/Meeting- Report-Buenos-Aires-2017-E2030-LAC-ENG.pdf 4) ‘Why Learning English Is So Important In The 21st Century’ retrieved from https://www.englishexplorer.com.sg/why-learning-english-is-so-important-in-the-21st- century/ 5) Thoughful Learning: What are the 4 C's of learning skills?(2016) retrieved from 6) https://newsroom.unl.edu/announce/csmce/5344/29195 5) Thoughful Learning: What are the 4 C's of learning skills?(2016) retrieved from 6) https://newsroom.unl.edu/announce/csmce/5344/2919 Page 22 Page 22 Srinivas University
W2581121765.txt	https://clinmedjournals.org/articles/ijnn/international-journal-of-neurology-and-neurotherapy-ijnn-3-043.pdf	en		Neuromyelitis Optica Spectrum Disorder in A Ninety-year-old Woman. A Case Report	International journal of neurology and neurotherapy	2,016	cc-by	1,323	International Journal of Romero-Delgado et al. Int J Neurol Neurother 2016, 3:043 DOI: 10.23937/2378-3001/3/2/1043 Volume 3 \| Issue 2 Neurology and Neurotherapy Case Report: Open Access ISSN: 2378-3001 Neuromyelitis Optica Spectrum Disorder in A Ninety-year-old Woman. A Case Report Fernando Romero-Delgado1, Juan-Pablo Cuello1, Jose-Manuel García-Domínguez1, Yolanda Higueras-Hernández1, Haydee Goicochea-Briceño1, Juan-Adán Guzmán-de-Villoria2 and Maria-Luisa Martínez-Ginés1 1 2 Department of Neurology, Hospital General Universitario Gregorio Marañón, Madrid, Spain Department of Radiology, Hospital General Universitario Gregorio Marañón, Madrid, Spain Corresponding author: Romero-Delgado Fernando, Department of Neurology, Hospital General Universitario Gregorio Marañón, Dr. Esquerdo 46, 28007, Madrid, Spain, Tel: 0034 914269610, E-mail: fernandoromerodelgado@gmail.com Keywords Neuromyelitis optica spectrum disorder (NMOSD), Optic neuritis, Longitudinally extensive myelitis lesion (LETM), Elderly, Nonagenarian, Anti-AQP4 antibody Introduction Neuromyelitis Optica (NMO) is an inflammatory demyelinating, necrotizing disease of the central nervous system, with a predilection for the optic nerves and spinal cord. Clinical, radiological and immunopathological characteristics distinguish it from multiple sclerosis (MS). Recently, new diagnostic criteria have been published, and the terms NMO and neuromyelitis optica spectrum disorder (NMOSD) have been unified, allowing a diagnosis of NMOSD on clinical grounds or on anti-AQP4 antibody status [1]. The onset age of NMO is commonly around the fourth decade of life, but the first attack may occur at any age from early childhood to elderly patients [2,3]. Case Report We report the case of a 90-year-old woman with medical history of atrial fibrillation, anticoagulated with acenocoumarol, valvular pathology resulting from rheumatic heart disease with severe mitral and tricuspid insufficiency, cardioembolic ischemic stroke of the left middle cerebral artery in 2006 with residual aphasia and right hemiparesis, pulmonary hypertension, and chronic renal failure. She was admitted to an emergency department due to subacute, painless loss of sight on her left eye. Fundus examination was normal. Brain Computerized Tomography (CT) scan showed a known residual ischemic lesion on the left temporal region. Laboratory exams were normal, including sedimentation rate. Her symptoms were interpreted as a non-Arteritic Anterior Ischemic Optic Neuropathy and she was discharged. A week later the patient returned to the emergency department with progressive weakness in her lower limbs and upwards hypoesthesia. On physical examination, she was afebrile, had left amaurosis with relative afferent pupillary defect, decreased pain ClinMed International Library and thermal sensation below dorsal four spinal cord level, moderate tetraparesis, urinary retention, generalized hyperreflexia and bilateral babinski sign. Spinal MRI revealed myelitis with a longitudinally extensive spinal cord lesion, extending from the medulla to C6-C7 level, consistent with myelitis of inflammatory origin (Figure 1). Cerebrospinal Fluid (CSF) analysis showed a protein count of 89mg/dL, 2/μL white cells, and a normal glucose. CSF cultures were all negative. No oligoclonal bands were found. Laboratory examination showed positive anti-AQP4 in the serum. Anti-AQP4 determination was performed by cell-based assay using transfected human embryonic kidney cells. Routine blood test, serological studies for viruses and a comprehensive autoimmunity panel (including tests for antinuclear, anti-double-stranded DNA, anti-neutrophil cytoplasmic, anti-smooth muscle antibodies and cryoglobulines) were negative. A whole body CT scan, mammography and neoplastic serological markers were normal. Onconeural antigens Antiamphiphysin, anti-CV2, anti-Hu, anti-Ma2/Ta, anti-Ri and antiYo were all negative. Brain Magnetic Resonance (MRI) revealed ischemic injury on the left middle cerebral artery and signs of mild supratentorial microangiopathy. A follow up CSF analysis showed a 57 mg/dL protein concentration; negative cell counts, negative immunophenotypic study and no oligoclonal bands. Our patient fulfilled classical Wingerchuk criteria for NMO, presenting with a Devic’s syndrome. A diagnosis of NMOSD was made on clinical and AQP4 seropositivity grounds. She was treated with two courses of methylprednisolone therapy (1 g/day during 5 days/course), but no clinical improvement was observed. Plasmapheresis and IV immunoglobulin treatment were considered, but they were rejected because of patient comorbidity. Two months after hospital discharge, the patient died due to decompensated heart failure complicated with bronchopneumonia. Discussion Since the discovery of anti-AQP4-IgG as the pathogenic marker for NMO, the spectrum of NMOSD has widened to include limited and unexpected forms of the disease [1,4]. Optic neuritis is more common in paediatric patients, while myelitis is more common in older patients [3]. The rate of onset above 80 years of age is around 1% for NMOSD [5]. Citation: Romero-Delgado F, Cuello JP, García-Domínguez JM, Higueras-Hernández Y, Goicochea-Briceño H, et al. (2016) Neuromyelitis Optica Spectrum Disorder in A Ninety-year-old Woman. A Case Report. Int J Neurol Neurother 3:043. doi.org/10.23937/2378-3001/3/2/1043 Received: February 01, 2016: Accepted: March 25, 2016: Published: March 28, 2016 Copyright: © 2016 Romero-Delgado F, et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited. DOI: 10.23937/2378-3001/3/2/1043 ISSN: 2378-3001 Figure 1: A) Sagittal T2-weighted MRI of cervical spinal cord demonstrates a longitudinally extensive myelitis lesion (LETM); B) Sagittal T1-weighted MRI with gadolinium shows tenuous irregular peripheral enhancement; C) Axial T2-weighted MRI of cervical spinal cord demonstrates a lesion involving the central cord. Screening for cancer is mandatory even in the anti-AQP4 antibody positive cases, for anti-AQP-4 antibodies have been reported in paraneoplastic context. Breast carcinoma is the most common neoplasm reported with paraneoplastic NMOSD. Lung, uterus, thymus, cervix, bladder, and thyroid carcinomas; seminoma; Romero-Delgado et al. Int J Neurol Neurother 2016, 3:043 ovarian teratoma; pituitary adenoma; lymphoma and monoclonal gammopathy have also been reported. In a minority of cases of paraneoplastic NMOSD, no cancer is found [6,7]. Patients with paraneoplastic NMOSD are generally older than 55 years of age. Only a few cases of NMOSD in nonagenarian patients have been • Page 2 of 3 • DOI: 10.23937/2378-3001/3/2/1043 described so far [5,8-10]. The presence of anti-AQP-4 antibodies, and the previous optic neuritis, prompted us to consider a diagnosis of NMOSD. The possibility of a paraneoplastic NMOSD could not be completely ruled out. Our report suggest that even in the very elderly, NMOSD must be considered in patients with classical features of NMO such as LETM or suspicion of Non-Arteritic Anterior Ischemic Optic Neuropathy without typical features. Anti-AQP4 antibodies should be determined in such cases. Ethical Statement The authors state that they have no Conflict of Interest (COI). References 1. Wingerchuk DM, Banwell B, Bennett JL, Cabre P, Carroll W, et al. (2015) International consensus diagnostic criteria for neuromyelitis optica spectrum disorders. Neurology 85: 177-189. 2. Jacob A, McKeon A, Nakashima I, Sato DK, Elsone L, et al. (2013) Current concept of neuromyelitis optica (NMO) and NMO spectrum disorders. J Neurol Neurosurg Psychiatry 84: 922-930. 3. Nagaishi A, Takagi M, Umemura A, Tanaka M, Kitagawa Y, et al. (2011) Romero-Delgado et al. Int J Neurol Neurother 2016, 3:043 ISSN: 2378-3001 Clinical features of neuromyelitis optica in a large Japanese cohort: comparison between phenotypes. J Neurol Neurosurg Psychiatry 82: 13601364. 4. Wingerchuk DM, Lennon VA, Lucchinetti CF, Pittock SJ, Weinshenker BG (2007) The spectrum of neuromyelitis optica. Lancet Neurol 6: 805-815. 5. Niino M, Uesugi H, Takahashi T, Fukazawa T, Minami N, et al. (2012) Recurrent Brainstem Lesions Mimicking Infarctions in an Elderly Patient with Neuromyelitis Optica Spectrum Disorder. Intern Med 51: 809-812. 6. Figueroa M, Guo Y, Tselis A, Pittock SJ, Lennon VA, et al. (2014) Paraneoplastic neuromyelitis optica spectrum disorder associated with metastatic carcinoid expressing aquaporin-4. JAMA Neurol 71: 495-498. 7. De Santis G, Caniatti L, De Vito A, De Gennaro R, Granieri E, et al. (2009) A possible paraneoplastic neuromyelitis optica associated with lung cancer. Neurol Sci 30: 397-400. 8. Quek AM, McKeon A, Lennon VA, Mandrekar JN, Iorio R, et al. (2012) Effects of age and sex on aquaporin-4 autoimmunity. Arch Neurol 69: 1039-1043. 9. Ohe Y, Fukuoka T, Deguchi I, Hayashi T, Tanahashi N (2014) Case of elderlyonset neuromyelitis optica spectrum disorder mimicking cervical spondylotic myelopathy. Neurology and Clinical Neuroscience 2: 18-19. 10. Kitazawa Y, Warabi Y, Bandoh M, Takahashi T, Matsubara S (2012) Elderly-Onset Neuromyelitis Optica which Developed after the Diagnosis of Prostate Adenocarcinoma and Relapsed after a 23-Valent Pneumococcal Polysaccharide Vaccination. Intern Med 51: 103-107. • Page 3 of 3 •
https://openalex.org/W2168647482	https://pubs.rsc.org/en/content/articlepdf/2016/ob/c5ob02037a	English	null	Evaluation of fluoropyruvate as nucleophile in reactions catalysed by N-acetyl neuraminic acid lyase variants: scope, limitations and stereoselectivity	Organic & biomolecular chemistry	2,016	cc-by	6,064	Evaluation of ﬂuoropyruvate as nucleophile in reactions catalysed by N-acetyl neuraminic acid lyase variants: scope, limitations and The catalysis of reactions involving ﬂuoropyruvate as donor by N-acetyl neuraminic acid lyase (NAL) variants was investigated. Under kinetic control, the wild-type enzyme catalysed the reaction between ﬂuoropyruvate and N-acetyl mannosamine to give a 90 : 10 ratio of the (3R,4R)- and (3S,4R)-conﬁgured products; after extended reaction times, equilibration occurred to give a 30 : 70 mixture of these products. The eﬃciency and stereoselectivity of reactions of a range of substrates catalysed by the E192N, E192N/ T167V/S208V and E192N/T167G NAL variants were also studied. Using ﬂuoropyruvate and (2R,3S)- or (2S,3R)-2,3-dihydroxy-4-oxo-N,N-dipropylbutanamide as substrates, it was possible to obtain three of the four possible diastereomeric products; for each product, the ratio of anomeric and pyranose/furanose forms was determined. The crystal structure of S. aureus NAL in complex with ﬂuoropyruvate was deter- mined, assisting rationalisation of the stereochemical outcome of C–C bond formation. Received 30th September 2015, Accepted 30th October 2015 DOI: 10.1039/c5ob02037a www.rsc.org/obc †Electronic supplementary information (ESI) available. See DOI: 10.1039/ c5ob02037a ‡Present address: Centre for Ultrafast Imaging, Institute of Nanostructure and Solid State Physics, University of Hamburg, Luruper Chausee 149, Hamburg 22761, Germany. aSchool of Chemistry, University of Leeds, Leeds, LS2 9JT, UK. E-mail: a.s.nelson@leeds.ac.uk bAstbury Centre for Structural Molecular Biology, University of Leeds, Leeds, LS2 9JT, UK. E-mail: a.berry@leeds.ac.uk cSchool of Molecular and Cellular Biology, University of Leeds, Leeds, LS2 9JT, UK dSchool of Chemistry, University of St Andrews, St Andrews, KY16 9ST, UK eChemical Development, AstraZeneca, Silk Road Business Park, Macclesfield, Cheshire, SK10 2NA, UK Organic & Biomolecular Chemistry Organic & Biomolecular Chemistry Open Access Article. Published on 05 November 2015. Downloaded on 10/24/2024 6:50:23 AM. This article is licensed under a Creative Commons Attribution 3.0 Unported Licence. Evaluation of ﬂuoropyruvate as nucleophile in reactions catalysed by N-acetyl neuraminic acid lyase variants: scope, limitations and stereoselectivity† Jennifer Stockwell,a,b Adam D. Daniels,b,c Claire L. Windle,b,c Thomas A. Harman,a,b Thomas Woodhall,a,b Tomas Lebl,d Chi H. Trinh,b,c Keith Mulholland,e Arwen R. Pearson,‡b,c Alan Berryb,c and Adam Nelsona,b Organic & Biomolecula PAPER Cite this: Org. Biomol. Chem., 2016, 14, 105 Received 30th September 2015, Accepted 30th October 2015 DOI: 10.1039/c5ob02037a www.rsc.org/obc Published on 05 November 2015. Downloaded on 10/24/2024 6:50:23 AM. icle is licensed under a Creative Commons Attribution 3.0 Unported Licence. Organic & Biomolecula PAPER Cite this: Org. Biomol. Chem., 2016, 14, 105 Received 30th September 2015, Accepted 30th October 2015 DOI: 10.1039/c5ob02037a www.rsc.org/obc Published on 05 November 2015. Downloaded on 10/24/2024 6:50:23 AM. icle is licensed under a Creative Commons Attribution 3.0 Unported Licence. Introduction The stereoselective synthesis of compounds with a fluorine- bearing stereocentre is a significant challenge. Most solutions to this problem rely on stereoselective C–F bond formation, for example by fluorination of allylic silanes.6 Some catalytic methods for enantioselective C–F bond formation have been developed: for example by organocatalytic α-fluorination of aldehydes7 or Pd-catalysed α-fluorination of β-keto phosphonates.8 The introduction of fluorine can have a profound eﬀect on bio- active molecules including their conformation, binding, bio- availability, metabolism, pharmacokinetics and pharmaco- dynamics.1 As a consequence, around 20% of prescribed drugs, and 30% of leading blockbuster drugs, contain at least one fluorine atom.2 Examples of fluorinated pharmaceuticals include the cholesterol-lowering drug Atorvastatin, and Sofosbuvir which is exploited in the treatment of Hepatitis C (Fig. 1).3,4 Moreover, fluorinated sugars can serve as valuable mechanism-based probes of carbohydrate-processing enzymes.5 Fig. 1 Examples of ﬂuorinated drugs. Fig. 1 Examples of ﬂuorinated drugs. Org. Biomol. Chem., 2016, 14, 105–112 \| 105 This journal is © The Royal Society of Chemistry 2016 Organic & Biomolecular Chemistry View Article Online View Article Online Paper Organic & Biomolecular Chemistry Scheme 1 Envisaged strategy for controlling F-bearing stereocentres by C–C bond formation. and E192N/T167V/S208V variants of NAL control the stereo- chemistry of C–C bond formation, and catalyse respectively the selective formation of the alternative diastereomeric products 3a and 3b (Scheme 2).14 Scheme 1 Envisaged strategy for controlling F-bearing stereocentres by C–C bond formation. Results and discussion Scheme 2 Stereoselective aldol reactions catalysed by aldolases gen- erated by directed evolution.14 The products are drawn in open chain form for clarity. Evaluation of wild-type NAL in the synthesis of fluorinated analogues of N-acetyl neuraminic acid Initially, the reaction between fluoropyruvate and ManNAc cat- alysed by wild-type NAL was investigated (Panel A, Scheme 3). Accordingly, the reaction was performed at 37 °C in an NMR tube (20 mM sodium fluoropyruvate and 100 mM N-acetyl mannosamine in 20 mM Tris-HCl pH 7.4 buﬀer) and followed by 19F NMR spectroscopy. After 500 min, the fluoropyruvate was >98% consumed, and a 90 : 10 mixture of the diastereo- meric products 4a and 4d had been formed (which vary only in their configuration at C-3). However, after a prolonged reaction time (∼5 weeks), with regular addition of more enzyme, the ratio of products had switched to 30 : 70 in favour of 4d. The alternative possible diastereomeric products (4b and 4c; Panel B) were not detected. These data suggest that 4a is the kinetic Scheme 2 Stereoselective aldol reactions catalysed by aldolases gen- erated by directed evolution.14 The products are drawn in open chain form for clarity. Scheme 3 Reaction between ﬂuoropyruvate and ManNAc catalysed by We envisaged a complementary catalytic approach in which a F-bearing stereocentre would be controlled by formation of a neighbouring C–C bond (Scheme 1). Aldolase-catalysed reaction involving fluoropyruvate and an aldehyde 1 would yield an aldol product 2 with two new stereogenic centres. This catalytic approach would complement enantioselective aldol reactions involving fluoroacetone.9 N-Acetyl neuraminic acid lyase (NAL) is a Class I aldolase that catalyses the reversible aldol reaction between pyruvate and N-acetyl mannosamine (ManNAc) to give N-acetyl neura- minic acid (Neu5Ac). A combination of mutagenesis, structural biology and computational chemistry has revealed insights into its catalytic mechanism.10 Despite a report that it is not a substrate,11 fluoropyruvate is a viable donor.12 However, diﬀering stereochemical outcomes have been reported for the NAL-catalysed reaction between fluoropyruvate and ManNAc.12 An initial aim of our study was, therefore, to clarify the stereo- chemical outcome of this reaction. We also sought to investigate the catalysed reactions between fluoropyruvate and alternative aldehyde acceptors. Here, we investigated the value of synthetically-useful NAL var- iants that we have previously generated using directed evol- ution.13,14 The E192N variant of NAL is an excellent catalyst of the poorly stereoselective reaction between pyruvate and the alternative substrate (2R,3S)-2,3-dihydroxy-4-oxo-N,N-dipropyl- butanamide, DHOB (Scheme 2).13 The structural basis of the modified substrate specificity of this variant has been gleaned using protein crystallography.15 In contrast, the E192N/T167G Scheme 3 Reaction between ﬂuoropyruvate and ManNAc catalysed by wild-type NAL. Evaluation of wild-type NAL in the synthesis of fluorinated analogues of N-acetyl neuraminic acid Table 2 Kinetic parameters for the cleavage of substrates catalysed by wild-type NALa Substrate kcat/min−1 KM/mM kcat/KM/min−1 mM−1 Neu5Ac 510 ± 10 2.0 ± 0.1 260 4a 0.91 ± 0.03 8.4 ± 0.7 0.11 4b NDb NDb a Determined using a coupled enzyme assay involving lactate dehydrogenase. b Not detectable. Table 2 Kinetic parameters for the cleavage of substrates catalysed by wild-type NALa product of the reaction, and that 4d is the thermodynamic product. The relative thermodynamic stability of 4d may stem from the stabilising gauche interaction between fluorine and vicinal electronegative atoms.16 As with previous studies, the reaction was found to yield selectively (4R)-configured pro- ducts. The diﬀerent ratios of products under kinetic and ther- modynamic control may account for the contrasting selectivities reported in previous studies.12 a Determined using a coupled enzyme assay involving lactate dehydrogenase. b Not detectable. The interpretation of the spectroscopic data was greatly assisted by the preparation of standard samples of the pro- ducts 4a and 4d.12a The reaction between ManNAc and sodium fluoropyruvate, catalysed by wild-type NAL, was con- ducted at 37 °C in 100 mM Tris-HCl pH 7.4 buﬀer, and the products purified by column chromatography. After 24 h reac- tion, the product 4a was obtained in 34% yield; whilst after reaction for >1 week, the product 4d was obtained in 43% yield. In both cases, a single pyranose anomer predominated (Table 1; Panel C, Scheme 3). The configuration of 4a and 4d was determined by careful analysis of vicinal coupling con- stants.17 In both pyranose anomers of 4a, there was a large coupling constant between fluorine and H-4 (∼30 Hz) and a small coupling constant between H-3 and H-4 (2.1 Hz in the major anomer) (Table 1). In contrast, in the major anomer of 4d, there was a small coupling constant between fluorine and H-4 (∼12 Hz) and a large coupling constant between H-3 and H-4 (8.8 Hz). Evaluation of wild-type NAL in the synthesis of fluorinated analogues of N-acetyl neuraminic acid A: The stereochemical outcome is determined by the reaction time. The products are depicted in open chain form for clarity. NAL was regularly added to the 5 week reaction. B: Diastereomeric pro- ducts that were not observed. C: Cyclised forms of 4a and 4d. Scheme 3 Reaction between ﬂuoropyruvate and ManNAc catalysed by wild-type NAL. A: The stereochemical outcome is determined by the reaction time. The products are depicted in open chain form for clarity. NAL was regularly added to the 5 week reaction. B: Diastereomeric pro- ducts that were not observed. C: Cyclised forms of 4a and 4d. 106 \| Org. Biomol. Chem., 2016, 14, 105–112 This journal is © The Royal Society of Chemistry 2016 View Article Online Organic & Biomolecular Chemistry Table 1 Spectroscopic data for the ﬂuorinated products of aldolase-catalysed reactions Product Form (proportion) δF/ ppm δ3H/ ppm δ4H/ ppm δ5H/ ppm δ6H/ ppm 2JHF a/ Hz 3JHF a/ Hz 3J3H–4H/ Hz 3J4H–5H/ Hz 3J5H–6H/ Hz 4a Major pyranose (98%) −208.1 4.65 3.93 4.07 3.87 49.3 30.0 2.1 10.6 10.6 Minor pyranose (2%) −217.9 NMb NMb NMb NMb 51.3 29.9 NMb NMb NMb 4d Major pyranose (96%) −199.3 4.47 ∼3.90 ∼3.90 ∼3.90 49.7 12.0 8.8 NMb NMb 16a Major pyranose (92%) −206.0 4.78 3.94 3.88 4.75 49.9 32.5 3.4 9.7 9.2 Minor pyranose (8%) −216.8 NMb NMb NMb NMb 51.4 32.8 NMb NMb NMb 16c Major pyranose (35%) −190.5 4.85 4.39 4.12 4.65 50.5 24.0 4.8 5.0 6.1 Major furanose (25%) −194.5 4.72 4.02 3.96 NMb 43.7 4.7 1.7 NMb NMb Minor pyranose (30%) −201.9 5.03 4.46 3.95 4.57 53.1 18.7 5.5 5.6 7.2 Minor furanose (10%) −207.4 4.86 4.30 4.18 4.75 48.5 10.1 7.3 NMb 5.4 ent-16d Pyranose (>98%) −199.8 4.60 3.95 3.78 4.62 49.3 13.3 9.3 9.3 9.7 17a Major pyranose (98%) −207.8 4.90 4.16 4.23 4.83 49.0 29.1 2.2 10.9 10.0 Minor pyranose (2%) −218.5 NMb NMb NMb NMb 50.2 28.8 NMb NMb NMb a Determined by analysis of the 296 MHz 19F NMR spectrum. b Not measured. Table 1 Spectroscopic data for the ﬂuorinated products of aldolase-catalysed reactions Open Access Article. Published on 05 November 2015. Downloaded on 10/24/2024 6:50:23 AM. This article is licensed under a Creative Commons Attribution 3.0 Unported Licence. §(2R,3S)-3-acetyl-2-hydroxy-4-oxo-N,N-dipropylbutanamide. This journal is © The Royal Society of Chemistry 2016 Preparation of substrate precursors Downloaded on 10/24/2024 6:50:23 AM. This article is licensed under a Creative Commons Attribution 3.0 Unported Licence. Open Access Article. Published on 05 November 2015. Downloaded on 10/24/2024 6:50:23 AM. This article is licensed under a Creative Commons Attribution 3.0 Unported Licence. Open Access Article. Published on 05 November 2015. Downloaded on 10/24/2024 6:50:2 This article is licensed under a Creative Commons Attribution 3.0 Unported Scheme 5 Possible diastereomeric products of aldolase-catalysed reactions with ﬂuoropyruvate as nucleophile. Scheme 4 Synthesis of precursors of aldehyde substrates. In addition, the alkene ent-15 was prepared from the enantiomeric lactone starting material 11; and 10 was prepared using an established route.16 Scheme 5 Possible diastereomeric products of aldolase-catalysed reactions with ﬂuoropyruvate as nucleophile. (10, ent-10, 15 or ent-16) was cleaved ozonolytically in methanol at −78 °C, and the reaction mixture quenched with dimethyl- sulfide, transferred into an NMR tube and evaporated; the reaction (20 mM sodium fluoropyruvate and 100 mM substrate in 50 mM Tris-HClpH 7.4 buﬀer) to give aldol products was then followed by 296 MHz 19F NMR spectroscopy. The results are presented in Table 3. following acetal hydrolysis, the β-amino amide derivative 14 was obtained in 88% yield. Finally, desulfonylation of 14, fol- lowed by acetylation, gave the required alkene 15. This journal is © The Royal Society of Chemistry 2016 Preparation of substrate precursors To enable evaluation of alternative potential substrates, a range of alkene precursors was prepared: ozonolysis of these alkenes (10, ent-10, 15 and ent-15) would yield the corres- ponding aldehydes (DHOB, ent-DHOB, AHOB§ and ent-AHOB). The alkene13a ent-10 was prepared using a route that was ana- logous to an established18 synthesis of 10 (Scheme 4). Thus, treatment of the lactone 5 (derived from lyxose19) with concen- trated hydrochloric acid in acetone gave the corresponding acetonide20 6. Treatment of 6 with iodine and triphenyl- phosphine gave the corresponding iodolactone 7, which was followed by reductive ring-opening to give the carboxylic acid ent-8 (whose enantiomer had been used to prepare18 10). Finally, amide formation (→9) and deprotection gave the required alkene ent-10. The catalysis of the cleavage of the reaction products 4a and 4d was also studied using an established coupled enzyme assay13b (Table 2). The cleavage of the fluorinated N-acetyl neuraminic acid analogue 4a was much less eﬃcient than that of Neu5Ac itself (kcat/KM: 0.11 min−1 mM−1 for 4a compared with 260 min−1 mM−1 for Neu5Ac). However, the catalysis of the cleavage of the diastereomeric fluorinated analogue 4d was even less eﬃcient and was not detectable under the conditions of the assay. This observation is consistent with (3R,4R)-config- ured 4a being the kinetic product of the NAL-catalysed reaction between fluoropyruvate and ManNAc. The alkenes 15 and ent-15 were prepared from the known21 enantiomerically pure lactones 11 and ent-11 (see Scheme 4 for the synthesis of 15). Treatment of the lactone 11 with LHMDS at −78 °C, and reaction with the N-sulfonyl imine 12, gave the product 13 as a 94 : 6 mixture of diastereomers; the relative configuration of the major diastereomer was determined by subsequent conversion into a cyclic derivative (see below). The lactone 13 was ring-opened by treatment with Pr2NAlMe2 and, Org. Biomol. Chem., 2016, 14, 105–112 \| 107 This journal is © The Royal Society of Chemistry 2016 View Article Online Scheme 5 Possible diastereomeric products of aldolase-catalysed reactions with ﬂuoropyruvate as nucleophile. Organic & Biomolecular Chemistry View Article Online Paper Organic & Biomolecular Chemistry Organic & Biomolecular Chemistry Scheme 4 Synthesis of precursors of aldehyde substrates. In addition the alkene ent-15 was prepared from the enantiomeric lactone startin material 11; and 10 was prepared using an established route.16 p This article is licensed under a Creative Commons Attribution 3.0 Unported Licence. ess Article. Published on 05 November 2015. Open Access Article. Published on 05 November 2015. Downloaded on 10/24/2024 6:50:23 AM. This article is licensed under a Creative Commons Attribution 3.0 Unported Licence. Open Access Article. Published on 05 November 2015. Downloaded on 10/24/2024 6:50:23 AM. This article is licensed under a Creative Commons Attribution 3.0 Unported Licence. a Prepared by ozonolysis of the corresponding alkene (10, ent-10, 15 or ent-15). b Consumption of fluoropyruvate (nmol min−1 per nmol protein) determined by 296 MHz 19F NMR spectroscopy. c Kinetic ratio of diastereomeric products determined by 296 MHz 19F NMR spectroscopy. d Not measured. e Not detectable. DHOB reacted to give ent-16d which was isolated in 52% yield. However, with the E192N variant, fluoropyruvate and DHOB were converted into a 40 : 60 mixture of 16a and 16c from which it was possible to obtain small samples of both products after reverse-phase HPLC purification. Similarly, using E192N, fluoropyruvate and AHOB reacted to give a diastereomeric mixture of products, from which a small sample of 17a could be obtained by mass-directed HPLC. The fluorinated products 16a, ent-16d and 17a existed in pyranose forms (Panel A, Fig. 2). In both pyranose anomers of 16a and 17a, there was a large coupling constant between fluo- rine and H-4 (∼30 Hz); in addition, in the major anomer of each compound, there was a small coupling constant between H-3 and H4 (16a: 3.4 Hz; 17a: 2.2 Hz) (Table 3). These data Table 4 Preparation of ﬂuorinated products of aldolase-catalysed reactions Substratea (eq.) Variant Productb Time/day (conversionc/ %) Yieldd/% (ratioe) DHOB (2 eq.) E192N 16a and 16c 2 (>99) 33 f (40 : 60) DHOB (2 eq.) E192N/T167 V/S208V 16c 2 (95) 41 ent-DHOB (1 eq.) E192N/T167 V/S208V ent-16d 1 (NMg) 52 AHOB (5 eq.) E192N 17a 5 (50) 7h a Prepared by ozonolysis of the corresponding alkene (10, ent-10, 15 or ent-15). b See Table 1 for details of ratios of anomers and pyranose/ furanose forms. c Determined by analysis of the crude product by 296 MHz 19F NMR spectroscopy. d Yield of purified product based on the limiting reactant. e Determined by 296 MHz 19F NMR spectroscopy. f Small samples of each diastereomer could be obtained by reverse- phase HPLC. g Not measured. h After purification by mass-directed HPLC. Table 4 Preparation of ﬂuorinated products of aldolase-catalysed reactions Catalysis by the E192N variant was significantly less eﬃcient with the other substrates investigated. Open Access Article. Published on 05 November 2015. Downloaded on 10/24/2024 6:50:23 AM. This article is licensed under a Creative Commons Attribution 3.0 Unported Licence. For example, with AHOB, in which the α-hydroxy group of DHOB has been replaced with an α-NHAc group, the rate of consumption of fluropyruvate was about 8-fold slower. Switching to the enan- tiomeric substrate series was also detrimental to catalysis: the rate of consumption of fluoropyruvate was about 20-fold slower with ent-DHOB (compared to DHOB) and about 15-fold slower with ent-AHOB (compared to AHOB). Substratea (eq.) Variant Productb Time/day (conversionc/ %) Yieldd/% (ratioe) DHOB (2 eq.) E192N 16a and 16c 2 (>99) 33 f (40 : 60) DHOB (2 eq.) E192N/T167 V/S208V 16c 2 (95) 41 ent-DHOB (1 eq.) E192N/T167 V/S208V ent-16d 1 (NMg) 52 AHOB (5 eq.) E192N 17a 5 (50) 7h In addition, the E192N/T167G and E192N/T167V/S208V var- iants are less eﬃcient catalysts than the E192N variant. For example, with DHOB as substrate, the rate of consumption of fluoropyruvate was about 15- and 150-fold slower with the E192N/T167G and E192N/T167V/S208V variants respectively than with the E192N variant. These variants were generated to catalyse complementary stereoselective reactions between pyru- vate and DHOB (Scheme 2): a reduction in the eﬃciency of cat- alysis (compared to the E192N variant) was also observed with pyruvate as the donor substrate.14 a Prepared by ozonolysis of the corresponding alkene (10, ent-10, 15 or ent-15). b See Table 1 for details of ratios of anomers and pyranose/ furanose forms. c Determined by analysis of the crude product by 296 MHz 19F NMR spectroscopy. d Yield of purified product based on the limiting reactant. e Determined by 296 MHz 19F NMR spectroscopy. f Small samples of each diastereomer could be obtained by reverse- phase HPLC. g Not measured. h After purification by mass-directed HPLC. This journal is © The Royal Society of Chemistry 2016 Evaluation of variant NALs in the catalysis of reactions involving fluoropyruvate The rate of consumption of fluoropyruvate was highest with the combination of DHOB and the E192N variant (9.1 nmol min−1 per nmol protein). This observation is, perhaps, unsur- prising given that E192N was obtained via a directed evolution approach that sought to optimise catalysis of cleavage to yield DHOB.13 However, it is notable that the E192N variant – in addition to the wild-type enzyme – accepts fluoropyruvate as an alternative donor. Eﬃciency of catalysis. The ability of NAL variants to catalyse reactions involving fluoropyruvate as donor was investigated. The aldehydes DHOB, ent-DHOB, AHOB and ent-AHOH were investigated as potential substrates for the E192N, E192N/ T167G and E192N/T167V/S208 V NAL variants (Scheme 5). The eﬃciency of catalysis was initially investigated by determining the rate of consumption of fluoropyruvate by 19F NMR spectroscopy. In each case, the corresponding alkene precursor 108 \| Org. Biomol. Chem., 2016, 14, 105–112 View Article Online Organic & Biomolecular Chemistry Table 3 Activity and selectivity of aldolase variants with ﬂuoropyruvate as donor Table 3 Activity and selectivity of aldolase variants with ﬂuoropyruvate as donor Substratea Variant Specific activityb/nmol min−1 nmol−1 Product Ratioc a : b : c : d (3R,4R) : (3S,4S) : (3R,4S) : (3S,4R)c DHOB E192N 9.1 16 40 : 0 : 50 : 10 40 : 0 : 50 : 10 DHOB E192N/T167V/S208V 0.56 16 0 : 0 : 100 : 0 0 : 0 : 100 : 0 DHOB E192N/T167G 0.06 16 30 : 0 : 70 : 0 30 : 0 : 70 : 0 ent-DHOB E192N 0.46 ent-16 10 : 0 : 0 : 90 0 : 10 : 90 : 0 ent-DHOB E192N/T167V/S208V 0.03 ent-16 0 : 0 : 0 : 100 0 : 0 : 100 : 0 ent-DHOB E192N/T167G 0.12 ent-16 20 : 0 : 0 : 80 0 : 20 : 80 : 0 AHOB E192N 1.1 17 60 : 0 : 40 : 0 60 : 0 : 40 : 0 AHOB E192N/T167V/S208V NDe — AHOB E192N/T167G 0.03 17 NMd ent-AHOB E192N 0.07 ent-17 NMd ent-AHOB E192N/T167V/S208V NDe — ent-AHOB E192N/T167G 0.03 ent-17 NMd a Prepared by ozonolysis of the corresponding alkene (10, ent-10, 15 or ent-15). b Consumption of fluoropyruvate (nmol min−1 per nmol protein) determined by 296 MHz 19F NMR spectroscopy. c Kinetic ratio of diastereomeric products determined by 296 MHz 19F NMR spectroscopy. d Not measured. e Not detectable. Preparation and characterisation of reaction products DHOB reacted to give ent-16d which was isolated in 52% yield. However, with the E192N variant, fluoropyruvate and DHOB were converted into a 40 : 60 mixture of 16a and 16c from which it was possible to obtain small samples of both products after reverse-phase HPLC purification. Similarly, using E192N, fluoropyruvate and AHOB reacted to give a diastereomeric mixture of products, from which a small sample of 17a could be obtained by mass-directed HPLC. DHOB reacted to give ent-16d which was isolated in 52% yield. However, with the E192N variant, fluoropyruvate and DHOB were converted into a 40 : 60 mixture of 16a and 16c from which it was possible to obtain small samples of both products after reverse-phase HPLC purification. Similarly, using E192N, fluoropyruvate and AHOB reacted to give a diastereomeric mixture of products, from which a small sample of 17a could be obtained by mass-directed HPLC. The determination of the stereoselectivity of the reactions was complicated by the possibility of four diastereomeric products, each of which might exist in diﬀerent anomeric and pyranose/ furanose forms. To assist analysis, selected reactions were con- ducted preparatively, and the products purified and character- ised (Table 4). In each case, the aldehyde substrate and sodium fluoropyruvate were dissolved in 50 mM Tris-HCl pH 7.4 buﬀer, and the relevant NAL variant added. The conversion of each reaction was determined by analysis of the crude product by 296 MHz 19F NMR spectroscopy. The fluorinated products 16a, ent-16d and 17a existed in pyranose forms (Panel A, Fig. 2). In both pyranose anomers of 16a and 17a, there was a large coupling constant between fluo- rine and H-4 (∼30 Hz); in addition, in the major anomer of each compound, there was a small coupling constant between H-3 and H4 (16a: 3.4 Hz; 17a: 2.2 Hz) (Table 3). These data imply that 16a and 17a are (3R,4R)-configured (and indirectly enabled determination of the relative configuration of 13). In contrast, ent-16d had a small coupling constant between the In two cases, the aldolase-catalysed reactions were highly diastereoselective, and >98 : <2 mixtures of diastereomeric pro- ducts were obtained after ion exchange chromatography. Thus, with the E192N/T167V/S208V variant, fluoropyruvate and DHOB reacted to give 16c which was isolated in 41% yield. Similarly, with the same NAL variant, fluoropyruvate and ent- This journal is © The Royal Society of Chemistry 2016 Org. Biomol. This journal is © The Royal Society of Chemistry 2016 110 \| Org. Biomol. Chem., 2016, 14, 105–112 Open Access Article. Published on 05 November 2015. Downloaded on 10/24/2024 6:50:2 This article is licensed under a Creative Commons Attribution 3.0 Unported Fig. 2 Forms of reaction products. Panel A: The products 16a, ent-16d and 17a exist predominantly in pyranose forms. Panel B: the product 16c exists as a mixture of pyranose and furanose anomers. equatorially-positioned fluorine and H-4 (13.3 Hz) and a large coupling constant between the axial protons H-3 and H-4 (9.3 Hz). The configuration of 16a and ent-16d was corrobo- rated by the observation of nOe interactions between the axial protons at H-4 and H-6. The analysis of 16c was hugely compli- cated by the existence of both pyranose and furanose anomers (Panel B, Fig. 2); however, 1H/19F HSQC-TOCSY spectroscopy enabled extraction of the 1H NMR spectra of each of the four species that were present (Table 1 and ESI†). The pyranose anomers of 16c have axially-oriented fluorine and 4-OH groups which cannot enjoy a stabilising gauche interaction.16 Stereoselectivity of reactions. The characterisation of the products enabled determination of the stereoselectivity of reactions catalysed by NAL variants (Table 3). With the E192N variant, DHOB and fluoropyruvate yielded a 40 : 0 : 50 : 10 mixture of 16a, 16b, 16c and 16d; this poor stereoselectivity parallels that observed with this enzyme in the reaction between pyruvate and DHOB.13a,14 In contrast, the E192N/T167V/S208V variant yielded selectively the (3R,4S)-con- figured product 16c. This variant was generated14 by directed evolution to yield selectively the 4S-configured product (3b) with pyruvate as nucleophile (Scheme 2): it is remarkable that the 4S selectivity is retained with an alternative nucleophile (fluoropyruvate). However, in contrast, selectivity for 4R-config- ured products was not observed with the E192N/T167G variant: with this variant, the reaction between DHOB and fluoropyruvate was very ineﬃcient, and a 30 : 70 mixture of 16a and 16c was obtained. The eﬀect of the structure of the aldehyde substrate on stereoselectivity was also investigated. AHOB has an α-NHAc group in place of the α-hydroxy group of DHOB; with AHOB and the E192N variant, similarly poor stereoselectivity was also observed with this substrate (17a : 17b : 17c : 17d 60 : 0 : 40 : 0). In the enantiomeric series, ent-DHOB gave predominantly the (3R,4S)-configured product ent-16d with all three NAL variants; here, the NAL variant had only a small eﬀect on the stereo- selectivity of the aldol reaction. Fig. 3 Rationale for stereoselectivity of NAL-catalysed reactions between ﬂuoropyruvate and aldehyde substrates. Panel A: S. Preparation and characterisation of reaction products Chem., 2016, 14, 105–112 \| 109 View Article Online Paper Organic & Biomolecular Chemistry Fig. 2 Forms of reaction products. Panel A: The products 16a, ent-16d and 17a exist predominantly in pyranose forms. Panel B: the product 16c exists as a mixture of pyranose and furanose anomers. . 3 Rationale for stereoselectivity of NAL-catalysed reactions Open Access Article. Published on 05 November 2015. Downloaded on 10/24/2024 6:50:23 AM. This article is licensed under a Creative Commons Attribution 3.0 Unported Licence. ccess Article. Published on 05 November 2015. Downloaded on 10/24/2024 6:50:23 AM. This article is licensed under a Creative Commons Attribution 3.0 Unported Licence. Open Access Article. Published on 05 November 2015. Downloaded on 10/24/2024 6:50:23 AM. This article is licensed under a Creative Commons Attribution 3.0 Unported Licence. Open Access Article. Published on 05 November 2015. Downloaded on 10/24/2024 6:50:2 This article is licensed under a Creative Commons Attribution 3.0 Unported aureus NAL in complex with ﬂuoropyruvate with general acid Y137 and residues that have key roles in recognition and stereocontrol (T167, E192 and S208) shown (PDB: 5A8G). The top face (as depicted) of the Z-conﬁgured enamine is poised to react with an aldehyde substrate. Panel B: Y137A variant of E. coli NAL in compex with 4-epi-Neu5Ac (PDB: 4BWL). Panel C: Possible stereochemical outcomes of the reaction of the top face of the Z-conﬁgured enamine with an aldehyde substrate. Fig. 3 Rationale for stereoselectivity of NAL-catalysed reactions between ﬂuoropyruvate and aldehyde substrates. Panel A: S. aureus NAL in complex with ﬂuoropyruvate with general acid Y137 and residues that have key roles in recognition and stereocontrol (T167, E192 and S208) shown (PDB: 5A8G). The top face (as depicted) of the Z-conﬁgured enamine is poised to react with an aldehyde substrate. Panel B: Y137A variant of E. coli NAL in compex with 4-epi-Neu5Ac (PDB: 4BWL). Panel C: Possible stereochemical outcomes of the reaction of the top face of the Z-conﬁgured enamine with an aldehyde substrate. Structural insights into stereoselectivity. It is notable that all combinations of substrates and NAL variants yielded 110 \| Org. Biomol. Chem., 2016, 14, 105–112 This journal is © The Royal Society of Chemistry 2016 View Article Online View Article Online Organic & Biomolecular Chemistry Paper 3R-configured products selectively under kinetic control. With DHOB/AHOB, the products 16a/17a [with (3R,4R) configur- ation] and/or 16c/17c [with (3R,4S) configuration] pre- dominated. With ent-DHOB, although a diﬀerent diastereomer (ent-16d) was formed selectively, its absolute configuration was still (3R,4S). In contrast, the stereoselectivity at C-4 could sometimes be altered by changing the enzyme variant used. 3R-configured products selectively under kinetic control. With DHOB/AHOB, the products 16a/17a [with (3R,4R) configur- ation] and/or 16c/17c [with (3R,4S) configuration] pre- dominated. With ent-DHOB, although a diﬀerent diastereomer (ent-16d) was formed selectively, its absolute configuration was still (3R,4S). In contrast, the stereoselectivity at C-4 could sometimes be altered by changing the enzyme variant used. reaction of this (Z)-configured enamine via the face that is presented to aldehyde substrates. Notes and references 1 (a) F. M. D. Ismail, J. Fluorine Chem., 2002, 118, 27; (b) K. Muller, C. Faeh and F. Diederich, Science, 2007, 317, 1881; (c) W. K. Hagmann, J. Med. Chem., 2008, 51, 4359; (d) E. P. Gillis, K. J. Eastman, M. D. Hill, D. J. Donnelly and N. A. Meanwell, J. Med. Chem., 2015, DOI: 10.1021/acs. jmedchem.5b00258. 1 (a) F. M. D. Ismail, J. Fluorine Chem., 2002, 118, 27; (b) K. Muller, C. Faeh and F. Diederich, Science, 2007, 317, 1881; (c) W. K. Hagmann, J. Med. Chem., 2008, 51, 4359; (d) E. P. Gillis, K. J. Eastman, M. D. Hill, D. J. Donnelly and N. A. Meanwell, J. Med. Chem., 2015, DOI: 10.1021/acs. jmedchem.5b00258. 2 D. O’Hagan, J. Fluorine Chem., 2010, 131, 1071. 2 D. O’Hagan, J. Fluorine Chem., 2010, 131, 1071. 3 B. D. Roth, Prog. Med. Chem., 2002, 40, 1. 3 B. D. Roth, Prog. Med. Chem., 2002, 40, 1. 4 E. Cholongitas and G. V. Papatheodoridis, Ann. Gastro- enterol., 2014, 27, 331. 5 For an example, see: J.-H. Kim, R. Resende, T. Wennekes, H.-M. Chen, N. Bance, S. Buchini, A. G. Watts, P. Pilling, V. A. Streltsov, M. Petric, R. Liggins, S. Barrett, J. L. McKimm-Breschkin, M. Nikura and S. G. Withers, Science, 2013, 340, 71. 6 (a) B. Greedy, J.-M. Paris, T. Vidal and V. Gouverneur, Angew. Chem., Int. Ed., 2003, 115, 3413; (b) G. T. Giuﬀredi, S. Purser, M. Sawicki, A. L. Thompson and V. Gouverneur, Tetrahedron: Asymmetry., 2009, 20, 910. Conclusions NAL variants can be useful catalysts of reactions between fluoro- pyruvate and aldehyde substrates. Wild-type NAL catalysed the reaction between fluoropyruvate and ManNAc, albeit much less eﬃciently than with pyruvate as donor. It was shown that a 90 : 10 ratio of (3R,4R)- and (3S,4R)-configured products was obtained under kinetic control; whilst a 30 : 70 mixture of these products was obtained at equilibrium. The switch between kinetic and thermodynamic control may account for previous apparently conflicting reports of the outcome of this reaction.12 7 (a) T. D. Beeson and D. W. C. MacMillan, J. Am. Chem. Soc., 2005, 127, 8826; (b) M. Marigo, D. Fielenbach, A. Braunton, A. Kjærsgaard and K. A. Jørgensen, Angew. Chem., Int. Ed., 2005, 44, 3703. 8 S. M. Kim, H. R. Kim and D. Y. Kim, Org. Lett., 2005, 7, 2309. 9 (a) G. Zhong, J. Fan and C. F. Barbas III, Tetrahedron Lett., 2004, 45, 5681; (b) X.-Y. Xu, Y.-Z. Wang and L.-Z. Gong, Org. Lett., 2007, 9, 4247; (c) X.-Y. Xu, Y.-Z. Wang, L.-F. Cun and L.-Z. Gong, Tetrahedon: Asymmetry., 2007, 18, 237. It was also shown that NAL variants are useful catalysts of reactions between fluoropyruvate and unnatural aldehyde sub- strates. The eﬃciency of catalysis varied widely, depending on the specific combination of NAL variant and aldehyde used. However, using the aldehyde DHOB or its enantiomer as sub- strate, three of the four possible diastereomeric products could be isolated. 10 A. D. Daniels, I. Campeotto, M. W. Van der Kamp, A. H. Bolt, C. H. Trinh, S. E. V. Phillips, A. R. Pearson, A. Nelson, A. J. Mulholland and A. Berry, ACS Chem. Biol., 2014, 9, 1025. 11 R. Gantt, S. Millner and S. B. Binkley, Biochemistry, 1964, 3, 1952. 12 12 (a) H. A. Chokhawala, H. Cao, H. Yu and X. Chen, J. Am. Chem. Soc., 2007, 129, 10630; (b) J. Beliczey, U. Kragl, A. Liese, C. Wandrey, K. Hamacher, H. H. Coenen and T. Tierling, US Patent, 635543, 2002; (c) A. G. Watts and S. G. Withers, Can. J. Chem., 2004, 82, 1581. It was noted that, under kinetic control, all productive NAL variant-catalysed reactions involving fluoropyruvate yielded (3R)-configured products selectively. The crystal structure of S. aureus NAL in complex with fluoropyruvate reveals the pres- ence of a (Z)-configured enamine. The (3R)-selectivity of NAL catalysed reactions may be rationalised in terms of selective 13 (a) T. Woodhall, G. J. This journal is © The Royal Society of Chemistry 2016 Acknowledgements We thank BBSRC, EPSRC and AstraZeneca for funding student- ships and EPSRC for funding core chemistry equipment (EP/K039202/1). To gain an insight into the structural basis of stereo- selectivity, the crystal structure of S. aureus NAL was deter- mined in complex with fluoropyruvate (PDB: 5A8G) (Panel A, Fig. 3); the structure and kinetic properties of S. aureus NAL have been previously shown to be extremely similar to those of E. coli NAL.22 The formation of a Z-configured enamine was observed, which presents only one face to aldehyde substrates. Reaction of this face of the (Z)-enamine intermediate would necessarily lead to the formation of 3R-configured products. The structure10 of an aldol product (4-epi-Neu5Ac) in complex with NAL (the Y137A variant of the E. coli enzyme) (PDB: 4BWL) is provided for comparison (Panel B, Fig. 3). Previous studies have shown that an analogue of DHOB – (2R,3S)-2,3-tri- hydroxy-4-oxo-N,N-dipropyl butanamide – can bind to the E192N variant of E. coli NAL in two distinct conformations (Panel B, Fig. 3).15 Aldehyde substrates may react via confor- mations that allow protonation by the general acid Y137.10 The facial selectivity of the reaction of the aldehyde determines the configuration of the product – (3R,4R) or (3R,4S) – obtained (Panel C, Fig. 3). In some cases, the ratio of C-4 epimers could be changed by altering the specific NAL variant used. This journal is © The Royal Society of Chemistry 2016 112 \| Org. Biomol. Chem., 2016, 14, 105–112 Conclusions Williams, A. Berry and A. Nelson, Angew. Chem., Int. Ed., 2005, 44, 2109; (b) G. J. Williams, Org. Biomol. Chem., 2016, 14, 105–112 \| 111 View Article Online Paper Organic & Biomolecular Chemistry T. Woodhall, A. Nelson and A. Berry, Protein Eng. Des. Sel., 2005, 18, 239. T. Woodhall, A. Nelson and A. Berry, Protein Eng. Des. Sel., 2005, 18, 239. 18 T. Woodhall, G. J. Williams, A. Berry and A. Nelson, Org. Biomol. Chem., 2005, 3, 1795. 19 S.-Y. Han, M. M. Joullie, V. V. Fokin and N. A. Petasis, Tetra- hedron: Asymmetry., 1994, 5, 2535. 14 G. J. Williams, T. Woodhall, L. M. Farnsworth, A. Nelson and A. Berry, J. Am. Chem. Soc., 2006, 128, 16238. 20 M. Godskesen, I. Lundt and I. Sotofte, Tetrahedron: Asym- metry., 2000, 11, 567. 15 I. Campeotto, A. H. Bolt, T. A. Harman, C. Dennis, C. H. Trinh, S. E. V. Phillips, A. Nelson, A. R. Pearson and A. Berry, J. Mol. Biol., 2010, 404, 56. 21 S. V. Ley, E. Diez, D. Dixon, R. T. Guy, P. Michel, G. L. Nattrass and T. D. Sheppard, Org. Biomol. Chem., 2004, 2, 3608. 16 D. O’Hagan, J. Org. Chem., 2012, 77, 3689. 22 N. Timms, C. L. Windle, A. Polyakova, J. R. Ault, C. H. Trinh, A. R. Pearson and A. Berry, ChemBioChem, 2013, 14, 474. 17 (a) M. Karplus, J. Am. Chem. Soc., 1963, 85, 2870; (b) W. R. Dolbier, Guide to Fluorine NMR for Organic Chemists, Wiley, New Jersey, 2009. 112 \| Org. Biomol. Chem., 2016, 14, 105–112
W4367624663.txt	https://www.mdpi.com/2076-2607/11/5/1186/pdf?version=1683271058	en		Overview of Diagnostic Methods, Disease Prevalence and Transmission of Mpox (Formerly Monkeypox) in Humans and Animal Reservoirs	Microorganisms	2,023	cc-by	20,184	microorganisms Review Overview of Diagnostic Methods, Disease Prevalence and Transmission of Mpox (Formerly Monkeypox) in Humans and Animal Reservoirs Ravendra P. Chauhan * , Ronen Fogel and Janice Limson * Biotechnology Innovation Centre, Rhodes University, Makhanda 6139, Eastern Cape, South Africa; r.fogel@ru.ac.za * Correspondence: ravendrachauhan@hotmail.com (R.P.C.); J.Limson@ru.ac.za(J.L.) Citation: Chauhan, R.P.; Fogel, R.; Limson, J. Overview of Diagnostic Methods, Disease Prevalence and Transmission of Mpox (Formerly Monkeypox) in Humans and Animal Reservoirs. Microorganisms 2023, 11, 1186. https://doi.org/10.3390/ microorganisms11051186 Academic Editor: Roger Frutos Received: 3 April 2023 Revised: 26 April 2023 Abstract: Mpox—formerly monkeypox—is a re-emerging zoonotic virus disease, with large numbers of human cases reported during multi-country outbreaks in 2022. The close similarities in clinical symptoms that Mpox shares with many orthopoxvirus (OPXV) diseases make its diagnosis challenging, requiring laboratory testing for confirmation. This review focuses on the diagnostic methods used for Mpox detection in naturally infected humans and animal reservoirs, disease prevalence and transmission, clinical symptoms and signs, and currently known host ranges. Using specific search terms, up to 2 September 2022, we identified 104 relevant original research articles and case reports from NCBI-PubMed and Google Scholar databases for inclusion in the study. Our analyses observed that molecular identification techniques are overwhelmingly being used in current diagnoses, especially real-time PCR (3982/7059 cases; n = 41 studies) and conventional PCR (430/1830 cases; n = 30 studies) approaches being most-frequently-used to diagnose Mpox cases in humans. Additionally, detection of Mpox genomes, using qPCR and/or conventional PCR coupled to genome sequencing methods, offered both reliable detection and epidemiological analyses of evolving Mpox strains; identified the emergence and transmission of a novel clade ‘hMPXV-1A’ lineage B.1 during 2022 outbreaks globally. While a few current serologic assays, such as ELISA, reported on the detection of OPXV- and Mpox-specific IgG (891/2801 cases; n = 17 studies) and IgM antibodies (241/2688 cases; n = 11 studies), hemagglutination inhibition (HI) detected Mpox antibodies in human samples (88/430 cases; n = 6 studies), most other serologic and immunographic assays used were OPXV-specific. Interestingly, virus isolation (228/1259 cases; n = 24 studies), electron microscopy (216/1226 cases; n = 18 studies), and immunohistochemistry (28/40; n = 7 studies) remain useful methods of Mpox detection in humans in select instances using clinical and tissue samples. In animals, OPXV- and Mpox-DNA and antibodies were detected in various species of nonhuman primates, rodents, shrews, opossums, a dog, and a pig. With evolving transmission dynamics of Mpox, information on reliable and rapid detection methods and clinical symptoms of disease is critical for disease management. Keywords: Mpox diagnostics; Mpox genome sequencing; Mpox host range; Mpox transmission; Mpox; orthopoxvirus; zoonosis Accepted: 28 April 2023 Published: 30 April 2023 1. Introduction Copyright: © 2023 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/). Mpox, previously known as monkeypox, is a re-emerging zoonotic virus pathogen belonging to the larger orthopoxvirus (OPXV) genus in the Poxviridae family [1,2]. Some of the most prominent symptoms of Mpox infections in humans include eruptive skin lesions [3–9], rash [3,5–7,10–14], fever [3–6,8,11–14], and lymphadenopathy [5,6,8,10–12,14]. These symptoms are largely similar to infections caused by other members of the OPXV genus, such as variola virus (VARV, the causative agent for smallpox disease) [15–19] and cowpox virus (CPXV, causing cowpox disease) [20–24], along with viral infections from other virus families, e.g., chickenpox caused by the varicella-zoster virus (VZV), a Microorganisms 2023, 11, 1186. https://doi.org/10.3390/microorganisms11051186 https://www.mdpi.com/journal/microorganisms Microorganisms 2023, 11, 1186 2 of 31 member of the Herpesvirus family [25–29]. The similarities in clinical manifestations of these diseases in humans make the diagnosis of Mpox disease by observing clinical disease symptoms alone, challenging. Therefore, the current World Health Organization (WHO) guidelines require the confirmation of Mpox diagnosis via reliable laboratory testing to identify infected individuals in order to manage its spread [30–32]. Members of the OPXV genus are known to infect a broad range of mammalian hosts [33–36]; however, the mechanisms behind their ability to transmit between host species are not well understood. Mpox is a highly-pathogenic poxvirus, carrying a human case fatality rate ranging between 1 and 11% [2,37–39]. Mpox infections may also result in numerous complications, such as miscarriage and intrauterine fetal death in pregnant women [40,41], vision loss due to corneal lesions [42,43], and encephalitis [44,45], among others. The 2022 multi-country outbreaks of Mpox reported an unprecedented high humanto-human transmission. The occurrence of these outbreaks outside the endemic regions in West and Central Africa [46,47] suggests that Mpox has further adapted to infect humans. Together with the associated case fatality rates and complications, this virus is a re-emerging disease of significant concern [48–50]. While the active infection of Mpox was suggested to be first identified in 1958 in cynomolgus monkeys (Macaca fascicularis) that were placed in captivity at the Statens Serum Institut in Copenhagen, Denmark, having been received from Singapore [51], a recent study using historical skin samples of African rope squirrel (Funisciurus sp.) stored in museums originally collected across Central Africa over the past 120 years provided retrospective molecular evidence of the Central African (formerly Congo Basin) clade of Mpox in specimens from the Democratic Republic of the Congo (DRC) and Central African Republic (CAR) from as early as 1899 [52]. Similarly, molecular evolutionary analysis also indicates the historic presence of Mpox: one study suggests that Mpox may have separated from Old World OPXV approximately 3500 years ago [53], while the separation of the West African Mpox subtype from other clades is estimated to have occurred between 600 years ago [53] and 1200 years ago [54]. These collectively suggest that Mpox disease may have been in circulation much earlier than its first identification. The clinical features of the Mpox disease in cynomolgus monkeys in 1958 exhibited skin lesions typical to the smallpox virus disease. The investigation using serologic assays determined the presence of OPXV-specific antibodies in sera samples. Electron microscopy (EM) investigation of isolates from these samples observed brick-shaped particles ranging between 100 and 150 nm in size, distinguishable from the other known members of the OPXV genus [51]. Due to its first detection in monkeys, the term ‘monkeypox virus’ was coined in 1958 and, in November 2022, was redesignated “Mpox” by the WHO [55,56]. The first documented case of Mpox infection in humans was in a nine-month-old male in the Democratic Republic of Congo (DRC), formerly Zaire; the patient, presenting fever and a rash, was admitted to a hospital on 1 September 1970 [57]. Similar to its first identification, the clinical manifestation of the disease appeared similar to smallpox—another OPXV-derived illness—which was previously in active circulation in the region [57,58]. Virus isolation from this patient successfully identified the pathogenic agent to be Mpox [58]. Subsequently, between September 1970 and May 1971, six more human cases initially suspected to be smallpox were reported across West Africa: in Liberia, Sierra Leone, and Nigeria. From these, four Mpox virus isolates were recovered and characterized using the reference strains of vaccinia virus (VACV), VARV, and Mpox [59]. From this study, the first distinguishable characteristics in the infection profile between Mpox and other members of OPXV were offered [59]. These distinctions were subtle: along with greater virulence in infected mice and chicken embryos, rabbits intradermally inoculated with isolated Mpox produced more severe illnesses compared to VARV and VACV [59]. Subsequently, various serologic investigations focused on the diagnosis of OPXV infection through the detection of OPXV-specific antibodies produced by infected humans and animals; some of these studies also combined virus isolation and EM study of clinical samples to arrive at a conclusive diagnosis of Mpox disease in humans [60–62]. Microorganisms 2023, 11, 1186 3 of 31 Historically, Mpox was endemic to some countries in West and Central Africa but has recently spread further [63]. The first human cases of Mpox disease outside the endemic regions were reported in 2003, presumably caused by a shipment of small rodents from Ghana to the Midwestern United States as part of the exotic pet trade. These were assumed to transmit Mpox to prairie dogs housed at a pet store in the United States through contact with imported rodents. The infected prairie dogs were purchased by several households in the Midwestern United States [64], resulting in zoonotic transmission of Mpox from the pet prairie dogs to the humans due to close contact within the households [65,66]. A total of 37 laboratory-confirmed cases of Mpox out of 71 suspected were reported in the United States between 15 May and 20 June 2003 [67]. More recently, the exportation of human cases of Mpox from Nigeria to other countries via air travelers was first reported in 2017 [68]. The frequency of imported Mpox cases outside Africa quickly increased in subsequent years [7,10,69–72], most likely due to the adaptation of Mpox to human hosts [73]. By April–May 2022, multi-country outbreaks of Mpox erupted, which witnessed an unprecedented increase in human infections, with many cases of human-to-human transmission reported due to sexual contact [48,74–79]. According to the WHO, between 1 January 2022 and 27 February 2023, a total of 86,173 laboratory-confirmed cases of Mpox, including 99 deaths, were reported from 110 countries [80]. Rapid, reliable, and scalable testing methods for Mpox detection are imperative for mitigating multi-country Mpox disease outbreaks by minimizing human-to-human transmission. Currently, Mpox disease diagnosis constitutes a combination of clinical examination and molecular detection with real-time PCR or conventional PCR followed by partial genome sequencing using clinical samples, such as surface lesion swabs or crusts [81]. This review discusses the technological advancements in Mpox detection in naturally infected human and animal populations, along with clinical manifestations and transmission dynamics of the disease. 2. Methodology 2.1. Article Search Criteria The guidelines of Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) [82] were followed to identify the relevant records for inclusion in the study (Figure 1). The original research articles and case reports documenting natural infections of Mpox in humans and animals were identified using Google Scholar and NCBI-PubMed databases. Search terms (Table 1) were entered one by one in the databases, and the titles of the suggested articles were screened to identify the relevant reports published up to 2 September 2022. To enhance record retrieval, search terms were not enclosed in quotation marks. The relevant articles thus identified were downloaded on a computer drive for further screening for inclusion in the study. 2.2. Article Inclusion and Exclusion Criteria The original research articles and case reports that documented the natural occurrence of Mpox disease in humans and animals up to 2 September 2022 were included in the study. Review articles, experimental studies, articles in a language other than English, abstracts, and conference proceedings were not included in the study. Microorganisms 2023, 11, 1186 4 of 31 Microorganisms 2023, 9, x FOR PEER REVIEW 4 of 31 Figure 1. PRISMA chart depicting the search criteria for identifying original research articles and Figure PRISMA chartnatural depicting the search criteria foranimals identifying original articles case 1. reports reporting infections of humans and/or with Mpox for research inclusion in the and casestudy. reports reporting naturalfull-text infections of humans and/or with Mpox for inclusion A total of 104 eligible research articles and case animals reports were included in the study. in the study. A total of 104 eligible full-text research articles and case reports were included in the study. Table 1. Search terms used for identifying the relevant research articles and case reports of Mpox natural infections in humans and animals upthe to 2relevant September 2022. articles and case reports of Mpox Table 1. Search terms used for identifying research natural infections in humans and animals up to 2 September 2022. in # Titles Screened in # Titles Screened Search Terms Used NCBI-PubMed Google Scholar Search Terms Used # Titles Screened in NCBI-PubMed # Titles Screened in Google Scholar Monkeypox virus outbreak 328 430 Monkeypox virus outbreak Detection of monkeypox virus 328 disease 430 390 83 240 Detection of monkeypox virus disease Monkeypox virus in Africa 83 390 330 Epidemiology of monkeypox virus disease 223 Monkeypox virus in Africa 240 330 270 Molecular detection of monkeypox virus 33 310 Epidemiology of monkeypox virus disease 223 270 Serological detection of monkeypox virus 10 270 Molecular detection of monkeypox virus 33 virus 310 210 Microarray for monkeypox 15 Real-time PCR for monkeypox 43 Serological detection of monkeypox virus 10 virus 270 360 Antigenic detection of monkeypox virus 29 Microarray for monkeypox virus 15 210 250 Genomic sequencing of monkeypox virus 83 220 Real-time PCR for monkeypox virus 43 360 Rapid detection of monkeypox virus 28 260 Antigenic detection of monkeypoxPoint virusof care detection of monkeypox 29 250 150 virus 3 ELISA for monkeypox virus83detection 16 Genomic sequencing of monkeypox virus 220 190 Monkeypox virus outbreak 2022 263 Rapid detection of monkeypox virus 28 260 490 Point of care detection of monkeypox virus 3. Results 3 150 ELISA for monkeypox virus detection 16 retrieval (Table 1) were not enclosed within 190 quotation Search terms used for record marks; this was intended to increase the variations of the terms used by the search engines Monkeypox virus outbreak 2022 263 490 when retrieving records. In total, 1397 records were screened in NCBI-PubMed and 5527 records in Google Scholar. These two databases routinely index the cases and reports of 3. Results Mpox disease globally, including the weekly Mpox disease updates from the WHO. Titles Search terms used for record retrieval (Table 1) were not enclosed within quotation marks; this was intended to increase the variations of the terms used by the search engines when retrieving records. In total, 1397 records were screened in NCBI-PubMed and 5527 records in Google Scholar. These two databases routinely index the cases and reports of Mpox disease globally, including the weekly Mpox disease updates from the WHO. Titles and abstracts were Microorganisms 2023, 11, 1186 Microorganisms 2023, 9, x FOR PEER REVIEW 5 of 31 5 of 31 manually read to evaluate the relevance for the further full-text reading, focusing strictly on and abstracts were manually read to evaluate the relevance for the further full-text readcase studies and original research articles that applied diagnostic technologies to the detection ing, focusing strictly on case studies and original research articles that applied diagnostic of Mpox were included: of the 6924 studies initially screened, only 104 relevant records were technologies to the detection of Mpox were included: of the 6924 studies initially screened, found eligible and thus included in this study (Figure 1). only 104 relevant records were found eligible and thus included in this study (Figure 1). 3.1. Trends of Mpox Disease Investigations 3.1. Trends of Mpox Disease Investigations A total of 104 full-text research articles and case reports published up to 2 September 2022 A total of 104 full-text research articles and case reports published up to 2 September documenting natural infections of Mpox in humans and animals were retrieved (Figure 1). 2022 documenting natural infections of Mpox in humans and animals were retrieved (FigThe chronological profile of these included studies is presented in Figure 2, where it is ure 1). The chronological profile of these included studies is presented in Figure 2, where compared to both Mpox case numbers which are included in this study, and to SCOPUSit is compared to both Mpox case numbers which are included in this study, and to SCOretrieved records for the term (“monkeypox” AND “human”, followed by either “detecPUS-retrieved records for the term (“monkeypox” AND “human”, followed by either “detion”, “characterisation” or “diagnosis”), as a general measure of research conducted in tection”, “characterisation” or “diagnosis”), as a general measure of research conducted this field. Figure 2 shows that, in general, the included studies used further in this pain this field. Figure 2 shows that, in general, the included studies used further in this paper per corresponded tooverall the overall literature being published surrounding topic corresponded wellwell to the literature being published surrounding this topicthis (SCO(SCOPUS records); in general, a larger number of included articles were retrieved when PUS records); in general, a larger number of included articles were retrieved when reresearch aroundthis thistopic topicincreased. increased. This indicates included studies are represensearch around This indicates thatthat the the included studies are representatative of the evolving manner by which Mpox was, and currently is, being monitored. tive of the evolving manner by which Mpox was, and currently is, being monitored. A A large wereeither eitherresearch researchreviews reviews existing literature largesubset subsetof oftotal totalrecords records retrieved retrieved were onon existing literature ororthe development of novel detection methods and experimental studies that were never the development of novel detection methods and experimental studies that were never applied to real-world samples of Mpox nor used in the diagnosis of infection. Therefore, applied to real-world samples of Mpox nor used in the diagnosis of infection. Therefore, the theoriginal originalresearch researcharticles articles and case reports thepresent presentstudy studyincluded included exclusively exclusively the and case reports thatthat detected ofMpox Mpoxininhuman human and/or animal populations. Theofaim detectednatural natural infections infections of and/or animal populations. The aim the of the study to compare various diagnostic technologies Mpox detection study waswas to compare various diagnostic technologies used forused Mpoxfor detection in clinical in clinical samples, analysis of evolving symptoms and signs ofin disease hosts, samples, analysis of evolving clinical clinical symptoms and signs of disease hosts, in and trans-and transmission dynamics. mission dynamics. Figure 2. Progression of Mpox disease surveillance in humans. The highest number of studies deFigure Progression Mpox diseaseduring surveillance in humans. Thein highest number of ofcases studies tecting2.Mpox infectionsofwere conducted multi-country outbreaks 2022. Estimates detecting Mpox infections were conducted multi-country in 2022. Estimates of were drawn from selected published reports,during including the followingoutbreaks citations [48,58,83–88] . cases were drawn from selected published reports, including the following citations [48,58,83–88]. Microorganisms 2023, 9, x FOR PEER REVIEW Microorganisms 2023, 11, 1186 6 of 31 6 of 31 Comparing research research interest interest (via (via SCOPUS-retrieved SCOPUS-retrieved records) records) with with case case incidences incidences Comparing (Figure 2) shows that a sharp rise in Mpox investigations occurred during the last two two (Figure 2) shows that a sharp rise in Mpox investigations occurred during the last decades after Mpox outbreaks in the Midwestern United States in 2003 [44,64–66,89–92], decades after Mpox outbreaks in the Midwestern United States in 2003 [44,64–66,89–92], i.e., the the reporting reporting of of human human cases cases of of Mpox Mpox in in aa non-endemic non-endemic region. region. This This is is again again evident evident i.e., in the the spike spike in in published published literature literature shortly shortly after after 2017, 2017, when when several several reports reports of of Mpox Mpox in in air air in travelers appeared appeared from from many many non-endemic non-endemic countries countries [4,68,69,71,93] and more more recently recently travelers [4,68,69,71,93] and during the themulti-country multi-country outbreaks of Mpox, starting in April–May 2022 [14,94–100]. during outbreaks of Mpox, starting in April–May 2022 [14,94–100]. These These are linked to the reportedly unprecedentedly high rates of human-to-human transare linked to the reportedly unprecedentedly high rates of human-to-human transmission, mission, in turn causing the highest number of Mpox human cases. in turn causing the highest number of Mpox human cases. Based on the Mpox genome sequences and clinical features features of the disease, disease, two two disdistinguishable clades of Mpox were identified circulating before the 2022 Mpox outbreaks. outbreaks. These were werethe theWest WestAfrican Africanand andCentral CentralAfrican African (formerly, Congo Basin) clades, based These (formerly, Congo Basin) clades, based on on their geographic origins [101,102]. phylogenomics analysis of available getheir geographic origins [101,102]. The The phylogenomics analysis of available MpoxMpox genomes nomes including the genomes of the 2022 Mpox outbreaks thatMpox the 2022 Mpox including the genomes of the 2022 Mpox outbreaks suggestedsuggested that the 2022 outbreaks outbreaks were a newly evolved clade termed ‘hMPXV-1A’ lineage B.1 [103]. were caused by acaused newly by evolved clade termed ‘hMPXV-1A’ lineage B.1 [103]. Historically, more frequent circulation the West African clade than the Central African clade was Historically, more frequentofcirculation of the West African clade than the Central African observed the before emergence of the 2022of Mpox outbreaks 3). (Figure 3). clade wasbefore observed the emergence the 2022 Mpox (Figure outbreaks Figure 3. Historically, a higher circulation frequency of the West African clade of Mpox occurred Figure 3. Historically, a higher circulation of the West clade African clade of Mpox The occurred [4,7,10,11,42,65,68,70,71,83,90,104–110] thanfrequency the Central African [6,9,13,111–115]. 2022 [4,7,10,11,42,65,68,70,71,83,90,104–110] than the Central African clade [6,9,13,111–115]. The Mpox outbreaks were caused by a newly emerged clade termed ‘hMPXV-1A’ lineage B.1 [103]2022 and Mpox outbreaks were caused by aofnewly clade termed lineage B.1 [103] and suggested the ongoing evolution Mpoxemerged (CAR = Central African‘hMPXV-1A’ Republic; DRC = Democratic Resuggested theCongo; ongoing Mpox (CAR African = Democratic public of the UKevolution = United of Kingdom; USA== Central United States ofRepublic; America).DRC Mpox 2022 multicountry outbreaks were caused by a newly emerged clade ‘hMPXV-1A’ B.1. Mpox 2022 Republic of the Congo; UK = United Kingdom; USA = United States lineage of America). multi-country outbreaks were caused by a newly emerged clade ‘hMPXV-1A’ lineage B.1. Microorganisms 2023, 9, x FOR PEER REVIEW Microorganisms 2023, 11, 1186 7 of 31 7 of 31 3.2. Methods Used for Mpox Detection in Clinical and Tissue Samples of Humans and Animals 3.2. Methods Used for Mpox Detection in Clinical and Tissue Samples of Humans and Animals over over the the Period Period and and Their Their Significance Significance Technological Technological advancement advancement over over the the decades decades since since Mpox Mpox was was first first characterized characterized has has strongly detection of of Mpox. Mpox instrongly influenced influenced the therange rangeofofavailable availablemethods methodsforforthe the detection Mpox. Mpox vestigations have used a avariety investigations have used varietyofofmethods methodsfor forthe theconfirmation confirmationof ofdisease disease (Figure (Figure 4); 4); most studies focused on a single—or a limited selection of separate methods—to process most studies focused on a single—or a limited selection of separate methods—to process samples, few studies studies used used aa combination combination of of various various techniques. techniques. These samples, but but aa few These have have been been separately categorized into approaches that diagnose Mpox in samples via the detection separately categorized into approaches that diagnose Mpox in samples via the detection of virus genomes genomes (overwhelmingly (overwhelmingly using using PCR-based PCR-based approaches), approaches), serological serological methods methods of virus aimed at detecting detectingvirus-specific virus-specificantibodies, antibodies,such suchasasEnzyme-Linked Enzyme-Linked Immunosorbent AsImmunosorbent Assay say (ELISA) hemagglutination inhibition assays, historic approaches (ELISA) andand hemagglutination inhibition (HI)(HI) assays, and and moremore historic approaches that that are being still being applied today, as immunohistochemistry (IHC), isolation, are still applied today, such such as immunohistochemistry (IHC), virusvirus isolation, and and electron microscopic identification of virus particles. electron microscopic identification of virus particles. Figure 4. A schematic representation of various methods used for Mpox detection in humans and Figure 4.over A schematic animals the periodrepresentation (1958–2022). of various methods used for Mpox detection in humans and animals over the period (1958–2022). The current WHO guidelines require molecular detection of Mpox using Mpox-speThe current WHO guidelines require molecular detection of Mpox using Mpox-specific cific real-time PCR or conventional PCR in combination with genome sequencing for epireal-time PCR or conventional PCR in combination with genome sequencing for epidemidemiological analysis [32]. The detection of viral genomes in samples using oligonucleoological analysis [32]. The detection of viral genomes in samples using oligonucleotide tide primers and a probe for real-time PCR oﬀering a rapid, sensitive, and specific detecprimers and a probe for real-time PCR offering a rapid, sensitive, and specific detection tion of the target has become the most popular method for Mpox disease confirmation of the target has become the most popular method for Mpox disease confirmation ususing clinical samples over the past 20 years (n = 48 studies; Figure 4), while conventional ing clinical samples over the past 20 years (n = 48 studies; Figure 4), while conventional PCR-based approaches enjoyed similar popularity (n = 32 studies). Several studies comPCR-based approaches enjoyed similar popularity (n = 32 studies). Several studies combined molecular detection using OPXV or Mpox-specific real-time PCR/conventional bined molecular detection using OPXV or Mpox-specific real-time PCR/conventional PCR-genome PCR-genome sequencing sequencing to to either either distinguish distinguish Mpox Mpox from from other other OPXVs OPXVs or or to to identify identify spespecific clades of Mpox. PCR amplification and Sanger sequencing of the amplicon cific clades of Mpox. PCR amplification and Sanger sequencing of the amplicon [9,89,109] [9,89,109] were used for for characterization characterizationand andtotodistinguish distinguishMpox Mpoxfrom fromother other OPXV using stretches were used OPXV using stretches of of its hemagglutinin (HA) gene [65,90,107,115,116]. Next-generation sequencing technolits hemagglutinin (HA) gene [65,90,107,115,116]. Next-generation sequencing technologies, ogies, such Ion Torrent PGM [106,116], Illumina [4,9,68,104,107,112], and Oxford Nasuch as Ion as Torrent PGM [106,116], Illumina [4,9,68,104,107,112], and Oxford Nanopore nopore MinION [4,68,117–119], are cutting-edge approaches being employed for Mpox MinION [4,68,117–119], are cutting-edge approaches being employed for Mpox diagnosis Microorganisms 2023, 11, 1186 8 of 31 (n = 16 studies). Next-generation sequencing, in particular, offers the opportunity for whole genome sequencing of Mpox in clinical samples, in turn, providing insight into the phylogenomics of circulating Mpox strains. Next-generation sequencing-based studies conducted during 2022 Mpox outbreaks identified the emergence of a new Mpox clade, termed ‘hMPXV-1A’ lineage B.1 [103]. The advances in the cost-effectiveness of next-generation sequencing are expected to further improve its routine application, in turn driving the molecular characterization of Mpox in patients and providing enhanced epidemiological and evolutionary data for analyses. In addition to molecular approaches, many studies have used serological approaches to identify Mpox. These largely monitor the prevalence of the virus in patients via the production of virus-specific antibodies during and post-infection, using secondary antibody-based systems to detect virus-specific antibodies. Diagnosis using either one or a combination of ELISAbased monitoring of the production of Mpox-specific IgG (n = 23 studies) [3,11,44,92,120–124] or IgM antibodies (n = 11 studies) [11,44,83,92,120,121,123,124], neutralization assay (n = 5 studies) [113,125,126], HI (n = 8 studies) [60–62,113,125], and/or complement fixation assays (n = 2 studies) [60] in blood samples have been reported. In general, molecular detection methods have several advantages over serological methods. Primarily, molecular detections are routinely capable of differentiating between Mpox and other species of OPXV in a multiplex assay [127,128], unlike most serological assays used, which due to immunological cross-reactivity between human OPXV, offer limited diagnostic value [129–131]. Although a few studies used Mpox-specific antibodies, they may lack reproducibility and are complex to execute [132–134]. Additionally, most serologic assays tend to offer only limited usage for the detection of acute infection due to the persistence of circulating antibodies post-infection [135], which may result in misleading diagnostic outcomes. The use of multiple sample types from a patient is suggested for IgG-based serologic investigations when determining recent exposure, comparing the levels of antibodies in those samples to achieve diagnosis [124]. IgM-detecting ELISA offers detection of recent OPXV, as these types of antibodies are markers of acute immune responses [124]. While IgG ELISA is useful in the detection of vaccine efficiency testing or correlating epidemiologically defined outbreaks, it tends to be less efficient in the context of species differentiation of OPXV [136]. An advantage that serological tests tend to offer over molecular detection is the longer timeframe in which exposure to Mpox can be detected. While PCR-based detection methods require the availability of sufficient virus genome copies (mainly confined to rash, fluid, or crust specimens), serologic detection, using blood, offers flexibility in terms of sample collection and can detect prior exposure after the clinical symptoms of disease subside, allowing for retrospective diagnosis. The final category (virus isolation and EM) has been consistently applied throughout the known existence of Mpox (Figure 4). Virus isolation, coupled with EM-based morphological identification of virus particles, was instrumental in the initial characterization of the virus and its first report of human infection [58]. While these are not the recommended test methods for routine diagnosis of Mpox in clinical laboratories [32], they have still been employed in recent studies (Figure 4). Virus isolation remains useful in virological investigations that report on the environmental persistence of the virus [137], identification and characterization of poxviruses [58,138], or investigation of the pathophysiology and virulence of viruses, including Mpox, under experimental settings [139]. Similarly, EM offers a great advantage for detecting emerging or novel viruses in clinical samples, as it does not rely on virus-specific reagents and can provide unbiased detection of viruses in general within a sample. For this reason, EM remains a method of choice for determining the size and shapes of poxvirus particles and distinguishing among poxviruses morphologically. Several studies during 2022 Mpox outbreaks used EM [14,96,140] and virus isolation [98,99,138,140] to confirm diagnoses, especially for patients where transmission may have occurred due to sexual contact; these patients presented with clinical disease symptoms similar to those of sexually-transmitted diseases. It can be observed from the studies that EM [8,105,110,140–142] and virus Microorganisms 2023, 9, x FOR PEER REVIEW Microorganisms 2023, 11, 1186 9 of 9 of 31 can be observed from the studies that EM [8,105,110,140–142] and virus isolat [8,12,13,58,95,99,105,113,120,140–142] haveconsistently been consistently over the decad isolation [8,12,13,58,95,99,105,113,120,140–142] have been used overused the decades within and outside the endemic regions for the confirmation of Mpox diagnosis. within and outside the endemic regions for the confirmation of Mpox diagnosis. methods making use of antibodies to detect portions of virus particles (Imm Other methodsOther making use of antibodies to detect portions of virus particles nographic methods, Figure 4) such as immunofluorescence (n = 8 studies) [125,126,13 (Immunographic methods, Figure 4) such as immunofluorescence (n = 8 studies) [125,126,130], immunohistochemistry (n = 7 studies) [40,44,64,92,97], radioimmunoassay (n = 5 studi immunohistochemistry (n = 7 studies) [40,44,64,92,97], radioimmunoassay (n = 5 stud[60,62], agar-gel test (n = 3 studies) [59,61], precipitation-in-gel test (n = 3 studies) [60], a ies) [60,62], agar-gel test (n = 3 studies) [59,61], precipitation-in-gel test (n = 3 studies) [60], Western blot (n = 3 studies) [113,143] were used historically, but have become less pre and Western blot (n = 3 studies) [113,143] were used historically, but have become less lent in recent years, except for immunohistochemistry (IHC) which oﬀers tissue-based prevalent in recent years, except for immunohistochemistry (IHC) which offers tissue-based agnostics with specificity, sensitivity, and great reproducibility [144]. The IHC remain diagnostics with specificity, sensitivity, and great reproducibility [144]. The IHC remains preferred method for investigating systemic Mpox infections in various human tiss a preferred method for investigating systemic Mpox infections in various human tissue samples in real-time, as well as retrospectively. Many studies in the recent past, includ samples in real-time, well as Mpox retrospectively. studiesdetected in the recent including duringasthe 2022 outbreaks,Many successfully Mpoxpast, antigen in clinical samp during the 2022using MpoxIHC outbreaks, successfully detected Mpox antigen in clinical samples [11,44,64,90,97]. using IHC [11,44,64,90,97]. Briefly, our analyses indicated that molecular detection of Mpox using PCR or re Briefly, our time analyses that molecular detection of Mpox using sequencing PCR or real-time PCRindicated in combination with Sanger and next-generation oﬀers rapid, s PCR in combination next-generation offers rapid, In sensitive, sitive,with and Sanger reliable and detection of Mpox insequencing human clinical samples. addition, standa and reliable detection of Mpox in human samples. addition, virological detecti virological methods, suchclinical as virus isolationInand EM, asstandard well as tissue-based methods, such assuch virus and commonly EM, as wellused as tissue-based detection, such as detection IHC, are and con as isolation IHC, are still in research laboratories for the still commonly used in of research laboratories for the detection and confirmation Mpox in remain mation Mpox in human clinical samples or tissues. The IgG andof IgM ELISA human clinical samples or tissues. The IgG and IgM ELISA remain the most used serologic most used serologic methods for Mpox or OPXV antibody detection in human sera. methods for Mpox or OPXV antibody detection in human sera. 3.3. Comparison of Detection Methods by Prevalence of Use in Human Mpox Diagnosis 3.3. Comparison of Detection Methods by Prevalence of Use in Human Mpox Diagnosis The majority of the selected studies (n = 104) investigated natural infections of Mp The majority the selected (nThe = 104) investigated natural infections Mpox were in of humans (n = 83studies studies). numbers of human-derived samplesofprocessed in humans (n = tracted 83 studies). numbersstudies of human-derived samples processed by were from The all included in this review and categorized theexdetection tracted from all included this review categorized the detection proaches studies outlinedinin Section 3.2and above. Figure 5by summarises theapproaches aggregated number outlined in Section 3.2 above. Figure 5 summarises aggregated number samples prosamples processed using the diﬀerentthe methods, as well as the of number of positive diag cessed using theses different methods, as well as the number positive diagnoses made with samp made with each method. While real-timeofPCR detected most Mpox-positive conventional PCR, virus isolation, EM, IgG and IgM ELISA, HI,conventional and neutralization assa each method. While real-time PCR detected most Mpox-positive samples, were other, more commonly used, methods than other serologic andother, immunograp PCR, virus isolation, EM, IgG and IgM ELISA, HI, and neutralization assays were more commonlymethods. used, methods than other serologic and immunographic methods. Figure 5. Representation of all the cases investigated for Mpox or OPXV detection using various methods in human clinical samples from eligible articles available up to 2 September 2022; n = number of studies. Microorganisms 2023, 11, 1186 10 of 31 Overwhelmingly, most samples processed for Mpox detection between 2010 and 2022 used either molecular or genomic detection of viral nucleic acid or antibodies using serological methods. Together, these two approaches comprise 15,140/18,060 (83.8%) patient samples processed in the selected articles and corresponded to 5720/6337 (90.2%) of all positive diagnoses made. Nearly half of the Mpox-suspected clinical samples were processed using either PCR- or real-time PCR-based methods, correlating to the larger number of recent reports that used these techniques (Figure 4) due to these methods’ rapidity, sensitivity, scalability, and efficiency for selectivity to Mpox (discussed in more detail in Section 3.7, below). The overall positivity rate of Mpox in tested human populations was much higher using real-time PCR-based detection (3982/7059; 56.41%) than conventional PCR-based detection (430/1830; 23.49%). For similar reasons of the rapidity of response and scalability, serological methods comprise the largest remaining group of samples processed. Many studies used only serologic detection of Mpox; most of these studies were conducted before the advent of PCR-based detection. Mpox antibodies were detected in human sera using IgG ELISA (652/1285; 50.74%), IgM ELISA (52/1077; 4.83%), neutralization assay (44/206; 21.36%), HI assay (15/195; 7.69%), and immunofluorescence assay (IFA) (15/164; 9.15%). Detection of anti-Mpox IgG via ELISA showed a higher apparent sensitivity (891/2801; 31.81%) compared to the other serological methods: IgM ELISA (241/2688; 8.96%), HI assay (88/430; 20.46%), and neutralization assay (85/285; 29.82%). The apparent higher sensitivity of IgG ELISA can be attributed to the presence of IgG antibodies post-infection for a longer period than IgM antibodies and can be detected retrospectively. Virus isolation (228/1259; 18.10%) and EM (216/1226; 17.61%) were more commonly used for Mpox or OPXV detection in human clinical samples compared to immunographic methods (Figure 5). Since most studies used more than one method for Mpox detection, their positivity rates for such combinations of methods are discussed in detail in the following section. Detailed information on Mpox- or OPXV-positive human samples from individual articles selected in this study can be seen in Supplementary Table S1. Briefly, studies that attempted only molecular detection of Mpox or OPXV resulted in a higher positivity rate for realtime PCR-based detection (3722/5893; 63.16%) than the conventional PCR-based detection (365/1167; 31.28%). Interestingly, conventional PCR-based methods are frequently included in approaches that use a combination of methods to detect Mpox or OPXV, such as genome sequencing. In a subset of studies, real-time PCR (9/9; 100%) and IHC (11/11; 100%) had higher positivity rates for Mpox detection in a limited sample size. In other studies, conventional PCR (10/14; 71.42%), EM (20/29; 68.96%), and virus isolation (10/22; 45.45%) resulted in a relatively lower positivity rate for Mpox detection in tested samples. On the other hand, few studies used a combination of serologic and molecular tools for Mpox detection in human samples. While real-time PCR yielded a higher positivity rate for Mpox detection in these samples (126/660; 19.09%) than conventional PCR (16/369; 4.34%), the IgG ELISA (75/736; 10.19%) was more frequently used serologic method for Mpox detection than IgM ELISA (3/8; 37.5%) and IFA (7/7; 100%). Many studies attempted virus isolation in combination with serologic and molecular investigations for Mpox detection. Such studies, collectively, reported Mpox detection using real-time PCR (122/494; 24.70%), conventional PCR (17/255; 6.67%), Western blot (65/90; 72.22%), IgG ELISA (97/195; 49.74%), IgM ELISA (27/458; 5.89%), neutralization assay (41/79; 51.90%), HI assay (54/77; 70.13%), radioimmunoassay (RIA) (5/23; 21.74%), and IFA (7/7; 100%). A few other studies, in combination with serologic and molecular methods, used virus isolation, IHC, and EM for Mpox detection. These studies reported the prevalence of Mpox in human samples using IgM ELISA (159/1145; 13.89%), IgG ELISA (67/585; 11.45%), virus isolation (201/1203; 16.71%), EM (196/1197; 16.37%), HI assay (19/158; 12.02%), precipitation-in-gel test (20/47; 42.55%), RIA (17/47; 36.17%), complement fixation test (3/47; 6.38%), IHC (17/29; 58.62%), conventional PCR (22/25; 88.0%), real-time PCR (3/3; 100%), agar-gel test (8/9; 88.89%), and IFA (1/1; 100%). Detailed Microorganisms 2023, 11, 1186 11 of 31 information on various aspects of Mpox disease, such as clinical disease symptoms, transmission status, and history of contact with animals, along with the number and types of samples and methods used for Mpox detection in naturally infected humans, is provided in Supplementary Table S1. Most serologic assays used ELISA for IgM and IgG detection in human sera. Interestingly, Mpox- specific IgM antibodies in human sera could be detected as early as 5 days after the onset of symptomatic rashes and as long as 56 days after symptom onset [12]; Mpox IgM titer was also detectable for the same duration [123]. In many studies, patients with elevated anti-OPXV IgM titer in blood collected within 56 days after the onset of rash illness were considered confirmed Mpox-positive [11,13]; however, a few sera samples had detectable anti-IgM titers up to 70 days post-onset of clinical symptoms [13]. The IgM ELISA could not detect anti-IgM titer in sera samples collected on day 1 and day 147 after the onset of the rash illness. Of note, sera obtained between 5 and 77 days after the onset of rash illness resulted in 92% sensitivity and 100% specificity for IgM detection [124]. It was intriguing to note that IgM response was suggested to be detectable up to six months in some of the primary vaccinees, based on anecdotal experiences with the Centers for Disease Control and Prevention (CDC) vaccines, based on unpublished data mentioned in one study [5]. Serologically, in human cases where anti-OPXV IgG ELISA was positive and antiOPXV IgM ELISA was negative, when a blood sample was collected at least 56 days after the onset of rash illness, given that the patient was not previously vaccinated for smallpox, was considered confirmed Mpox-positive [121]. While IgG ELISA failed to detect anti-IgG titer in sera samples that were collected within 3 days and on the 8th day after the onset of rash illness [124], anti-IgG titer in blood at day 12 after the onset of rash illness was detected to be 1:800, which increased four-fold to 1:3200 on day 55 of rash onset [109]. Karem et al. 2005 [124] suggested that the sera samples collected at ≥ 5 days for IgM detection after the onset of rash illness were most efficient for Mpox detection. Of note, the sera samples collected less than 14 days after the onset of rash illness may be problematic for IgG detection in unvaccinated individuals. Of note, sera samples collected after 14 days of rash onset resulted in 100% sensitivity and 88.5% specificity for IgG detection [124]. On the other hand, OPXV- and Mpox-specific real-time PCR failed to detect Mpox DNA in blood samples collected between 61 and 90 days after the onset of rash illness [107]. While most PCR-based studies successfully detected Mpox DNA in clinical samples or blood collected up to day 12 after rash onset [109], only one study reported the detection of Mpox DNA in skin lesions, swabs, and blood obtained from one patient, after 41 days of rash onset [107]. While Mpox whole genome sequences could be obtained using clinical samples with Cq values < 30 [96], one study determined that clinical swabs with Cq values ≥ 35 had very low or no infectivity in tissue culture [145]. Samples obtained during the first nine days after the onset of clinical symptoms had higher viral loads, with Cq < 25 [96]. Many studies used Sanger sequencing for Mpox partial genome sequencing, including HA gene, for confirmation and clade identification; various recent studies used next-generation sequencing methods, such as ION Torrent PGM [116], Illumina sequencing [14,68], or Oxford Nanopore MinION [68,117] for Mpox whole genome sequencing, for confirmation and characterization. The virus isolation [68,95,98,105,110,138] and EM [105] remain useful for confirmation of diagnosis, including those cases where Mpox transmission occurred via sexual contact, and based on clinical symptoms, the cases were misdiagnosed [14]. The overwhelming majority of Mpox human cases were symptomatic; only one study reported two patients with no clinical symptoms [95]. Novel and unusual clinical symptoms were observed in many cases during the 2022 Mpox outbreaks [48]—themselves uncharacteristic of historical Mpox—such as severe anal pain [99,146], proctitis [48,147], and genital ulcers [117]. Of note, these patients were largely thought to have contracted Mpox via human-to-human transmission during sexual contact [14,99,146,148]. Microorganisms 2023, 11, 1186 12 of 31 While Mpox is endemic in the regions of West and Central Africa, the first report of human Mpox outside its endemic regions appeared in the United States in 2003 when a child was detected as Mpox positive [66]. The outbreak in the United States was the zoonotic transmission of Mpox from infected prairie dogs to humans. Of note, the first human-mediated exportation of Mpox from Africa to outside occurred via air travel in 2017, when four people (two Nigerian nationals, one British, and one Israeli citizen) traveled outside Nigeria [68]. A sharp increase in human cases of Mpox was seen outside the endemic regions, starting after 2017, and culminated in an epidemic starting in early 2022. 3.4. Prevalence of Mpox Disease in Animals Active infection of Mpox was first detected, confirmed, and characterized using lesion swabs obtained from sick cynomolgus monkeys, using HI assay, virus isolation, and EM [51]. Of the selected 104 reports, a substantial minority (n = 21 studies) [44,51, 58,64,89,91,106,107,111,113,120,122,125,126,130,149–154] detailed the diagnosis of Mpox in animal hosts. The animal samples were processed using the different approaches detailed above, with some notable biases evident depending on the species of the host. We observed that only a few animal species, such as prairie dogs [44,64,91], a few species of nonhuman primates [51,106,107], a dead squirrel, and a dog [151,154] had visible external clinical signs of disease at the time of investigation. Many animals under investigation, including various species of rodents [111,113,120,122,130,150,153] and some nonhuman primates [125,149,150], did not exhibit clinical signs of disease. This appears to have led to the selection of a significantly smaller number of animal samples that were collected for Mpox or OPXV detection over the period. Rather than presenting aggregate numbers of Mpox- or OPXV-positive animal samples detected via different techniques, we categorized the animal species under study in separate groups to compare the Mpox or OPXV positivity in various animal hosts. Real-time PCR (1/189), conventional PCR (62/492), virus isolation (36/54), EM (3/3), IgG ELISA (12/81), HI assay (28/206), neutralization assay (5/14), and IFA (8/17) detected Mpox or OPXV infections in various species of nonhuman primates (Figure 6A; Supplementary Table S2). The IgG ELISA (65/736), real-time PCR (34/503), and conventional PCR (26/211) were more frequently used methods to detect Mpox infections or antibodies in rodents than other methods, such as IFA (12/160), neutralization assay (13/43), virus isolation (23/31), EM (3/3), and IHC (2/2) (Figure 6B). While IgG ELISA (17/93) and neutralization assay (2/4) detected Mpox antibodies in the sera of several mammals, including shrews and opossums, real-time PCR (3/70) and virus isolation (1/1) detected active Mpox infections only in a few samples tested (Figure 6C). Intriguingly, Mpox DNA was detected, using PCR, in some of the flies and maggots (5/89) [107] that were feeding upon a dead chimpanzee who had typical Mpox lesions externally (Figure 6D). Detailed information on Mpox prevalence, animal species naturally infected, samples used for detection, methods used, and the clinical signs of disease observed in various animal species is provided in Supplementary Table S2. Most studies that investigated Mpox in animals sampled multiple species of nonhuman primates and rodents, which are known to serve the animal reservoirs of Mpox, irrespective of clinical signs of disease. Intriguingly, many species of rodents with no external clinical signs of disease were found positive for Mpox DNA using samples obtained from multiple internal organs, suggesting systemic infections [111,120]. Mpox DNA was also detected in the multiple internal organs (necropsy tissues) of nonhuman primates with clinical signs of disease [106,107]. In addition, anti-OPXV IgG antibodies were detected in the blood samples of rodents [122,130,150] and nonhuman primates with no external lesions [58,125,126,149,150,152]. Detection of Mpox DNA using PCR, occasionally in combination with virus isolation, using multiple organ tissues and clinical samples obtained from nonhuman primates and rodents such as prairie dogs, wild squirrels, dormice, and shrews suggested the evidence of systemic Mpox infections in these animals. This may justify why nonhuman primates and various species of rodents are considered reservoirs Microorganisms 2023, 11, 1186 13 of 31 Microorganisms 2023, 9, x FOR PEER REVIEW for Mpox 13signs of 31 infections. The next sections will discuss the known host ranges, clinical and symptoms of Mpox disease, and the transmission dynamics. Figure 6. Detection of Mpox in animals. (A) More Mpox infections in nonhuman primates were Figure 6. Detection of Mpox in animals. (A) More Mpox infections in nonhuman primates were detected using conventional PCR than serologic methods and virus isolation. (B) Mpox infections detected using conventional thanspecies serologic and virus isolation. Mpox infections were successfully detected inPCR various of methods rodents using molecular and(B) serologic methods, were successfully detected in various species of rodents using molecular and serologic methods, virus isolation, IHC, and EM. (C) Mpox was also detected in some other mammals, including virus isolation, IHC, and EM. (C) Mpox was also detected in some other mammals, including shrews, opossums, a pig, and a dog using real-time PCR, serologic assays, and virus isolation. (D) Conventional PCR detected Mpox infection some of PCR, the flies and maggots uponisolation. a chimshrews, opossums, a pig, and a dog usinginreal-time serologic assays,feeding and virus panzee with external Mpox lesions found dead ininthe forest. (D) Conventional PCR detected Mpox infection some of the flies and maggots feeding upon a chimpanzee with external Mpox lesions found dead in the forest. Most studies that investigated Mpox in animals sampled multiple species of nonhu3.5. Ranges of Mpox manHost primates and rodents, which are known to serve the animal reservoirs of Mpox, irrespective of from clinical signs (Homo of disease. Intriguingly, species of rodents with animal no external Apart humans sapiens), Mpox has many a well-documented and broad host range, various were species of nonhuman primates other mammals, as rodents, clinicalincluding signs of disease found positive for Mpox and DNA using samplessuch obtained from shrews, opossums (Supplementary Table S2). Studies investigating nonhuman hosts inmultipleand internal organs, suggesting systemic infections [111,120]. Mpox DNA was also cluded: humans and prairie dogs (n = 1 study) [44]; humans and rodents (n = 2) [113,120]; a hudetected in the multiple internal organs (necropsy tissues) of nonhuman primates with man andsigns a dogof(ndisease = 1) [154]; rodentsIn(naddition, = 8 studies) [64,89,91,111,122,130,151,153]; nonhuman clinical [106,107]. anti-OPXV IgG antibodies were detected in primates = 7) [51,58,106,125,126,149,152]; nonhuman primates and rodents = 1) [150]: the blood(nsamples of rodents [122,130,150] and nonhuman primates with no(nexternal lenonhuman primates and insects Detection (n = 1) [107]. methods, as molecular, serosions [58,125,126,149,150,152]. of Using Mpox various DNA using PCR,such occasionally in comlogic, andwith immunographic approaches, these studies natural infections ofobtained Mpox in bination virus isolation, using multiple organidentified tissues and clinical samples cynomolgus monkey (Makaka [126,152], chimpanzee (Pan troglodytes verus) [107], from nonhuman primates andfascicularis) rodents such as prairie dogs, wild squirrels, dormice, and red colobus monkey (Colobus badius) [125], sooty mangabey (Cercocebus atys) [106], langur shrews suggested the evidence of systemic Mpox infections in these animals. This monmay key (Semnopithecus spp.) [152], vervet monkey (Chlorocebus white-nosed monjustify why nonhuman primates and various species ofpygerythrus), rodents arelesser considered reservoirs key (Cercopithecus petaurista) [125], Philippine philippinensis) [152], for Mpox infections. The next sections willlong-tailed discuss themacaque known (Macaca host ranges, clinical signs chacma baboon of (Papio ursinus) [150], and yellow baboon (Papio cynocephalus). Other mamand symptoms Mpox disease, and thethe transmission dynamics. mals identified as naturally infected with Mpox include domestic pig (Sus scrofa domesticus) [113], dog (Canis lupus familiaris) [154], gray short-tailed opossum (Monodelphis domestica) [153], south- Microorganisms 2023, 11, 1186 14 of 31 ern opossum (Didelphis marsupialis) [153], shrew (Crocidura littoralis) [150], and elephant shrew (Petrodromus tetradactylus) [122]. Natural Mpox infections have been widely reported in several species of rodents, including African hedgehog (Atelerix spp.) [153], African giant pouched rat (Cricetomys sp.) [150], giant-pouched rat (Cricetomys emini) [3], African dormouse (Graphiurus lorraineus) [153], target rat (Stochomys longicaudatus) [111], rusty-nosed rat (Oenomys hypoxanthus), wild mice (Mus musculus) [130], wild rats (Rattus norvegicus and Rattus rattus) [130], gerbil (Gerbillus sp.) [89], prairie dog (Cynomys sp.) [44,89], black-tailed prairie dog (Cynomys ludovicianus) [153], chinchilla (Chinchilla sp.) [89], jerboa (Jaculus spp.) [153], hamster (Cricetus spp.) [89], woodchuck (Marmota monax) [153], and several species of squirrels, such as Thomas’s rope squirrel (Funisciurus anerythrus) [111], Lunda rope squirrel (Funisciurus bayonii) [111], Kuhl’s tree squirrel (Funisciurus congicus) [113], and sun squirrel (Heliosciurus rufobrachium) [113,122]. Interestingly, Mpox DNA was also detected in maggots and flies that were feeding upon a dead chimpanzee (Pan troglodytes verus) in the wild, reported in one study [107]. 3.6. Clinical Manifestations of Mpox Disease in Humans and Animals Many studies reported clinical disease symptoms and clinical signs that occurred due to Mpox infections in humans and animals, respectively. Of note, rash (n = 64), fever (n = 57), skin lesions (n = 48), lymphadenopathy (n = 38), headache (n =23), myalgia (n = 17), sore throat (n = 15), chills (n = 12), malaise (n = 10), and fatigue (n = 10) were some of the common disease symptoms in humans (Figure 7A; Supplementary Table S1). The Mpox disease resulted in clinical signs of anorexia, lethargy, wasting, and influenza-like illness, among other clinical signs in prairie dogs (Cynomys sp.) [64,65,91] (Figure 7B; Supplementary Table S2). Interestingly, no clinical signs of disease were reported in other Mpox-positive rodent species and shrews. Clinical signs of rash, vesicular lesions, respiratory illness exhibited by coughing, breathing with mouth open, and nasal and ocular discharge were observed in some of the chimpanzees (Pan troglodytes verus) [107]. A sooty mangabey (Cercocebus atys) that was found dead was Mpox-positive and had skin lesions [106]. Imported cynomolgus monkeys (Makaka fascicularis) from Singapore, in captivity at the Statens Serum Institut, Copenhagen, Denmark, where the Mpox disease was first detected in animals in 1958, had maculopapular rash and skin lesions [51] (Figure 7C; Supplementary Table S2). A four-year-old greyhound pet dog (Canis lupus familiaris) had mucocutaneous lesions, abdomen pustules, and anal ulcers [154]. Skin eruptions were observed in a wild squirrel (Funisciurus anerythrus) that was found dead near a village in DRC. Interestingly, Mpox was confirmed with virus isolation; the first declared Mpox isolation from a dead squirrel [151]. 3.7. Oligonucleotide Primers Used for Mpox Detection in PCR and Real-Time PCR Assays Some of the commonly reported oligonucleotide primers and probes that were used for the molecular detection of Mpox, targeting specific genes, using real-time PCR or conventional PCR in previous studies, are summarized in Table 2. Microorganisms 2023, 9, x FOR PEER REVIEW 15 of 31 Microorganisms 2023, 11, 1186 15 of 31 Mpox was confirmed with virus isolation; the first declared Mpox isolation from a dead squirrel [151]. Figure 7. Number of studies that reported (A) the clinical symptoms of Mpox disease in humans, Figure 7. Number ofMpox studies that in reported (A) the symptoms of Mpox disease in humans, (B) clinical signs of disease prairie dogs, andclinical (C) clinical signs of Mpox disease in nonhu(B) man clinical signs of Mpox disease in prairie dogs, and (C) clinical signs of Mpox disease in primates. nonhuman primates. 3.7. Oligonucleotide Primers Used for Mpox Detection in PCR and Real-Time PCR Assays Target (Specificity) Some of the commonly reported oligonucleotide primers and probes that were used Table Oligonucleotide primers and probes usedspecific for Mpox detection clinical PCR samples. for2.the molecular detection of Mpox, targeting genes, usingin real-time or conventional PCR in previous studies, are summarized in Table 2. ORF or Genomic References Region/Primer and Probe Oligonucleotide Sequence 50 –30 Table 2. Oligonucleotide primers and probes used for Mpox detection in clinical samples. (Earliest Citation) ID ORF or Genomic General OPXV-detecting real-time PCR Region/Primer and Oligonucleotide Sequence 5′–3′ TAATACTTCGATTGCTCATCCAGG P4A/Forward Probe ID ACTTCTCACAAATGGATTTGAAAATC P4A/Reverse General OPXV-detecting real-time PCR FAM-TCCTTTACGTGATAAATCAT-NFQ P4A/Probe MGB P4A/Forward TAATACTTCGATTGCTCATCCAGG P4A: Major core protein 4a E9L: DNA polymerase TCAACTGAAAAGGCCATCTATGA P4A/Reverse ACTTCTCACAAATGGATTTGAAAATC E9L/Forward geneprecursor GAGTATAGAGCACTATTTCTAAATCCCA FAM-TCCTTTACGTGATAAATCAT-NFQ E9L/Reverse P4A/Probe (OPXV(OPXV family;family) TET-CCATGCAATATACGTACAAGATAMGB E9L/NVAR-Probe non-variola) GTAGCCAAC-QSY7 E9L/Forward TCAACTGAAAAGGCCATCTATGA E9L: DNA polymerase Mpox-specific real-time PCR gene E9L/Reverse GAGTATAGAGCACTATTTCTAAATCCCA GGAAAATGTAAAGACAACGAATACAG G2R/G-Forward G2R: TNF receptor gene GCTATCACATAATCTGGAAGCGTA G2R/G-Reverse (Mpox-generic) FAM-AAGCCGTAATCTATGTTGTCTATCG2R/G-Probe GTGTCC-BHQ1 Target (Specificity) P4A: Major core protein 4a precursor (OPXV family) References (Earliest Citation) [155] [155] [13,64,114,153,156] [13,64,114,153,156] [3,8,114,157,158] Microorganisms 2023, 11, 1186 16 of 31 Table 2. Cont. Target (Specificity) ORF or Genomic Region/Primer and Probe ID Oligonucleotide Sequence 50 –30 References (Earliest Citation) G2R: TNF receptor gene (West African-specific Mpox clade) G2R/WA-Forward G2R/WA-Reverse G2R/WA-Probe CACACCGTCTCTTCCACAGA GATACAGGTTAATTTCCACATCG FAM-AACCCGTCGTAACCAGCAATACATTT-BHQ1 [3,8,114,157,158] C3L/CB-Forward TGTCTACCTGGATACAGAAAGCAA C3L/CB-Reverse GGCATCTCCGTTTAATACATTGAT FAM-CCCATATATGCTAAATGTACCGGT -ACCGGA-BHQ1 C3L: Complement binding protein Central African (formerly Congo Basin)-specific Mpox clade B6R: Envelope protein gene (Mpox) F3L: Interferon resistance gene (Mpox) N3R: Conserved (unclassified function) (Mpox) C3L/CB-Probe ATTGGTCATTATTTTTGTCACAGGAACA AATGGCGTTGACAATTATGGGTG MGB-AGAGATTAGAAATA-FAM B6R/Forward B6R/Reverse B6R/Probe F3L/F290-Forward F3L/R396-Reverse F3L/p333S-MGB-Probe N3R/F319-Forward N3R/R457-Reverse N3R/p352S-MGB-Probe CTCATTGATTTTTCGCGGGATA GACGATACTCCTCCTCGTTGGT FAM-CATCAGAATCTGTAGGCCGT -MGB NFQ AACAACCGTCCTACAATTAAACAACA CGCTATCGAACCATTTTTGTAGTCT FAM-TATAACGGCGAAGAATATACT -MGB NFQ [3,8,114,157,158] [13,64,114,153,156] [40,89,109,159] [40,89,109,159] Conventional Mpox PCR E5R: Nucleic acid-independent nucleoside triphosphatase (Mpox) ATI: A-type inclusion body protein (Mpox) E5R/Forward ATGTTGATATTAATAATCGTATTGTGGTT E5R/Reverse AAAGTCAATACACTCTTAAAGATTCTCAA ATI/Gabon 1- Forward ATI/Gabon 2- Reverse GAGAGAATCTCTTGATAT ATTCTAGATTGTAATC [4,160] [142,161] Primer-conjugated fluorophores and quenchers appear at 5’ and 3’ ends, respectively, in the above sequences, with the exception of B6R (envelope protein gene) of Mpox. Primer-bound fluorophores: FAM—6-carboxyfluoresceine; TET—tetrachlorofluoresceine. Primer-bound fluorophore quenchers: MGB—Minor groove binding; QSY7—quencher succinimidyl ester; BHQ1—blackhole quencher 1. While PCR-based detection offers a great advantage over serological assays, often regarding sensitivity and specificity, for the detection and identification of infectious pathogens in test samples, it also has some serious limitations. For example, PCR-based detection of infectious pathogens, including Mpox, requires post-PCR visual quantification with agarose gel electrophoresis. Independently of the initial concentration of the DNA template used for amplification of the target, the conventional PCR assay might fail to amplify enough of the DNA required for visual quantification on the agarose gel, which might result in a false negative [162]. Most conventional three-step PCR assays use an elongation temperature of 72 ◦ C, which is the recommended temperature for optimum activity of Taq polymerase enzyme in a PCR assay. The conventional PCR assay is usually designed to generate relatively large sizes of amplicons, required for post-run quantification on the agarose gel. The higher elongation temperature in PCR should ideally melt any secondary structures that might be formed in the long DNA template formed during amplification and might block the extension of the amplified product [162]. Such limitations have been greatly overcome by the use of probe-based, such as TaqMan, real-time PCR for highly sensitive and specific detection of the target in a test sample. While several TaqMan real-time PCR assays may use a lower extension temperature, such Microorganisms 2023, 11, 1186 17 of 31 as 60 ◦ C [163], they are also designed to generate much shorter amplicons than the conventional PCR, thus minimizing the probability of forming secondary structures during amplification and extension. Another great advantage that TaqMan real-time PCR detection of Mpox offers over conventional PCR is its higher specificity and sensitivity due to the use of fluorescently labeled probes. The real-time PCR assays have used various fluorescent probes for Mpox detection, which have covalently bonded fluorescent-labeled reporter dye at the 5-prime end, such as tetrachlorofluorescein (TET) [13,64,114,153,156] or 6-carboxyfluorescein (FAM) [3,8,114,157,158] and a quencher, such as blackhole quencher 1 (BHQ1) [3,8,114,157,158], quencher succinimidyl ester (QSY7) [13,64,114,153,156], or minor groove binding (MGB) [40,89,109,159] at the 3-prime end. In the event of the amplification of the target molecule during the reaction, the given probe anneals to the target, followed by its cleavage by the Taq polymerase enzyme [164]. The cleavage or degradation of the fluorescent probe results in the emission of a fluorescence signal based on the principle of fluorescence resonance energy transfer (FRET) [165]. In the lack of amplification of the target molecule, the fluorescent probe remains intact, and as a result, its proximity with the quencher significantly decreases the emitted fluorescence. Therefore, by monitoring the fluorescence intensity during a reaction, real-time PCR determines the amplification of the target [164]. Interestingly, the MGB probe, due to its chemistry, offers a higher melting temperature and increased specificity to the target compared to other probes used [166]. Often, real-time PCR quantifies the concentration of the target as the cycle threshold (Ct) value by determining the lowest PCR amplification cycle at which significant amounts of fluorescent signals are produced. Higher concentrations of targets thus correspond to lower Ct values and vice versa; in general, lower Ct values correspond to increased infectivity and higher viral load [167]. Paran et al. 2022, reported that Ct values ≥ 35 in clinical samples predict very low or no infectivity of Mpox and correlates with noninfectious virus [145]. Lapa et al. 2022, isolated the replication-competent Mpox from the seminal fluid with a Ct value of 22.7 [138]. In another study, Ma et al. 2022, reported isolation of Mpox from 19 human clinical samples that had Ct values in the range of 15.3 to 29.0 [168]. The selected studies for Mpox detection in humans used a wide range of clinical samples; however, the preferred human samples included skin lesions, lesion swabs or exudates, vesicular lesions, vesicular fluids or swabs, crusts, and pustule swabs for molecular detection and blood or sera for serologic detection. Detailed information on the sample types and Mpox positivity is provided in Supplementary Table S1 for human samples and Supplementary Table S2 for animal samples. Mpox comprises a double-stranded DNA genome of approximately 197 kb [4,169] and encodes 190 open reading frames (ORFs) [169]. Stretches of nucleotide sequences in some of the gene segments of orthopoxviruses, for example, P4A, a major core protein 4a precursor, is a suitable target, offers pan-orthopoxvirus detection, using clinical samples [155]. Another most used target includes conserved sequences of E9L, the DNA polymerase gene, which offers non-variola orthopoxvirus detection in clinical samples with real-time PCR [13,64,114,153,156]. Table 2 summarizes the most used oligonucleotide primers and probes that offer sensitive and specific detection of Mpox in clinical samples. In fact, the confirmatory diagnosis of Mpox requires the detection of Mpox DNA in clinical samples using Mpox-specific real-time PCR or conventional Mpox-specific PCR in combination with partial genome sequencing for clade determination and epidemiological analysis [32]. While most studies adopted a two-step approach to Mpox detection, using OPXV-specific real-time PCR for orthopoxvirus detection followed by Mpox-specific real-time PCR or Mpox-specific conventional PCR with partial genome sequencing for confirmation of Mpox DNA, a study used a novel multiplex real-time PCR assay, included specific-Mpox signature encoded in Mpox envelop gene (B6R) and the OPXV-specific signature encoded in the DNA polymerase gene (E9L, non-variola) [64]. Interestingly, an automated Mpox testing system termed GeneXpert MPX/OPX platform, though in its initial phase of development [128], can detect Mpox DNA using OPXV-generic and Mpox-specific multiplex Microorganisms 2023, 11, 1186 18 of 31 detection approach with high sensitivity and 100% specificity [128]. The availability of such a high-throughput and automated detection system would certainly facilitate scalable detection of Mpox under field outbreak settings with the added advantage of minimal manual sample handling to reduce the risk of exposure to the operator. 3.8. Transmission Dynamics of Mpox Disease Reports studying the zoonotic and interspecies transmission of Mpox are summarised in Figure 8A. The research articles and case reports investigated in the present study identified 27 studies, mostly before 2020, that documented human-to-human household transmission of Mpox presumed due to close, extended contact between patients [3,5,7,9,12,13,39,42, 44,60–62,68,71,83,85,87,104,113,121,142,143,158,170–173]. In contrast to these reports, the Microorganisms 2023, 9, x FOR PEER REVIEW 19 of 31 2022 multi-country outbreaks of Mpox were dominated by human-to-human transmission via sexual contact, as reported in 18 studies [14,31,48,86,94–96,99,116,138,146,148,174–179], raising concerns about Mpox’s potential as a sexually-transmitted infection. Co-infection [5,11,44,66,91,92]. Contact with, or feeding monkeysin and otherduring nonhuman with sexually-transmitted infections wasupon, also studied detail thisprimates outbreak: 18 was another common risk factor forPCR Mpox in five separate studies studies, using either Mpox-specific or zoonosis, real-timereported PCR, collectively identified 606 pa[114,126,141,173,182]. Similarly, hunting and eating squirrels can transmit Mpox hu- and tients that were infected solely with Mpox, a further 228 patients had dualtoMpox mans (n = 3) [109,113,158]. Additionally, onewith studyMpox, each reported thatSARS-CoV2, human contactand witha final HIV infections, 1 patient was co-infected HIV, and pet rodents [66], rodent carcasses [4], and dead squirrels [151] were some of the risk factors patient was co-infected with Mpox, HIV, and syphilis [180]. An overwhelming majority of for Mpox zoonosis (Figure 8B). case-patients were male, with a median age of 35 years [86]. The detection of Mpox DNA Animal-to-animal transmission of Mpox may be facilitated by close contact between in semen samples [99,138,178], along with the reported isolation of replication-competent animals in their natural habitat [122,152] or due to housing within a restricted space [153]. Mpox from seminal fluidproximity, [138], suggested possibility ofthat sexual transmission of Mpox In addition to the close one studythe [107] determined feeding or scavenging between however, further studies are required. by flieshumans, and maggots upon dead animals or carcasses may transmit Mpox (Figure 8C). Figure 8. Transmission dynamics of Mpox. (A) Zoonotic or reverse zoonotic transmission of Mpox Figure 8. Transmission dynamics of Mpox. (A) Zoonotic or transmission reverse zoonotic transmission of Mpox in in humans and animals. (B) Risk factors involved in zoonotic of Mpox. (C) Risk factors humans andinanimals. (B) Risk factors involved in zoonotic transmission of Mpox. (C) Risk factors involved the animal-to-animal transmission of Mpox. involved in the animal-to-animal transmission of Mpox. Collectively, the studies reporting zoonotic and interspecies transmission of Mpox reinforce that this is a contagious that combines atransmission, high potential of zoonotic spillover Considering minor modes ofvirus human-to-human three studies [5,100,181] with a well-documented wide host range. Hosts include various species of nonhuman reported hospital-acquired transmission of Mpox, which might have occurred due to primates and small mammals, especially rodents. While the ‘West African’ of Mpox visits, sharing of hospital beds, exposure to contaminated equipment(s) duringclade the hospital dominated the circulation within and outside the endemic regions more than the ‘Central and/or spillover to the healthcare workers during inpatient treatments. One study reported African’, formerly known as ‘Congo Basin’, clade, the multi-country Mpox outbreaks of mother-to-fetus transmission of Mpox [40]. 2022 were suggested to have occurred due to the emergence of a new clade, termed ‘hMPXV-1A’ lineage B.1, which appeared to have diverged from the ‘West African’ clade [101]. Skin rash, fever, lesions, and lymphadenopathy were some of the most commonly occurring clinical symptoms of Mpox disease in humans; however, various other symptoms may also be present. As observed during 2022 Mpox outbreaks in non-endemic countries, clinical disease symptoms, such as proctitis and anogenital ulcers, were a few of the unusual symptoms observed in Mpox infections in humans (Supplementary Table Microorganisms 2023, 11, 1186 19 of 31 Twenty-three studies reported animal-to-human zoonotic transmission of Mpox [3– 6,11,61,65,66,83,85,92,109,110,112–114,120,126,134,141,151,153,158], which was suggested to have occurred due to several factors, including exposure to or feeding upon dead wild animals or occasionally from pets within the household. Five studies suggested animal-to-animal, and one study identified human-to-dog transmission of Mpox due to close contact or extended sharing of space [51,64,89,91,153,154] (Figure 8A). Some studies (n = 6) reported that close contact with pet prairie dogs might transmit Mpox to humans [5,11,44,66,91,92]. Contact with, or feeding upon, monkeys and other nonhuman primates was another common risk factor for Mpox zoonosis, reported in five separate studies [114,126,141,173,182]. Similarly, hunting and eating squirrels can transmit Mpox to humans (n = 3) [109,113,158]. Additionally, one study each reported that human contact with pet rodents [66], rodent carcasses [4], and dead squirrels [151] were some of the risk factors for Mpox zoonosis (Figure 8B). Animal-to-animal transmission of Mpox may be facilitated by close contact between animals in their natural habitat [122,152] or due to housing within a restricted space [153]. In addition to the close proximity, one study [107] determined that feeding or scavenging by flies and maggots upon dead animals or carcasses may transmit Mpox (Figure 8C). Collectively, the studies reporting zoonotic and interspecies transmission of Mpox reinforce that this is a contagious virus that combines a high potential of zoonotic spillover with a well-documented wide host range. Hosts include various species of nonhuman primates and small mammals, especially rodents. While the ‘West African’ clade of Mpox dominated the circulation within and outside the endemic regions more than the ‘Central African’, formerly known as ‘Congo Basin’, clade, the multi-country Mpox outbreaks of 2022 were suggested to have occurred due to the emergence of a new clade, termed ‘hMPXV1A’ lineage B.1, which appeared to have diverged from the ‘West African’ clade [101]. Skin rash, fever, lesions, and lymphadenopathy were some of the most commonly occurring clinical symptoms of Mpox disease in humans; however, various other symptoms may also be present. As observed during 2022 Mpox outbreaks in non-endemic countries, clinical disease symptoms, such as proctitis and anogenital ulcers, were a few of the unusual symptoms observed in Mpox infections in humans (Supplementary Table S1). Close contact between humans and animals, as well as customary feeding upon wild animals, were observed to be major risk factors triggering Mpox zoonosis. While earlier studies detected Mpox seroprevalence using various serologic techniques, the virologic investigations were based on virus isolation, EM, and histologic examination using IHC. In later years, the introduction of DNA amplification-based detection, such as real-time PCR, became the choice of the molecular detection of Mpox in clinical specimens and successfully detected Mpox DNA in a larger number of clinical samples and organ tissues than any other methods used. Further, Sanger and the next-generation sequencing techniques, in combination with conventional PCR, greatly facilitated Mpox genome sequencing for epidemiological investigations and monitoring the evolution of the Mpox genome and the clade determination. 4. Discussion The clinical symptoms of Mpox in humans may be similar to those caused by other members of the OPXV genus; thus, the molecular detection of Mpox is required to confirm the diagnosis. Interestingly, based on clinical disease symptoms and detection techniques used, human cases of Mpox can be categorized into three groups: suspected, probable, and confirmed [117]. According to the recently revised definitions of Mpox cases, suspect cases are those with a febrile prodrome with at least one or more clinical manifestations such as skin rash, headache, backache, asthenia, lymphadenopathy, myalgia, or sudden onset of anogenital complaints developed after 15 March 2022, unexplained by other differential diagnoses [117]. Mpox probable cases are those that meet the Mpox suspect case criteria and have traveled to the Mpox endemic region or had contact with an Mpox suspected, probable, or confirmed case within the past 21 days before the onset of clinical symptoms. Microorganisms 2023, 11, 1186 20 of 31 Additionally, sexual contact with anonymous or multiple partners within 21 days before the onset of symptoms and fulfilling the definition of a Mpox suspected case classify the cases into the Mpox probable case category. Furthermore, cases where the clinical specimens have tested positive in a laboratory using Mpox- specific real-time PCR or genome sequencing are considered Mpox-confirmed cases [117]. The earliest documented Mpox outbreaks in the endemic regions (rain forests in West and Central Africa) were caused by either West African or Central African clades, based on phylogenomics and clinical manifestations of the disease [101]. While the West African clade has been reported to inflict a mild disease, the symptoms caused by the Central African clade are relatively severe [101]. The first human cases of Mpox outside endemic regions in Africa were zoonotically transmitted from Mpox-infected pet prairie dogs. Interestingly, the prairie dogs became infected upon contact with other imported species of rodents at pet shops in the Midwestern United States in 2003 that originated from Ghana as part of the exotic pet trade. While more recently, the exportation of Mpox human cases outside of Africa started in 2017, transmitted via air travelers [69], and the multicountry outbreaks of Mpox, which began in April–May 2022, appear to have been driven by sexual contact [98,99,138]. These multi-country outbreaks of Mpox have been reported to have occurred due to a newly emerged clade termed ‘hMPXV-1A’ lineage B.1 [103], which may have diverged from the West African clade of Mpox. It has been determined that certain adaptive mutations in the Mpox 2022 isolates facilitated its human-to-human transmission [103]. The incubation period of Mpox is generally between seven to fourteen days [81]; however, it has been observed to be up to 20 [48] or 21 [183] days, depending on the route of exposure [184]. Rash, fever, skin lesions, and lymphadenopathy were some of the most observed clinical symptoms of Mpox disease in humans, caused by West African and Central African clades [5,112]. Intriguingly, the 2022 Mpox outbreaks driven by the ‘hMPXV1A’ lineage B.1, attributed to sexual contact, appear to have additional clinical symptoms which were never reported in previous infections. Some of these unusual and characteristic symptoms related to the 2022 Mpox outbreaks include proctitis (severe anal pain and/or bleeding), anorectal lesions, anal ulcers, perianal ulcers, and penile ulcers [146,174,178]. Some of these symptoms may coincide with sexually transmitted diseases in humans, such as syphilis and gonorrhoea, and thus might complicate the diagnosis [48,185]. Therefore, it is prudent to investigate and understand the epidemiology and clinical manifestations of the 2022 Mpox epidemic to mitigate further transmission. One of the major challenges in the effective and timely diagnosis of the Mpox epidemic is the lack of an approved Mpox detection kit. One probable reason behind this would be that the human cases of Mpox were primarily restricted to endemic regions in Africa until recently and were routinely detected using OPXV-specific conventional PCR followed by Mpox-specific real-time PCR for confirmation [85]. It was only after 2017 that Mpox human cases started appearing outside Africa and were traced to travel history [70,105]. With a limited number of sporadic case reports appearing outside Africa with limited transmission, not enough attention was given to exploring the evolution and adaptation of Mpox in humans, and, therefore, Mpox surveillance outside Africa remained neglected. Mpox can be transmitted through close contact between humans in households [3,83]. Additionally, studies have also documented the transmission through contaminated surfaces within the household [105]. Animal-to-human transmission may occur through close contact with pets or rodents [153]. Hunting and feeding upon wild animals and exposure to dead animals or carcasses are also risk factors for Mpox infections [109]. Intriguingly, maggots and flies that were observed feeding upon a dead chimpanzee in the wild were positive for Mpox DNA and suggested that scavenging may be involved in the transmission of Mpox [107]. Whether maggots and flies can serve as vectors of Mpox remains a question for further investigation. Most cases of Mpox in animals were detected in nonhuman primates and rodents; only one study each reported Mpox DNA in a pet dog (Canis lupus familiaris) [154] and Microorganisms 2023, 11, 1186 21 of 31 antibodies in a domestic pig (Sus scrofa domesticus) [113]. While domestic pigs are reported to harbor several virus pathogens, including RNA [186–192] and DNA [193,194] viruses, the occurrence of Mpox antibodies in a domestic pig indicates the possibility of natural infection of domestic pigs with Mpox and warrants the surveillance of Mpox in domestic pig populations to investigate their status and significance. Earlier serologic studies detected OPXV antibodies in human sera, which supplemented the diagnosis in the presence of the typical poxvirus disease symptoms, such as the presence of maculopustular rash, skin lesions, fever, lymphadenopathy, etc. While IgG ELISA was one of the most used serologic methods to detect past exposure to Mpox in humans, IgM ELISA detected more recent infections in humans. Most studies that used ELISA detected OPXV antibodies and, based on clinical symptoms, suspected previous exposures to Mpox [123,195]. Most of these earlier studies, solely based on serologic detection, could only suggest the occurrence of Mpox probable cases but, in most cases, were not able to confirm the presence of Mpox because some of the other poxviruses, such as smallpox, vaccinia, and variola, are known to cause similar clinical symptoms in humans [123]. Thus, more sensitive and reliable detection methods were required to make a conclusive diagnosis. In the absence of DNA-based detection, several of those earlier studies used EM [11–13] and virus isolation [11,12] methods to confirm the Mpox disease diagnosis. While virus isolation is a gold standard for virological investigations [37], success depends on multiple factors, including virus titer in the sample and sample handling, such as collecting and storing samples. Similarly, EM is a traditional gold standard for the morphological identification of emerging and novel viruses [37] and has been proven quite useful for distinguishing isolated Mpox particles from other OPXV particles [60,61]. In addition, histological methods such as IHC were used in some studies to detect Mpox antigen in stained tissues obtained from an autopsy. IHC is a useful technique for understanding the distribution of virus particles in infected tissues [196]. The advent of PCR techniques in later years greatly facilitated the targeted detection of Mpox DNA in various clinical and environmental samples. While real-time PCR detected the highest number of Mpox infections in human populations, some of the molecular studies used genome sequencing utilizing Sanger or next-generation sequencing techniques, which greatly facilitated the monitoring of the Mpox genome evolution [4,9]. One of the greatest challenges of detecting Mpox in suspected human cases was the lack of approved or recommended Mpox detection kits by authorities such as the WHO or CDC. The scale and magnitude of the 2022 Mpox outbreaks require the availability of an automated Mpox testing system, such as the GeneXpert MPX/OPX platform, which is reported to be under development [128]. This system can detect Mpox using an OPXV-generic and Mpox-specific multiplex detection approach with high sensitivity and 100% specificity and has been tested under laboratory and field conditions in the initial phase [128]. The availability of a highthroughput and scalable, automated system for Mpox detection in an epidemic setting would greatly reduce the turnaround time of testing with the added advantage of minimal manual sample handling to reduce the risk of exposure to laboratory personnel. In the current scenario, real-time PCR-based detection of Mpox appears cost-effective and accessible for routine diagnostic testing. Real-time PCR is the most common method used for Mpox detection and is also considered a gold standard in diagnostic laboratories. While various real-time PCR assays may vary in the gene targets, specificity, and sensitivity, they still identified a much higher number of Mpox-positive human cases than any other methods used. The advantage of molecular testing of Mpox over serologic testing is that varied sample types can be used for PCR-based detection of Mpox. For example, a large number of studies successfully detected Mpox in skin lesion swabs, crusts, and blood [6,86,88,116,170], but on the other hand, serologic investigations could only use blood for the detection of OPXV antibodies [124,133,195]. A rapid, reliable, and costeffective molecular assay for Mpox detection would provide more sensitive and specific detection over serologic methods. In addition, genomic surveillance of Mpox using genome Microorganisms 2023, 11, 1186 22 of 31 sequencing will generate Mpox genomes which will help update the molecular diagnostic tools and existing vaccines to safeguard human and veterinary health. Development of a rapid, sensitive, OPXV and Mpox-specific multiplex real-time PCR is prudent for scalable detection of Mpox while detecting and differentiating among poxviruses in clinical samples. The increasing use of next-generation sequencing methods in recent years, such as Illumina and Oxford Nanpore MinION, appeared promising regarding genomic surveillance of Mpox, generated complete Mpox genomes using clinical samples, and therefore, were instrumental in determining the emergence of a new Mpox clade, hMPXV-1A lineage B.1, during 2022 Mpox outbreaks in non-endemic countries. 5. Conclusions Real-time PCR offers rapid and reliable detection of Mpox in clinical samples; however, there is scope for developing a multiplex real-time PCR assay for simultaneous detection of commonly circulating poxviruses and for OPXV-generic and Mpox-specific detection in a single assay. It would be interesting to see if sequencing platform(s) could offer costeffective detection of poxviruses based on genome sequences in the coming years. Mpox genome sequencing would be informative and useful for Mpox epidemic preparedness, for updating diagnostic methods such as oligonucleotide primers, probes, and antibodies, for superior sensitivity and specificity of assays in the detection of (re)emerging Mpox strains or variants. Mpox genome sequencing will enable more efficient and real-time tracking of the novel variants and emerging strains for accurate epidemiological analysis and will assist in identifying vaccine candidates. Though not recommended by the WHO for routine diagnosis of Mpox in clinical laboratories, virus isolation remains particularly useful for detecting and investigating the viability, pathogenicity, and virulence of Mpox in clinical samples. Moreover, the EM remains useful for the differentiation and morphological characterization of poxviruses. The IHC remains another useful method for investigating Mpox systemic infection and antigen localization in various animal tissues and organs. The IgG and IgM ELISA remain useful for serologic detection investigating recent or past Mpox infections while determining vaccine efficiencies in a given population, especially in Mpox endemic regions. Supplementary Materials: The following supporting information can be downloaded at: https://www.mdpi.com/article/10.3390/microorganisms11051186/s1, Supplementary Table S1: Data on natural Mpox infections in humans, reported in selected articles and case reports, published up to 2 September 2022. Supplementary Table S2: Data on Mpox disease in animals, including the prevalence, samples, methods used for detection, and clinical signs. Author Contributions: Conceptualization, J.L., R.P.C. and R.F.; Data Collection, R.P.C.; Methodology, R.P.C. and R.F.; Formal Analysis, R.P.C. and R.F.; Investigation, R.P.C. and R.F.; Preparation of Figures and Tables, R.P.C. and R.F.; Writing—Original Draft Preparation, R.P.C. and R.F.; Writing—Review and Editing, R.P.C., R.F. and J.L.; Supervision, J.L.; Funding Acquisition, J.L. All authors have read and agreed to the published version of the manuscript. Funding: This research was made possible in part through research grant funding of the Department of Science and Innovation (DSI)/ National Research Foundation South African Research Chair in Biotechnology Innovation & Engagement (UID 95319) and the DSI Nano-Micro Manufacturing Facility. Institutional Review Board Statement: Not applicable. Informed Consent Statement: Not applicable. Data Availability Statement: Not applicable. Conflicts of Interest: The authors declare no conflict of interest. Microorganisms 2023, 11, 1186 23 of 31 References 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. Silva, N.I.O.; de Oliveira, J.S.; Kroon, E.G.; Trindade, G.S.; Drumond, B.P. Here, There, and Everywhere: The Wide Host Range and Geographic Distribution of Zoonotic Orthopoxviruses. Viruses 2020, 13, 43. [CrossRef] [PubMed] Kipkorir, V.; Dhali, A.; Srichawla, B.; Kutikuppala, S.; Cox, M.; Ochieng, D.; Nyaanga, F.; Găman, M.A. The re-emerging monkeypox disease. Trop. Med. Int. Health 2022, 27, 961–969. [CrossRef] [PubMed] Besombes, C.; Gonofio, E.; Konamna, X.; Selekon, B.; Grant, R.; Gessain, A.; Berthet, N.; Manuguerra, J.C.; Fontanet, A.; Nakouné, E. Intrafamily Transmission of Monkeypox Virus, Central African Republic, 2018. Emerg. Infect. Dis. 2019, 25, 1602–1604. [CrossRef] [PubMed] Cohen-Gihon, I.; Israeli, O.; Shifman, O.; Erez, N.; Melamed, S.; Paran, N.; Beth-Din, A.; Zvi, A. Identification and Whole-Genome Sequencing of a Monkeypox Virus Strain Isolated in Israel. Microbiol. Resour. Announc. 2020, 9, e01524-19. [CrossRef] [PubMed] Fleischauer, A.T.; Kile, J.C.; Davidson, M.; Fischer, M.; Karem, K.L.; Teclaw, R.; Messersmith, H.; Pontones, P.; Beard, B.A.; Braden, Z.H.; et al. Evaluation of human-to-human transmission of monkeypox from infected patients to health care workers. Clin. Infect. Dis. 2005, 40, 689–694. [CrossRef] Froeschl, G.; Kayembe, P.K. Pox-like lesions and haemorrhagic fever in two concurrent cases in the Central African Republic: Case investigation and management in difficult circumstances. Pan Afr. Med. J. 2015, 22, 23. [CrossRef] Hobson, G.; Adamson, J.; Adler, H.; Firth, R.; Gould, S.; Houlihan, C.; Johnson, C.; Porter, D.; Rampling, T.; Ratcliffe, L.; et al. Family cluster of three cases of monkeypox imported from Nigeria to the United Kingdom, May 2021. Eurosurveillance 2021, 26, 2100745. [CrossRef] [PubMed] Hughes, C.M.; Liu, L.; Davidson, W.B.; Radford, K.W.; Wilkins, K.; Monroe, B.; Metcalfe, M.G.; Likafi, T.; Lushima, R.S.; Kabamba, J.; et al. A Tale of Two Viruses: Coinfections of Monkeypox and Varicella Zoster Virus in the Democratic Republic of Congo. Am. J. Trop. Med. Hyg. 2020, 104, 604–611. [CrossRef] Kugelman, J.R.; Johnston, S.C.; Mulembakani, P.M.; Kisalu, N.; Lee, M.S.; Koroleva, G.; McCarthy, S.E.; Gestole, M.C.; Wolfe, N.D.; Fair, J.N.; et al. Genomic variability of monkeypox virus among humans, Democratic Republic of the Congo. Emerg. Infect. Dis. 2014, 20, 232–239. [CrossRef] Costello, V.; Sowash, M.; Gaur, A.; Cardis, M.; Pasieka, H.; Wortmann, G.; Ramdeen, S. Imported Monkeypox from International Traveler, Maryland, USA, 2021. Emerg. Infect. Dis. 2022, 28, 1002–1005. [CrossRef] Croft, D.R.; Sotir, M.J.; Williams, C.J.; Kazmierczak, J.J.; Wegner, M.V.; Rausch, D.; Graham, M.B.; Foldy, S.L.; Wolters, M.; Damon, I.K.; et al. Occupational risks during a monkeypox outbreak, Wisconsin, 2003. Emerg. Infect. Dis. 2007, 13, 1150–1157. [CrossRef] Yinka-Ogunleye, A.; Aruna, O.; Dalhat, M.; Ogoina, D.; McCollum, A.; Disu, Y.; Mamadu, I.; Akinpelu, A.; Ahmad, A.; Burga, J.; et al. Outbreak of human monkeypox in Nigeria in 2017-18: A clinical and epidemiological report. Lancet Infect. Dis. 2019, 19, 872–879. [CrossRef] [PubMed] Formenty, P.; Muntasir, M.O.; Damon, I.; Chowdhary, V.; Opoka, M.L.; Monimart, C.; Mutasim, E.M.; Manuguerra, J.C.; Davidson, W.B.; Karem, K.L.; et al. Human monkeypox outbreak caused by novel virus belonging to Congo Basin clade, Sudan, 2005. Emerg. Infect. Dis. 2010, 16, 1539–1545. [CrossRef] [PubMed] Hammerschlag, Y.; MacLeod, G.; Papadakis, G.; Adan Sanchez, A.; Druce, J.; Taiaroa, G.; Savic, I.; Mumford, J.; Roberts, J.; Caly, L.; et al. Monkeypox infection presenting as genital rash, Australia, May 2022. Eurosurveillance 2022, 27, 2200411. [CrossRef] [PubMed] Parrino, J.; Graham, B.S. Smallpox vaccines: Past, present, and future. J. Allergy Clin. Immunol. 2006, 118, 1320–1326. [CrossRef] [PubMed] Melamed, S.; Israely, T.; Paran, N. Challenges and Achievements in Prevention and Treatment of Smallpox. Vaccines 2018, 6, 8. [CrossRef] Li, Y.; Carroll, D.S.; Gardner, S.N.; Walsh, M.C.; Vitalis, E.A.; Damon, I.K. On the origin of smallpox: Correlating variola phylogenics with historical smallpox records. Proc. Natl. Acad. Sci. USA 2007, 104, 15787–15792. [CrossRef] Berche, P. Life and death of smallpox. La Presse Médicale 2022, 51, 104117. [CrossRef] Olson, V.A.; Shchelkunov, S.N. Are We Prepared in Case of a Possible Smallpox-Like Disease Emergence? Viruses 2017, 9, 242. [CrossRef] Nitsche, A.; Kurth, A.; Pauli, G. Viremia in human Cowpox virus infection. J. Clin. Virol. 2007, 40, 160–162. [CrossRef] Baxby, D.; Bennett, M.; Getty, B. Human cowpox 1969–93: A review based on 54 cases. Br. J. Dermatol. 1994, 131, 598–607. [CrossRef] Pether, J.V.S.; Trevains, P.H.; Harrison, S.R.B.; Baxby, D.; Bennett, M.; Gibb, A.P.R. Cowpox from cat to man. Lancet 1986, 1, 38–39. [CrossRef] [PubMed] Ninove, L.; Domart, Y.; Vervel, C.; Voinot, C.; Salez, N.; Raoult, D.; Meyer, H.; Capek, I.; Zandotti, C.; Charrel, R.N. Cowpox virus transmission from pet rats to humans, France. Emerg. Infect. Dis. 2009, 15, 781–784. [CrossRef] Campe, H.; Zimmermann, P.; Glos, K.; Bayer, M.; Bergemann, H.; Dreweck, C.; Graf, P.; Weber, B.K.; Meyer, H.; Büttner, M.; et al. Cowpox virus transmission from pet rats to humans, Germany. Emerg. Infect. Dis. 2009, 15, 777–780. [CrossRef] [PubMed] Mueller, N.H.; Gilden, D.H.; Cohrs, R.J.; Mahalingam, R.; Nagel, M.A. Varicella zoster virus infection: Clinical features, molecular pathogenesis of disease, and latency. Neurol. Clin. 2008, 26, 675–697. [CrossRef] [PubMed] Tan, M.G.; Beecker, J. Chickenpox in an elderly man. Can. Fam. Physician 2020, 66, e213–e215. Microorganisms 2023, 11, 1186 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48. 49. 24 of 31 Vairo, F.; Di Bari, V.; Panella, V.; Quintavalle, G.; Torchia, S.; Serra, M.C.; Sinopoli, M.T.; Lopalco, M.; Ceccarelli, G.; Ferraro, F.; et al. An outbreak of chickenpox in an asylum seeker centre in Italy: Outbreak investigation and validity of reported chickenpox history, December 2015–May 2016. Eurosurveillance 2017, 22, 17-00020. [CrossRef] Egawa, G.; Egawa, K.; Kabashima, K. Case of chickenpox in which varicella zoster virus genotype E was identified for the first time in Japan. J. Dermatol. 2020, 47, 54–57. [CrossRef] Andrade, S.M.C.; Haslett, M.I.C.; Malta, J.; Renoiner, E.I.M.; Lucena, A.R.F.; Fantinato, F.F.S.; Cruz, V.O.; Costa, C.S.D.; Santos, E.D.D. Chickenpox outbreak among Venezuelan immigrants housed in shelters and occupancies in the state of Roraima, Brazil, 2019: A descriptive study. Epidemiol. Serv. Saude 2021, 30, e2021156. [CrossRef] WHO. Monkeypox. 2022. Available online: https://www.who.int/news-room/fact-sheets/detail/monkeypox (accessed on 23 April 2023). Minhaj, F.S.; Ogale, Y.P.; Whitehill, F.; Schultz, J.; Foote, M.; Davidson, W.; Hughes, C.M.; Wilkins, K.; Bachmann, L.; Chatelain, R.; et al. Monkeypox Outbreak—Nine States, May 2022. MMWR Morb. Mortal. Wkly. Rep. 2022, 71, 764–769. [CrossRef] WHO. Laboratory Testing for the Monkeypox Virus: Interim Guidance. 23 May 2022. Available online: https://apps. who.int/iris/bitstream/handle/10665/354488/WHO-MPX-Laboratory-2022.1-eng.pdf?sequence=1&isAllowed=y (accessed on 1 September 2022). Reynolds, M.G.; Guagliardo, S.A.J.; Nakazawa, Y.J.; Doty, J.B.; Mauldin, M.R. Understanding orthopoxvirus host range and evolution: From the enigmatic to the usual suspects. Curr. Opin. Virol. 2018, 28, 108–115. [CrossRef] [PubMed] Balamurugan, V.; Venkatesan, G.; Bhanuprakash, V.; Singh, R.K. Camelpox, an emerging orthopox viral disease. Indian J. Virol. 2013, 24, 295–305. [CrossRef] [PubMed] Kinnunen, P.M.; Henttonen, H.; Hoffmann, B.; Kallio, E.R.; Korthase, C.; Laakkonen, J.; Niemimaa, J.; Palva, A.; Schlegel, M.; Ali, H.S.; et al. Orthopox virus infections in Eurasian wild rodents. Vector Borne Zoonotic Dis. 2011, 11, 1133–1140. [CrossRef] [PubMed] Pauli, G.; Blümel, J.; Burger, R.; Drosten, C.; Gröner, A.; Gürtler, L.; Heiden, M.; Hildebrandt, M.; Jansen, B.; Montag-Lessing, T.; et al. Orthopox Viruses: Infections in Humans. Transfus. Med. Hemother. 2010, 37, 351–364. [CrossRef] [PubMed] Beer, E.M.; Rao, V.B. A systematic review of the epidemiology of human monkeypox outbreaks and implications for outbreak strategy. PLoS Negl. Trop. Dis. 2019, 13, e0007791. [CrossRef] [PubMed] Benites-Zapata, V.A.; Ulloque-Badaracco, J.R.; Alarcon-Braga, E.A.; Hernandez-Bustamante, E.A.; Mosquera-Rojas, M.D.; Bonilla-Aldana, D.K.; Rodriguez-Morales, A.J. Clinical features, hospitalisation and deaths associated with monkeypox: A systematic review and meta-analysis. Ann. Clin. Microbiol. Antimicrob. 2022, 21, 36. [CrossRef] [PubMed] Besombes, C.; Mbrenga, F.; Schaeffer, L.; Malaka, C.; Gonofio, E.; Landier, J.; Vickos, U.; Konamna, X.; Selekon, B.; Dankpea, J.N.; et al. National Monkeypox Surveillance, Central African Republic, 2001–2021. Emerg. Infect. Dis. 2022, 28, 2435–2445. [CrossRef] Mbala, P.K.; Huggins, J.W.; Riu-Rovira, T.; Ahuka, S.M.; Mulembakani, P.; Rimoin, A.W.; Martin, J.W.; Muyembe, J.T. Maternal and Fetal Outcomes among Pregnant Women with Human Monkeypox Infection in the Democratic Republic of Congo. J. Infect. Dis. 2017, 216, 824–828. [CrossRef] D'Antonio, F.; Pagani, G.; Buca, D.; Khalil, A. Monkeypox infection in pregnancy: A systematic review and metaanalysis. Am. J. Obstet. Gynecol. MFM 2023, 5, 100747. [CrossRef] Adler, H.; Gould, S.; Hine, P.; Snell, L.B.; Wong, W.; Houlihan, C.F.; Osborne, J.C.; Rampling, T.; Beadsworth, M.B.; Duncan, C.J.; et al. Clinical features and management of human monkeypox: A retrospective observational study in the UK. Lancet Infect. Dis. 2022, 22, 1153–1162. [CrossRef] Ogoina, D.; Iroezindu, M.; James, H.I.; Oladokun, R.; Yinka-Ogunleye, A.; Wakama, P.; Otike-Odibi, B.; Usman, L.M.; Obazee, E.; Aruna, O.; et al. Clinical Course and Outcome of Human Monkeypox in Nigeria. Clin. Infect. Dis. 2020, 71, e210–e214. [CrossRef] [PubMed] Sejvar, J.J.; Chowdary, Y.; Schomogyi, M.; Stevens, J.; Patel, J.; Karem, K.; Fischer, M.; Kuehnert, M.J.; Zaki, S.R.; Paddock, C.D.; et al. Human monkeypox infection: A family cluster in the midwestern United States. J. Infect. Dis. 2004, 190, 1833–1840. [CrossRef] [PubMed] Shafaati, M.; Zandi, M. Monkeypox virus neurological manifestations in comparison to other orthopoxviruses. Travel. Med. Infect. Dis. 2022, 49, 102414. [CrossRef] [PubMed] Forni, D.; Molteni, C.; Cagliani, R.; Sironi, M. Geographic structuring and divergence time frame of monkeypox virus in the endemic region. J. Infect. Dis. 2022, 227, 742–751. [CrossRef] Yeh, T.Y.; Hsieh, Z.Y.; Feehley, M.C.; Feehley, P.J.; Contreras, G.P.; Su, Y.C.; Hsieh, S.L.; Lewis, D.A. Recombination shapes the 2022 monkeypox (mpox) outbreak. Med (N. Y.) 2022, 3, 824–826. [CrossRef] Thornhill, J.P.; Barkati, S.; Walmsley, S.; Rockstroh, J.; Antinori, A.; Harrison, L.B.; Palich, R.; Nori, A.; Reeves, I.; Habibi, M.S.; et al. Monkeypox Virus Infection in Humans across 16 Countries—April–June 2022. N. Engl. J. Med. 2022, 387, 679–691. [CrossRef] Kohli, R.M.; Isaacs, S.N. Mpox Evolution: Has the Current Outbreak Revealed a Pox on “U”? J. Infect. Dis. 2023, 227, 828–830. [CrossRef] Microorganisms 2023, 11, 1186 50. 51. 52. 53. 54. 55. 56. 57. 58. 59. 60. 61. 62. 63. 64. 65. 66. 67. 68. 69. 70. 71. 72. 73. 74. 75. 25 of 31 Isidro, J.; Borges, V.; Pinto, M.; Sobral, D.; Santos, J.D.; Nunes, A.; Mixão, V.; Ferreira, R.; Santos, D.; Duarte, S.; et al. Phylogenomic characterization and signs of microevolution in the 2022 multi-country outbreak of monkeypox virus. Nat. Med. 2022, 28, 1569–1572. [CrossRef] Magnus, P.v.; Andersen, E.K.; Petersen, K.B.; Birch-Andersen, A. A pox-like disease in Cynomolgus monkeys. Acta Pathol. Microbiol. Scand. 1959, 46, 156–176. [CrossRef] Tiee, M.S.; Harrigan, R.J.; Thomassen, H.A.; Smith, T.B. Ghosts of infections past: Using archival samples to understand a century of monkeypox virus prevalence among host communities across space and time. R. Soc. Open. Sci. 2018, 5, 171089. [CrossRef] Babkin, I.V.; Babkina, I.N.; Tikunova, N.V. An Update of Orthopoxvirus Molecular Evolution. Viruses 2022, 14, 388. [CrossRef] [PubMed] Babkin, I.V.; Babkina, I.N. A retrospective study of the orthopoxvirus molecular evolution. Infect. Genet. Evol. 2012, 12, 1597–1604. [CrossRef] [PubMed] WHO. WHO Recommends New Name for Monkeypox Disease; WHO: Geneva, Switzerland, 28 November 2022. Graham, F. Daily briefing: Mpox-a new name for monkeypox. Nature 2022. [CrossRef] Ladnyj, I.D.; Ziegler, P.; Kima, E. A human infection caused by monkeypox virus in Basankusu Territory, Democratic Republic of the Congo. Bull. World Health Organ. 1972, 46, 593–597. [PubMed] Marennikova, S.S.; Seluhina, E.M.; Mal’ceva, N.N.; Cimiskjan, K.L.; Macevic, G.R. Isolation and properties of the causal agent of a new variola-like disease (monkeypox) in man. Bull. World Health Organ. 1972, 46, 599–611. Lourie, B.; Bingham, P.G.; Evans, H.H.; Foster, S.O.; Nakano, J.H.; Herrmann, K.L. Human infection with monkeypox virus: Laboratory investigation of six cases in West Africa. Bull. World Health Organ. 1972, 46, 633–639. Breman, J.G.; Kalisa, R.; Steniowski, M.V.; Zanotto, E.; Gromyko, A.I.; Arita, I. Human monkeypox, 1970–1979. Bull. World Health Organ. 1980, 58, 165–182. Foster, S.O.; Brink, E.W.; Hutchins, D.L.; Pifer, J.M.; Lourie, B.; Moser, C.R.; Cummings, E.C.; Kuteyi, O.E.; Eke, R.E.; Titus, J.B.; et al. Human monkeypox. Bull. World Health Organ. 1972, 46, 569–576. Jezek, Z.; Marennikova, S.S.; Mutumbo, M.; Nakano, J.H.; Paluku, K.M.; Szczeniowski, M. Human monkeypox: A study of 2,510 contacts of 214 patients. J. Infect. Dis. 1986, 154, 551–555. [CrossRef] Levine, R.S.; Peterson, A.T.; Yorita, K.L.; Carroll, D.; Damon, I.K.; Reynolds, M.G. Ecological niche and geographic distribution of human monkeypox in Africa. PLoS ONE 2007, 2, e176. [CrossRef] Guarner, J.; Johnson, B.J.; Paddock, C.D.; Shieh, W.J.; Goldsmith, C.S.; Reynolds, M.G.; Damon, I.K.; Regnery, R.L.; Zaki, S.R. Monkeypox transmission and pathogenesis in prairie dogs. Emerg. Infect. Dis. 2004, 10, 426–431. [CrossRef] [PubMed] Reed, K.D.; Melski, J.W.; Graham, M.B.; Regnery, R.L.; Sotir, M.J.; Wegner, M.V.; Kazmierczak, J.J.; Stratman, E.J.; Li, Y.; Fairley, J.A.; et al. The detection of monkeypox in humans in the Western Hemisphere. N. Engl. J. Med. 2004, 350, 342–350. [CrossRef] [PubMed] Anderson, M.G.; Frenkel, L.D.; Homann, S.; Guffey, J. A case of severe monkeypox virus disease in an American child: Emerging infections and changing professional values. Pediatr. Infect. Dis. J. 2003, 22, 1093–1096; discussion 1096–1098. [CrossRef] [PubMed] Cunha, B.E. Monkeypox in the United States: An occupational health look at the first cases. AAOHN J. 2004, 52, 164–168. [CrossRef] Mauldin, M.R.; McCollum, A.M.; Nakazawa, Y.J.; Mandra, A.; Whitehouse, E.R.; Davidson, W.; Zhao, H.; Gao, J.; Li, Y.; Doty, J.; et al. Exportation of Monkeypox Virus From the African Continent. J. Infect. Dis. 2022, 225, 1367–1376. [CrossRef] Rao, A.K.; Schulte, J.; Chen, T.H.; Hughes, C.M.; Davidson, W.; Neff, J.M.; Markarian, M.; Delea, K.C.; Wada, S.; Liddell, A.; et al. Monkeypox in a Traveler Returning from Nigeria—Dallas, Texas, July 2021. MMWR Morb. Mortal. Wkly. Rep. 2022, 71, 509–516. [CrossRef] Yong, S.E.F.; Ng, O.T.; Ho, Z.J.M.; Mak, T.M.; Marimuthu, K.; Vasoo, S.; Yeo, T.W.; Ng, Y.K.; Cui, L.; Ferdous, Z.; et al. Imported Monkeypox, Singapore. Emerg. Infect. Dis. 2020, 26, 1826–1830. [CrossRef] Vaughan, A.; Aarons, E.; Astbury, J.; Balasegaram, S.; Beadsworth, M.; Beck, C.R.; Chand, M.; O’Connor, C.; Dunning, J.; Ghebrehewet, S.; et al. Two cases of monkeypox imported to the United Kingdom, September 2018. Eurosurveillance 2018, 23, 1800509. [CrossRef] Laiton-Donato, K.; Álvarez-Díaz, D.A.; Franco-Muñoz, C.; Ruiz-Moreno, H.A.; Rojas-Estévez, P.; Prada, A.; Rosales, A.; Ospina, M.L.; Mercado-Reyes, M. Monkeypox virus genome sequence from an imported human case in Colombia. Biomedica 2022, 42, 541–545. [CrossRef] Alcamí, A. Pathogenesis of the circulating mpox virus and its adaptation to humans. Proc. Natl. Acad. Sci. USA 2023, 120, e2301662120. [CrossRef] Vusirikala, A.; Charles, H.; Balasegaram, S.; Macdonald, N.; Kumar, D.; Barker-Burnside, C.; Cumiskey, K.; Dickinson, M.; Watson, M.; Olufon, O.; et al. Epidemiology of Early Monkeypox Virus Transmission in Sexual Networks of Gay and Bisexual Men, England, 2022. Emerg. Infect. Dis. 2022, 28, 2082–2086. [CrossRef] Patel, A.; Bilinska, J.; Tam, J.C.H.; Da Silva Fontoura, D.; Mason, C.Y.; Daunt, A.; Snell, L.B.; Murphy, J.; Potter, J.; Tuudah, C.; et al. Clinical features and novel presentations of human monkeypox in a central London centre during the 2022 outbreak: Descriptive case series. BMJ 2022, 378, e072410. [CrossRef] [PubMed] Microorganisms 2023, 11, 1186 76. 77. 78. 79. 80. 81. 82. 83. 84. 85. 86. 87. 88. 89. 90. 91. 92. 93. 94. 95. 96. 97. 98. 99. 26 of 31 Philpott, D.; Hughes, C.M.; Alroy, K.A.; Kerins, J.L.; Pavlick, J.; Asbel, L.; Crawley, A.; Newman, A.P.; Spencer, H.; Feldpausch, A.; et al. Epidemiologic and Clinical Characteristics of Monkeypox Cases—United States, May 17–July 22, 2022. MMWR Morb. Mortal. Wkly. Rep. 2022, 71, 1018–1022. [CrossRef] Vivancos, R.; Anderson, C.; Blomquist, P.; Balasegaram, S.; Bell, A.; Bishop, L.; Brown, C.S.; Chow, Y.; Edeghere, O.; Florence, I.; et al. Community transmission of monkeypox in the United Kingdom, April to May 2022. Eurosurveillance 2022, 27, 2200422. [CrossRef] [PubMed] Girometti, N.; Byrne, R.; Bracchi, M.; Heskin, J.; McOwan, A.; Tittle, V.; Gedela, K.; Scott, C.; Patel, S.; Gohil, J.; et al. Demographic and clinical characteristics of confirmed human monkeypox virus cases in individuals attending a sexual health centre in London, UK: An observational analysis. Lancet Infect. Dis. 2022, 22, 1321–1328. [CrossRef] Heskin, J.; Belfield, A.; Milne, C.; Brown, N.; Walters, Y.; Scott, C.; Bracchi, M.; Moore, L.S.; Mughal, N.; Rampling, T.; et al. Transmission of monkeypox virus through sexual contact—A novel route of infection. J. Infect. 2022, 85, 334–363. [CrossRef] WHO. 2022-23 Mpox (Monkeypox) Outbreak: Global Trends; WHO: Geneva, Switzerland, 28 February 2023. Altindis, M.; Puca, E.; Shapo, L. Diagnosis of monkeypox virus—An overview. Travel. Med. Infect. Dis. 2022, 50, 102459. [CrossRef] [PubMed] Page, M.J.; McKenzie, J.E.; Bossuyt, P.M.; Boutron, I.; Hoffmann, T.C.; Mulrow, C.D.; Shamseer, L.; Tetzlaff, J.M.; Akl, E.A.; Brennan, S.E.; et al. The PRISMA 2020 statement: An updated guideline for reporting systematic reviews. BMJ 2021, 372, n71. [CrossRef] [PubMed] Yinka-Ogunleye, A.; Aruna, O.; Ogoina, D.; Aworabhi, N.; Eteng, W.; Badaru, S.; Mohammed, A.; Agenyi, J.; Etebu, E.N.; Numbere, T.W.; et al. Reemergence of Human Monkeypox in Nigeria, 2017. Emerg. Infect. Dis. 2018, 24, 1149–1151. [CrossRef] [PubMed] Ligon, B.L. Monkeypox: A review of the history and emergence in the Western hemisphere. Semin. Pediatr. Infect. Dis. 2004, 15, 280–287. [CrossRef] Whitehouse, E.R.; Bonwitt, J.; Hughes, C.M.; Lushima, R.S.; Likafi, T.; Nguete, B.; Kabamba, J.; Monroe, B.; Doty, J.B.; Nakazawa, Y.; et al. Clinical and Epidemiological Findings from Enhanced Monkeypox Surveillance in Tshuapa Province, Democratic Republic of the Congo During 2011–2015. J. Infect. Dis. 2021, 223, 1870–1878. [CrossRef] [PubMed] Iñigo Martínez, J.; Gil Montalbán, E.; Jiménez Bueno, S.; Martín Martínez, F.; Nieto Juliá, A.; Sánchez Díaz, J.; García Marín, N.; Córdoba Deorador, E.; Nunziata Forte, A.; Alonso García, M.; et al. Monkeypox outbreak predominantly affecting men who have sex with men, Madrid, Spain, 26 April to 16 June 2022. Eurosurveillance 2022, 27, 2200471. [CrossRef] [PubMed] Osadebe, L.; Hughes, C.M.; Shongo Lushima, R.; Kabamba, J.; Nguete, B.; Malekani, J.; Pukuta, E.; Karhemere, S.; Muyembe Tamfum, J.J.; Wemakoy Okitolonda, E.; et al. Enhancing case definitions for surveillance of human monkeypox in the Democratic Republic of Congo. PLoS Negl. Trop. Dis. 2017, 11, e0005857. [CrossRef] Hoff, N.A.; Morier, D.S.; Kisalu, N.K.; Johnston, S.C.; Doshi, R.H.; Hensley, L.E.; Okitolonda-Wemakoy, E.; Muyembe-Tamfum, J.J.; Lloyd-Smith, J.O.; Rimoin, A.W. Varicella Coinfection in Patients with Active Monkeypox in the Democratic Republic of the Congo. Ecohealth 2017, 14, 564–574. [CrossRef] [PubMed] Kulesh, D.A.; Loveless, B.M.; Norwood, D.; Garrison, J.; Whitehouse, C.A.; Hartmann, C.; Mucker, E.; Miller, D.; Wasieloski, L.P., Jr.; Huggins, J.; et al. Monkeypox virus detection in rodents using real-time 30 -minor groove binder TaqMan assays on the Roche LightCycler. Lab. Investig. 2004, 84, 1200–1208. [CrossRef] Bayer-Garner, I.B. Monkeypox virus: Histologic, immunohistochemical and electron-microscopic findings. J. Cutan. Pathol. 2005, 32, 28–34. [CrossRef] Langohr, I.M.; Stevenson, G.W.; Thacker, H.L.; Regnery, R.L. Extensive lesions of monkeypox in a prairie dog (Cynomys sp). Vet. Pathol. 2004, 41, 702–707. [CrossRef] Kile, J.C.; Fleischauer, A.T.; Beard, B.; Kuehnert, M.J.; Kanwal, R.S.; Pontones, P.; Messersmith, H.J.; Teclaw, R.; Karem, K.L.; Braden, Z.H.; et al. Transmission of monkeypox among persons exposed to infected prairie dogs in Indiana in 2003. Arch. Pediatr. Adolesc. Med. 2005, 159, 1022–1025. [CrossRef] Jang, Y.R.; Lee, M.; Shin, H.; Kim, J.W.; Choi, M.M.; Kim, Y.M.; Lee, M.J.; Kim, J.; Na, H.K.; Kim, J.Y. The First Case of Monkeypox in the Republic of Korea. J. Korean Med. Sci. 2022, 37, e224. [CrossRef] Patrocinio-Jesus, R.; Peruzzu, F. Monkeypox Genital Lesions. N. Engl. J. Med. 2022, 387, 66. [CrossRef] Moschese, D.; Pozza, G.; Mileto, D.; Giacomelli, A.; Cutrera, M.; Cossu, M.V.; Matone, M.; Beltrami, M.; Salari, F.; Antinori, S.; et al. Isolation of viable monkeypox virus from anal and urethral swabs, Italy, May to July 2022. Eurosurveillance 2022, 27, 2200675. [CrossRef] [PubMed] Orviz, E.; Negredo, A.; Ayerdi, O.; Vázquez, A.; Muñoz-Gomez, A.; Monzón, S.; Clavo, P.; Zaballos, A.; Vera, M.; Sánchez, P.; et al. Monkeypox outbreak in Madrid (Spain): Clinical and virological aspects. J. Infect. 2022, 85, 412–417. [CrossRef] [PubMed] Hernandez, L.E.; Jadoo, A.; Kirsner, R.S. Human monkeypox virus infection in an immunocompromised man: Trial with tecovirimat. Lancet 2022, 400, e8. [CrossRef] [PubMed] De Baetselier, I.; Van Dijck, C.; Kenyon, C.; Coppens, J.; Michiels, J.; de Block, T.; Smet, H.; Coppens, S.; Vanroye, F.; Bugert, J.J.; et al. Retrospective detection of asymptomatic monkeypox virus infections among male sexual health clinic attendees in Belgium. Nat. Med. 2022, 28, 2288–2292. [CrossRef] Noe, S.; Zange, S.; Seilmaier, M.; Antwerpen, M.H.; Fenzl, T.; Schneider, J.; Spinner, C.D.; Bugert, J.J.; Wendtner, C.M.; Wölfel, R. Clinical and virological features of first human monkeypox cases in Germany. Infection 2022, 51, 265–270. [CrossRef] Microorganisms 2023, 11, 1186 27 of 31 100. Meduri, E.; Malclès, A.; Kecik, M. Conjunctivitis with Monkeypox Virus Positive Conjunctival Swabs. Ophthalmology 2022, 129, 1095. [CrossRef] 101. Likos, A.M.; Sammons, S.A.; Olson, V.A.; Frace, A.M.; Li, Y.; Olsen-Rasmussen, M.; Davidson, W.; Galloway, R.; Khristova, M.L.; Reynolds, M.G.; et al. A tale of two clades: Monkeypox viruses. J. Gen. Virol. 2005, 86, 2661–2672. [CrossRef] 102. Nakazawa, Y.; Mauldin, M.R.; Emerson, G.L.; Reynolds, M.G.; Lash, R.R.; Gao, J.; Zhao, H.; Li, Y.; Muyembe, J.J.; Kingebeni, P.M.; et al. A phylogeographic investigation of African monkeypox. Viruses 2015, 7, 2168–2184. [CrossRef] 103. Luna, N.; Ramírez, A.L.; Muñoz, M.; Ballesteros, N.; Patiño, L.H.; Castañeda, S.A.; Bonilla-Aldana, D.K.; Paniz-Mondolfi, A.; Ramírez, J.D. Phylogenomic analysis of the monkeypox virus (MPXV) 2022 outbreak: Emergence of a novel viral lineage? Travel. Med. Infect. Dis. 2022, 49, 102402. [CrossRef] 104. Yadav, P.D.; Reghukumar, A.; Sahay, R.R.; Sudeep, K.; Shete, A.M.; Raman, A.; Vk, P.; Abraham, P.; Benson, R.; Sm, S.; et al. First two cases of Monkeypox virus infection in travellers returned from UAE to India, July 2022. J. Infect. 2022, 85, e145–e148. [CrossRef] 105. Atkinson, B.; Burton, C.; Pottage, T.; Thompson, K.A.; Ngabo, D.; Crook, A.; Pitman, J.; Summers, S.; Lewandowski, K.; Furneaux, J.; et al. Infection-competent monkeypox virus contamination identified in domestic settings following an imported case of monkeypox into the UK. Environ. Microbiol. 2022, 24, 4561–4569. [CrossRef] [PubMed] 106. Radonić, A.; Metzger, S.; Dabrowski, P.W.; Couacy-Hymann, E.; Schuenadel, L.; Kurth, A.; Mätz-Rensing, K.; Boesch, C.; Leendertz, F.H.; Nitsche, A. Fatal monkeypox in wild-living sooty mangabey, Côte d'Ivoire, 2012. Emerg. Infect. Dis. 2014, 20, 1009–1011. [CrossRef] [PubMed] 107. Patrono, L.V.; Pléh, K.; Samuni, L.; Ulrich, M.; Röthemeier, C.; Sachse, A.; Muschter, S.; Nitsche, A.; Couacy-Hymann, E.; Boesch, C.; et al. Monkeypox virus emergence in wild chimpanzees reveals distinct clinical outcomes and viral diversity. Nat. Microbiol. 2020, 5, 955–965. [CrossRef] 108. Sadeuh-Mba, S.A.; Yonga, M.G.; Els, M.; Batejat, C.; Eyangoh, S.; Caro, V.; Etoundi, A.; Carniel, E.; Njouom, R. Monkeypox virus phylogenetic similarities between a human case detected in Cameroon in 2018 and the 2017–2018 outbreak in Nigeria. Infect. Genet. Evol. 2019, 69, 8–11. [CrossRef] [PubMed] 109. Ye, F.; Song, J.; Zhao, L.; Zhang, Y.; Xia, L.; Zhu, L.; Kamara, I.L.; Ren, J.; Wang, W.; Tian, H.; et al. Molecular Evidence of Human Monkeypox Virus Infection, Sierra Leone. Emerg. Infect. Dis. 2019, 25, 1220–1222. [CrossRef] [PubMed] 110. Erez, N.; Achdout, H.; Milrot, E.; Schwartz, Y.; Wiener-Well, Y.; Paran, N.; Politi, B.; Tamir, H.; Israely, T.; Weiss, S.; et al. Diagnosis of Imported Monkeypox, Israel, 2018. Emerg. Infect. Dis. 2019, 25, 980–983. [CrossRef] [PubMed] 111. Marien, J.; Laudisoit, A.; Patrono, L.; Baelo, P.; van Vredendaal, R.; Akawa, P.; Tungaluna, G.-C.; Mande, C.; Ngoy, S.; Mussaw, M.; et al. Monkeypox viruses circulate in distantly-related small mammal species in the Democratic Republic of the Congo. Biology 2021. [CrossRef] 112. Berthet, N.; Descorps-Declère, S.; Besombes, C.; Curaudeau, M.; Nkili Meyong, A.A.; Selekon, B.; Labouba, I.; Gonofio, E.C.; Ouilibona, R.S.; Simo Tchetgna, H.D.; et al. Genomic history of human monkey pox infections in the Central African Republic between 2001 and 2018. Sci. Rep. 2021, 11, 13085. [CrossRef] 113. Hutin, Y.J.; Williams, R.J.; Malfait, P.; Pebody, R.; Loparev, V.N.; Ropp, S.L.; Rodriguez, M.; Knight, J.C.; Tshioko, F.K.; Khan, A.S.; et al. Outbreak of human monkeypox, Democratic Republic of Congo, 1996 to 1997. Emerg. Infect. Dis. 2001, 7, 434–438. [CrossRef] 114. McCollum, A.M.; Nakazawa, Y.; Ndongala, G.M.; Pukuta, E.; Karhemere, S.; Lushima, R.S.; Ilunga, B.K.; Kabamba, J.; Wilkins, K.; Gao, J.; et al. Human Monkeypox in the Kivus, a Conflict Region of the Democratic Republic of the Congo. Am. J. Trop. Med. Hyg. 2015, 93, 718–721. [CrossRef] 115. Rimoin, A.W.; Kisalu, N.; Kebela-Ilunga, B.; Mukaba, T.; Wright, L.L.; Formenty, P.; Wolfe, N.D.; Shongo, R.L.; Tshioko, F.; Okitolonda, E.; et al. Endemic human monkeypox, Democratic Republic of Congo, 2001–2004. Emerg. Infect. Dis. 2007, 13, 934–937. [CrossRef] [PubMed] 116. Antinori, A.; Mazzotta, V.; Vita, S.; Carletti, F.; Tacconi, D.; Lapini, L.E.; D’Abramo, A.; Cicalini, S.; Lapa, D.; Pittalis, S.; et al. Epidemiological, clinical and virological characteristics of four cases of monkeypox support transmission through sexual contact, Italy, May 2022. Eurosurveillance 2022, 27, 2200421. [CrossRef] [PubMed] 117. Perez Duque, M.; Ribeiro, S.; Martins, J.V.; Casaca, P.; Leite, P.P.; Tavares, M.; Mansinho, K.; Duque, L.M.; Fernandes, C.; Cordeiro, R.; et al. Ongoing monkeypox virus outbreak, Portugal, 29 April to 23 May 2022. Eurosurveillance 2022, 27, 2200424. [CrossRef] 118. Vandenbogaert, M.; Kwasiborski, A.; Gonofio, E.; Descorps-Declère, S.; Selekon, B.; Nkili Meyong, A.A.; Ouilibona, R.S.; Gessain, A.; Manuguerra, J.-C.; Caro, V.; et al. Nanopore sequencing of a monkeypox virus strain isolated from a pustular lesion in the Central African Republic. Sci. Rep. 2022, 12, 10768. [CrossRef] [PubMed] 119. Claro, I.M.; Romano, C.M.; Candido, D.D.S.; Lima, E.L.; Lindoso, J.A.L.; Ramundo, M.S.; Moreira, F.R.R.; Barra, L.A.C.; Borges, L.M.S.; Medeiros, L.A.; et al. Shotgun metagenomic sequencing of the first case of monkeypox virus in Brazil, 2022. Rev. Inst. Med. Trop. Sao Paulo 2022, 64, e48. [CrossRef] 120. Reynolds, M.G.; Carroll, D.S.; Olson, V.A.; Hughes, C.; Galley, J.; Likos, A.; Montgomery, J.M.; Suu-Ire, R.; Kwasi, M.O.; Jeffrey Root, J.; et al. A silent enzootic of an orthopoxvirus in Ghana, West Africa: Evidence for multi-species involvement in the absence of widespread human disease. Am. J. Trop. Med. Hyg. 2010, 82, 746–754. [CrossRef] Microorganisms 2023, 11, 1186 28 of 31 121. Learned, L.A.; Reynolds, M.G.; Wassa, D.W.; Li, Y.; Olson, V.A.; Karem, K.; Stempora, L.L.; Braden, Z.H.; Kline, R.; Likos, A.; et al. Extended interhuman transmission of monkeypox in a hospital community in the Republic of the Congo, 2003. Am. J. Trop. Med. Hyg. 2005, 73, 428–434. [CrossRef] 122. Doty, J.B.; Malekani, J.M.; Kalemba, L.N.; Stanley, W.T.; Monroe, B.P.; Nakazawa, Y.U.; Mauldin, M.R.; Bakambana, T.L.; Liyandja Dja Liyandja, T.; Braden, Z.H.; et al. Assessing Monkeypox Virus Prevalence in Small Mammals at the Human-Animal Interface in the Democratic Republic of the Congo. Viruses 2017, 9, 283. [CrossRef] 123. MacNeil, A.; Reynolds, M.G.; Carroll, D.S.; Karem, K.; Braden, Z.; Lash, R.; Moundeli, A.; Mombouli, J.V.; Jumaan, A.O.; Schmid, D.S.; et al. Monkeypox or varicella? Lessons from a rash outbreak investigation in the Republic of the Congo. Am. J. Trop. Med. Hyg. 2009, 80, 503–507. [CrossRef] 124. Karem, K.L.; Reynolds, M.; Braden, Z.; Lou, G.; Bernard, N.; Patton, J.; Damon, I.K. characterization of acute-phase humoral immunity to monkeypox: Use of immunoglobulin M enzyme-linked immunosorbent assay for detection of monkeypox infection during the 2003 North American outbreak. Clin. Diagn. Lab. Immunol. 2005, 12, 867–872. [CrossRef] 125. Breman, J.G.; Bernadou, J.; Nakano, J.H. Poxvirus in West African nonhuman primates: Serological survey results. Bull. World Health Organ. 1977, 55, 605–612. 126. Gispen, R.; Brand-Saathof, B.B.; Hekker, A.C. Monkeypox-specific antibodies in human and simian sera from the Ivory Coast and Nigeria. Bull. World Health Organ. 1976, 53, 355–360. [PubMed] 127. Maksyutov, R.A.; Gavrilova, E.V.; Shchelkunov, S.N. Species-specific differentiation of variola, monkeypox, and varicella-zoster viruses by multiplex real-time PCR assay. J. Virol. Methods 2016, 236, 215–220. [CrossRef] [PubMed] 128. Li, D.; Wilkins, K.; McCollum, A.M.; Osadebe, L.; Kabamba, J.; Nguete, B.; Likafi, T.; Balilo, M.P.; Lushima, R.S.; Malekani, J.; et al. Evaluation of the GeneXpert for Human Monkeypox Diagnosis. Am. J. Trop. Med. Hyg. 2017, 96, 405–410. [CrossRef] [PubMed] 129. Esposito, J.J.; Obijeski, J.F.; Nakano, J.H. Serological relatedness of monkeypox, variola, and vaccinia viruses. J. Med. Virol. 1977, 1, 35–47. [CrossRef] 130. Douglas, K.O.; Cayol, C.; Forbes, K.M.; Samuels, T.A.; Vapalahti, O.; Sironen, T.; Gittens-St Hilaire, M. Serological Evidence of Multiple Zoonotic Viral Infections among Wild Rodents in Barbados. Pathogens 2021, 10, 663. [CrossRef] 131. Gilchuk, I.; Gilchuk, P.; Sapparapu, G.; Lampley, R.; Singh, V.; Kose, N.; Blum, D.L.; Hughes, L.J.; Satheshkumar, P.S.; Townsend, M.B.; et al. Cross-Neutralizing and Protective Human Antibody Specificities to Poxvirus Infections. Cell 2016, 167, 684–694.e689. [CrossRef] 132. Walls, H.H.; Ziegler, D.W.; Nakano, J.H. Characterization of antibodies to orthopoxviruses in human sera by radioimmunoassay. Bull. World Health Organ. 1981, 59, 253–262. 133. Macneil, A.; Abel, J.; Reynolds, M.G.; Lash, R.; Fonnie, R.; Kanneh, L.D.; Robert, W.; Lungay, V.K.; Goba, A.; Moses, L.M.; et al. Serologic evidence of human orthopoxvirus infections in Sierra Leone. BMC Res. Notes 2011, 4, 465. [CrossRef] 134. Guagliardo, S.A.J.; Monroe, B.; Moundjoa, C.; Athanase, A.; Okpu, G.; Burgado, J.; Townsend, M.B.; Satheshkumar, P.S.; Epperson, S.; Doty, J.B.; et al. Asymptomatic Orthopoxvirus Circulation in Humans in the Wake of a Monkeypox Outbreak among Chimpanzees in Cameroon. Am. J. Trop. Med. Hyg. 2020, 102, 206–212. [CrossRef] 135. Stienlauf, S.; Shoresh, M.; Solomon, A.; Lublin-Tennenbaum, T.; Atsmon, Y.; Meirovich, Y.; Katz, E. Kinetics of formation of neutralizing antibodies against vaccinia virus following re-vaccination. Vaccine 1999, 17, 201–204. [CrossRef] [PubMed] 136. Jezek, Z.; Nakano, J.H.; Arita, I.; Mutombo, M.; Szczeniowski, M.; Dunn, C. Serological survey for human monkeypox infections in a selected population in Zaire. J. Trop. Med. Hyg. 1987, 90, 31–38. [PubMed] 137. Leland, D.S.; Ginocchio, C.C. Role of cell culture for virus detection in the age of technology. Clin. Microbiol. Rev. 2007, 20, 49–78. [CrossRef] [PubMed] 138. Lapa, D.; Carletti, F.; Mazzotta, V.; Matusali, G.; Pinnetti, C.; Meschi, S.; Gagliardini, R.; Colavita, F.; Mondi, A.; Minosse, C.; et al. Monkeypox virus isolation from a semen sample collected in the early phase of infection in a patient with prolonged seminal viral shedding. Lancet Infect. Dis. 2022, 22, 1267–1269. [CrossRef] 139. Saijo, M.; Ami, Y.; Suzaki, Y.; Nagata, N.; Iwata, N.; Hasegawa, H.; Iizuka, I.; Shiota, T.; Sakai, K.; Ogata, M.; et al. Virulence and pathophysiology of the Congo Basin and West African strains of monkeypox virus in non-human primates. J. Gen. Virol. 2009, 90, 2266–2271. [CrossRef] [PubMed] 140. Kim, J.W.; Lee, M.; Shin, H.; Choi, C.H.; Choi, M.M.; Kim, J.W.; Yi, H.; Yoo, C.K.; Rhie, G.E. Isolation and identification of monkeypox virus MPXV-ROK-P1-2022 from the first case in the Republic of Korea. Osong Public. Health Res. Perspect. 2022, 13, 308–311. [CrossRef] 141. Mutombo, M.; Arita, I.; Jezek, Z. Human monkeypox transmitted by a chimpanzee in a tropical rain-forest area of Zaire. Lancet 1983, 1, 735–737. [CrossRef] 142. Meyer, H.; Perrichot, M.; Stemmler, M.; Emmerich, P.; Schmitz, H.; Varaine, F.; Shungu, R.; Tshioko, F.; Formenty, P. Outbreaks of disease suspected of being due to human monkeypox virus infection in the Democratic Republic of Congo in 2001. J. Clin. Microbiol. 2002, 40, 2919–2921. [CrossRef] 143. Mukinda, V.B.; Mwema, G.; Kilundu, M.; Heymann, D.L.; Khan, A.S.; Esposito, J.J. Re-emergence of human monkeypox in Zaire in 1996. Monkeypox Epidemiologic Working Group. Lancet 1997, 349, 1449–1450. [CrossRef] 144. Sukswai, N.; Khoury, J.D. Immunohistochemistry Innovations for Diagnosis and Tissue-Based Biomarker Detection. Curr. Hematol. Malig. Rep. 2019, 14, 368–375. [CrossRef] Microorganisms 2023, 11, 1186 29 of 31 145. Paran, N.; Yahalom-Ronen, Y.; Shifman, O.; Lazar, S.; Ben-Ami, R.; Yakubovsky, M.; Levy, I.; Wieder-Feinsod, A.; Amit, S.; Katzir, M.; et al. Monkeypox DNA levels correlate with virus infectivity in clinical samples, Israel, 2022. Eurosurveillance 2022, 27, 2200636. [CrossRef] [PubMed] 146. Pfäfflin, F.; Wendisch, D.; Scherer, R.; Jürgens, L.; Godzick-Njomgang, G.; Tranter, E.; Tober-Lau, P.; Stegemann, M.S.; Corman, V.M.; Kurth, F.; et al. Monkeypox in-patients with severe anal pain. Infection 2022, 51, 483–487. [CrossRef] [PubMed] 147. Dimitrakoff, J. Monkeypox Virus Infection across 16 Countries—April–June 2022. N. Engl. J. Med. 2022, 387, e69. [CrossRef] [PubMed] 148. Paparizos, V.; Nicolaidou, E.; Tryfinopoulou, K.; Papa, A.; Rigopoulos, D.; Tsiodras, S.; Stratigos, A. Monkeypox virus infection: First reported case in Greece in a patient with a genital rash. J. Eur. Acad. Dermatol. Venereol. 2022, 37, e350–e351. [CrossRef] [PubMed] 149. Goldberg, T.L.; Chapman, C.A.; Cameron, K.; Saj, T.; Karesh, W.B.; Wolfe, N.D.; Wong, S.W.; Dubois, M.E.; Slifka, M.K. Serologic evidence for novel poxvirus in endangered red Colobus monkeys, Western Uganda. Emerg. Infect. Dis. 2008, 14, 801–803. [CrossRef] 150. Orba, Y.; Sasaki, M.; Yamaguchi, H.; Ishii, A.; Thomas, Y.; Ogawa, H.; Hang’ombe, B.M.; Mweene, A.S.; Morikawa, S.; Saijo, M.; et al. Orthopoxvirus infection among wildlife in Zambia. J. Gen. Virol. 2015, 96, 390–394. [CrossRef] 151. Khodakevich, L.; Jezek, Z.; Kinzanzka, K. Isolation of monkeypox virus from wild squirrel infected in nature. Lancet 1986, 1, 98–99. [CrossRef] 152. Marennikova, S.S.; Seluhina, E.M.; Mal’ceva, N.N.; Ladnyj, I.D. Poxviruses isolated from clinically ill and asymptomatically infected monkeys and a chimpanzee. Bull. World Health Organ. 1972, 46, 613–620. 153. Hutson, C.L.; Lee, K.N.; Abel, J.; Carroll, D.S.; Montgomery, J.M.; Olson, V.A.; Li, Y.; Davidson, W.; Hughes, C.; Dillon, M.; et al. Monkeypox zoonotic associations: Insights from laboratory evaluation of animals associated with the multi-state US outbreak. Am. J. Trop. Med. Hyg. 2007, 76, 757–768. [CrossRef] 154. Seang, S.; Burrel, S.; Todesco, E.; Leducq, V.; Monsel, G.; Le Pluart, D.; Cordevant, C.; Pourcher, V.; Palich, R. Evidence of human-to-dog transmission of monkeypox virus. Lancet 2022, 400, 658–659. [CrossRef] 155. Schroeder, K.; Nitsche, A. Multicolour, multiplex real-time PCR assay for the detection of human-pathogenic poxviruses. Mol. Cell. Probes 2010, 24, 110–113. [CrossRef] [PubMed] 156. Li, Y.; Olson, V.A.; Laue, T.; Laker, M.T.; Damon, I.K. Detection of monkeypox virus with real-time PCR assays. J. Clin. Virol. 2006, 36, 194–203. [CrossRef] [PubMed] 157. Li, Y.; Zhao, H.; Wilkins, K.; Hughes, C.; Damon, I.K. Real-time PCR assays for the specific detection of monkeypox virus West African and Congo Basin strain DNA. J. Virol. Methods 2010, 169, 223–227. [CrossRef] [PubMed] 158. Kalthan, E.; Tenguere, J.; Ndjapou, S.G.; Koyazengbe, T.A.; Mbomba, J.; Marada, R.M.; Rombebe, P.; Yangueme, P.; Babamingui, M.; Sambella, A.; et al. Investigation of an outbreak of monkeypox in an area occupied by armed groups, Central African Republic. Med. Mal. Infect. 2018, 48, 263–268. [CrossRef] 159. La Rosa, G.; Mancini, P.; Veneri, C.; Ferraro, G.B.; Lucentini, L.; Iaconelli, M.; Suffredini, E. Detection of Monkeypox Virus DNA in Airport Wastewater, Rome, Italy. Emerg. Infect. Dis. 2023, 29, 193–196. [CrossRef] [PubMed] 160. Shchelkunov, S.N.; Gavrilova, E.V.; Babkin, I.V. Multiplex PCR detection and species differentiation of orthopoxviruses pathogenic to humans. Mol. Cell. Probes 2005, 19, 1–8. [CrossRef] [PubMed] 161. Neubauer, H.; Reischl, U.; Ropp, S.; Esposito, J.J.; Wolf, H.; Meyer, H. Specific detection of monkeypox virus by polymerase chain reaction. J. Virol. Methods 1998, 74, 201–207. [CrossRef] [PubMed] 162. Kubista, M.; Andrade, J.M.; Bengtsson, M.; Forootan, A.; Jonák, J.; Lind, K.; Sindelka, R.; Sjöback, R.; Sjögreen, B.; Strömbom, L.; et al. The real-time polymerase chain reaction. Mol. Aspects Med. 2006, 27, 95–125. [CrossRef] 163. Holland, P.M.; Abramson, R.D.; Watson, R.; Gelfand, D.H. Detection of specific polymerase chain reaction product by utilizing the 50 —-30 exonuclease activity of Thermus aquaticus DNA polymerase. Proc. Natl. Acad. Sci. USA 1991, 88, 7276–7280. [CrossRef] 164. Arya, M.; Shergill, I.S.; Williamson, M.; Gommersall, L.; Arya, N.; Patel, H.R. Basic principles of real-time quantitative PCR. Expert. Rev. Mol. Diagn. 2005, 5, 209–219. [CrossRef] 165. Cardullo, R.A.; Agrawal, S.; Flores, C.; Zamecnik, P.C.; Wolf, D.E. Detection of nucleic acid hybridization by nonradiative fluorescence resonance energy transfer. Proc. Natl. Acad. Sci. USA 1988, 85, 8790–8794. [CrossRef] [PubMed] 166. Kutyavin, I.V.; Afonina, I.A.; Mills, A.; Gorn, V.V.; Lukhtanov, E.A.; Belousov, E.S.; Singer, M.J.; Walburger, D.K.; Lokhov, S.G.; Gall, A.A.; et al. 30 -minor groove binder-DNA probes increase sequence specificity at PCR extension temperatures. Nucleic Acids Res. 2000, 28, 655–661. [CrossRef] 167. Martins-Filho, P.R.; Tanajura, D.M.; Alves Dos Santos, C. Polymerase chain reaction positivity and cycle threshold values in biological samples from patients with monkeypox: A meta-analysis. Travel. Med. Infect. Dis. 2022, 50, 102448. [CrossRef] 168. Ma, A.; Langer, J.; Hanson, K.E.; Bradley, B.T. Characterization of the Cytopathic Effects of Monkeypox Virus Isolated from Clinical Specimens and Differentiation from Common Viral Exanthems. J. Clin. Microbiol. 2022, 60, e0133622. [CrossRef] [PubMed] 169. Shchelkunov, S.N.; Totmenin, A.V.; Safronov, P.F.; Mikheev, M.V.; Gutorov, V.V.; Ryazankina, O.I.; Petrov, N.A.; Babkin, I.V.; Uvarova, E.A.; Sandakhchiev, L.S.; et al. Analysis of the monkeypox virus genome. Virology 2002, 297, 172–194. [CrossRef] [PubMed] Microorganisms 2023, 11, 1186 30 of 31 170. Dumont, C.; Irenge, L.M.; Magazani, E.K.; Garin, D.; Muyembe, J.J.; Bentahir, M.; Gala, J.L. Simple technique for in field samples collection in the cases of skin rash illness and subsequent PCR detection of orthopoxviruses and varicella zoster virus. PLoS ONE 2014, 9, e96930. [CrossRef] [PubMed] 171. Nolen, L.D.; Osadebe, L.; Katomba, J.; Likofata, J.; Mukadi, D.; Monroe, B.; Doty, J.; Hughes, C.M.; Kabamba, J.; Malekani, J.; et al. Extended Human-to-Human Transmission during a Monkeypox Outbreak in the Democratic Republic of the Congo. Emerg. Infect. Dis. 2016, 22, 1014–1021. [CrossRef] [PubMed] 172. Aguilera-Alonso, D.; Alonso-Cadenas, J.A.; Roguera-Sopena, M.; Lorusso, N.; Miguel, L.G.S.; Calvo, C. Monkeypox virus infections in children in Spain during the first months of the 2022 outbreak. Lancet Child. Adolesc. Health 2022, 6, e22–e23. [CrossRef] 173. Jezek, Z.; Arita, I.; Mutombo, M.; Dunn, C.; Nakano, J.H.; Szczeniowski, M. Four generations of probable person-to-person transmission of human monkeypox. Am. J. Epidemiol. 1986, 123, 1004–1012. [CrossRef] 174. de Nicolas-Ruanes, B.; Vivancos, M.J.; Azcarraga-Llobet, C.; Moreno, A.M.; Rodriguez-Dominguez, M.; Berna-Rico, E.D.; Garcia-Mouronte, E.; Carron-Herrero, A.; McGee, A.; Galan, J.C.; et al. Monkeypox virus case with maculopapular exanthem and proctitis during the Spanish outbreak in 2022. J. Eur. Acad. Dermatol. Venereol. 2022, 36, e658–e660. [CrossRef] 175. Nolasco, S.; Vitale, F.; Geremia, A.; Tramuto, F.; Maida, C.M.; Sciuto, A.; Coco, C.; Manuele, R.; Frasca, E.; Frasca, M.; et al. First case of monkeypox virus, SARS-CoV-2 and HIV co-infection. J. Infect. 2023, 86, e21–e23. [CrossRef] [PubMed] 176. Tarín-Vicente, E.J.; Alemany, A.; Agud-Dios, M.; Ubals, M.; Suñer, C.; Antón, A.; Arando, M.; Arroyo-Andrés, J.; Calderón-Lozano, L.; Casañ, C.; et al. Clinical presentation and virological assessment of confirmed human monkeypox virus cases in Spain: A prospective observational cohort study. Lancet 2022, 400, 661–669. [CrossRef] [PubMed] 177. Pipitò, L.; Cascio, A. Monkeypox virus infection and creatine phosphokinase increase: A case from Italy. Travel. Med. Infect. Dis. 2022, 50, 102412. [CrossRef] 178. Peiró-Mestres, A.; Fuertes, I.; Camprubí-Ferrer, D.; Marcos, M.; Vilella, A.; Navarro, M.; Rodriguez-Elena, L.; Riera, J.; Català, A.; Martínez, M.J.; et al. Frequent detection of monkeypox virus DNA in saliva, semen, and other clinical samples from 12 patients, Barcelona, Spain, May to June 2022. Eurosurveillance 2022, 27, 2200503. [CrossRef] 179. Vallée, A.; Farfour, E.; Zucman, D. Monkeypox virus: A novel sexually transmitted disease? A case report from France. Travel. Med. Infect. Dis. 2022, 49, 102394. [CrossRef] [PubMed] 180. Boesecke, C.; Monin, M.B.; van Bremen, K.; Schlabe, S.; Hoffmann, C. Severe monkeypox-virus infection in undiagnosed advanced HIV infection. Infection 2022, 50, 1633–1634. [CrossRef] 181. Vaughan, A.; Aarons, E.; Astbury, J.; Brooks, T.; Chand, M.; Flegg, P.; Hardman, A.; Harper, N.; Jarvis, R.; Mawdsley, S.; et al. Human-to-Human Transmission of Monkeypox Virus, United Kingdom, October 2018. Emerg. Infect. Dis. 2020, 26, 782–785. [CrossRef] 182. Okareh, O.T.; Morakinyo, O. Monkeypox in Nigeria: A case report of re-emerged disease outbreak. J. Microbiol. Exp. 2018, 6, 89–91. [CrossRef] 183. Miura, F.; van Ewijk, C.E.; Backer, J.A.; Xiridou, M.; Franz, E.; Op de Coul, E.; Brandwagt, D.; van Cleef, B.; van Rijckevorsel, G.; Swaan, C.; et al. Estimated incubation period for monkeypox cases confirmed in the Netherlands, May 2022. Eurosurveillance 2022, 27, 2200448. [CrossRef] 184. Reynolds, M.G.; Yorita, K.L.; Kuehnert, M.J.; Davidson, W.B.; Huhn, G.D.; Holman, R.C.; Damon, I.K. Clinical manifestations of human monkeypox influenced by route of infection. J. Infect. Dis. 2006, 194, 773–780. [CrossRef] 185. Turco, M.; Mancuso, F.R.; Pisano, L. A monkeypox virus infection mimicking primary syphilis. Br. J. Dermatol. 2022, 187, e194–e195. [CrossRef] [PubMed] 186. Chauhan, R.P.; Gordon, M.L. A Systematic Review Analyzing the Prevalence and Circulation of Influenza Viruses in Swine Population Worldwide. Pathogens 2020, 9, 355. [CrossRef] [PubMed] 187. Chauhan, R.P.; Gordon, M.L. Deciphering transmission dynamics and spillover of avian influenza viruses from avian species to swine populations globally. Virus Genes 2021, 57, 541–555. [CrossRef] [PubMed] 188. Chauhan, R.P.; Gordon, M.L. A systematic review of influenza A virus prevalence and transmission dynamics in backyard swine populations globally. Porc. Health Manag. 2022, 8, 10. [CrossRef] 189. Chauhan, R.P.; Gordon, M.L. Review of genome sequencing technologies in molecular characterization of influenza A viruses in swine. J. Vet. Diagn. Investig. 2022, 34, 177–189. [CrossRef] 190. Chauhan, R.P.; Gordon, M.L. An overview of influenza A virus genes, protein functions, and replication cycle highlighting important updates. Virus Genes 2022, 58, 255–269. [CrossRef] 191. Chauhan, R.P.; San, J.E.; Gordon, M.L. Metagenomic Analysis of RNA Fraction Reveals the Diversity of Swine Oral Virome on South African Backyard Swine Farms in the uMgungundlovu District of KwaZulu-Natal Province. Pathogens 2022, 11, 927. [CrossRef] 192. Chauhan, R.P.; Gordon, M.L. Characterization of a Near Full-Length Hepatitis E Virus Genome of Subtype 3c Generated from Naturally Infected South African Backyard Pigs. Pathogens 2022, 11, 1030. [CrossRef] 193. Blome, S.; Franzke, K.; Beer, M. African swine fever—A review of current knowledge. Virus Res. 2020, 287, 198099. [CrossRef] 194. Rudova, N.; Buttler, J.; Kovalenko, G.; Sushko, M.; Bolotin, V.; Muzykina, L.; Zinenko, O.; Stegniy, B.; Dunaiev, Y.; Sytiuk, M.; et al. Genetic Diversity of Porcine Circovirus 2 in Wild Boar and Domestic Pigs in Ukraine. Viruses 2022, 14, 924. [CrossRef] Microorganisms 2023, 11, 1186 31 of 31 195. Lederman, E.R.; Reynolds, M.G.; Karem, K.; Braden, Z.; Learned-Orozco, L.A.; Wassa-Wassa, D.; Moundeli, O.; Hughes, C.; Harvey, J.; Regnery, R.; et al. Prevalence of antibodies against orthopoxviruses among residents of Likouala region, Republic of Congo: Evidence for monkeypox virus exposure. Am. J. Trop. Med. Hyg. 2007, 77, 1150–1156. [CrossRef] [PubMed] 196. Zaucha, G.M.; Jahrling, P.B.; Geisbert, T.W.; Swearengen, J.R.; Hensley, L. The pathology of experimental aerosolized monkeypox virus infection in cynomolgus monkeys (Macaca fascicularis). Lab. Investig. 2001, 81, 1581–1600. [CrossRef] [PubMed] Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.
https://openalex.org/W4284990034	https://www.biorxiv.org/content/biorxiv/early/2022/07/06/2022.07.04.498740.full.pdf	English	null	Modification Mechanism of Phenol-formaldehyde Resin with Crude Bio-oil by Model Compound Method	bioRxiv (Cold Spring Harbor Laboratory)	2,022	cc-by	8,077	. CC-BY 4.0 International license made available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprint this version posted July 6, 2022. ; https://doi.org/10.1101/2022.07.04.498740 doi: bioRxiv preprint . CC-BY 4.0 International license made available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprint this version posted July 6, 2022. ; https://doi.org/10.1101/2022.07.04.498740 doi: bioRxiv preprint . CC-BY 4.0 International license made available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprint this version posted July 6, 2022. ; https://doi.org/10.1101/2022.07.04.498740 doi: bioRxiv preprint . CC-BY 4.0 International license made available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprint this version posted July 6, 2022. ; https://doi.org/10.1101/2022.07.04.498740 doi: bioRxiv preprint Modification Mechanism of Phenol-formaldehyde Resin with Crude Bio-oil by Model Compound Method Compound Method Yuxiang Yu 1,2 , Xiaoqian Qiu 1,2, Chao Li 1,2, Jianmin Chang 3, Defu Bao 1,2 1 College of Art and Design, Zhejiang Sci-Tech University, 928 Seconded Avenue, Xiasha High Education Zone, Hangzhou 310018, China Education Zone, Hangzhou 310018, China 2 Lab of Material Innovation Design and Intelligent Interaction, Zhejiang Sci-Tech University, 928 Seconded Avenue, Xiasha High Education Zone, Hangzhou 310018, China 3 College of Materials Science and Technology, Beijing Forestry University, 35 Qinghua East Road, Haidian District, Beijing, 100083, China 2 Lab of Material Innovation Design and Intelligent Interaction, Zhejiang Sci-Tech University, 928 Seconded Avenue, Xiasha High Education Zone, Hangzhou 310018, China 3 College of Materials Science and Technology, Beijing Forestry University, 35 Qinghua East Road, Haidian District, Beijing, 100083, China Corresponding authors E-mail:yuyuxiang0612@zstu.edu.cn(YY) . CC-BY 4.0 International license made available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprint this version posted July 6, 2022. ; https://doi.org/10.1101/2022.07.04.498740 doi: bioRxiv preprint . CC-BY 4.0 International license made available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprint this version posted July 6, 2022. ; https://doi.org/10.1101/2022.07.04.498740 doi: bioRxiv preprint Abstract To clarify the modification mechanism of bio-oil for phenol-formaldehyde resin with crude bio-oil (BPF), the bio-oil compounds were simplified by model compound method according to the component distribution. The phenol-formaldehyde resin with bio-oil model compounds (BMPF) were prepared and their basic performance, bonding strength and aging characteristics were determined. The changes on the microstructure and chemical bonds of BMPF were also analyzed by scanning electron microscope, Fourier transform infrared spectroscopy, and nuclear magnetic resonance analysis. Results showed that the components of bio-oil had different influence on the performance and microstructure of BMPF, especially phenols. Structural analysis indicated that the phenols and ketones of bio-oil had positive effects on the synthesis of BMPF, while the aldehydes and acids had negative effects. But all components of bio-oil could improve the aging resistance of BMPF inordinately. These results could provide a basis for the modification of BPF. Keywords: Modification mechanism; Phenol-formaldehyde resin; Crude bio-oil; Model compound method Introduction Phenol-formaldehyde resin (PF) is widely used in wood-based panel because of its excellent bonding strength, thermal stability, and heat resistance, etc. [1,2]. Phenol, as the main raw material for preparing PF, however, comes from fossil sources. With the promotion of global policies on resource shortage and environmental protection, seeking or developing a renewable resource to substitute phenol for preparing PF is becoming more important. Over the . CC-BY 4.0 International license made available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprint this version posted July 6, 2022. ; https://doi.org/10.1101/2022.07.04.498740 doi: bioRxiv preprint . CC-BY 4.0 International license made available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprint this version posted July 6, 2022. ; https://doi.org/10.1101/2022.07.04.498740 doi: bioRxiv preprint past decade, researchers have successfully produced bio-based PF by replacing phenol with sustainable materials such as lignin [3-6], tannins [7,8], cardanol [9], furfural [10], glyoxal [11] and bio-oil [12-15]. past decade, researchers have successfully produced bio-based PF by replacing phenol with sustainable materials such as lignin [3-6], tannins [7,8], cardanol [9], furfural [10], glyoxal [11] and bio-oil [12-15]. Bio-oil, the main product of the rapid pyrolysis of biomass, has been successfully used in the preparation of PF resin in recent years [16-18]. Compared with petroleum-based PF, bio-oil phenol-formaldehyde resin (BPF) has similar chemical and physical properties [13,19], but a better price competitiveness [20,21]. However, the components of bio-oil are complex, including hundreds of organic components such as phenols, ketones, aldehydes, acids and sugars [17,22]. Some organic components in the bio-oil can participate in the reaction to change the structure of resin, thus affecting the properties [23]. Other components that are not involved in the reaction also have influence on the resin properties as fillers [24,25]. In order to enhance the beneficial effects and reduce the negative effects of bio-oil on resin synthesis, it is of great significance to clarify the specific reaction mechanism of the different components in bio-oil. Model compound method is a way to use artificial compounds with same or similar structure and function to simplify the complex compound, such as bio-oil. Fortunate et al. Introduction [26] used 2-hydroxybenzaldehyde as bio-oil model compound to study the delineate effect of process variables on the catalytic hydrodeoxygenation of bio-oil by computational fluid dynamics. Tran et al. [27] studied the influence of Fe/AC and Ni/gamma-Al2O3 catalysts on the hydrodeoxygenation of woody bio-oil using guaiacol as model compound. Wang et al. [28] tried to better understand the aging mechanism of bio-oil by evaluating the aging performance for 39 kinds of bio-oil model compounds, and found that acids played an important role in the aging process, because it could be the reactant in the esterification reaction as well as the catalyst for the polymerization reaction of phenol and aldehyde. In this study, the bio-oil was simplified into five groups, including phenols, ketones, aldehydes, acids and sugars, according to the component distribution of bio-oil by gas . CC-BY 4.0 International license made available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprint this version posted July 6, 2022. ; https://doi.org/10.1101/2022.07.04.498740 doi: bioRxiv preprint . CC-BY 4.0 International license made available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprint this version posted July 6, 2022. ; https://doi.org/10.1101/2022.07.04.498740 doi: bioRxiv preprint chromatographic-mass spectrometric (GC-MS) analysis. And the bio-oil model compounds and its component model compounds were formulated by model compound method. The phenol- formaldehyde resin with bio-oil model compounds (BMPF) were prepared and their basic performance, bonding strength and aging characteristics were determined. The changes on the microstructure and chemical bonds of BMPF were analyzed by scanning electron microscope (SEM), Fourier transform infrared spectroscopy (FT-IR), and solid state nuclear magnetic resonance (NMR) analysis. mm × 1.5 mm) were provided by Xinda wooden Co., Ltd., Hebei, China. Materials Bio-oil, an acid liquid (pH 3.5), was obtained by the fast pyrolysis of Larix gmelinii (Rupr.) Kuzen in a fluidized bed at 550 ℃ for 2-3 s by the Lab of Fast Pyrolysis of Biomass and Productive Utilization (Beijing Forestry University, Beijing, China). And the organic components of bio-oil were displayed in Table 1 by GC-MS. Phenol, formaldehyde (aqueous solution, 37 wt.%), sodium hydroxide (NaOH), guaiacol, catechol butanone, cyclopentane dione, vanillin, furfural, D-glucose, acetic acid, phenylacetic acid were supplied by Xilong Chemical Industries, Guangdong, China. Poplar veneers (8 % moisture content, 400 mm × 400 ) id d b i d d d b i hi mm × 1.5 mm) were provided by Xinda wooden Co., Ltd., Hebei, China. . CC-BY 4.0 International license made available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprint this version posted July 6, 2022. ; https://doi.org/10.1101/2022.07.04.498740 doi: bioRxiv preprint . CC-BY 4.0 International license made available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprint this version posted July 6, 2022. ; https://doi.org/10.1101/2022.07.04.498740 doi: bioRxiv preprint Table 1. GC-MS of bio-oil. Compounds Molecular Formula Peak Area (%) Phenols 33.42 Guaiacol C7H8O2 6.52 4-methylguaiacol C8H10O2 5.72 Phenol C6H6O 3.16 4-methylcatechol C7H8O2 3.65 Catechol C6H6O2 2.64 P-methylphenol C7H8O 1.44 2-methoxy-4-propylphenol C10H14O2 1.39 Isoeugenol C10H12O2 1.39 2-cresols C7H8O 1.20 4-ethylguaiacol C9H12O2 1.04 4-ethylresorcinol C8H10O2 0.90 3,4-xylenol C10H12O2 0.88 4-methyl-2- methoxy-6 propenyl phenol C10H10O3 0.82 Eugenol C10H12O2 0.74 4-Ethylguaiacol C9H12O2 0.69 2-methyl-1,4catechol C7H8O2 0.52 2,6-dimethylphenol C8H10O 0.39 Trimethylhydroquinone C9H12O2 0.33 Ketones 29.56 2-Butanone C4H8O 6.25 3-Methyl-1,2-cyclopentanedione C6H8O2 5.89 Compounds . CC-BY 4.0 International license made available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprint this version posted July 6, 2022. ; https://doi.org/10.1101/2022.07.04.498740 doi: bioRxiv preprint . CC-BY 4.0 International license made available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprint this version posted July 6, 2022. Materials ; https://doi.org/10.1101/2022.07.04.498740 doi: bioRxiv preprint 4,6-dimethyl-2-pyrone C7H8O2 4.22 4-methyl-2(5H)-furanone C5H6O2 3.81 2-propyl-2-ethylcyclohexanone C11H20O 3.30 5-methyl-2-furanone C5H6O2 2.70 Acetovanillone C9H10O3 1.54 3-ethyl-2-hydroxy-2-cyclopenten-1-one C7H10O2 1.36 2-hydroxy-6-methoxyacetophenone C9H10O3 0.49 Aldehydes 13.45 Vanillin C8H8O3 7.03 Furfural C5H4O2 2.87 5-acetonyl-2-furaldehyde C8H8O4 2.34 Butanal C4H8O 1.21 Acids 9.33 Acetic Acid C2H4O2 7.13 4-hydroxy-3-methoxyphenylacetic acid C9H10O4 1.29 2,2-dimethylglutaric acid C7H12O4 0.75 Camphoric acid C10H16O4 0.16 Sugars 10.05 D-mannose C6H12O6 9.26 β-D-allose C16H12O6 0.79 Others 4.19 2-ethyne C6H10 1.43 1,9-decadiene C10H18 1.35 2,4-diacetoxypentanel C9H16O4 0.27 Hepty lacetat C8H10O5 0.37 . CC-BY 4.0 International license made available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprint this version posted July 6, 2022. ; https://doi.org/10.1101/2022.07.04.498740 doi: bioRxiv preprint . CC-BY 4.0 International license made available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprint this version posted July 6, 2022. ; https://doi.org/10.1101/2022.07.04.498740 doi: bioRxiv preprint 2,5-dimethyl-3-hexanol C8H12O 0.34 N-methylpiperidine C6H13N 0.43 2.2 Preparation 2.2.1 Preparation of model compounds 2.2.2 Preparation of resins The BMPF were synthesized according to Yu [13]. The molar ratio of phenol (including bio-oil or bio-oil model compounds) to formaldehyde was 1:2. The substitute rate of bio-oil to phenol was 20 wt.%, and the addition amount of NaOH was 20 wt.% of the mass of phenol. The BPF with bio-oil was denoted as BPF0, and the other BPF with bio-oil model compounds were denoted as BMPF1, BMPF2, BMPF3, BMPF4 and BMPF5 according to the name of bio- oil model compounds. 2.2.1 Preparation of model compounds According to the results of GC-MS analysis of bio-oil (Table 1), the model compounds of bio-oil were configured (Table 2). The moisture content of bio-oil was calculated as 30 %. The bio-oil and its component model compounds were prepared according to the corresponding proportions in the bio-oil. The formula and naming of bio-oil model compounds were detailed in Table 3. Table 2. Composition and ratio of Bio-oil model compounds. Model compound Composition Ratio Phenols Phenol; guaiacol; catechol 4:1:2 Ketones Butanone; cyclopentane dione 1:1 Aldehydes Vanillin; furfural 7:3 Sugars D-glucose —— Acids Acetic acid; phenylacetic acid 5:3 Table 2. Composition and ratio of Bio-oil model compounds. Model compound Composition Ratio Phenols Phenol; guaiacol; catechol 4:1:2 Ketones Butanone; cyclopentane dione 1:1 Aldehydes Vanillin; furfural 7:3 Sugars D-glucose —— Acids Acetic acid; phenylacetic acid 5:3 Table 2. Composition and ratio of Bio-oil model compounds. . CC-BY 4.0 International license made available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprint this version posted July 6, 2022. ; https://doi.org/10.1101/2022.07.04.498740 doi: bioRxiv preprint . CC-BY 4.0 International license made available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprint this version posted July 6, 2022. ; https://doi.org/10.1101/2022.07.04.498740 doi: bioRxiv preprint Table 3. Preparation of bio-oil model compounds. Model compounds Phenols Ketones Aldehydes Acids SugarsWater Name Bio-oil model 23% 21% 9% 6% 7% 30% M1 Phenol-free bio-oil model —— 21% 9% 6% 7% 30% M2 Ketone-free bio-oil model 23% —— 9% 6% 7% 30% M3 Aldehyde-free bio-oil model 23% 21% —— 6% 7% 30% M4 Acid-free bio-oil model 23% 21% 9% —— 7% 30% M5 Table 3. Preparation of bio-oil model compounds. 2.2.3 Preparation of samples The BMPF plywood and films were prepared according to Yu [13]. After keeping for 48 h in the room, the plywood was sawed into samples with dimension of 100 mm × 25 mm × 3 mm, and 5.3 ± 0.3 g samples were selected as plywood samples. The cured resin films (80 mm × 10 mm × 2 mm) were kept in temperature humidity chamber at 20 ± 2 ℃ and 65 ± 5 % relative humidity for 48 h. The sample with a mass of 1.7 ± 0.2 g was selected as the resin film sample. . CC-BY 4.0 International license made available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprint this version posted July 6, 2022. ; https://doi.org/10.1101/2022.07.04.498740 doi: bioRxiv preprint . CC-BY 4.0 International license made available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprint this version posted July 6, 2022. ; https://doi.org/10.1101/2022.07.04.498740 doi: bioRxiv preprint . CC-BY 4.0 International license made available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprint this version posted July 6, 2022. ; https://doi.org/10.1101/2022.07.04.498740 doi: bioRxiv preprint 2.3.3 Characterization The scanning electron microscope (SEM) analysis of resin resins were measured by SU8010 SEM (Hitachi, Tokyo, Japan) at 5.0 kV accelerating voltage. The Fourier transform infrared spectroscopy (FT-IR) analysis of resin resins were tested by a Nicolet iS5 FT-IR (Nicolet, Wisconsin, USA) over the range of 400 to 4000 cm−1 with a 4 cm−1 resolution and 64 scans. The solid state nuclear magnetic resonance (NMR) analysis of resins was acquired at a frequency 100 MHz using JNM-ECZ600R (JEOL, Tokyo, Japan). 2.3.2 Aging test The plywood and resin film samples were examined by a UV accelerated weathering tester (Yiheng Co., Ltd., Shanghai, China) according to the ASTM G 154. Each 12 h weathering cycle consisted of 8 h of UV exposure at 60 ℃ and 4 h condensation at 50 ℃. The plywood sample number was 6, and the resin film sample number was 3. 2.3.1 Basic performance test The GC-MS analysis of bio-oil was recorded on a GC/MS-QP system (Shimadzu, Kyoto, Japan) with 20 ℃/min heating rate, 3 min time-duration. An inlet temperature of 250 ℃, a He gas injection volume of 0.5 μL and a flow split-ratio of 50:1 were used for the GC. A junction temperature of 260 ℃, an ion temperature of 220 ℃, an EI source electron energy of 70 eV and a scan range of 29-500 amu were used for MS. The moisture content of bio-oil was calculated by V30 Karl-Fischer moisture meter (Mettler Toledo, Zurich, Switzerland). The pH, viscosity, and solid content of resins were tested by China National Standards GB/T 14074-2013. Each experiment was conducted at least three times under the same conditions, and their average values were reported. The formaldehyde emission and bonding strength of plywood was evaluated on the basis of China National Standard GB/T 17657-2013. In the test of formaldehyde emission, 10 test specimens (150mm×50 mm), cut from the plywood, were positioned in a 10 L glass desiccator. The emitted formaldehyde was absorbed by a Petri dish filled with 300 ml distilled water, and tested by the ultraviolet spectrophotometer (UV) at 412 nm wavelength. In the test of bonding strength, 12 plywood specimens with dimension of 25 mm × 10 mm × 3 mm were dipped into boiling water for 4 h, dried in 60 ± 3 °C for 20 h, and dipped into boiling water for 4 h, then submersed in cold water for 1 h. Then, the samples were tested with a speed at 10.0 mm/min of the cross head. . CC-BY 4.0 International license made available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprint this version posted July 6, 2022. ; https://doi.org/10.1101/2022.07.04.498740 doi: bioRxiv preprint . CC-BY 4.0 International license made available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprint this version posted July 6, 2022. ; https://doi.org/10.1101/2022.07.04.498740 doi: bioRxiv preprint 3.1 Basic performance The basic performance of BPF and BMPF is shown in Table 4. The pH value of BMPF were all increased in comparison with BPF0, indicating that the acidity of bio-oil model compound was lower than that of bio-oil. BMPF5 had the highest pH value because of the lack of acid compounds. Compare with BPF0, the viscosity of BMPF decreased, which might be due to the less reactions. The bio-oil model compounds had less substances compare to bio-oil, . CC-BY 4.0 International license made available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprint this version posted July 6, 2022. ; https://doi.org/10.1101/2022.07.04.498740 doi: bioRxiv preprint . CC-BY 4.0 International license made available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprint this version posted July 6, 2022. ; https://doi.org/10.1101/2022.07.04.498740 doi: bioRxiv preprint which would decrease the number of reactions. Besides, the fewer long-chain substances of bio-oil model compounds also weakened the drag effect. The low viscosity of BMPF2 was due to the lack of phenolic substances, which reduced the crosslinking degree of resin. However, BMPF3 had the lowest viscosity because the ketones were kind of good solvent and could promote the reaction between substances. which would decrease the number of reactions. Besides, the fewer long-chain substances of bio-oil model compounds also weakened the drag effect. The low viscosity of BMPF2 was due to the lack of phenolic substances, which reduced the crosslinking degree of resin. However, BMPF3 had the lowest viscosity because the ketones were kind of good solvent and could promote the reaction between substances. As shown in Table 4, the solid content of BMPF1 was higher than that of BPF0, indicating that the bio-oil had a greater negative effect than the bio-oil model compounds. BMPF2 had the smallest solid content, because the phenols in M2 were reduced resulting in the lower polymerization degree. The solid content of BMPF3 was smaller than that of BMPF1, which meant that ketones had a positive effect on the resin polyreaction. The increase in solid content of BMPF4 and BMPF5 indicated that the aldehydes and acids of bio-oil had a negative effect. 3.1 Basic performance The acids in bio-oil cloud fall the pH value of system, leading to the decreasing number of multi-substituted hydroxymethyl phenol in the resin addition process, and further reducing the resin crosslinking degree. The active site of formaldehyde is 2, while the reaction activity of other aldehydes was lower than that of formaldehyde. This meant that the other aldehydes would become the terminator of reaction, thus reducing the resin solid content. It can be seen from Table 4 that the formaldehyde emission of BMPF1 was higher than that of BPF0, indicating that many substances in bio-oil were conducive to reduce the formaldehyde emission. In other words, there were many substances in bio-oil that could react with or adsorb formaldehyde. Besides, the formaldehyde emission of BMPF2-5 was also higher than that of BPF0, indicating that the different components of bio-oil played different positive effect on reducing the formaldehyde emission. Such as phenols can polymerized with formaldehyde, . CC-BY 4.0 International license made available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprint this version posted July 6, 2022. ; https://doi.org/10.1101/2022.07.04.498740 doi: bioRxiv preprint . CC-BY 4.0 International license made available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprint this version posted July 6, 2022. ; https://doi.org/10.1101/2022.07.04.498740 doi: bioRxiv preprint acids can be esterified with formaldehyde, and ketones can promote the reaction of other substances with formaldehyde. Table 4. Basic performance of BPF and BMPF. Characteristics Resins pH (25℃) Viscosity (25 ℃, mPa·s) Solid content (%) Formaldehyde emission (mg/L) BPF0 11.10±0.13 322±54 45.93±0.22 0.272±0.021 BMPF1 11.28±0.20 136±33 46.38±0.07 0.387±0.018 BMPF2 11.33±0.07 101±12 43.40±0.10 0.594±0.024 BMPF3 11.32±0.03 98±18 44.80±0.12 0.452±0.032 BMPF4 11.32±0.05 112±25 46.16±0.13 0.539±0.039 BMPF5 11.40±0.04 139±34 46.68±0.30 0.438±0.028 Table 4. Basic performance of BPF and BMPF. Characteristics 3.2 Bonding strength It can be seen from Figure 1 that after 960 h of aging, the loss rate of bonding strength of BMPF1 plywood was higher than that of BPF0 plywood, indicating that except the bio-oil model compounds, the other substances in bio-oil also played a positive role in aging resistance. It can be seen from Figure 1 that after 960 h of aging, the loss rate of bonding strength of BMPF1 plywood was higher than that of BPF0 plywood, indicating that except the bio-oil model compounds, the other substances in bio-oil also played a positive role in aging resistance. The biggest bonding strength loss rate of BMPF2 plywood was due to the decrease of phenolic content. Compared with BMPF1 plywood, the bonding strength loss rate of BMPF3 plywood was higher, which was due to the long-chain flexible groups in ketones. As all known, the long- chain flexible groups could effectively improve the toughness of resin. Besides, the increased toughness would reduce the resin layer fractures caused by the stress generated and water- corrosion during aging, thereby improving the aging resistance of resin. The decrease of bonding strength loss rate of BMPF4 and BMPF5 plywood was due to the negative effect from aldehydes and pH value on resin synthesis reaction. However, the bonding strength loss rate of BMPF5 plywood was higher than that of BMPF4 plywood, because the lower pH value cloud reduced the resin solubility, which slightly improved the water resistance. 3.2 Bonding strength The bonding strength and its loss rate of BPF and BMPF plywood prepared are shown in Figure 1 [18]. The bonding strength of BMPF1 plywood was lower than that of BPF0 plywood, indicating that the bio-oil had less influence on the resin system than the bio-oil model compounds. BMPF2 plywood had the lowest bonding strength because the core of PF resin was formed by phenol and formaldehyde, but the total amount of phenolic substances of M2 was the lowest, resulting in a serious decrease of resin polymerization degree. Compared with BMPF1 plywood, BMPF3 plywood had a lower bonding strength, indicating that ketones could . CC-BY 4.0 International license made available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprint this version posted July 6, 2022. ; https://doi.org/10.1101/2022.07.04.498740 doi: bioRxiv preprint . CC-BY 4.0 International license made available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprint this version posted July 6, 2022. ; https://doi.org/10.1101/2022.07.04.498740 doi: bioRxiv preprint indeed promote the resin reaction. The increased bonding strength of BMPF4 resin plywood meant that the aldehydes in the bio-oil had a negative effect on the bonding strength. This might be due to 1) the aldehydes in bio-oil will compete with formaldehyde for reactive sites; 2) The aldehydes in bio-oil basically have only one active site, which blocked the further increase of resin molecular chain. Compared with BMPF1 plywood, the bonding strength of BMPF5 plywood increased, which was due to the decrease of pH value. indeed promote the resin reaction. The increased bonding strength of BMPF4 resin plywood meant that the aldehydes in the bio-oil had a negative effect on the bonding strength. This might be due to 1) the aldehydes in bio-oil will compete with formaldehyde for reactive sites; 2) The aldehydes in bio-oil basically have only one active site, which blocked the further increase of resin molecular chain. Compared with BMPF1 plywood, the bonding strength of BMPF5 plywood increased, which was due to the decrease of pH value. Fig 1. Bonding strength and its loss rate of BPF and BMPF plywood. . CC-BY 4.0 International license made available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprint this version posted July 6, 2022. ; https://doi.org/10.1101/2022.07.04.498740 doi: bioRxiv preprint . CC-BY 4.0 International license made available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprint this version posted July 6, 2022. ; https://doi.org/10.1101/2022.07.04.498740 doi: bioRxiv preprint 3.3 SEM analysis The apparent morphology changes of BPF and BMPF films before and after aging are shown in Figure 2 [14,19]. The BMPF films before aging had a smooth surface, and the degree of smoothness was higher than that of BPF0. The reason was that there were some solid particles in the bio-oil, which reduced the smoothness of its solidified surface. Besides, it also showed that the selection of bio-oil model was reasonable, which produced better synthesis reaction and formed a smooth surface after curing. Compared with BMPF2-5, the smoothness of BMPF3 film surface was lower because ketones can act as a good solvent and promote the resin to form a better surface. It can be seen from Figure 2 that the surface roughness of BPF and BMPF increased obviously after aging for 960 h, and defects such as holes and cracks appeared. Compared with BPF0, BMPF1 film had a higher degree of surface aging. Compared with BMPF1-5, BMPF3 and BMPF2 had a greater aging degree. The reduction of phenolic substances reduced the degree of resin polymerization, thereby weakening the aging resistance. Meanwhile, the lack of ketones reduced the oily substances in the system, resulting in a lower water resistance and a higher degree of surface aging of BMPF2 and BMPF3. 3.4 FT-IR analysis The FT-IR curves of BPF and BMPF before and after aging for 960 h are displayed in Figure 3, and the attribution of characteristic absorption peaks is shown in Table 5 [12,13,15,29]. . CC-BY 4.0 International license made available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprint this version posted July 6, 2022. ; https://doi.org/10.1101/2022.07.04.498740 doi: bioRxiv preprint . CC-BY 4.0 International license made available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprint this version posted July 6, 2022. ; https://doi.org/10.1101/2022.07.04.498740 doi: bioRxiv preprint As shown in Figure 3, the peak of methylene (CH2) groups at the region of 2924 cm−1 and 2854 cm−1. The content of CH2 groups can indirectly represent the polymerization degree of PF resin [30,31]. The CH2 groups peak of BMPF1 was stronger than that of BPF0, because the phenolic substance activity of bio-oil model was stronger than that of bio-oil. Compared with the BMPF1, the CH2 groups peak of BMPF2 was weaker because CH2 was generated by the reaction of phenols and aldehydes, and the reduction of phenols reduced the CH2 content in the resin. The weakening of CH2 groups peak of BMPF3 also confirmed the promoting effect of ketones on resin polymerization. The increased of CH2 groups peak of BMPF4 and BMPF5 meant that aldehydes and acids in bio-oil have a negative effect on the synthesis of BPF resin. The peak of ether bond (C—O—C) groups at the region of 2924 cm−1 and 2854 cm−1. Ether bond is another connecting structure between molecular chains of PF resin, which can also indirectly represent the degree of polymerization of PF resin. Compared with BPF0, the change on ether bond peak of BMPF1-5 was similar to those of CH2 peak. After aging for 960 h, the peak of CH2 and C—O—C groups decreased significantly, indicating that the polymerization degree of resin decreased after aging. Compared with BPF0, the CH2 peak of BMPF1 decreased significantly. It showed that other substances in bio-oil could reduce the polymerization degree of resin synthesis, but improved the aging property of resin. 3.4 FT-IR analysis Compared with BMPF1, the CH2 groups peak of BMPF2 decreased significantly after aging, which might because resin failed to form a good network structure due to the reduction of phenols, resulting in the reduction of aging resistance. The CH2 groups peak of BMPF3 weakened the biggest. It might be due to the lack of ketones, which increased the contact area of external environment (water, light and other aging factors), and aggravated the aging degree. . CC-BY 4.0 International license made available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprint this version posted July 6, 2022. ; https://doi.org/10.1101/2022.07.04.498740 doi: bioRxiv preprint . CC-BY 4.0 International license made available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprint this version posted July 6, 2022. ; https://doi.org/10.1101/2022.07.04.498740 doi: bioRxiv preprint Fig 3. FTIR curves of BPF and BMPF before and after aging for 960 h. Table 5. Peaks and assignment of FT-IR spectra for BPF. Wave number (cm-1) Vibration Assignment 3408 ν(—OH) a Phenolic OH and aliphatic OH stretching vibration 2923, 2852 ν(CH2) Aliphatic CH2 asymmetric stretching vibration 1604, 1452 ν(C=C) C=C aromatic ring stretching vibration 1047 ν(C—O—C) Phenolic C—O—C stretching vibration aν: Stretching vibration. Table 5. Peaks and assignment of FT-IR spectra for BPF. 3.5 NMR analysis 13C NMR images of BPF before and after aging for 960 h are displayed in Figure 4. The chemical shifts of carbon atoms in each group of BPF resin are shown in Table 6 [18,32]. The carbon spectra of BMPF1-5 and BPF0 were basically similar, indicating that the preparation of bio-oil model compound was reasonable, and the conclusion was the same as that of FT-IR analysis. BPF0 had more substances with C=O structure than BMPF1-5 because there were more kinds of C=O substances in bio-oil. Compared with BMPF1, the values of substituted ortho and para aromatic carbons (139-120 ppm) and methylene bridges (45-27 ppm) of BPF0 were smaller, indicating that other components of bio-oil would still affect the resin synthesis process. However, the value of dimethylene ether bridges (77-68 ppm) increased, which might be due to the presence of ethers in bio-oil. Compared with BMPF1, the values of methylene bridges of BMPF2 and BMPF3 decreased, indicating that phenols and ketones in bio-oil played a positive role in synthesis reaction, while the values of methylene bridges of BMPF4 and . CC-BY 4.0 International license made available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprint this version posted July 6, 2022. ; https://doi.org/10.1101/2022.07.04.498740 doi: bioRxiv preprint . CC-BY 4.0 International license made available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprint this version posted July 6, 2022. ; https://doi.org/10.1101/2022.07.04.498740 doi: bioRxiv preprint BMPF5 increased, indicating that aldehydes and acids in bio-oil had opposite effects. BMPF5 increased, indicating that aldehydes and acids in bio-oil had opposite effects. After aging for 960 h, the intensity of C=O from carboxylic acids and esters (186-168 ppm) increased, indicating that the aged resin had undergone oxidative aging. The intensity of C=O from aldehydes (220-200 ppm) decreased, indicating that some aldehyde groups were converted to carboxyl or ester groups. The reduction of the substituted ortho and para aromatic carbons meant that the resin network structure was broken after aging, and small molecular substances were formed. The decrease of the intensity of methylene bridges indicated that the molecular chain of resin was broken after aging. Compared with BMPF1, the reduction of the methylene bridge of BPF0 was lower, indicating that the improvement on aging resistance of BPF was the result of the overall interaction of bio-oil molecules. The reduction of methylene bridge of BMPF2-5 was higher than that of BMPF1, which also confirmed this inference. Fig 4. 13C NMR images of BPF and BMPF before and after aging for 960 h. Table 6. 13C NMR assignment and quantitative analysis of chemical groups for BPF before and after aging for 960 h. Calculation Before aging After aging Chemical group structure Chemical shirt (ppm) 0a 1 2 3 4 5 0 1 2 3 4 5 C=O from aldehydes/ketones 220-200 1.161.111.090.790.911.05 0.95 0.970.820.650.880.94 C=O from carboxylic acids 186-176 0.060.050.060.060.060.04 0.08 0.090.100.110.080.05 C=O from esters 176-168 0.040.020.030.030.020.01 0.09 0.040.080.090.120.08 Phenolic carbon 164-145 1.001.001.001.001.001.00 1.00 1.001.001.001.001.00 Fig 4. 13C NMR images of BPF and BMPF before and after aging for 960 h. Fig 4. C NMR images of BPF and BMPF before and after aging for 960 h. Table 6. 13C NMR assignment and quantitative analysis of chemical groups for BPF before and after aging for 960 h. Calculation Before aging After aging Chemical group structure Chemical shirt (ppm) 0a 1 2 3 4 5 0 1 2 3 4 5 C=O from aldehydes/ketones 220-200 1.161.111.090.790.911.05 0.95 0.970.820.650.880.94 C=O from carboxylic acids 186-176 0.060.050.060.060.060.04 0.08 0.090.100.110.080.05 C=O from esters 176-168 0.040.020.030.030.020.01 0.09 0.040.080.090.120.08 Phenolic carbon 164-145 1.001.001.001.001.001.00 1.00 1.001.001.001.001.00 . CC-BY 4.0 International license made available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprint this version posted July 6, 2022. ; https://doi.org/10.1101/2022.07.04.498740 doi: bioRxiv preprint . CC-BY 4.0 International license made available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprint this version posted July 6, 2022. ; https://doi.org/10.1101/2022.07.04.498740 doi: bioRxiv preprint Substituted ortho and para aromatic carbons 139-120 4.725.434.524.205.325.60 4.12 4.524.244.034.695.04 Dimethylene ether bridges 77-68 0.500.460.220.450.390.43 0.49 0.410.420.340.370.36 Methylol groups 68-60 0.610.580.430.530.600.53 0.57 0.380.390.300.260.26 Benzylamines 60-45 1.191.051.010.901.041.08 1.12 0.920.920.710.930.73 Methylene bridges 45-27 2.552.632.442.472.652.87 2.43 2.492.322.302.512.72 a 0: BPF0; 1: BMPF1; 2: BMPF2; 3: BMPF3; 4: BMPF4; 5: BMPF5 Conclusions Crude bio-oil was used as a phenol substitution to synthesize the BPF resin, and its performance and modification mechanism were analyzed by model compound method. The bio-oil components had different influence on performance and microstructure of BPF. Compared with BMPF1, the bonding strength of BMPF2 and BMPF3 plywood decreased, while the bonding strength of BMPF4 and BMPF5 increased. Structural analysis showed that the CH2 peak in FT-IR and the methylene bridges intensity in NMR of BMPF2 and BMPF3 were lower than that of BMPF1, while the results for BMPF4 and BMPF5 were opposite. These indicated that the phenols and ketones of bio-oil had positive effects, while the aldehydes and acids had negative effects. After aging for 960 h, compared with BMPF1, the CH2 peak of BMPF2-5 decreased significantly and the reduction of methylene bridge from BMPF2-5 was higher, which meant that all components of bio-oil could improve the aging resistance of BMPF inordinately. These results could provide a basis for the modification of BPF. . CC-BY 4.0 International license made available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprint this version posted July 6, 2022. ; https://doi.org/10.1101/2022.07.04.498740 doi: bioRxiv preprint . CC-BY 4.0 International license made available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprint this version posted July 6, 2022. ; https://doi.org/10.1101/2022.07.04.498740 doi: bioRxiv preprint . CC-BY 4.0 International license made available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprint this version posted July 6, 2022. ; https://doi.org/10.1101/2022.07.04.498740 doi: bioRxiv preprint Acknowledgements This work was supported by the Talent Introduction Program of Zhejiang Sci-Tech University (19082425-Y) and Philosophy and Social Science Planning Project of Zhejiang Province (21NDQN234YB). References 1. Sarika PR, Nancarrow P, Khansaheb A, Ibrahim T. Bio-based alternatives to phenol and formaldehyde for the Production of Resins. Polymers. 2020 Sep;12(10):2237. https://doi.org/polym12102237 1. Sarika PR, Nancarrow P, Khansaheb A, Ibrahim T. Bio-based alternatives to phenol and formaldehyde for the Production of Resins. Polymers. 2020 Sep;12(10):2237. https://doi.org/polym12102237 2. Xu Y, Guo L, Zhang H, Zhai H, Ren H. Research status, industrial application demand and prospects of phenolic resin. RSC Adv. 2019 Aug;9:28924-28935. https://doi.org/10.1039/C9RA06487G 3. Gong X, Meng Y, Lu J, Tao Y, Cheng Y, Wang H. A review on lignin‐based phenolic resin adhesive. Macromol. Chem. Phys. 2022;2100434. https://doi.org/10.1002/macp.202100434 4. Chen Y, Gong X, Yang G, Li Q, Zhou N. Preparation and characterization of a nanolignin phenol formaldehyde resin by replacing phenol partially with lignin nanoparticles. RSC Adv. 2019 Sep;9: 29255-29262. https://doi org/10 1039/C9RA04827H 5. Solt P, Rößiger B, Konnerth J, Van Herwijnen H. Lignin phenol formaldehyde resoles using base-catalysed depolymerized Kraft lignin. Polymers. 2018 Sep;10:1162. https://doi.org/10.3390/polym10101162 . CC-BY 4.0 International license made available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprint this version posted July 6, 2022. ; https://doi.org/10.1101/2022.07.04.498740 doi: bioRxiv preprint . CC-BY 4.0 International license made available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprint this version posted July 6, 2022. ; https://doi.org/10.1101/2022.07.04.498740 doi: bioRxiv preprint . CC-BY 4.0 International license made available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprint this version posted July 6, 2022. ; https://doi.org/10.1101/2022.07.04.498740 doi: bioRxiv preprint 6. Yang W, Rallini M, Natali M, Kenny J, Puglia D. Preparation and properties of adhesives based on phenolic resin containing lignin micro and nanoparticles: A comparative study. Mater. Design. 2019 Jan;2019,161:55-63. https://doi.org/10.1016/j.matdes.2018.11.032 7. Li JJ, Zhu WJ, Zhang SF, Gao Q, Xia CL, Zhang W, Li JZ. Depolymerization and characterization of Acacia mangium tannin for the preparation of mussel-inspired fast- curing tannin-based phenolic resins. Chem. Eng. J. 2019 Aug;370:420-431. https://doi.org/10.1016/j.cej.2019.03.211 8. Hoyos-Martínez PL, Issaoui H, Herrera R, Labidi J, Charrier-El F. Wood fireproofing coatings based on biobased phenolic resins. ACS Sustain. Chem. References Eng. 2021 Jan;9(4):1729-1740. https://doi.org/10.1016/10.1021/acssuschemeng.0c07505 9. Zhang WZ, Jiang N, Zhang TT, Li TS.Thermal stability and thermal degradation study of phenolic resin modified by cardanol. Emerg. Mater. Res. 2020 Mar;9(1):180-185. https://doi.org/10.1680/jemmr.18.00133 10. Sui GH, Cheng YY, Yang XM, Wang XF, Wang ZC. Use of sustainable glucose and furfural in the synthesis of formaldehyde-free phenolic resole resins. J. Appl. Polym. Sci. 2019 Mar;136:47733. https://doi.org/10.1002/app.47732 11. Ballerini A, Despres A, Pizzi A. Non-toxic, zero emission tannin-glyoxal adhesives for wood panels. Hoiz Roh Werkst. 2005 Nov;63:477-478. https://doi.org/10.1007/s00107-005-0048-x 12. Kokten ES, Özbay G, Ayrilmis N. Synthesis of biobased phenolic resins using catalytic pyrolysis oil and its effect on oriented strand board performance. J. Adhesion. 2018 . CC-BY 4.0 International license made available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprint this version posted July 6, 2022. ; https://doi.org/10.1101/2022.07.04.498740 doi: bioRxiv preprint . CC-BY 4.0 International license made available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprint this version posted July 6, 2022. ; https://doi.org/10.1101/2022.07.04.498740 doi: bioRxiv preprint Jun:1-15. https://doi.org/10.1080/00218464.2018.1481400 13. Yu YX, Xu PP, Chang MM, Chang JM. Aging properties of phenol-formaldehyde resin modified by bio-oil using UV weathering. Polymers. 2018 Oct;10:1183. https://doi.org/10.3390/polym10111183 14. Cui Y, Hou XP, Wang WL, Chang JM. Synthesis and characterization of bio-oil phenol formaldehyde resin used to fabricate phenolic based materials. Materials. 2017 Jun;10 (6):668-677. https://doi.org/10.3390/ma10060668 15. Dong FG, Wang MC, Wang ZH. Bio-oil as substitute of phenol for synthesis of resol- type phenolic resin as wood adhesive. Int. J. Chem. React. Eng. 2017 Dec;16:1-7. https://doi.org/10.1515/ijcre-2017-0107 16. Li DB, Zhu LY, Berruti F. Autothermal fast pyrolysis of waste biomass for wood adhesives. Industrial Crops and Products. 2021 Oct;170:113711. https://doi.org/10.1016/j.indcrop.2021.113711 17. Kim J. Production, separation and applications of phenolic-rich bio-oil — a review 17. Kim J. Production, separation and applications of phenolic-rich bio-oil — a review. Bioresource Technol. 2015 Feb;178:90–98. https://doi.org/10.1016/j.biortech.2014.08.121 Bioresource Technol. 2015 Feb;178:90–98. https://doi.org/10.1016/j.biortech.2014.08.121 18. Celikbag Y, Nuruddin M, Biswas M, Asafu-Adjaye O, Via BK. Bio-oil-based phenol- formaldehyde resin: comparison of weight-and molar-based substitution of phenol with bio-oil. J. Adhes. Sci. Technol. 2020 Jun;34:1-12. https://doi.org/10.1080/01694243.2020.1784540 https://doi.org/10.1080/01694243.2020.1784540 19. Aalan M, Özbay G, Ayrilmis N. Adhesive characteristics and bonding performance of phenol formaldehyde modified with phenol-rich fraction of crude bio-oil. J. Adhes. Sci, . CC-BY 4.0 International license made available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprint this version posted July 6, 2022. ; https://doi.org/10.1101/2022.07.04.498740 doi: bioRxiv preprint Technol. 2015 Aug;29:2679-2691. https://doi.org/10.1080/01694243.2015.1080474 20. Choi G, Oh S, Lee S, Kim J. Production of bio-based phenolic resin and activated carbon from bio-oil and biochar derived from fast pyrolysis of palm kernel shells. Bioresource Technol. 2014 Feb;178:99-107. https://doi.org/10.1016/j.biortech.2014.08.053 21. Domínguez JC, Oliet M, Alonso MV, Rojo E, Rodríguez F. Structural, thermal and rheological behavior of a bio-based phenolic resin in relation to a commercial resol resin. Ind. Crop. Prod. 2013 Jun;42:308-314. https://doi.org/10.1016/j.indcrop.2012.06.004 22. Yogalakshmi KN, Devi TP, Sivashanmugam P, Kavitha S, Ravi YK, Varjani SJ, et al. Lignocellulosic biomass-based pyrolysis: A comprehensive review. Chemosphere. 2021 Aug;286:131827. https://doi.org/10.1016/j.chemosphere.2021.131824 23. Usino DO, Ylitervo P, Moreno A, Sipponen MH, Richards T. Primary interactions of biomass components during fast pyrolysis. J. Anal. Appl. Pyrol. 2021 Aug;7:10529. https://doi.org/10.1016/j.jaap.2021.105297 24. Zhang J, Choi YS, Yoo CG, Kim TH, Brown RC, Shanks BH. Cellulose– Hemicellulose and cellulose–lignin interactions during fast pyrolysis. ACS Sustain. Chem. Eng. 2015 Dec;3:293-301. https://doi.org/10.1021/sc500664h 25. Jun:1-15. https://doi.org/10.1080/00218464.2018.1481400 Zhou H, Wu CF, Meng AH, Zhang YG, Williams PT. Effect of interactions of biomass constituents on polycyclic aromatic hydrocarbons (PAH) formation during fast pyrolysis. J. Anal. Appl. Pyrol. 2014 Sep;110:264-269. . CC-BY 4.0 International license made available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprint this version posted July 6, 2022. ; https://doi.org/10.1101/2022.07.04.498740 doi: bioRxiv preprint . CC-BY 4.0 International license made available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprint this version posted July 6, 2022. ; https://doi.org/10.1101/2022.07.04.498740 doi: bioRxiv preprint https://doi.org/10.1016/j.jaap.2014.09.007 ; https://doi.org/10.1101/2022.07.04.498740 CC-BY 4.0 Internation made available under a (which was not certified by peer review) is the author/funder, who has granted bioRxi this version posted ; https://doi.org/10.1101/2022.07.04.498740 doi: bioRxiv preprint (w b al license a license to display the preprint in perpetuity. It is The copyright holder for this preprint uly 6, 2022. https://doi.org/10.1016/j.jaap.2014.09.007 26. Fortunate O, Kishore N. Computational fluid dynamics investigation on catalytic hydrodeoxygenation of a bio-oil model compound in a fluidized bed reactor. J. Therm. Sci. Eng. Appl. 2021 Dec;13:061018. https://doi.org/10.1115/1.4050565 27. Tran QK, Ly HV, Kwon B, Kim SS, Kim J. Catalytic hydrodeoxygenation of guaiacol as a model compound of woody bio-oil over Fe/AC and Ni/gamma-Al2O3 catalysts. Renew. Energ. 2021 Aug;173:886-895. https://doi.org/10.1016/j.renene.2021.03.138 28. Wang R, Ben H. Accelerated aging process of bio-oil model compounds: A mechanism study. Front. Energy Res. 2020 May;8:79. https://doi.org/10.3389/fenrg.2020.00079 29. Yu YX, Xu PP, Chen C, Chang JM, Li LF. Formaldehyde emission behavior of plywood with phenol-formaldehyde resin modified by bio-oil under radiant floor heating condition. Build. Environ. 2018 Aug;144:565-572. https://doi.org/10.1016/j.buildenv.2018.08.025 30. Strzemiecka B, Zieba-Palus J, Voelkel A, Lachowicz T, Socha E. Examination of the chemical changes in cured phenol-formaldehyde resins during storage. J. Chromatogr. A. 2016 Apr;1441:106-115. https://doi.org/10.1016/j.chroma.2016.02.051 31. Pethrick RA. Design and ageing of adhesives for structural adhesive bonding–A review. P. I. Mech. Eng. L-J. Mat. 2014 Feb;229:1-31. h //d i /0 1177/1464420714522981 32. Fan DB, Li GY, Qin TF, Chu FX. Synthesis and structure characterization of phenol- urea-formaldehyde resins in the presence of magnesium oxide as catalyst. Polymers 32. Fan DB, Li GY, Qin TF, Chu FX. Synthesis and structure characterization of phenol- urea-formaldehyde resins in the presence of magnesium oxide as catalyst. Polymers . CC-BY 4.0 International license made available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprint this version posted July 6, 2022. ; https://doi.org/10.1101/2022.07.04.498740 doi: bioRxiv preprint . CC-BY 4.0 International license made available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprint this version posted July 6, 2022. ; https://doi.org/10.1101/2022.07.04.498740 doi: bioRxiv preprint 2014 Aug;6(8):2221-2231. https://doi.org/10.3390/polym6082221 . CC-BY 4.0 International license made available under a ed by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprint this version posted July 6, 2022.
https://openalex.org/W4309140274	https://zenodo.org/records/7326551/files/s11846-022-00570-4.pdf	English	null	Critical HR capabilities in agile organisations a cross-case analysis in Swiss SMEs	Zenodo (CERN European Organization for Nuclear Research)	2,022	cc-by	10,538	Review of Managerial Science https://doi.org/10.1007/s11846-022-00570-4 Review of Managerial Science https://doi.org/10.1007/s11846-022-00570-4 ORIGINAL PAPER Stephanie Kaudela-Baum Stephanie.Kaudela@hslu.ch Abstract The current agile management literature is missing insight about the challenges agile organisations face regarding human resource management (HRM/HR) – and how they may overcome them. Based on an exploratory case study design, we investigate the managerial challenges in seven pioneering companies, all of them medium-sized firms (SME) from the IT sector in Switzerland. The majority of the qualitative data gathered stems from interviews, that was coded along emerging themes. The results are divided into three sections: a proposed (1) typology for Business Agilists, (2) general challenges and (3) emergent agile HR capabilities. Following the proposition of agile HR as a distributed capability, we discuss sev eral theoretical and practical implications. In essence, particular attention is to be placed on a cultural fit between employees and the agile working environment, which demands individuals to be highly self-reliant and autonomous. This entails the need for a flexible support structure to grow employees’ skills accordingly. The conclusion emphasises addressing the shared responsibility for HR work and the corresponding broad capability development of different role holders, resulting in the suggestion to replace the term HR with people management as a more inclusive bracket for integrative HR, culture and leadership work in agile organisations. Keywords HRM · People management · Agile · Business agility · Case Study Keywords HRM · People management · Agile · Business agility · Case Study Critical HR capabilities in agile organisations a cross-case analysis in swiss SMEs Léonie S. Mollet1 · Stephanie Kaudela-Baum1 Received: 31 March 2021 / Accepted: 24 June 2022 © The Author(s) 2022 1 Lucerne University of Applied Sciences, Luzern, Switzerland 1 Introduction Agile frameworks are increasingly being introduced to transform whole organisations, in a quest to deal with complexity and uncertainty. According to Charbonnier-Voirin (2011), organisational agility is associated with attributes such as responsiveness, 1 3 3 L. S. Mollet, S. Kaudela-Baum quickness and learning. When agile principles are incorporated throughout a whole organisation, we refer to these organisations as Business Agilists. The drivers are manifold. Externally, organisations are confronted with disrupting business mod els and dynamic markets, pluralistic and diverse value patterns, and digitalisation (Meyer et al. 2017, p. 534; Harraf et al. 2015, p. 675; Saha et al. 2017, p. 326). Inter nally, employees demand mobile-flexible, meaningful and fulfilling work (Coates 2017, p. 63; Rigby et al. 2016). Agile transformations are not merely a methodological issue, but involve numer ous HR challenges, as they essentially require designing an organisation around human needs to learn, share and connect (Tolchinsky 2015, p. 60). Existing research also questions the viability of staff functions such as HR, which traditionally orga nise people aspects (Laloux 2014, p. 71). In contrast, current HRM literature is still strongly influenced by Ulrich’s business partner model that emphasises the strategic role of HR professionals in coordinating the whole employee lifecycle (Brockbank and Ulrich 2009; Ulrich 1997). Despite high practitioner interest, there is a distinct lack of academic activity to “address the science-practice divide” concerning the role of HR in agile organisations (McMackin and Heffernan 2020, p. 11). This study aims to identify the main characteristics that typify agile transformations, with a focus on HR in knowledge-intensive SME in the Swiss IT sector. To address this research gap, our paper concentrates on the following questions: how do organisa tions use HR practices to foster agility? And what are the implications for HR agents in terms of requirements, roles or focus? This paper is structured as follows: after setting the context of agile organisa tions and the implications for HR work, the research design is discussed. The result chapter then describes HR practices in the companies studied, interprets them against the chosen heuristic framework and condenses management challenges associated. Finally, the article discusses theoretical and practical implications for the incorpora tion of agility into HR capabilities and practices. 2 Organisational agility and HR capabilities Despite a growing number of academic papers, the theoretical framing of agile organ isations, and the corresponding role of HR, remains difficult. The different strands of the emerging literature may be condensed into a threefold purpose HR fulfills in agile organisations. Firstly, regarding organisational development, HR facilitates the structural and cultural conditions for promoting agility – in particular by promoting self-organisation, shared values and collective decision-making, and by supporting the organisation in mobilising resources flexibly (see Chaps. 2.2 and 2.3). Secondly, HR supports continuous, collective learning and adaptation processes, and an open exchange of information (see Chap. 2.3). Thirdly, HR expedites participative devel opment of the business strategy, scans strategic influencing factors, and empowers leaders to make courageous, fast decisions (see Chap. 2.4). 3 1 Critical HR capabilities in agile organisations a cross-case analysis in… 2.1 Maturity models of organisational agility The emerging literature differentiates between “doing agile” and “being agile” (Fer nandez and Fernandez 2008, p. 16). These descriptions represent whether organisa tions simply implement tools and terminology, or incorporate the underlying values and principles as well. This value-orientation can be used as a decisive factor for classifying organisations into maturity levels. Accordingly, Werder and Maedche (2018) separate three states of maturity in agile organisations, i.e. (1) starting or (2) being in the middle of their transition, and (3) being more mature in their organisa tional agility. Aghina et al. (2020) developed the following criteria to comprehen sively assess agility maturity in large organisations: (1) a shared purpose and vision embodied across the organisation, (2) a network of empowered teams, (3) rapid deci sion and learning cycles, (4) dynamic people model that ignites passion, (5) next-gen eration enabling technology. Our study focuses on mature Business Agilists, in order to streamline resources towards companies that are likely to exhibit HR practices that may be characteristic for agile organisations. Only few studies have examined the resulting agile organisational culture (see e.g. Siakas and Siakas 2007). Based on a literature analysis, Rebentisch et al. (2018) have derived the following eight dimensions of agile culture: (1) fast, team driven decision-making; (2) willingness to continuously learn and improve; (3) autonomy and empowerment of people; (4) supportive and collaborative management; (5) team orientation; (6) intensified personal communication; (7) open information sharing and (8) comfort with change and uncertainty. SMEs seem particularly numerous amongst Business Agilists. Already marginally bureaucratic and hierarchic, and with a certain openness to change and pragmatism, SMEs show a strong disposition for agile frameworks (Damanpour 2010, p. 997). Business agility feeds into SME-typical attributes, such as their resourcefulness, ability to make bold yet viable decisions or fluidly reconfigure resources accord ingly (Arbussa et al. 2016, p. 289). Rauch and Hatak (2016, p. 487) directly con nect an SME’s performance to its ability to enhance employee skills, motivation, and empowerment, concepts that seem to be essential in agile frameworks. With limited resources, however, experimenting with HR practices comes at a high price (Meyer et al. 2017, p. 542). 2.2 Organisational agility and the concept of self-organisation How can agility be promoted throughout the whole organisation? The management framework Holacracy provides a practical answer to this question. This sociocrati cally (Endenburg and Pearson 1998) informed concept goes back to Robertson (2015), who designed an elaborate blueprint for distributed decision-making. In the following years, an increasing number of companies started to experiment with simi lar approaches to self-organisation, in an attempt to cope with their ever-changing environment (Atuahene-Gima 2003; Kettunen et al. 2019). While agile management frameworks currently seem to gain traction, their underlying concepts are far from new. The organisational development possibilities of abolishing hierarchical struc tures and decision-making have been discussed by systems theorists since the 1990s. 1 3 3 3 L. S. Mollet, S. Kaudela-Baum Luhmann (1995), for instance, describes organisations as social systems that can only change from within themselves. As such, organising through planned intervention is only possible to a restricted extent (Malik and Probst 1984; Probst 1987) describes this self-generation of order as an emergent phenomenon in a social system. This order is characterised by a spontaneous and unintentional character and is the result of multiple experiences, interactions and decisions. Self-organisation implicates a polycentric system with self-coordination and reciprocal, participatory adaptation and change of behaviour. The principle of distributed authority is an operational formula that promotes emergent decision-making processes. Against this backdrop, the Holacracy movement mentioned above may be interpreted as a contemporary attempt to support self-organisation by influencing communication and decision- making systems, rather than individual functionaries. So far, we have recognised that Business Agilists show characteristics of self- organising social systems and that SMEs seem to be typical first movers. This sys temic-holistic perspective emphasises the usefulness of culture- and context-bound interventions. However, the existing literature fails to provide implications for the development of the agile HR capabilities associated. To reflect on these, we turn to established management frameworks that offer approaches for coping with complex ity and change. 1 3 2.3 Management capabilities for a dynamic-agile environment They link agility to dynamic capabilities, which can be divided into three clusters: (1) Sensing of unknown futures (i.e. identification, co-development, and assessment of opportunities and threats in relationship to customer needs). (2) Seizing (i.e. mobilisation of resources to address needs and opportunities and capture value). (3) Shifting or transforming (i.e. continued renewal). Dynamic capabilities are deeply embedded in the organisation, hard to acquire or imitate, and require entrepreneurial management qualities – and thus largely depend on knowledge workers and experts (Teece 2017, p. 698). Laloux (2014) examined companies that are economically successful through agile principles (i.e. focus on current problems instead of goals or decentralisation). His case studies highlight the capability to empower self-organised teams in agile transformations. While this is old news, organisations still find it difficult to switch from a hierarchical structure to a context where teams assume responsibility for their own decisions (Moravec 1999).fl The approaches introduced above offer a good basis for reflection, especially at a strategic management, leadership and team development level – yet exclude concrete implications for HR. 2.3 Management capabilities for a dynamic-agile environment At the team management level, agility emerged in the 1990s as a bottom-up counter- movement to traditional project management methods, whose rigid up-front plan ning failed to serve evolving customer needs. Agile project management can be described as a set of values and principles that act as a compass for decision-making and aligning practice on a team level. Primarily, these principles define teams as cross-functional and self-organised, while being trusted to accomplish the job. Orien tated towards customer needs, these self-regulated teams adjust actions frequently to changing requirements and new insight, ensuring continuous improvement and col lective learning (Parker et al. 2015, p. 112–113; Polley and Ribbens 1998). According to Moravec (1999), introducing self-organised teams requires building mutual under standing and allowing all team members to develop leadership skills. Agile practices are nowadays mainstream in software development, and are increasingly applied to other functional areas (Kettunen et al. 2019). Unsurprisingly, IT companies are at the forefront of business agility – often as an organic spill-over from project management. These knowledge-intensive service organisations depend on maximising their flexibility, ability to collaborate and innovate on a competitive global market (Nielsen and Montemari 2012, pp. 145–146). This makes them depen dent on sourcing talent in a labour market where a lack of skilled workers poses a major challenge (Spring Professional 2020). The principles described have many references to the dynamic capabilities on a strategic management level. Peters and Austin (1985) and Moss Kanter (1985) first used the term agility in the 1980s, describing the ability of excellent companies to adapt to market changes – distinguishing them from rigid bureaucratic organisations. Across the following decades, their ideas were picked up by others. According to 1 3 Critical HR capabilities in agile organisations a cross-case analysis in… Doz and Kosonen (2008, p.96) strategic agility results from the combination of three major meta-capabilities: Doz and Kosonen (2008, p.96) strategic agility results from the combination of three major meta-capabilities: (1) Strategic sensitivity (i.e. a strong externally oriented and internally participative strategy process, a high level of attentiveness, and an intense, and open internal dialogue). (2) Leadership unity (i.e. the ability of the leadership team to make fast, courageous decisions).li ) (3) Resource fluidity (i.e. the internal capability to reconfigure business systems and redeploy resources rapidly). The last aspect particularly relates to the research findings of Teece et al. (2016, p.18). 2.4 HR capabilities for a dynamic-agile environment Our research looks at Business Agilists through an HR lens. In agile organisations, the HR focus shifts from individuals to whole networks of people in cross-functional relationships (Appelbaum et al. 2017, p. 12). These highly collaborative networks define an organisation’s “capacity of transformation and innovation” (Charbonnier- Voirin 2011, p. 148). While people development and leadership therefore become a major focus when designing agile organisations, HR is still under pressure to prove itself as a driver of agility (Saha et al. 2017, p. 326).ff In this context, McMackin and Heffernan (2021) differentiate between “HR for Agile” and “Agile for HR”. The first concept refers to the design of HR systems to support agile organisational development, while the second refers to applying agile principles to existing HR functions, as a purely operational strategy. This paper 1 3 3 L. S. Mollet, S. Kaudela-Baum focuses on “HR for Agile”: highlighting the critical role of HR capabilities for agile transformation (Bonavia and Marin-Garcia 2011). The result of “HR for Agile” is an agile workforce built around dynamic talent that adapts easily to change and delivers skills in a dynamic way (Heilmann et al. 2020). This is especially important for organisations co-creating together with their business partners (Nielsen and Montemari 2012). An agile workforce enables scal ability by guaranteeing both flexibility and strategic alignment of the skills needed in the organisation (Wright and Snell 1998). It all depends on “the speed and ease with which transitions are made from one aligned human resource configuration to another” (Dyer and Ericksen 2006, p. 12). Accordingly, the relevance of dynamic capability development in HR work cannot be overstated, especially considering the strong tendency of Business Agilists to rely on self-organisation. Based on these theoretical perspectives, the role of the HR profession faces a new set of expecta tions – especially considering the diffusion of staff functions and the embracement of role-based approaches. So, how can these critical capabilities be fostered in agile organisations – if HR does not want to stop at merely adopting “agile lite” (Cappelli and Tavis 2018, p. 47)? 3 Research design As outlined earlier, the existing literature provides little insight into HR aspects in agile organisations of any size (McMackin and Heffernan 2020, p. 11). Chapter 1 established our research interest as identifying which HR practices contribute to busi ness agility and what knowledge potential can be derived from pioneering companies. To illuminate this matter, we opted for an exploratory, data rich design (Eisenhardt and Graebner 2007; Yin 2018). Multiple-case studies (Eisenhardt 1989; Yin 2018) are suitable for studying such a contemporary phenomenon in its real-life context, especially when variables, bound aries and context remain unclear (Eisenhardt and Graebner 2007; Yin 2018). How we selected our cases, gathered and analysed the data is described in the research process below (see Fig. 1 for an overview).i Firstly, we defined the case selection strategy. The companies were chosen accord ing to a purposeful sampling strategy, used for identifying information-rich cases as an effective use of limited resources (Patton 2002). Thus, we were able to pinpoint informants that are experienced with the phenomenon of interest (Elo et al. 2014, p. 4) and willing to communicate practices in a reflective manner (Bernard 2002).i The first two of the following criteria generated a pool of organisations that increased our chances of being able to swiftly identify cases that matched the third criterion – our main area of interest: ● Medium-sized enterprises: Chap. 2.1 showed that SMEs tend to adopt agile frame works quicker than larger firms. The Northern American criteria define business with fewer than 500 employees as SME (Small Business Administration 2020).i ● Medium-sized enterprises: Chap. 2.1 showed that SMEs tend to adopt agile frame works quicker than larger firms. The Northern American criteria define business with fewer than 500 employees as SME (Small Business Administration 2020). F IT i Ch 2 3 f d h IT lik l i l d fi with fewer than 500 employees as SME (Small Business Administration 2020). ● Focus on IT services: Chap. 2.3 found the IT sector likely to include first movers towards business agility. ● Focus on IT services: Chap. 2.3 found the IT sector likely to include first movers towards business agility. 1 3 Critical HR capabilities in agile organisations a cross-case analysis in… p g g y ● State of agility: focus on mature Business Agilists (according to Werder and Maedche’s 2018, see Chap. 2.1). Fig. 1 Overview of the research design Fig. 3 Research design 1 Overview of the research design Fig. 1 Overview of the research design ● State of agility: focus on mature Business Agilists (according to Werder an Maedche’s 2018, see Chap. 2.1). Due to the chosen spotlight on Swiss IT SMEs and the tight-knit community within that field, identifying potential cases was relatively straightforward. Companies on a long-list were screened via openly accessible information, including active participa tion in agile communities (e.g. conferences or publications). Next, we approached promising organisations on the remaining short-list directly, disclosed our research interest and subsequently determined their suitability in informal conversations. Albeit being developed for larger organisations, we found the characteristics of mature organisational agility, as summarised by Aghina et al. (2020) (see Chap. 2.1.), to be useful in this process. Within the scope of this research, we thus selected seven organisations as illumi native or extreme cases (Yin 2018; Eisenhardt and Graebner 2007; Ongwuegbuzie and Leech 2007, p. 113). As mature Business Agilists and reflective practitioners, they are pushing the boundaries of people management by applying agile principles to HR and leadership – showing how it “could be” (Suri 2011, 67). The rareness of such cases makes them suitable for explorative endeavours (Seawright and Gerring 2008, p. 301–302). As our case companies are engaged in the same professional com munities, our sampling strategy minimises variation and aims at replicating findings within (Eisenhardt and Graebner 2007; Guba and Lincoln 1989). At the same time, we maximise case variation through contrary replication (Guba and Lincoln 1989). We do so in including cases with different motives for pursuing business agility, thus illuminating constructs and identifying relationships concerning the phenomenon under investigation (Ridder 2017). 1 3 L. S. Mollet, S. Kaudela-Baum L. S. Mollet, S. Kaudela-Baum Secondly, the study relies on data gathered between January 2017 and February 2019, in no particular order. With interviews being our main data source, we inter viewed a total of 27 board members, HR specialists and leaders who were willing to share insight into their organisational development and people management practices. Within each organisation, at least two different perspectives were obtained: top man agement, and team leader or expert. The interviews can be characterised as problem- centred (Witzel 2000). They followed semi-structured, open-ended guidelines and took the form of guided conversations rather than structured inquiry (Groeben 1990). 3 Research design The communication strategies included a conversational entry, general and specific prompting, and ad hoc questions (Flick 2018). Interviewees were provided with brief information about the research interest two weeks prior to the interview date. The interviews generally lasted 60–90 min, were audio-recorded and subsequently tran scribed – resulting in over 360 pages of text. Additionally, we collected documents (e.g. presentations, guidelines, programmes) from each company. According to Yin (2018), triangulation of data is important to strengthen validity. A detailed overview of our sample can be found in Table 1, with names excluded to ensure confidentiality. Thirdly, the data analysis started with a “first cycle coding” procedure (Miles et al. 2020) after conducting the first interview. Subsequently, the key factors found in all single cases were marked with a series of codes. The simultaneous collection and pre liminary analysis of data helped us to recognise whether data saturation was achiev able within the chosen sample (Elo et al. 2014, p. 5; Francis et al. 2009 p. 1230). Table 1 Within-case analysis of companies and drivers for agile transformation Table 1 Within-case analysis of companies and drivers for agile transformation Table 1 Within-case analysis of companies and drivers for agile transformation Secondly, the study relies on data gathered between January 2017 and February 2019, in no particular order. With interviews being our main data source, we inter viewed a total of 27 board members, HR specialists and leaders who were willing to share insight into their organisational development and people management practices. Within each organisation, at least two different perspectives were obtained: top man agement, and team leader or expert. The interviews can be characterised as problem- centred (Witzel 2000). They followed semi-structured, open-ended guidelines and took the form of guided conversations rather than structured inquiry (Groeben 1990). The communication strategies included a conversational entry, general and specific prompting, and ad hoc questions (Flick 2018). Interviewees were provided with brief information about the research interest two weeks prior to the interview date. The interviews generally lasted 60–90 min, were audio-recorded and subsequently tran scribed – resulting in over 360 pages of text. Additionally, we collected documents (e.g. presentations, guidelines, programmes) from each company. According to Yin (2018), triangulation of data is important to strengthen validity. 1 3 3 Research design A detailed overview of our sample can be found in Table 1, with names excluded to ensure confidentiality.i Thirdly, the data analysis started with a “first cycle coding” procedure (Miles et al. 2020) after conducting the first interview. Subsequently, the key factors found in all single cases were marked with a series of codes. The simultaneous collection and pre liminary analysis of data helped us to recognise whether data saturation was achiev able within the chosen sample (Elo et al. 2014, p. 5; Francis et al. 2009 p. 1230). 1 3 Critical HR capabilities in agile organisations a cross-case analysis in… able 2 Coding and categories Table 2 Coding and categories Upon completion of all data collection, the initial codes were grouped into catego ries (i.e. collection of codes of similar content) in a “second cycle coding” procedure (Miles et al. 2020) to make them more usable. These categories were grouped along four analytical dimensions: (1) motives to pursue business agility (Chap. 4.1), (2) challenges of organising agility (Chap. 4.2), (3) emerging HR roles and practices, and (4) emerging tensions during the agilisation of HR work (Chap. 4.3). In each dimen sion, we first looked for patterns in individual case data, before comparing them in a cross-case analysis (Cresswell 2007; Yin 2018). While the within-case analysis was performed using MAXQDA, the cross-case analysis was based on a tabular analysis by hand. Throughout this process, we used other sources of data to verify whether our observations could be grounded. Themes that could not be confirmed across all seven companies were dropped from further analysis. This form of content analysis allowed us to see and interpret the data within its context (Kohlbacher 2006). Thus, the finding of patterns and commonalities was the main aim of our analysis. Table 2 shows the initial codes, how they are grouped into categories and also indicates codes that were not pursued further. At this stage of the analysis, we “enfolded” (Eisenhardt 1989) literature to link the identified challenges to theoretical and practical implications (Chaps. 4 and 5), in an endeavour to go beyond a mere description of the emergent practice. 1 3 3 L. S. Mollet, S. Kaudela-Baum 4 Empirical findings After proposing our Business Agilist typology, we summarise general challenges of managing agile transformation, HR roles and practices. While doing so, we bear in mind that any observation can be only preliminary in the study of this recent phenom enon – which is why Chap. 5 includes a further research outlook. Finally, this chapter outlines tensions that surface in the wake of agile HR. 4.1 Business agility typology: identifying patterns for pursuing agility Albeit being considered Business Agilists and exhibiting mutual challenges, the cases show unique agility drivers. These cultural patterns may be considered when creating frameworks for agile HR. 4.1.1 Learners: from agile methods to the ultimate agile mindset for close collaboration, continuous reflection and improvement) and externally (e.g. for networking and (employer) branding purposes). 4.1.1 Learners: from agile methods to the ultimate agile mindset Learners include companies A, B, C and E, who consciously acquired agile methods and tools and are an active part of the agile community. Learners strive for continu ous improvement and broad implementation of agile principles. Their enthusiasm is grounded in anti-hierarchy, anti-status thinking and an anti-bureaucracy movement Learners include companies A, B, C and E, who consciously acquired agile methods and tools and are an active part of the agile community. Learners strive for continu ous improvement and broad implementation of agile principles. Their enthusiasm is grounded in anti-hierarchy, anti-status thinking and an anti-bureaucracy movement – rooted in agile values. The agile transformation is driven by learning agile methods and exchanging experiences. Agility may be understood more as a management trend rather than a strategic response to complex challenges (Teece et al. 2016). The focus is on maximising customer centricity and, above all, being attractive to a new genera tion of employees (i.e. generation Y). Learners participate intensively in publications and forums, wanting to stay avant-garde. Community-building is at the core of HR practices, which includes helping employees to find their voice internally (e.g. for close collaboration, continuous reflection and improvement) and externally (e.g. for networking and (employer) branding purposes). g y g y – rooted in agile values. The agile transformation is driven by learning agile methods and exchanging experiences. Agility may be understood more as a management trend rather than a strategic response to complex challenges (Teece et al. 2016). The focus is on maximising customer centricity and, above all, being attractive to a new genera tion of employees (i.e. generation Y). Learners participate intensively in publications and forums, wanting to stay avant-garde. Community-building is at the core of HR practices, which includes helping employees to find their voice internally (e.g. for close collaboration, continuous reflection and improvement) and externally (e.g. for networking and (employer) branding purposes). – rooted in agile values. The agile transformation is driven by learning agile methods and exchanging experiences. Agility may be understood more as a management trend rather than a strategic response to complex challenges (Teece et al. 2016). The focus is on maximising customer centricity and, above all, being attractive to a new genera tion of employees (i.e. generation Y). Learners participate intensively in publications and forums, wanting to stay avant-garde. Community-building is at the core of HR practices, which includes helping employees to find their voice internally (e.g. 4.1.2 Entrepreneurs: agile frameworks strengthen autonomy and accountability For Entrepreneurs (company D), agility is a means of promoting entrepreneurship by empowering individuals – essentially turning them into a swarm of CEOs. Dis tributed and entrepreneurial management qualities directly tune into the dynamic capabilities of an organisation (Teece et al. 2016; Meyer et al. 2017). Being an active part of the agile community is not a driver, which makes the Entrepreneur much less ideological. After all, the entrepreneurial vision follows economic reasons. As a manager from company D elaborates: “we approached it very simple. We didn’t want to turn it into a religious debate”. Continuous innovation through collaboration and self-reliance ensures the company’s market position. The HR focus lies on support ing employees in dealing with extreme transparency, taking over responsibility and conflict management. In recruitment, attention is paid to resilience, independence, and initiative, while agility is also used to convey employer branding. 1 3 3 Critical HR capabilities in agile organisations a cross-case analysis in… 4.2.1 Challenge 1: shared commitment The importance of involving all organisational layers was present across all inter views: current decision-makers must support the journey towards agility and demand that everyone takes part. Selective implementation amplifies tensions and contra dicts the need for alignment via value and principles. This all-or-nothing attitude is reflected by extensive training in order to establish a common language: “If you want to make that transformation, you must have a commitment across the board” (C, I1). 4.1.3 Values-driven: purpose and agile values aligned Values-driven companies F and G are interested in ideological, human-centred aspects, rather than perfecting agile methods. Mindful of the expectations of a new generation of employees, they prioritise flexibility, participation, co-creation, free dom, fairness, and respect. The shared sense of purpose and a network of empowered teams (Aghina et al. 2020) are at the core of these organisations. A manager from company G explains this new meaning of flexibility: " Gen Y, they’re different… As a Swiss or Central European, I can quit today, go abroad for 6 months, come back and if I’m reasonably well educated, there’s a 99% chance I’ll have a good job again. We also have more part-time employment, where you pursue other passions”. The agile mindset coincides with these values, creating a symbiosis between these companies’ own value orientation and the agile movement. Ensuring a value-fit is an important aspect of their recruitment process, as well as creating a human-centric work environ ment that lives up to their declared values. 4.2 Organising Business agility: identifying patterns of challenges From a systems theory perspective, the challenges these organisations face can be linked to replacing hierarchical structures with self-organisation – opening up oppor tunities for HR agents to co-create possible solutions, as explained in Chap. 4.3. 4.3 Overcoming challenges: incorporating agile capabilities into HR capabilities As another common theme, tasks and topics commonly associated with HR are realised as a shared responsibility between employees, leaders, and HR specialists. In short, distributed HR capabilities contribute to creating the right cultural environ ment for mechanisms such as self-organisation and shared decision-making to come into play. 4.2.2 Challenge 2: committing to transparency and openness Transparency and openness represent cornerstones of implementing business agil ity, as reflected in the following statement: “what must be certain is an openness to new things. And with it, interest, and willingness to learn” (B, I3). On an indi vidual level, transparency means communicating openly and being authentic. On an organisational level, transparency manifests in an error culture of sharing mistakes and learnings. Transparency also means forgoing information-asymmetry – a crucial aspect of self-organisation (Parker et al. 2015). Individuals can only make decisions in the company’s interest with access to all relevant information, as the head of HR of company F explains. On a normative level, transparency becomes a “weapon” that disarms individuals and empowers the collective. Consequently, transparency is a condition for employees to become co-entrepreneurs (Teece 2017). 1 3 L. S. Mollet, S. Kaudela-Baum 4.2.3 Challenge 3: building (trusting) communities All interviewees stressed the importance of employees proactively networking and contributing to the development of the company. In this context, the desired culture is often compared to a family. This entails building trusting relationships, a prerequisite to self-organised teams (Polley and Ribbens 1998): “Trust. I believe this is the basis for agility. Because in agility, you try to get rid of as much overhead as possible” (E, I1). Employees should not compete for status, but instead support each other in mak ing an impact. As authority shifts, individuals must also be coached and advised in exercising their power. 4.2.4 Challenge 4: shaping fast yet positive changes Establishing a strong learning culture is linked to embracing failure as a part of the process, while rapid prototyping fosters problem-solving and innovation. This prag matic approach is an anti-thesis to bureaucracy and status thinking. For a senior man ager at company G, experimenting is based around the criterion of “safe enough to try”. Following continuous improvement necessitates a high degree of self-reflection and vulnerability – a dimension of agile culture that Rebentisch et al. (2018) omit. 4.3.1 Agile HR as distributed capability and shared responsibility Self-attribution of leadership includes being a coach, enabler, facilitator, consultant – all in line with servant leader ship, a concept that upends traditional power relationships sees leaders in supporting roles. Many organisations report this role-shift poses a major challenge for lead ers who were socialised differently. Leadership role-holders contribute to individual and collective self-organisation, aiming to generate fewer interfaces and enable fast decision-making close to the expertise. Above all, this requires relationships at eye level, as a founder of company D stresses. A manager from company A states that “it is a matter of creating a framework. Within in, you’re simply there for questions” (A, I3). This conception exceeds the pre-existing definitions for agile leadership as a mere tool for accelerating decision-making (Doz and Kosonen 2008). 4.3.1.3 Role attribution of HR specialists: expert knowledge owners and network ers In agile organisations, HR specialists are expected to equally coach all organ isational members – a stark contrast to HR business partners that solely serve line management (Ulrich 1997). An HR specialist in company D speaks of offering a pool of expertise that the organisation can tap into, without forcing it. In line with their diverse customers, this new multifaceted role might include knowledge management, stages along the employee lifecycle, community-building or organisational develop ment. These aspects centre around nourishing a culture of learning and sharing, with an error culture that reframes mistakes as learning opportunities. Routine tasks are sought to be automatised and integrated into a seamless employee experience. 4.3.1 Agile HR as distributed capability and shared responsibility The incorporation of agile capabilities into HR work takes place in a distributed way, in a triangle of responsibility between employee, leader, and HR specialist (see Fig. 2). The term HR specialist, in our understanding, includes organisational devel opment competencies, as these two fields of expertise seem inextricably linked in all companies studied. Compared to previous models (Appelbaum et al. 2017) responsi bility largely shifts to individual employees: while leaders and HR specialists might offer to coach or moderate, it is ultimately up to every organisational member to take initiative. Against this background, we suggest a terminology change from HR to (agile) people management (PM) – reflecting the more comprehensive, integrative and collaborative approach to distributed HR work. 4.3.1.1 Role attribution of employees: intrapreneurs With a high level of responsi bility for shaping their role(s) and individual development, employees are required to communicate and engage with feedback openly and constructively. An interviewee from company B mentions: “the co-responsibility of employees is a crucial aspect” (B, I1). Individuals need to be actively supported in developing their decision-mak ing, self-reflection and communication skills. 1 3 Critical HR capabilities in agile organisations a cross-case analysis in… 4.3.1.2 Role attribution of leaders: on-demand coaches The image of the heroic leader is replaced with role-based leadership-systems. Self-attribution of leadership includes being a coach, enabler, facilitator, consultant – all in line with servant leader ship, a concept that upends traditional power relationships sees leaders in supporting roles. Many organisations report this role-shift poses a major challenge for lead ers who were socialised differently. Leadership role-holders contribute to individual and collective self-organisation, aiming to generate fewer interfaces and enable fast decision-making close to the expertise. Above all, this requires relationships at eye level, as a founder of company D stresses. A manager from company A states that “it is a matter of creating a framework. Within in, you’re simply there for questions” (A, I3). This conception exceeds the pre-existing definitions for agile leadership as a mere tool for accelerating decision-making (Doz and Kosonen 2008). Fig. 2 Agile people manage ment as shared responsibility Fig. 2 Agile people manage ment as shared responsibility 4.3.1.2 Role attribution of leaders: on-demand coaches The image of the heroic leader is replaced with role-based leadership-systems. 4.3.2 Critical HR capabilities for agile organisations As a next step, we describe key approaches of how companies overcome agile man agement challenges in the responsibility triangle introduced above. 3 3 L. S. Mollet, S. Kaudela-Baum 4.3.2.1 Recruiting and employer branding In their proactive employer branding, agile stands for a modern working environment and plays a vital part in finding employees with the best cultural fit in a dried-up labour market. Company D, for example, calls itself “a closed group” interested in the values of potential joiners. Positively connotated agile values such as personal responsibility, autonomy and participation are marketed to candidates (matching new employee expectations, see Chap. 1), whereas attributes such as discipline and reliability are less highlighted – presenting the agile culture (Rebentisch et al. 2018) in a starkly positive light. Hiring needs and recruiting decisions are often established collectively. 4.3.2.2 Agile (self-)learning These organisations fully embrace that their agil ity ultimately depends on the ability of their members to transform and innovate (Charbonnier-Voirin 2011). Handing over budget and responsibility for individual development allows individuals to tailor their learning to their current needs and interests, for instance via internal academies, ideation hubs or other learning com munities. A manager from company A explains that “you have to get away from tradi tional HR development concepts, you have to move in the direction of self-learning” (A, I3). These approaches follow the maxim that the autonomy of self-organising social systems cannot be orchestrated (Malik and Probst 1984; Luhmann 19956) – but that its members can be supported in navigating uncertainty, sharing insight and community-building. 4.3.2.3 Agile people development and career planning These organisations depend on their knowledge workers to continuously expand their expertise. Accordingly, they have a strong interest in facilitating alternative career paths that allow individuals to develop tomorrow’s skills and follow new opportunities). Instead of rigid, pre-defined career paths, agile organisations understand careers as individual, strength- and inter est-based, self-driven and everchanging portfolios of roles. Accordingly, companies implement flexible working conditions (e.g. unpaid leave, remote or part-time work). Whereas job labels might be very limited in such organisations, the opportunity for individual skill development is abundant. 4.3.3 Tensions when strengthening agile people management 4.3.3.1 Fear of losing control Whereas line management legitimises itself through hierarchy, agile leadership is more fluid and leading by virtue of competence. Being stripped off their title, many former line managers fear to become irrelevant, failing to see the opportunities suddenly available (e.g. new tasks or coaching roles). 4.3.3.2 Overwhelmed employees Especially in early stages of a transformation, employees can feel rather overwhelmed. While not uncommon for any change pro cess, this might be amplified by the new maxim of self-organisation and increased decision-making capacity. The significant psychological adaption presupposes a high degree of self-reflection. Some employees leave the organisation as a result, underlin ing the importance of recruiting for cultural fit. A founder of company D highlights that some individuals need to be supported in learning how to navigate self-organ isation, without changing the system itself – respecting its autonomy. The new career paradigm can also be frustrating, as many organisations are only starting to explore the systematics of organising and recognising alternative career paths. 4.3.3.1 Fear of losing control Whereas line management legitimises itself through hierarchy, agile leadership is more fluid and leading by virtue of competence. Being stripped off their title, many former line managers fear to become irrelevant, failing to see the opportunities suddenly available (e.g. new tasks or coaching roles). 4.3.3.2 Overwhelmed employees Especially in early stages of a transformation, employees can feel rather overwhelmed. While not uncommon for any change pro cess, this might be amplified by the new maxim of self-organisation and increased decision-making capacity. The significant psychological adaption presupposes a high degree of self-reflection. Some employees leave the organisation as a result, underlin ing the importance of recruiting for cultural fit. A founder of company D highlights that some individuals need to be supported in learning how to navigate self-organ isation, without changing the system itself – respecting its autonomy. The new career paradigm can also be frustrating, as many organisations are only starting to explore the systematics of organising and recognising alternative career paths. 1 3 Critical HR capabilities in agile organisations a cross-case analysis in… 4.3.3.3 In search of the human factor In an agile context, individuals need to con stantly weigh their decisions against values and principles, instead of being able to follow a simple set of rules. Somewhat paradox, a high degree of self-organisation hence requires a high degree of alignment to ensure coherence. 5 Implications and conclusions While looking at agile organisations through a systems theory lens revealed benefi cial cultural and structural interventions, there were no implications for the develop ment of agile HR capabilities. Literature on existing agile management approaches discusses the strategic and team level, yet again largely overlooks the role of HR. In order to close this gap, key management challenges across different pioneering Business Agilists were condensed, and critical HR capabilities across all types were described and interpreted. The theoretical and practical implications are now sum marised below. 4.3.3 Tensions when strengthening agile people management A senior manager in company G acknowledges that many current agile management systems omit human aspects, and that it is therefore vital to embedded them in the culture and leadership system, regardless of the chosen framework. 5.1.1 A proposed typology for business agilists Despite our focus on common patterns in first-movers, we noticed slight nuances in their manifestation of agile HR capabilities. The proposed typology offers the oppor tunity to expand the emerging theory beyond the mere labelling of organisations as agile. Learners are highly invested in creating a new HR narrative that is com patible with their agile vocabulary, essentially developing new roles around people development while actively unlearning hierarchical behaviours. Entrepreneurs adopt a radical stance, subordinating all decisions to the maxim of entrepreneurship. HR specialist roles often vanish completely, with every organisational member emerg ing as a self-organising unit. Value-driven organisations are predominantly interested in cultural development. Their focus lies on employee wellbeing and motivation, on creating a supportive environment where people feel valued. Removing the HR function is not a priority, yet there is often a noticeable shift towards role-based and distributed approaches. 5.1.2 HR as a (radical) distributed practice Practice has shown that business agility simultaneously challenges on a team man agement level and a strategic management level, requiring all organisational mem bers to engage with agile values and principles – a crucial task that can only be fulfilled by strengthening HR capabilities. The development of these HR capabilities takes place as a collective knowledge development process (Laloux 2014; Maximini 2018). An important step is the decentralisation of HR work from functional special 3 3 L. S. Mollet, S. Kaudela-Baum ists to teams, employees, leaders, and tech solutions alike. One might therefore speak of an HR sharing approach. HR is no longer a strategic servant, but – as a shared practice – an inherent part of organisational, cultural and innovation development (Worley and Pillans 2019). By supporting and scaling the skills development needed to sense and seize opportunities, HR capabilities directly contribute to ensuring con tinuous transformation in the organisation (Teece et al. 2016; see Chap. 2.3). 5.2 Practical implications The cases studied show how Business Agilists may fulfil the threefold purpose of agile HR by (1) contributing to the structural and cultural conditions around self- organisation, (2) supporting continuous individual and collective learning, and (3) fostering intra- and entrepreneurial behaviour. In short, there is no best-practice approach to “HR for Agile” (McMackin and Heffernan 2020), and every organisa tion is urged to co-create context-specific HR practices. However, to overcome the identified challenges, there are crucial elements all organisations may benefit from incorporating. 5.1.4 Dynamic (self-)learning processes over HR programmes Capable individuals, with their willingness and ability to learn, are at the centre of organisational development. HR capabilities may support them in this endeavour in a flexible way. The strategic role of HR (Brockbank and Ulrich 2009; Ulrich 1997) gives way to a pragmatic, development-focused role – in line with a renaissance of the concept of organisational learning (Argyris and Schön 1978). While HR exper tise is only involved when needed, daily HR work is done by highly autonomous teams, empowered individuals, and servant leaders (Luhmann 1995; Probst 1987). Providing direct support instead of being solely programmatic, HR is no longer alien ated from the business. 5.1.3 Network and role-based approach to HR Agile organisations often choose a role-based approach, where self-reliant individ uals are encouraged to curate their own path of professional and personal devel opment. This stresses the need for a re-definition of occupational profiles and role descriptions. Even the term HR specialist may be critically reviewed in the transition towards a more meritocratic or distributed approach. Furthermore, for many SME, the creation of dedicated in-house roles might not prove feasible. 1 3 5.2.1 Transparency as a precondition in the agility-stability paradox It soon becomes evident that many of the tensions correspond to what is known as the stability-agility paradox (Smith and Lewis 2011). Constantly re-calibrating itself to current needs, an agile organisation needs to continuously balance exploring with exploitation in a highly dynamic environment, rendering over-simplified cookbook recipes impossible. Considering this, transparency becomes even more important, 1 3 Critical HR capabilities in agile organisations a cross-case analysis in… also in the sense of shared values, goals and intensions and fostering leadership skills in all members of the organisations (Moravec 1999). HR work can only be distributed across the organisation if information is shared openly and thus allows decision- making in line with company interests (Laanti et al. 2011). 5.2.3 Clarifying roles and considering network and role-based approaches In line with the need for a re-definition of HR roles, there are many opportunities for organisations to experiment with distributing HR work. HR specialists might take on roles involved in client work or be a part of an inter-organisational people man agement ecosystem, alongside their in-house HR accountabilities. In doing so, they not only broaden their own competences, but gain insights into customer needs and strengthen relationships. Especially for SMEs, this might be a way to mitigate the financial pressure of having in-house HR expertise. 5.2.4 Integrating HR in servant leadership practices The maxim of self-reliance may amplify interpersonal conflict. Therefore, HR capa bilities must adjust to dealing with contradiction, responsibility and independence. This aligns with established change management role in terms around resistance to change (Senge et al. 1999). The co-development of HR capabilities in the HR respon sibility triangle also goes together with the idea of servant leadership, where leaders enable individuals and teams to navigate autonomy instead of micromanaging them (Parker et al. 2015, p. 119). Leadership becomes a major contributing factor to creat ing an organisation where individuals are seen as trustworthy, resourceful and intrin sically motivated (Appelbaum et al. 2017, p. 12). As a result, leadership roles may incorporate many accountabilities traditionally associated with HR. 5.2.2 Integrating the human factor into the system When agile organisations are overly focused on structure or methods (Fernandez and Fernandez 2008), they fail to include the human nature into their system – for instance relying completely on employee proactivity. Business Agilists benefit from developing human-centric concepts that support individuals in unleashing their full potential. These concepts are characterised by individuals experimenting with role(s) based on their competencies and interest in taking over responsibility in a certain area. Pairing support with granting autonomy in turn acts as an incentive for individu als to show initiative and ultimately behave as co-entrepreneurs. 5.3 Conclusions This study addresses two gaps identified in the existing literature: firstly, the lack of implications for developing agile HR capabilities from a systemic-holistic perspec tive, and secondly, the lack of HR concepts in the emerging agile management lit erature. The framing of distributed HR as a relevant dimension of strengthening the dynamic capabilities within agile organisations is a major contribution of this paper. 3 3 L. S. Mollet, S. Kaudela-Baum Following this idea might lead to an establishment of people management ecosystems (Denning 2015), alongside empowering individuals to embrace own HR capabilities and integrating HR into servant management practices. This servant attitude does not render agile HR a mere service provider, but instead a vital contributor to organ isational development – a force of entrepreneurship and creativity. Furthermore, the strategic role of HR in implementing business agility strategies clearly revolves around supporting learning processes at every organisational level. Following this idea might lead to an establishment of people management ecosystems (Denning 2015), alongside empowering individuals to embrace own HR capabilities and integrating HR into servant management practices. This servant attitude does not render agile HR a mere service provider, but instead a vital contributor to organ isational development – a force of entrepreneurship and creativity. Furthermore, the strategic role of HR in implementing business agility strategies clearly revolves around supporting learning processes at every organisational level. 6 Limitations and future research As outlined in the theoretical background, the IT sector is often at the forefront of agile transformation, which led to our purposeful case selection. The application of the study’s results to other sectors therefore may be investigated in further research. The same applies for countries that are not comparable to the economic and cultural parameters of Switzerland. The transferability of the results is therefore limited. Fur thermore, we focused on SME ranging from 150 to 300 employees. Considering SME’s scarcity of management resources, their often people-centric business mod els, reliance on leaders as well as often tacit HR knowledge (Klaas et al. 2012), it can be concluded that the results would have to be verified for other organisation sizes. However, emerging results, such as the potential of HR capabilities in assisting community-building within the organisation, might be just as interesting for larger organisations. The relatively small number of cases in this explorative study might also be expanded in a follow-up study, which could then provide the basis for a dif ferentiated assignment of HR capabilities to the proposed Business Agility types. Moreover, amplifying the perspective of employees may prove fruitful in further research, quasi as recipients of the agile initiatives. Our analysis also excludes topics that did not show a pattern, as we were interested in typical issues relevant to all the companies studied in our search for critical capabilities. Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/ licenses/by/4.0/. Aghina W, Handscomb C, Ludolph J et al (2020) Enterprise agility: Buzz or business impact. McKinsey (March 20). https://www.mckinsey.com/business-functions/organization/our-insights/enterprise- agilitybuzz-or-business-impact. Accessed 15 December 2021 Funding Open access funding provided by Lucerne University of Applied Sciences and Arts. References 1 3 1 3 Critical HR capabilities in agile organisations a cross-case analysis in… Appelbaum SH, Calla R, Desautels D, Hasan L (2017) The challenges of organisational agility (part 1). Industrial and Commercial Training 49:6–14 g Arbussa A, Bikfalvi A, Marquès P (2016) Strategic agility-driven business model renewal: the case of an SME. Manag Decis 55:271–193 SME. Manag Decis 55:271–193 Argyris C, Schön D (1978) Organizational Learning - A Theory of Action Perspective. Addison-Wesley, Readingf Atuahene-Gima K (2003) The effects of centrifugal and centripetal forces on product development speed and quality: how does problem solving matter? Acad Manag J 46:359–373 Bernard HR (2002) Research methods in anthropology: Qualitative and quantitative approaches, 3rd edn. Alta Mira Press, Walnut Creek Bernard HR (2002) Research methods in anthropology: Qualitative and quantitative approaches, 3rd edn. Alta Mira Press, Walnut Creek Alta Mira Press, Walnut Creek Bonavia T, Marin-Garcia JA (2011) Integrating human resource management into lean production an their impact on organizational performance. Int J Manpow 32:923–938 Brockbank W, Ulrich D (2009) The HR business-partner model: Past learnings and future challenge People & Strategy 32:5–8 p gy Cappelli P, Tavis A (2018) HR goes agile. Harvard Business Rev 96:46–52 Charbonnier-Voirin A (2011) The development and partial testing of the psychometric properties of a measurement scale of organisational agility. M@n@gement 14:120–156 Coates TKL (2017) Hearing the voices of Generation Y employees: a hermeneutic phenomenologic study. Hum Resource Dev Int 20:37–67i Cresswell JW (2007) Qualitative inquiry and research method: Choosing among five approaches, 2nd ed Sage, Thousand Oaksifi g Damanpour F (2010) An integration of research findings of effect of firm size and market competition on product and process innovations. Br J Manag 21:996–1010 p p g Denning S (2015) Agile: it’s time to put it to use to manage business complexity. Strategy & Leadership 43:10–17 Doz Y, Kosonen M (2008) The dynamics of strategic agility: Nokia’s rollercoaster experience. Ca Manag Rev 50:95–118 Dyer L, Ericksen J (2006) Dynamic Organizations: Achieving Marketplace Agility Through Workforce Scalability, in: CAHRS Working Paper:6–12 Elo S, Kääriäinen M, Kanste O, Pälkki, Utriainen K, Kyngäs H (2014) Qualitative content analysis: a focus o trustworthiness. SAGE Open January-March 2014:1–10. https://doi.org/10.1177/215824401452263 Eisenhardt KM (1989) Building theories from case study research. Acad Manage Rev 14:532–55 Eisenhardt KM, Graebner ME (2007) Theory building from cases: opportunities and challenges. References Ac Manag J 50:25–32 Endenburg G, Pearson M (1998) Sociocracy as social design: Its characteristics and course of develop ment, as theoretical design and practical project. Eburon, Delft Fernandez DJ, Fernandez JD (2008) Agile Project Management: Agilism versus traditional approaches. J Comput Inform Syst 49:10–17 Flick U (2018) An introduction to qualitative research. Sage, London Fernandez DJ, Fernandez JD (2008) Agile Project Management: Agilism versus traditional approaches Comput Inform Syst 49:10–17 Flick U (2018) An introduction to qualitative research. Sage, London Francis JJ, Johnston M, Robertson C, Glidewell L, Entwistle V, Eccles MP, Grimshaw JM (2009) What is an adequate sample size? Operationalising data saturation for theory-based interview studies. Psy chol Health 25:1229–1245 Groeben N (1990) Subjective Theories and the Explanation of Human Action. In: Semin GR, Gergen KJ ( d ) E d U d di S i l d S i ifi I li i S L d 19 44 Groeben N (1990) Subjective Theories and the Explanation of Human Action. In: Semin GR, Gergen KJ (eds) Everyday Understanding: Social and Scientific Implications. Sage, London, pp 19–44 ( ) j p , g (eds) Everyday Understanding: Social and Scientific Implications. Sage, London, pp 19–44 Guba E, Lincoln YS (1989) Fourth Generation Evaluation. Sage, Newbury Park Harraf A, Wanasika I, Tate K, Talbott K (2015) Organisational Agility. J Appl Bus Res 31:675–686 Harraf A, Wanasika I, Tate K, Talbott K (2015) Organisational Agility. J Appl Bus Res 31:675–686 Heilmann P, Forsten-Astikainen R, Kultalahti S (2020) Agile HRM practices of SMEs. J Small Bus Man age 58(6):1291–1306 Heilmann P, Forsten-Astikainen R, Kultalahti S (2020) Agile HRM practices of SMEs. J Small Bus Ma age 58(6):1291–1306 Kettunen P, Laanti M, Fagerholm F, Mikkonen T (2019) Agile in the Era of Digitalization: A Finnish Survey Study. In: Franch X, Männistö T, Martínez-Fernández S (eds) Product-Focused Software Process Improvement. PROFES 2019. Lecture Notes in Computer Science, vol 11915. Springer, Cham. https://doi.org/10.1007/978-3-030-35333-9_28 p g _ aas BS, Semadeni M, Klimchak M, Ward A-K (2012) High-performance work system implementation in small and medium enterprises: A knowledge-creation perspective. Hum Resour Manag 51:487–510 Klaas BS, Semadeni M, Klimchak M, Ward A-K (2012) High-performance work system implementation in small and medium enterprises: A knowledge-creation perspective. Hum Resour Manag 51:487–510 Kohlbacher F (2006) The use of qualitative content analysis in case study research. References Int J Productivity Perform Manage 64:112–128 Patton MQ (2002) Qualitative research and evaluation methods, 3rd edn. Sage, Thousand Oaksf Peters TJ, Austin N (1985) A passion for excellence. The leadership difference. Random House, New York Peters TJ, Austin N (1985) A passion for excellence. The leadership difference. Random House, New York Polley D, Ribbens B (1998) Sustaining self-managed teams: a process approach to team wellness. Team Perform Manage 4:3 21 Polley D, Ribbens B (1998) Sustaining self-managed teams: a process approach to team wellness. Team Perform Manage 4:3–21 Probst G (1987) Selbst-Organisation. Ordnungsprozesse in sozialen Systemen aus ganzheitlicher Sich Parey, Berlin y Rauch A, Hatak I (2016) A meta-analysis of different HR-enhancing practices and performance of small and medium sized firms. J Bus Ventur 31:485–504 Ridder HG (2017) The theory contribution of case study research designs. Bus Res 10:281–3 er HG (2017) The theory contribution of case study r Rigby DK, Sutherland J, Takeuchi H (2016) Embracing agile. Harvard Business Rev 94:40–50 Rebentisch E, Schuh G, Dölle C, Mattern C, Abel H (2018) Defining Agile Culture Using Topic Model ling. DS 92: Proceedings of the DESIGN 2018, 15th International Design Conference: 2111–2120 Robertson BJ (2015) Holacracy: The revolutionary management system that abolishes hierarchy Penguin Rebentisch E, Schuh G, Dölle C, Mattern C, Abel H (2018) Defining Agile Culture Using Topic Model ling. DS 92: Proceedings of the DESIGN 2018, 15th International Design Conference: 2111–2120 ling. DS 92: Proceedings of the DESIGN 2018, 15th International Design Conference: 2111–2120 Robertson BJ (2015) Holacracy: The revolutionary management system that abolishes hierarchy. Penguin UK, London Robertson BJ (2015) Holacracy: The revolutionary management system that abolishes hierarchy. Penguin UK, Londoni Saha N, Gregar A, Sáha P (2017) Organisational Agility and HRM strategy: Do they really enhance firms’ competitiveness? Int J Organ Leadersh 6:323–334 Seawright J, Gerring J (2008) Case selection techniques in case study research: a menu of qualitative and quantitative options. Polit Res Q 61:294–308 Senge PM Roberts C Ross R Roth G Smith B (1999) The Dance of Change Nicholas Brealey London q p nge PM, Roberts C, Ross R, Roth G, Smith B (1999) The Dance of Change. Nicholas Brealey, London g , , , , ( ) g y, Siakas KV, Siakas E (2007) The agile professional culture: a source of agile quality. Softw Process Improv Pract 12:597–610. References Forum: Qualitative Social Research 7:21 Kohlbacher F (2006) The use of qualitative content analysis in case study research. Forum: Qualitative Social Research 7:21 Laanti M, Salo O, Abrahamsson P (2011) Agile methods rapidly replacing traditional methods at Nokia: A survey of opinions on agile transformation. ‎Inf Softw Technol 53:276–290 1 3 L. S. Mollet, S. Kaudela-Baum Laloux F (2014) Reinventing organizations: A guide to creating organizations inspired by the next stage in human consciousness. Nelson Parker, Brussels Laloux F (2014) Reinventing organizations: A guide to creating organizations inspired by the next stage in human consciousness. Nelson Parker, Brussels , Luhmann N (1995) Social Systems. Stanford University Press, Stanford alik F, Probst G (1984) Evolutionary management. In: Ulrich H, Probst (eds) G. Self-organization and management of social systems. Springer, Berlin/Heidelberg, pp 105–120f cMackin J, Heffernan M (2020) Agile for HR: Fine in practice, but will it work in theory? Hum Resource Manage Rev 31:100791 Maximini D (2018) Scrum Culture. Introducing agile methods in organizations, 2nd edn. Springer, Chami Maximini D (2018) Scrum Culture. Introducing agile methods in organizations, 2nd edn. Springer, Cham Meyer CR Cohen DG Nair S (2017) Some have to and some want to: Why firms adopt a post-industrial Maximini D (2018) Scrum Culture. Introducing agile methods in organizations, 2nd edn. Springer, Cham Meyer CR, Cohen DG, Nair S (2017) Some have to, and some want to: Why firms adopt a post-industrial form. J Manage Gov 21:533–559 Meyer CR, Cohen DG, Nair S (2017) Some have to, and some want to: Why firms adopt a post-industrial form. J Manage Gov 21:533–559 Miles MB, Huberman MA, Saldana J (2020) Qualitative data analysis: a methods sourcebook, 4th ed Sage, Thousand Oaks Moravec M (1999) Self-managed teams: All team members can show leadership skills. Exec Excellen 16:18–18 Moss Kanter (1985) Change masters. Free Press, New York Moss Kanter (1985) Change masters. Free Press, New York ( ) g Nielsen C, Montemari M (2012) The role of human resources in business model performance. The network-based companies. J Hum Resource Costing Acc 16:142–164 p g Ongwuegbuzie A, Leech NL (2007) A call for qualitative power analyses. Qual Quant 41:105–121 r DW, Holesgrove M, Pathak R (2015) Improving productivity with self-organised teams and agile eadership Int J Productivity Perform Manage 64:112 128 Parker DW, Holesgrove M, Pathak R (2015) Improving productivity with self-organised teams and a leadership. Werder K, Maedche A (2018) Explaining the emergence of team agility: a complex adaptive systems per spective. Inform Technol People 31:819–844 Witzel A(2000) The problem-centered interview. Forum: Qualitative Social Research 1(1):Art. 22. http:// nbn-resolving.de/urn:nbn:de:0114-fqs0001228. Accessed 28 December 2020 Worley C, Pillans G (2019) Organization Agility, Performance, and the Role of the HR Function. In: Olbert S, Prodoehl HG (eds) Überlebenselixier Agilität. Springer, Wiesbaden, pp 121–142 Wright PM, Snell SA (1998) Towards a Unifying Framework for Exploring Fit and Flexibility in Strategic Human Resource Management. Acad Manage Rev 23:756–772 Yin R (2018) Case Study Research: Design and Methods, 6th edn. Sage, Los Angeles Publisher’s Note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations Critical HR capabilities in agile organisations a cross-case analysis in… Publisher’s Note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. , ( ) p g g spective. Inform Technol People 31:819–844 References https://doi.org/10.1002/spip.344 Smith WK, Lewis MW (2011) Toward a Theory of Paradox: A dynamic Equilibrium Model of Organizing. Acad Manage Rev 36:381–403 Small Business Administration (2020) Table of Small Business Standards Matched to North American Industry Classification Codes. https://www.sba.gov/sites/default/files/2019-08/SBA%20Table%20 of%20Size%20Standards_Effective%20Aug%2019%2 C%202019_Rev.pdf. Accessed 2 January 2021 Small Business Administration (2020) Table of Small Business Standards Matched to North American Industry Classification Codes. https://www.sba.gov/sites/default/files/2019-08/SBA%20Table%20 of%20Size%20Standards Effective%20Aug%2019%2 C%202019 Rev pdf Accessed 2 January yi p gi of%20Size%20Standards_Effective%20Aug%2019%2 C%202019_Rev.pdf. Accessed 2 January 2021 Spring Professional (2020) Swiss Skills Shortage Index 2020. Available at https://ssi.springprofessional. ch/en/swiss-skills-shortage-index-2020/. Accessed 7 January 2021 ch/en/swiss skills shortage index 2020/. Accessed 7 January 2021 Suri H (2011) Purposeful sampling in qualitative research synthesis. Qualitative Res J 11:63–75i Suri H (2011) Purposeful sampling in qualitative research synthesis. Qualitative Res J 11:63–75 Teece D (2017) Towards a capability theory of (innovating) firms: implications for management and pol Suri H (2011) Purposeful sampling in qualitative research synthesis. Qualitative Res J 11:63–75 T D (2017) T d bilit th f (i ti ) fi i li ti f t d ( ) p p g q y Q Teece D (2017) Towards a capability theory of (innovating) firms: implications for management and po icy. Camb J Econ 41:693–720 Teece D, Peteraf M, Leih S (2016) Dynamic Capabilities and Organizational Agility: Risk, Uncertaint Teece D, Peteraf M, Leih S (2016) Dynamic Capabilities and Organizational Agility: Risk, Uncertaint and Strategy in the Innovation Economy. Calif Manag Rev 58:13–35 Tolchinsky PDP (2015) Accelerating Change: New Ways of Thinking about Engaging the Whole System. Organisation Dev J 33:45–63 Ulrich D (1997) Human Resource Champions. The Next Agenda for Adding Value and Delivering Results. Harvard Business School Press, Boston 1 3 Critical HR capabilities in agile organisations a cross-case analysis in… 1 3
https://openalex.org/W4288385502	https://zenodo.org/records/6845089/files/Thesis.pdf	English	null	Supporting and Verifying Transient Behavior Specifications in Chaos Engineering - Supplementary Materials	Zenodo (CERN European Organization for Nuclear Research)	2,022	cc-by	32,213	Institute of Software Engineering Software Quality and Architecture University of Stuttgart Universitätsstraße 38 D–70569 Stuttgart Masterarbeit Supporting and Verifying Transient Behavior Specifications in Chaos Engineering Denis Zahariev Course of Study: Softwaretechnik Examiner: Dr.-Ing. André van Hoorn Supervisor: Dr.-Ing. André van Hoorn, Sebastian Frank, M.Sc. Mir Alireza Hakamian, M.Sc. Commenced: January 28, 2022 Completed: July 28, 2022 University of Stuttgart Universitätsstraße 38 D–70569 Stuttgart Abstract Context. Chaos Engineering is an approach for investigating the resilience of software systems, i.e. their ability to withstand unexpected events, adapt accordingly, and continue providing functionality. An integral part of the approach is continuous experimentation, expressed in continuously executing so-called Chaos Experiments. A Chaos Experiment consists of two crucial elements, namely a steady-state hypothesis and an anomaly injection. Traditionally, during the experimentation process, the steady-state hypothesis is verified at the start of an experiment, after which the anomaly is injected, followed by a second evaluation of the steady-state hypothesis. This Chaos Experimentation process is well suited for identifying whether a system is in a steady state after a failure is introduced. Problem. When applied, the traditional Chaos Experimentation approach can only verify whether the system is in a steady state without providing any information about the time between the state changes, e.g. the recovery of the system. The experimentation process does not explicitly allow the specification of hypotheses regarding the transient behavior, i.e. the behaviors experienced during the transition between steady states after a failure has been introduced. As a result, the experimentation process also does not support explicit verification of requirements on the transient behavior during the experiment, e.g, whether the response time stays below a given threshold all the time or whether a circuit breaker opens within a given time. Knowledge about such transient behaviors is beneficial for the stakeholders of the system. For example, assuming that high availability is of utmost importance for the business model of an application, a long period of recovery during which the application is unavailable after an unexpected failure could lead to considerable losses for various stakeholders.i Objective. The first objective of the thesis is to examine how a transient behavior requirement can be specified in the context of Chaos Engineering and Chaos Experiments. The next goal of the thesis is to study how the Chaos Experimentation process can include a verification of transient behavior requirements. A further goal is to create a concept for an extended Chaos Engineering approach, which supports the specification of transient behavior hypotheses and their verification, and to implement the concept into a working prototype. After the prototype is developed, the last objective of the thesis is to evaluate the extended Chaos Engineering approach and its implementation.i Method. Supporting and Verifying Transient Behavior Specifications in Chaos Engineering Denis Zahariev the extended approach, allowing them to have stricter guarantees regarding the resilience of the application. Moreover, the prototype implementing the extended approach allows software engineers to easily adopt it and possibly extend it. Kurzfassung Kontext. Chaos-Engineering ist ein Ansatz zur Analyse der Resilienz von Softwaresystemen, d.h. ihrer Fähigkeit, unerwarteten Ereignissen zu widerstehen, sich entsprechend anzupassen und weiterhin ihre Funktionalität bereitzustellen. Ein wesentlicher Bestandteil des Ansatzes ist das kontinuierliche Experimentieren, das sich in der kontinuierlichen Durchführung sogenannter Chaos- Experimente manifestiert. Ein Chaos-Experiment besteht aus zwei wesentlichen Elementen, nämlich einer Steady-State-Hypothese und einer Anomalie-Injektion. Traditionell wird zu Beginn eines Experiments die Steady-State-Hypothese verifiziert, danach wird die Anomalie injiziert, gefolgt von einer zweiten Bewertung der Steady-State-Hypothese. Dieses Chaos-Experimentierverfahren ist gut geeignet, um festzustellen, ob sich ein System nach dem Auftreten einer Störung in einem stabilen Zustand befindet. i Problemstellung. Bei Anwendung des traditionellen Chaos-Engineerings kann nur überprüft werden, ob sich das System in einem stabilen Zustand befindet, ohne Informationen über die Zeit zwischen den Zustandsänderungen, z. B. die Wiederherstellung des Systems, zu liefern. Der Experimentierprozess erlaubt keine explizite Spezifizierung von Hypothesen über das transiente Verhalten, d. h. das Verhalten während des Übergangs zwischen den stabilen Zuständen, nachdem ein Fehler aufgetreten ist. Infolgedessen unterstützt der Experimentierprozess auch keine explizite Verifikation von Anforderungen an das transiente Verhalten während des Experiments, z. B. ob die Antwortzeit die ganze Zeit unter einem bestimmten Wert bleibt oder ob ein Curcuit-Breaker innerhalb einer bestimmten Zeit geöffnet wird. Das Wissen über ein solches transientes Verhalten ist für die Stakeholder des Systems von Vorteil. Geht man beispielsweise davon aus, dass eine hohe Verfügbarkeit für das Geschäftsmodell einer Anwendung von größter Bedeutung ist, könnte eine lange Wiederherstellungsphase, in der die Anwendung nach einem unerwarteten Ausfall nicht verfügbar ist, zu erheblichen Verlusten für verschiedene Stakeholder führen. Zielsetzung. Das erste Ziel dieser Arbeit ist es, zu untersuchen, wie Anforderungen an das transiente Verhalten im Rahmen von Chaos-Engineering und Chaos-Experimenten spezifiziert werden können. Das nächste Ziel der Arbeit ist es, zu untersuchen, wie der Chaos-Experimentierprozess eine Verifikation von Anforderungen an das transiente Verhalten ermöglichen kann. Ein weiteres Ziel ist es, ein Konzept für einen erweiterten Chaos-Engineering-Ansatz zu erstellen, der die Spezifikation von Anforderungen an das transiente Verhalten und deren Verifikation unterstützt, und das Konzept in einen funktionierenden Prototyp umzusetzen. Nach der Entwicklung des Prototyps ist das letzte Ziel der Arbeit die Evaluation des erweiterten Chaos-Engineering-Ansatzes und seiner Implementierung. Methode. Um das erste Ziel der Arbeit zu erreichen, werden Formalismen, die transientes Verhalten beschreiben können, im Hinblick auf ihre Integration in Chaos-Experimente untersucht. Abstract In order to achieve the first objective of the thesis, formalisms capable of describing transient behaviors are examined with regard to their integration into Chaos Experiments. To accom- plish the second objective, state-of-the-art Chaos Engineering tools are studied, and additionally, three expert interviews are conducted in order to elicit the requirements for such an extension of the Chaos Engineering approach. To accomplish the goal of creating a concept for the approach, the research results from the previous goals and the requirements elicited during the interviews are combined into a concept which is then implemented into a prototype. To reach the last goal of the thesis, the approach and the prototype are examined in three separate types of evaluation. Result. First, the results of the thesis include the requirements for an extended Chaos Engineering approach, supporting the specification and verification of transient behavior requirements. Further- more, the results also include a concept for the realization of the extension. Moreover, the results also comprise a functioning prototype of the proposed approach and its evaluation. Conclusion. The proposed extension of the Chaos Engineering approach allows a deeper and more precise analysis of the resilience of software systems by enabling the specification and evaluation of more detailed and strict resilience requirements including transient behavior specifications. Furthermore, various stakeholders such as customers, operators, and developers benefit from iii the extended approach, allowing them to have stricter guarantees regarding the resilience of the application. Moreover, the prototype implementing the extended approach allows software engineers to easily adopt it and possibly extend it. iv Contents 1 Introduction 1 2 Foundations and Related Work 7 2.1 Foundations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7 2.2 Related work . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13 3 Elicitation of Requirements for the Extended Chaos Engineering Approach 15 3.1 Goals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15 3.2 Method . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15 3.3 Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21 4 Extended Chaos Engineering Approach 25 4.1 Transient Behavior Hypothesis Specification . . . . . . . . . . . . . . . . . . . 26 4.2 Transient Behavior Verification . . . . . . . . . . . . . . . . . . . . . . . . . . 28 4.3 Verification Result Visualization . . . . . . . . . . . . . . . . . . . . . . . . . 28 4.4 Graphical User Interface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29 4.5 Transient Behavior Monitoring . . . . . . . . . . . . . . . . . . . . . . . . . . 29 4.6 Interaction of the Proposed Concepts . . . . . . . . . . Contents . . . . . . . . . . . . . 29 5 Prototype Implementation 31 5.1 Chaos Engineering Tool . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31 5.2 Transient Behavior Specification . . . . . . . . . . . . . . . . . . . . . . . . . 32 5.3 Transient Behavior Verification Tool . . . . . . . . . . . . . . . . . . . . . . . 39 6 Evaluation 47 6.1 Correctness Evaluation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47 6.2 Chaos Experiment Evaluation . . . . . . . . . . . . . . . . . . . . . . . . . . . 49 6.3 Evaluation of the Compatibility With Other Approaches . . . . . . . . . . . . . 56 6.4 Evaluation of Research Questions . . . . . . . . . . . . . . . . . . . . . . . . . 61 6.5 Threats to Validity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62 6.6 Limitations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62 7 Conclusion and Future Work 65 7.1 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 65 7.2 Future Work . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Kurzfassung Um das zweite Ziel zu erreichen, werden die state-of-the-art Chaos-Engineering-Tools untersucht und zusätzlich drei Experteninterviews durchgeführt, um die Anforderungen an eine solche Erweiterung des Chaos-Engineering-Ansatzes zu erheben. Zur Erreichung des Ziels, ein Konzept für den Ansatz zu erstellen, werden die Forschungsergebnisse aus den vorangegangenen Zielen und die in den Interviews erhobenen Anforderungen zu einem Konzept zusammengefasst, das anschließend in einem Prototyp umgesetzt wird. Um das letzte Ziel der Arbeit zu erreichen, werden der Ansatz und der Prototyp in drei verschiedenen Evaluationen untersucht. Ergebnis. Die Ergebnisse der Arbeit beinhalten zum einen die Anforderungen an einen erweiterten Chaos-Engineering-Ansatz, der die Spezifikation und Verifikation von Anforderungen an das transiente Verhalten unterstützt. Des Weiteren beinhalten die Ergebnisse auch ein Konzept zur v Realisierung der Erweiterung. Weiterhin umfassen die Ergebnisse auch einen funktionierende Prototyp des vorgeschlagenen Ansatzes und dessen Evaluation. Fazit. Die vorgeschlagene Erweiterung des Chaos-Engineering-Ansatzes ermöglicht eine tiefere und präzisere Analyse der Resilienz von Softwaresystemen, indem sie die Spezifikation und Evaluation von detaillierteren und strengeren Resilienzanforderungen mit Anforderungen an das transiente Verhalten ermöglicht. Darüber hinaus profitieren verschiedene Stakeholder wie Kunden, Betreiber und Entwickler von dem erweiterten Ansatz, da sie strengere Garantien hinsichtlich der Resilienz der Anwendung erhalten können. Außerdem ermöglicht der Prototyp, der den erweiterten Ansatz implementiert, Software-Ingenieuren, ihn ohne weiteres zu verwenden und ihn möglicherweise zu erweitern. vi A.2 Compatibility Evaluation Resources . . . . . . . . . . . . . . . . . . . . . . . 78 A.2 Compatibility Evaluation Resources List of Figures 2.1 Property Specification Patterns - Original Hierarchy [DAC99] . . . . . . . . . . 11 2.2 Property Specification Patterns - Scopes [DAC99] . . . . . . . . . . . . . . . . . 11 2.3 Property Specification Patterns - Extended Catalog [AGL+15]. . . . . . . . . . . 12 4.1 Extended Chaos Engineering Approach - Concept Overview. . . . . . . . . . . . 26 4.2 Transient Behavior Hypothesis Specification Concept. . . . . . . . . . . . . . . . 27 4.3 Transient Behavior Verification - Sequence Diagram. . . . . . . . . . . . . . . . 30 5.1 Prototype Components - Overview. . . . . . . . . . . . . . . . . . . . . . . . . . 31 5.2 Component Diagram of the Verification Tool. . . . . . . . . . . . . . . . . . . . 39 5.3 Result Plot Example. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43 5.4 Graphical User Interface - Home Page. . . . . . . . . . . . . . . . . . . . . . . . 44 5.5 Graphical User Interface - Behavior Specification Using Metric Temporal Logic (MTL). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44 5.6 Graphical User Interface - Behavior Specification Using Property Specification Patterns (PSPs). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45 5.7 Graphical User Interface - Experiment Creation Page. . . . . . . . . . . . . . . 45 5.8 Graphical User Interface - Stand-alone Verification Page. . . . . . . . . . . . . . Contents 65 Bibliography 67 A Evaluation Resources 71 A.1 Chaos Experiment Evaluation - Chaos Experiments . . . . . . . . . . . . . . . 72 vii A.2 Compatibility Evaluation Resources . . . . . . . . . . . . . . . . . . . . . . . 78 viii List of Figures 46 5.9 Graphical User Interface - Verification Results Page. . . . . . . . . . . . . . . . 46 6.1 Goal Question Metric (GQM) Models Overview. . . . . . . . . . . . . . . . . . 47 6.2 Test System - Architecture. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51 6.3 Chaos Experiment 1 - Verification Results. . . . . . . . . . . . . . . . . . . . . . 53 6.4 Chaos Experiment 2 - Verification Results. . . . . . . . . . . . . . . . . . . . . . 54 6.5 Chaos Experiment 3 - Verification Results. . . . . . . . . . . . . . . . . . . . . . 55 6.6 MiSim Simulated Architecture - Component Diagram. . . . . . . . . . . . . . . 57 6.7 Simulated Response Time. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 58 6.8 Simulated Instance Counts. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 58 6.9 Compatibility Evaluation - Verification Result Plot 1. . . . . . . . . . . . . . . . 59 6.10 Compatibility Evaluation - Verification Result Plot 2. . . . . . . . . . . . . . . . 60 ix List of Tables 5.1 Chaos Engineering Tools - Comparison . . . . . . . . . . . . . . . . . . . . . . 32 5.2 MTL Formula Conversion Example [AGL+15; Ulu19a] . . . . . . . . . . . . . . 41 5.3 Future and Past Example Trace. . . . . . . . . . . . . . . . . . . . . . . . . . . . 42 6.1 GQM Model for the Correctness Evaluation. . . . . . . . . . . . . . . . . . . . . 48 6.2 Correctness Evaluation - Test Suite Results. . . . . . . . . . . . . . . . . . . . . 49 6.3 GQM Model for the Chaos Experiment Evaluation. . . . . . . . . . . . . . . . . 50 6.4 Executed Chaos Experiments - Overview. . . . . . . . . . . . . . . . . . . . . . 51 6.5 GQM Model for the Evaluation of the Compatibility With Other Approaches. . . 57 xi xi List of Listings 2.1 Chaos Toolkit Example Experiment [Chac] . . . . . . . . . . . . . . . . . . . . 9 5.1 Transient Behavior Specification - MTL & CSV Example. . . . . . . . . . . . . 34 5.2 Transient Behavior Specification - PSP & InfluxDB Example. . . . . . . . . . . 35 5.3 Chaos Experiment Containing a Transient Behavior Specification - Steady-State Hypothesis (Without Probes) and Method. (Continued in Listing 5.4) . . . . . . 37 5.4 Chaos Experiment Containing a Transient Behavior Specification - Probes in the Steady-State Hypothesis (Continuation of Listing 5.3). . . . . . . . . . . . . . . 38 5.5 Predicate Functions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40 A.1 Evaluation - Chaos Experiment 1 (Without Transient Behavior Hypothesis Probe). 72 A.2 Evaluation - Chaos Experiment 1 (Transient Behavior Hypothesis Probe). . . . . 73 A.3 Evaluation - Chaos Experiment 2 (Without Transient Behavior Hypothesis Probe). 74 A.4 Evaluation - Chaos Experiment 2 (Transient Behavior Hypothesis Probe). . . . . 75 A.5 Evaluation - Chaos Experiment 3 (Without Transient Behavior Hypothesis Probe). 76 A.6 Evaluation - Chaos Experiment 3 (Transient Behavior Hypothesis Probe). . . . . 77 A.7 MiSim Architecture Definition. . . . . . . . . . . . . . . . . . . . . . . . . . . . 78 A.8 MiSim Architecture Definition (Continuation). . . . . . . . . . . . . . . . . . . 79 A.9 MiSim Experiment Definition. . . . . . . . . . . . . . . . . . . . . . . . . . . . 79 A.10 Compatibility Evaluation - Transient Behavior Specification 1. . . . . . . . . . . 80 A.11 Compatibility Evaluation - Transient Behavior Specification 2. . . . . . . . . . . 81 xiii xiii 1 Introduction In the last decade, there has been a huge shift from the monolithic architectural and deployment styles to distributed and cloud-native approaches [GD18]. One reason for this shift is that the distributed and cloud-native approaches are inherently more resilient to failures, whereas monolithic applications are often prone to having single points of failure [VGC+15]. Although inherently more resilient and not subjected to single points of failure, the distributed and cloud-native applications are still not protected from failures of individual computing resources and potentially cascading failures [New15]. Additionally, the distributed nature of these applications makes the identification of such failures and the assessment of their magnitude more complex [New15]. One approach, aiming to support software engineers in analyzing and preventing such failures, is Chaos Engineering. The Chaos Engineering approach assesses the resilience of software systems by deliberately injecting failures or anomalies as part of experiments. The so-called Chaos Experiments are combining the injection of anomalies with validations of hypotheses regarding a steady state of the system’s operation before and after the insertion of the anomaly. Experimentation is also a defining property of the approach, as experimentation creates new knowledge compared to testing which only verifies existing knowledge [RJ20]. Furthermore, Chaos Engineering should be applied iteratively and continuously, so that it can assess the resilience of the software after changes have been introduced and at various points during the software development life cycle [RJ20]. The anomalies injected as part of Chaos Engineering cause the target systems to experience transient behaviors. The term transient behavior is originating from the field of electrical engineering and is referring to the change of states when a system is transitioning from one steady state to another [VFH14]. By deliberately injecting anomalies such as stopping services, introducing sudden workload changes, and triggering auto-scaling policies, the Chaos Experiments are triggering the resilience mechanisms of the target systems and are causing them to change states and experience transient behaviors. The state-changing behaviors of adaptive systems have already been studied and PSPs [DAC99] and temporal logic [AGL+15] have been identified as capable of formally specifying such behaviors [CL12]. Even though these behaviors are already known and studied, the behaviors and their verification have not been studied with regard to Chaos Engineering and in particular the Chaos Experimentation process. Acronyms ATAM Architecture Trade-Off Analysis Method. 14 CSV Comma-Separated Values. 22 CTL Computational Tree Logic. 10 DTMC Discrete-Time Markov Chain. 13 EPL Event Processing Language. 11 GQM Goal Question Metric. ix, 47 GUI Graphical User Interface. 26 HTTP Hypertext Transfer Protocol. 17 JSON JavaScript Object Notation. 8 LTL Linear Temporal Logic. 10 MTL Metric Temporal Logic. ix, 10 PCTL Probabilistic Computation Tree Logic. 13 PSP Property Specification Pattern. ix, 7, 10 QoS Quality of Service. 14 YAML YAML Ain’t Markup Language. 8 ATAM Architecture Trade-Off Analysis Method. 14 CSV Comma-Separated Values. 22 CTL Computational Tree Logic. 10 DTMC Discrete-Time Markov Chain. 13 EPL Event Processing Language. 11 GQM Goal Question Metric. ix, 47 GUI Graphical User Interface. 26 HTTP Hypertext Transfer Protocol. 17 JSON JavaScript Object Notation. 8 LTL Linear Temporal Logic. 10 MTL Metric Temporal Logic. ix, 10 PCTL Probabilistic Computation Tree Logic. 13 PSP Property Specification Pattern. ix, 7, 10 QoS Quality of Service. 14 YAML YAML Ain’t Markup Language. 8 JSON JavaScript Object Notation. 8 xv xv Problem Statement The Chaos Experimentation process, in particular the verification of the steady-state hypothesis before and after the introduction of anomalies, is capable of providing insight into the resilience of an application and whether it is able to adapt to failures and continue operating properly after such are introduced. However, this process is conceptually limited as the Chaos Experiments 1 1 Introduction only allow the specification of steady-state hypotheses, meaning that a verification of the transient behaviors exhibited during state changes is not explicitly considered. Even though different Chaos Engineering tools offer the possibility to configure the timing of the steady-state hypothesis checks and even to carry out such continuously during the whole experiment, the Chaos Experiments are still limited since the steady-state hypotheses are meant to make statements only regarding the stable state of the system. Additionally, the transient states might have different characteristics and acceptance thresholds. To better illustrate this, consider an example, in which there is a system consisting of multiple services, a steady-state hypothesis that the response time of the system is under 1 second, and a failure injection which suddenly stops one of the services. It is also assumed that the response time of the system is an important quality for the users of the system and a sudden increase in it could lead to the loss of users which is also a loss for various stakeholders of the application. Therefore, the stakeholders of the application require guarantees regarding the transient phases in which they could lose users. As a result, the stakeholders define the transient behavior requirement that after a failure of a service, the response time of the system could increase to 2 seconds for a time period not exceeding 30 seconds, and the stopped instance should be restarted during these 30 seconds. By using the traditional Chaos Engineering approach, first, it would not be possible to define the transient behavior requirement as part of the Chaos Experiment using the supported constructs, and second, a verification of the transient behavior requirement would not be possible even if the steady-state hypothesis check is performed continuously. On the contrary, the extended Chaos Engineering approach proposed in this thesis allows the specification of transient behavior hypotheses and their verification. As modern software systems are often consisting of numerous highly distributed components, failures or anomalies of single components are not uncommon. Problem Statement As a result, the system spends a certain time transitioning between states, i.e. experiencing transient behaviors. Additionally, such behaviors and state changes can be triggered by various stimuli. Examples of such are rolling deployments when both old and new versions of a system are running simultaneously, failures of components when requests are rerouted to stable components and the failed components are restarted, and workload changes when resources have to be dynamically added or removed [Bec20]. Since these behaviors can have an influence on the overall usage of the system, a closer look into the transient behaviors is beneficial to the stakeholders of the software. Another problem is that a formal method for specifying transient behaviors in Chaos Engineering has not been researched yet. Formal methods such as temporal logic have been already studied with regards to specifying temporal behaviors, however, the applicability of these methods to specifying transient behaviors as part of Chaos Experiments and an approach for integrating them into Chaos Engineering have not been studied yet. Furthermore, introducing support for the analysis of formal specifications of such behaviors as part of the Chaos Experimentation process would allow a deeper insight into the resilience of software applications and also enable the specification and verification of more fine-grained resilience requirements. Objective The first objective of the thesis is to study how a transient behavior requirement specified using a formalism capable of describing transient behaviors and their temporal properties can be incorporated into Chaos Experiments. The results regarding this objective will also provide the answers to the first research question of the thesis which is the following: 2 RQ1 How can a transient behavior requirement be specified in the context of Chaos Engineering and Chaos Experiments? The second objective of this thesis is to investigate how the Chaos Experimentation process can be extended to support the verification of transient behavior requirements. In addition, this objective also includes eliciting requirements regarding such an extension of the Chaos Experimentation process from individuals with Chaos Engineering expertise. The findings regarding this objective will respectively answer the following research question: RQ2 How can the Chaos Experimentation process incorporate a verification of transient behavior requirement specifications? After defining how the Chaos Engineering approach can be extended and gathering requirements for this extension, a further objective of the thesis is to create a final concept for the proposed extension and to develop a prototype with the goal of evaluating the proposed extension. This is reflected in the following research question: RQ3 What does a concept for an extended Chaos Engineering approach contain and how can it be implemented into a prototype? RQ3 What does a concept for an extended Chaos Engineering approach contain and how can it b implemented into a prototype? A further goal of the thesis is to evaluate the proposed extended Chaos Engineering approach and its implementation with regard to different criteria, namely the correctness of the verification of the transient behavior specifications, the execution of Chaos Experiments, and the compatibility with other approaches. The results of the evaluation will provide answers to the following research questions: RQ4 Is the extended Chaos Engineering approach able to correctly verify different transient behavior requirements? RQ4 Is the extended Chaos Engineering approach able to correctly verify different transient behavior requirements? RQ5 Is the extended Chaos Engineering approach capable of correctly executing Chaos Experiments and verifying transient behavior requirements as part of the experiments? RQ5 Is the extended Chaos Engineering approach capable of correctly executing Chaos Experiments and verifying transient behavior requirements as part of the experiments? RQ6 Is the extended Chaos Engineering approach compatible with other approaches and tools? Results First, the extension of the Chaos Experiments with a transient behavior specification using formalisms such as temporal logic and PSPs, and the extension of the Chaos Experimentation process by the inclusion of runtime verification are studied. The resulting initial concepts from this research are combined into an initial prototype which is used during the three expert interviews that successfully elicited requirements for the extended Chaos Engineering approach. After these are taken into consideration, a final concept for the extension of the Chaos Engineering approach, which among others includes a transient behavior hypothesis section in the Chaos Experiment and an additional transient behavior verification step in the experiment execution is created. The concept is then successfully implemented into a prototype with the aim to evaluate the proposed extension. All of the planned evaluations are successfully carried out and their results prove that the extension of the Chaos Engineering approach achieves its purpose and is a viable extension to the traditional Chaos Engineering approach. Method To achieve the first goal of the thesis, the temporal logic formalisms that have already been identified as capable of describing transient behaviors are studied with regard to how they can be incorporated as transient behavior requirements into Chaos Experiments. To achieve the second goal of finding out how the Chaos Experiments can support the verification of transient behavior requirements, state-of-the-art Chaos Engineering tools are examined. Additionally, the requirements for such an extension of the Chaos Experimentation process are gathered in three interviews with people with Chaos Engineering experience. After completing the interviews, in order to accomplish the third goal of the thesis, the requirements elicited during them and the information gathered during the research regarding the previous objectives are used to develop a final concept for the extended Chaos Engineering approach. After the concept is created, it is implemented into a prototype. Once the implementation of the prototype is completed, it is evaluated with regard to various criteria. First, the prototype is evaluated with regard to the correctness of the verification of transient behavior 3 1 Introduction specifications. For this, a large variety of transient behavior specifications is verified using the prototype in order to find out how many of these get correctly verified. Afterward, the ability to carry out this verification as part of the Chaos Experimentation process is evaluated by executing a variety of Chaos Experiments including transient behavior hypotheses on a test system. Furthermore, the compatibility of the proposed approach with other tools and approaches is studied by using the verification of transient behavior specification on the data produced by a tool capable of simulating microservice architectures and their resilience. Chapter 7 – Conclusion and Future Work: This chapter concludes the thesis and discusses pos- sible future work extending this thesis. Furthermore, all supplementary materials such as the source code of the prototype and all evaluation resources are publicly available [Zah22]. 2 Foundations and Related Work This chapter discusses the foundations of this thesis and related research. 2.1 Foundations In the following sections, the theoretical foundations of this work, such as Chaos Engineering, temporal logic, Property Specification Patterns (PSPs), and runtime verification, as well as the technical background required for a proper understanding of the work are described in detail. Thesis Structure In the following, the chapters of this thesis proposal are listed and their content is described. Chapter 2 – Foundations and Related Work: This chapter provides valuable information about the theoretical foundations, the technical background of the thesis, and related research in this area. Chapter 3 – Elicitation of Requirements for the Extended Chaos Engineering Approach: This chapter presents the interview process, initial concepts presented to the interviewees, the interview questions, and the results of the interviews. Chapter 3 – Elicitation of Requirements for the Extended Chaos Engineering Approach: This chapter presents the interview process, initial concepts presented to the interviewees, the interview questions, and the results of the interviews. Chapter 4 – Extended Chaos Engineering Approach: In this chapter, the concept for the ex- tended Chaos Engineering approach is defined and thoroughly described. Chapter 4 – Extended Chaos Engineering Approach: In this chapter, the concept for the ex- tended Chaos Engineering approach is defined and thoroughly described. Chapter 4 – Extended Chaos Engineering Approach: In this chapter, the concept for the ex- tended Chaos Engineering approach is defined and thoroughly described. Chapter 5 – Prototype Implementation: In this chapter, the technical details regarding the imple- mentation of the concept into a working prototype are discussed. Chapter 5 – Prototype Implementation: In this chapter, the technical details regarding the imple- mentation of the concept into a working prototype are discussed. Chapter 6 – Evaluation: In this chapter, the extended Chaos Engineering approach and its imple- mentation are evaluated with regard to the correctness of the verification of transient behavior specifications, the integration of the verification into the Chaos Experimentation process, and the compatibility with other approaches. 4 4 Furthermore, all supplementary materials such as the source code of the prototype and all evaluation resources are publicly available [Zah22]. 5 5 2.1.1 Chaos Engineering Chaos Engineering is an approach pioneered by Netflix, in which failures are deliberately caused while the system is operating in a normal production environment, in order to observe if the system is resilient to such failures [RJ20]. The Chaos Team at Netflix also came up with a definition and a blueprint of the Chaos Engineering approach called "The Principles" [Pri] in which the approach is formally defined as "the discipline of experimenting on a distributed system in order to build confidence in the system’s capability to withstand turbulent conditions in production" [RJ20]. Additionally, the five steps of the Chaos Experimentation process are defined as follows: (1) create a hypothesis regarding the steady-state behavior of the system, i.e. a steady-state hypothesis, (2) select experiment variables reflecting real-world events, (3) run experiments in production, (4) automate experiment execution, and (5) minimize the blast radius [RJ20]. As mentioned, two of the elements of a Chaos Experiment are the steady-state hypothesis and the injection of an anomaly or a failure. The steady-state hypothesis defines a statement regarding the steady state of the application, i.e. the state of its normal and stable operation. One example of a steady-state hypothesis could be the following statement: "The /getUsers URL responds with a 200 status code, within 1 second" [Mil19]. An example of a failure that could be defined as part of a Chaos Experiment could be the following: "Disconnect the database cluster from the network" [Mil19]. The execution of a Chaos Experiment often starts with a verification of the steady-state hypothesis before the failure is injected. This is done in order to verify that the system is stable before any disruptions are introduced. If this holds, the anomaly is introduced, i.e. the failure is injected. Afterward, the steady-state hypothesis is verified once again, in order to examine if the system is behaving normally after the failure was introduced. 7 2 Foundations and Related Work 2.1.3 Transient Behaviors The terms transient behavior and transient state originate from the field of electrical engineering. Venikov [VFH14] defines transient states as “occurring when the system is changing from one steady state to another“. Similarly to electrical systems, software systems are also able to transition from one steady state into another while experiencing certain behaviors during this transition. As a result, the terms transient behavior and transient state can also be applied to the behaviors occurring in software systems. These terms can also be particularly useful in approaches like Chaos Engineering that are examining the steady states and their changes when anomalies are introduced. The causes for transient behavior in software systems can vary widely. Some possible causes for such behaviors have been discussed in the work of Beck [Bec20] and are presented in this paragraph. One of the possible causes can be rolling deployments, during which both the old and the new version of the software are simultaneously running, and as a result, the system is in an undefined state of transient behavior during the deployment. Another cause can be the failure of components and the triggering of resilience mechanisms, e.g. when the requests to a failed instance are rerouted to another one while the failed one gets restarted. During this time the system would again be exhibiting transient behaviors. A further cause could be sudden changes in user behavior and respectively in the workload, as a result of which the autoscaling of the application might initiate the process of starting or stopping instances, during which the system would be experiencing transient behaviors. 2.1.2 Chaos Toolkit Chaos Toolkit [Chac] is a tool that enables the execution of automated Chaos Experiments [Mil19]. Its main advantages are that it is open-source and has a variety of extensions that allow the injections of faults in multiple layers, namely platform, network, and application [Chac]. Chaos Toolkit supports Chaos Experiments specified as either JavaScript Object Notation (JSON) or YAML Ain’t Markup Language (YAML) files. Chaos Toolkit’s Chaos Experiments usually contain two components, a steady-state hypothesis, and a fault injection, also called method. The steady-state hypothesis usually contains probes, which are calls retrieving information from the system under test, while the method usually contains actions, which respectively perform operations on the system under test [Chac]. In Listing 2.1 an example experiment specified using Chaos Toolkit’s syntax is presented. In this example, the experiment contains a simple steady-state hypothesis, namely that the microservices of the system remain healthy, and a simple fault injection, which is expressed in killing a random microservice instance. Additionally, the probe in the steady-state hypothesis and the action in the method are making use of the Kubernetes extension for Chaos Toolkit, which integrates various Kubernetes operations into Chaos Toolkit’s probes and actions [Chac]. 2.1.4 Temporal Logic The term temporal logic covers a variety of logical formalisms that enhance the traditional propositional logic with temporal modalities to specify and reason about timing properties of logical expressions [AH92]. Two of the most common temporal modalities present in most temporal 8 2.1 Foundations Listing 2.1 Chaos Toolkit Example Experiment [Chac] Listing 2.1 Chaos Toolkit Example Experiment [Chac] 1 { 2 "title": "Do we remain available in face of pod going down?", 3 "description": "We expect Kubernetes to handle the situation gracefully when a pod goes ↩→down", 4 "tags": ["kubernetes"], 5 "steady-state-hypothesis": { 6 "title": "Verifying service remains healthy", 7 "probes": [ 8 { 9 "name": "all-our-microservices-should-be-healthy", 10 "type": "probe", 11 "tolerance": true, 12 "provider": { 13 "type": "python", 14 "module": "chaosk8s.probes", 15 "func": "microservice_available_and_healthy", 16 "arguments": { 17 "name": "myapp" 18 } 19 } 20 } 21 ] 22 }, 23 "method": [ 24 { 25 "type": "action", 26 "name": "terminate-db-pod", 27 "provider": { 28 "type": "python", 29 "module": "chaosk8s.pod.actions", 30 "func": "terminate_pods", 31 "arguments": { 32 "label_selector": "app=my-app", 33 "name_pattern": "my-app-[0-9]$", 34 "rand": true 35 } 36 }, 37 "pauses": { 38 "after": 5 39 } 40 } 41 ] 42 } } 9 9 2 Foundations and Related Work logics are the following operators: ♢- eventually (eventually in the future), □- always (now and forever in the future), and 𝑈- until [BK08]. Some of these temporal logics can also be extended with past operators such as the following: ♢−1 - sometime in the past, and □−1 - always in the past [BK08]. There are two main types of temporal logics, namely linear-time and branching-time logics [AH92]. In linear-time logics, the logical statements are interpreted over linear structures of states, meaning that there is a single linear trace of states. On the contrary, in branching-time logics, the statements are interpreted over tree structures of states, where there are multiple different paths, any one of which might be realized [AH92]. One example of a temporal logic based on linear time is the Linear Temporal Logic (LTL). An example logical expression for the response property “every request will eventually lead to a response“ specified using LTL is the following: □(𝑟𝑒𝑞𝑢𝑒𝑠𝑡→♢𝑟𝑒𝑠𝑝𝑜𝑛𝑠𝑒) [BK08]. Another example, this time for a temporal logic based on branching time, is Computational Tree Logic (CTL) [BK08]. The response property introduced above specified using CTL is the following: ∀□(𝑟𝑒𝑞𝑢𝑒𝑠𝑡→∀♢𝑟𝑒𝑠𝑝𝑜𝑛𝑠𝑒) [BK08]. 2.1.4 Temporal Logic Even though the two examples might look similar, as LTL and CTL share a lot of logical operators, the expressiveness of LTL and CTL is incomparable [BK08]. 2.1.5 Metric Temporal Logic Metric Temporal Logic (MTL) is an extension of LTL in which the temporal operators are time-bounded by the introduction of timing constraints in the form of time intervals [Ulu19a]. Additionally, MTL supports past and future operators. As a result, MTL is often split semantically into two subvariants, namely past-MTL and future-MTL, depending on the utilized temporal operators. In the following, only the syntax of past-MTL and examples regarding this variant are presented, since this is the MTL variant that is primarily used in the remainder of the thesis. Past-MTL uses the following temporal operators: 𝑝𝑎𝑠𝑡𝑒𝑣𝑒𝑛𝑡𝑢𝑎𝑙𝑙𝑦(♦𝐼), 𝑝𝑎𝑠𝑡𝑎𝑙𝑤𝑎𝑦𝑠(■𝐼), and 𝑠𝑖𝑛𝑐𝑒(𝑆𝐼) [Ulu19b]. Given a finite set 𝑃of atomic propositions, past-MTL formulas are defined by the following grammar [Ulu19b]: 𝜑= ⊤\| 𝑝\| ¬𝜑\| 𝜑1 ∨𝜑2 \| 𝜑1 S𝐼𝜑2 where 𝑝∈𝑃and 𝐼⊆[0, ∞). In the above-presented notation, the subscript 𝐼at each temporal operator represents the time interval, i.e. the timing constraint of the given operator. For example, the MTL formula ♦[0,5]𝑄means that 𝑄will hold eventually within 5 time units. The satisfaction relation (𝑤, 𝑡) ⊢𝜑indicates that the temporal behavior 𝑤: T →B𝑃satisfies the formula 𝜑at time point 𝑡∈T as follows [Ulu19b]: (𝑤, 𝑡) ⊢𝑝⇔𝑤𝑝(𝑡) = ⊤ (𝑤, 𝑡) ⊢¬𝜑⇔(𝑤, 𝑡) ⊬𝜑 (𝑤, 𝑡) ⊢𝜑1 ∨𝜑2 ⇔(𝑤, 𝑡) ⊢𝜑1 or (𝑤, 𝑡) ⊢𝜑2 (𝑤, 𝑡) ⊢𝜑1S𝐼𝜑2 ⇔∃𝑡′ ∈(0, 𝑡). (𝑤, 𝑡′) ⊢𝜑2 and ∀𝑡′′ ∈(𝑡′, 𝑡). (𝑤, 𝑡′′) ⊢𝜑1 and 𝑡−𝑡′ ∈𝐼 (𝑤, 𝑡) ⊢𝑝⇔𝑤𝑝(𝑡) = ⊤ (𝑤, 𝑡) ⊢¬𝜑⇔(𝑤, 𝑡) ⊬𝜑 (𝑤, 𝑡) ⊢𝜑1 ∨𝜑2 ⇔(𝑤, 𝑡) ⊢𝜑1 or (𝑤, 𝑡) ⊢𝜑2 (𝑤, 𝑡) ⊢𝜑1S𝐼𝜑2 ⇔∃𝑡′ ∈(0, 𝑡). (𝑤, 𝑡′) ⊢𝜑2 and ∀𝑡′′ ∈(𝑡′, 𝑡). (𝑤, 𝑡′′) ⊢𝜑1 and 𝑡−𝑡′ ∈𝐼 (𝑤, 𝑡) ⊢𝑝⇔𝑤𝑝(𝑡) = ⊤ (𝑤, 𝑡) ⊢¬𝜑⇔(𝑤, 𝑡) ⊬𝜑 (𝑤, 𝑡) ⊢𝜑1 ∨𝜑2 ⇔(𝑤, 𝑡) ⊢𝜑1 or (𝑤, 𝑡) ⊢𝜑2 (𝑤, 𝑡) ⊢𝜑1S𝐼𝜑2 ⇔∃𝑡′ ∈(0, 𝑡). (𝑤, 𝑡′) ⊢𝜑2 and ∀𝑡′′ ∈(𝑡′, 𝑡). (𝑤, 𝑡′′) ⊢𝜑1 and 𝑡−𝑡′ ∈𝐼 (𝑤, 𝑡) ⊢𝑝⇔𝑤𝑝(𝑡) = ⊤ (𝑤, 𝑡) ⊢¬𝜑⇔(𝑤, 𝑡) ⊬𝜑 (𝑤, 𝑡) ⊢𝜑1S𝐼𝜑2 ⇔∃𝑡′ ∈(0, 𝑡). (𝑤, 𝑡′) ⊢𝜑2 and ∀𝑡′′ ∈(𝑡′, 𝑡). (𝑤, 𝑡′′) ⊢𝜑1 and 𝑡−𝑡′ ∈𝐼 (𝑤, 𝑡) ⊢𝜑1S𝐼𝜑2 ⇔∃𝑡′ ∈(0, 𝑡). (𝑤, 𝑡′) ⊢𝜑2 and ∀𝑡′′ ∈(𝑡′, 𝑡). (𝑤, 𝑡′′) ⊢𝜑1 and 𝑡−𝑡′ ∈𝐼 2.1.6 Property Specification Patterns Property Specification Patterns (PSPs), introduced by Dwyer et al. [DAC99], are a system of recurring solutions, dealing with the temporal intricacies of reactive systems [AGL+15]. These patterns provide general rules that help practitioners to qualify order and occurrence, quantify time 10 2.1 Foundations Property Patterns Occurrence Order Chain Precedence Bounded Existence Chain Response Existence Response Universality Precedence Absence Figure 2.1: Property Specification Patterns - Original Hierarchy [DAC99] After Q until R Between Q and R After Q Before Q Global State Sequence Q R Q Q R Q Figure 2.2: Property Specification Patterns - Scopes [DAC99] Property Patterns perty Patterns Occurrence Order Bounded Existence Chain Precedence Chain Response Response Universality bsence Precedence Universality Absence Existence Figure 2.1: Property Specification Patterns - Original Hierarchy [DAC99] Figure 2.1: Property Specification Patterns - Original Hierarchy [DAC99] Before Q Figure 2.2: Property Specification Patterns - Scopes [DAC99] bounds, and express probabilities of events [AGL+15]. Figure 2.1 depicts an overview of the pattern hierarchy. Each of the patterns describes a commonly occurring requirement on the event sequences in a finite-state system model [DAC99]. In the following, three example patterns are listed, the Existence and Bounded Existence patterns from the Occurrence category and the Response pattern from the Order category: Existence A given event must occur within a scope. Existence A given event must occur within a scope. Bounded Existence A given event must occur 𝑘times within a scop Bounded Existence A given event must occur 𝑘times within a scope Response A given event 𝑃must always be followed by an event 𝑆within a scope. Response A given event 𝑃must always be followed by an event 𝑆within a scope. As can be seen in the three example patterns listed above, the patterns also contain a scope. Dwyer et al. [DAC99] define the scopes presented in Figure 2.2. Additionally, for each pattern Dwyer et al [DAC99] specify mappings for the pattern into the previously introduced temporal logics, namely LTL and CTL. A paper by Czepa and Zdun [CZ18] compared the PSPs to LTL and Event Processing Language (EPL) with regard to their under- standability. The authors found evidence supporting their hypothesis that PSPs are more easily understandable than the rest of the studied formalisms [CZ18]. Throughout the years, additional patterns have been introduced into the pattern catalog and new mappings into additional temporal logic formalisms have been introduced [AGL+15]. The new extended version of the PSP catalog is 11 2 Foundations and Related Work Figure 2.3: Property Specification Patterns - Extended Catalog [AGL+15]. Figure 2.3: Property Specification Patterns - Extended Catalog [AGL+15]. presented in Figure 2.3. Furthermore, a study by Grunske et al. [AGL+15] introduced a software tool called PSP Wizard, allowing its users to easily create PSP specifications in natural language and transform these into expressions in various temporal logic formalisms. 2.1.7 Propositional Functions and Predicates Throughout the thesis, the terms predicate and predicate function are used, thus they must be defined. A predicate or also called a propositional function is an expression that contains a variable and becomes a statement when a constant is substituted into the variable [CCR13]. A predicate normally has 𝑡𝑟𝑢𝑒and 𝑓𝑎𝑙𝑠𝑒substitution instances [CCR13]. For example, when 𝑃represents a human, 𝑠 represents Socrates, and 𝑐represents Chicago, then 𝑃(𝑠) is 𝑡𝑟𝑢𝑒while 𝑃(𝑐) is 𝑓𝑎𝑙𝑠𝑒[CCR13]. In the remainder of the thesis, the term predicate is used instead of the term propositional function. Furthermore, predicates are used in MTL formulas and PSPs similarly to how propositions are used. 2.1.8 Runtime Verification Leucker and Schallhart [LS09] define runtime verification as a discipline dealing with the techniques enabling checking whether a run of a system satisfies or violates a given correctness property. Leucker and Schallhart also define a component called a monitor, which is responsible for reading a finite event trace and delivering a certain 𝑦𝑒𝑠/𝑛𝑜or 𝑡𝑟𝑢𝑒/ 𝑓𝑎𝑙𝑠𝑒verdict. Even though runtime verification has its origins in model checking, the two approaches are inherently different according to the authors. Model-checking typically examines all executions of a system in order to answer 12 2.2 Related work whether it satisfies a given property. On the contrary, runtime verification considers only a single execution. Additionally, model checking usually considers infinite traces while runtime verification considers finite ones [LS09]. Runtime verification has also been studied with regards to verifying properties specified using temporal logic formalisms such as LTL and MTL, and approaches to achieve this verification and to build the required monitors have already been proposed [BLS07; Ulu19b]. Ulus [Ulu19b] proposes an approach for creating monitors for the past variant of MTL using sequential networks. The sequential networks used in the approach are abstract machines with finite state spaces, a set of update rules, and an output function [Ulu19b]. Even though sequential networks are functionally identical to finite automata, they differ in their structure [Ulu19b]. These structural differences make the sequential networks particularly suitable to create runtime verification monitors [Ulu19b]. The approach proposed by Ulus considers only the past fragment of MTL because past-oriented monitoring is inherently less costly than future-oriented monitoring and because otherwise, the monitor would have to be able to delay the output or output 𝑢𝑘𝑛𝑜𝑤𝑛[Ulu19b]. Furthermore, Ulus presents an implementation of his approach and compares it with other tools providing past-MTL monitors. The results of this comparison indicate that Ulus’ approach handles large time constraints over long sequences more efficiently than the other monitoring tools [Ulu19b]. 2.2 Related work In the following sections, related research regarding the specification and verification of transient behaviors using the already introduced formalisms, an extension of the Chaos Engineering approach, and an approach to visualize transient behaviors are discussed. 2.2.1 Specifying and Verifying Transient Behaviors Using Temporal Logic Camara and De Lemos [CL12] propose an approach for verification of self-adaptive systems and resilience properties based on stimulation and probabilistic model-checking [CL12]. The approach of the authors is to stimulate the system’s environment to exercise its adaptive capabilities and to capture data about the system’s reaction to the stimulus [CL12]. The collected data is then transformed into Discrete-Time Markov Chains (DTMCs). The authors also specify the resilience properties they want to verify as Probabilistic Computation Tree Logic (PCTL) expressions using the response pattern from the PSPs. The researchers then verified the PCTL expressions against the DTMCs using model-checking. The logical language chosen by the researchers is a variant of the CTL language, which is suitable for expressing temporal properties, typical for transient behaviors. The DTMC models, however, are not representative of the techniques utilized by Chaos Engineering, since DTMCs are a probabilistic approach, used to generalize the results of multiple executions. The aim of this thesis is to enable exact statements regarding transient behaviors in a single experiment run. This thesis also differs from the work of Camara and De Lamos [CL12] as it focuses on the verification of transient behaviors as part of the Chaos Experimentation process, excluding the assumptions and the probabilistic properties of the DTMC approach. 13 2 Foundations and Related Work 2.2.2 Extending the Chaos Engineering Approach Frank et al. [FHW+21] conducted a study with an industry partner, the aim of which was to explore the application and adoption of the scenario-based Architecture Trade-Off Analysis Method (ATAM) for the elicitation of resilience requirements. The authors conducted an ATAM-based workshop with the industry partners and collected resilience scenarios which were then turned into Chaos Experiments. The Chaos Experiments were executed on a mock system provided by the industry partner and the authors were able to identify weaknesses. The above-mentioned goal and employed methods are not particularly concerning this thesis, however, the experimental setup used by the researchers is of interest for this thesis. The authors used Chaos Toolkit [Chac], which is a tool allowing software engineers to perform Chaos Experiments, together with a load generator, generating a steady amount of requests to the system under test during the whole experiment. The response times of the requests were also captured and saved in an InfluxDB database. Additionally, the authors also implemented their own hypothesis validator. This experimental setup allowed the researchers to measure detailed information about the system under test’s response times during the whole duration of the Chaos Experiments, i.e. the approach could capture the transient behaviors regarding the response time. Even though the approach was able to capture these transient behaviors, it did not support a verification of the experienced behaviors against a given transient behavior requirement. This is where this thesis differs from the work of Frank et al. since the Chaos Engineering approach that it proposes includes the specification of transient behavior requirements as part of the Chaos Experiments and provides explicit verification of the specified requirements. 2.2.3 Visualizing Transient Behaviors Frank et al. [BFH+21] conducted another study, as part of which the authors present TransVis, an approach to support software architects and DevOps engineers to visually specify and assess transient behaviors in microservice systems [BFH+21]. The approach includes a dashboard, integrating visualizations of the microservice architecture, the transient behaviors requirement specifications, and the actually experienced transient behaviors. The researchers also integrated a chatbot into their approach, enabling interactions with the visualizations through natural language. After introducing TransVis, the authors conducted an exploratory expert study aiming to examine how well the approach supports software architects and DevOps engineers in the specification and verification of transient behaviors [BFH+21]. The results of the study indicated that the developed visualizations are effective, especially for exploring transient behavior and comparing a service’s Quality of Service (QoS) with a specified behavior, however, the chatbot did not prove to be significantly beneficial [BFH+21]. Even though the study by Frank et al. [BFH+21] is not concerning the area of Chaos Engineering, it proposes a way to specify transient behavior using a formalism, namely a resilience triangle [BCE+03] consisting of three fixed metrics. This is also where the approach for specifying transient behaviors of Frank et al. differs from the one proposed by this thesis, which utilizes temporal logic and PSPs. Additionally, the approach proposed by Frank et al. allows only a manual verification in the form of comparisons of the transient behavior specifications and the actual behaviors which is where this thesis also differs from the work of Frank et al. as the approach proposed in this thesis automates this verification as part of the Chaos Experimentation process. 14 3 Elicitation of Requirements for the Extended Chaos Engineering Approach In addition to identifying a means to specify transient behavior requirements as part of Chaos Experiments and how such requirements can be verified during the execution of the experiments, a further goal of the thesis is to gather requirements for an extended Chaos Engineering approach incorporating these specifications and their verification. For this purpose, a number of interviews were conducted with people that are practicing the Chaos Engineering techniques in industry and research. 3.1 Goals The first goal of the interviews was to explore whether people that have experience with Chaos Engineering have already tried to examine transient behaviors using Chaos Engineering and have experienced difficulties while specifying the behaviors that they wanted to verify. Another goal of the interviews was to find out whether the interviewees agree with the concepts and ideas that can be used to realize the extended Chaos Engineering approach. The last goal of the interviews was to elicit the requirements of the interviewees regarding the extended Chaos Engineering approach, including requirements on the conceptual level as well as the technical realization of the implementation. 3.2 Method The candidates that were invited to participate in an interview were either people involved in the implementation of state-of-the-art Chaos Engineering tools, people working in the industry and practicing Chaos Engineering techniques, or people involved in Chaos Engineering research. When the candidates were contacted they also received an invitation document containing a short description of the interview and a consent form. In total 7 people were contacted of which 4 answered with interest in participating, however, one of them was not able to schedule an interview. As a result, there were three participants and respectively three interviews in total. As arranged with the interviewees, all collected data was kept confidential, so any personal information that could reveal the identity of the interviewees has been obscured in the remainder of the thesis. The first interviewee is a senior software architect with a doctorate degree in computer science. The second interviewee is a master’s student who has done research in the domain of Chaos Engineering and is working as a student research assistant at the time of writing. The third interviewee is also a master’s student involved in Chaos Engineering research, working and applying Chaos Engineering in the industry. 15 3 Elicitation of Requirements for the Extended Chaos Engineering Approach The interviews were conducted online as video calls and in a semi-structured manner. Each interview lasted between 60 and 90 minutes. Additionally, the interviews were recorded with the participant’s consent. The interviews consisted of three parts which were structured as follows: The interviews were conducted online as video calls and in a semi-structured manner. Each interview lasted between 60 and 90 minutes. Additionally, the interviews were recorded with the participant’s consent. The interviews consisted of three parts which were structured as follows: 1. Introduction and Motivation: In this part, the interviewees were given a short introduction and were asked a series of introductory questions with the goal of finding out whether they have experienced the problems that the approach proposed in this thesis aims to solve. 2. Initial Ideas and Concepts: In this part, the interviewees were presented with some initial ideas and concepts regarding the extended Chaos Engineering approach. Furthermore, they were also presented with some technologies and approaches that could be used to implement the concepts into an initial prototype. In addition, this part included a video demonstration of the initial prototype. 3. 3.2.1 Initial Concepts and Prototype The research regarding the first two goals of the thesis, namely to identify how transient behavior requirements can be specified as part of Chaos Experiments and how a verification of such requirements can be carried out during the execution of Chaos Experiments, resulted in the gathering of initial concepts and ideas for the realization of the extended Chaos Engineering approach. These ideas were combined into an initial prototype with the goal of finding out whether the ideas are suitable for the creation of the final concept of the extended Chaos Engineering approach and its implementation. In the sections below, the initial ideas used to construct the initial prototype presented in the interviews are described. 3.2 Method Requirements Elicitation: In this part, the interviewees’ requirements regarding the extended Chaos Engineering approach were elicited with the help of a predefined question catalog. The question catalog was split into three parts (excluding the introductory questions in the first part of the interview), namely conceptual, technical, and non-functional requirements. The conceptual part contained questions regarding the theoretical concepts in the extended Chaos Engineering approach. The technical part consisted of questions regarding the technical realization of the concept. The non-functional requirements part included questions targeting aspects like usability, performance, and extensibility. In the following sections, the contents of the individual parts of the interview are described. Since the introduction to the thesis and the motivation behind it were already covered in the previous chapters, the following sections focus mainly on the initial ideas and concepts that were part of the initial prototype shown to the interviewees and the requirements elicitation questions that the interviewees were asked. In the following sections, the contents of the individual parts of the interview are described. Since the introduction to the thesis and the motivation behind it were already covered in the previous chapters, the following sections focus mainly on the initial ideas and concepts that were part of the initial prototype shown to the interviewees and the requirements elicitation questions that the interviewees were asked. Transient Behavior Specification Formalism As already mentioned in Chapter 2, there are two types of temporal logics, namely linear-time logics, which are interpreted over a single linear trace of events, and branching-time logics, which are interpreted over tree structures of states with multiple possible paths. To identify which interpretation of time is better suited for the needs of Chaos Engineering, the properties of the Chaos Engineering approach itself have to be considered. Each Chaos Experiment that gets executed is 16 3.2 Method producing a single execution run of the system under test. In these runs, there are no multiple possible states in different branches and the observed state traces can be considered linear. As a result, the branching-time logics can be discarded because of their tree-like state interpretation and linear-time logics are the more suitable type of temporal logics. Two prominent examples of linear-time temporal logics have already been introduced in Chapter 2, namely LTL and MTL. Since MTL is an extension of LTL that adds temporal constraints to the temporal operators of LTL, which are particularly useful for specifying requirements concerning the transient phases of the Chaos Experiments while maintaining the original syntax of LTL, MTL proves to be the most viable temporal logic formalism for the proposed extension of the chaos engineering approach. Moreover, the PSPs that were also introduced in Chapter 2 are another promising formalism for expressing transient behavior specifications, especially considering the fact that the users of the extended Chaos Engineering approach might not have prior experience with temporal logic. Therefore, a particularly useful property of the PSPs is that there are already existing mappings of the patterns into temporal logic formalisms including MTL. This additionally removes the need to learn how to express complex behaviors with the logical syntax of MTL and allows an easier definition using only the patterns and then transforming the patterns into MTL expressions. As a result, the PSPs are also considered for the extended Chaos Engineering approach as an abstraction layer for the complexity of MTL. Transient Behavior Verification As already mentioned in Chapter 2, there are already existing approaches and implementations of runtime verification monitors for MTL expressions. Nevertheless, the integration of such monitors with the existing Chaos Engineering tools has to be considered. Namely, a method for triggering the runtime verification and exchanging information between the Chaos Engineering tools and the runtime verification monitor has to be identified. One way to do this is by utilizing the Hypertext Transfer Protocol (HTTP) protocol which is supported by most Chaos Engineering tools as it is used to perform various actions during the execution of Chaos Experiments. Transient Behavior Monitoring The approach of runtime verification requires an event trace of the occurring behaviors in order to be able to deliver verification results. Therefore, an extended Chaos Engineering approach requires the introduction of an additional component, the purpose of which is the monitoring of the events and the creation of the event traces. 3.2.2 Hypotheses In this section, several hypotheses are stated that should be either confirmed or rejected after the interviews are conducted and the interviewees’ answers are collected. The purpose of the hypotheses is to assess whether the respondents agree with the problems that the extended Chaos Engineering approach is aiming to solve and the concepts that will be included in the extended approach. The results regarding each hypothesis and whether the hypotheses are confirmed or rejected will affect the final concept of the extended approach by determining what will be included in it. 𝐻1: The interviewees have experienced situations in which they were not able to define the behaviors that they wanted to verify. 𝐻1: The interviewees have experienced situations in which they were not able to define the behaviors that they wanted to verify. 𝐻2: The interviewees have tried to analyze behaviors that were not regarding the steady state and they were unable to achieve this. 𝐻3: The interviewees think that the proposed extension to the Chaos Engineering approach will provide a deeper insight into the resilience of software systems and view it as a meaningful extension to the traditional Chaos Engineering approach. 𝐻4: The interviewees would like to be able to use the verification of transient behavior specifications outside the scope of Chaos Engineering, i.e., as a stand-alone analysis. 𝐻5: The concept regarding the specification of transient behavior requirements using MTL and PSPs is sufficient and complete. 𝐻6: The concept regarding the verification of transient behavior requirements using runtime verification monitors is sufficient and complete. 𝐻7: A visualization containing the verification results and the monitoring data is a meaningful output of the approach. 𝐻8: Usability, extensibility, and performance are important non-functional requirements for the proposed approach and its prototype. 𝐻8: Usability, extensibility, and performance are important non-functional requirements for the proposed approach and its prototype. 𝐻8: Usability, extensibility, and performance are important non-functional requirements for the proposed approach and its prototype. The Initial Prototype Combining the ideas presented above results in the creation of an initial prototype, used during the interviews with the experts in order to present them with the initial conceptual ideas and demonstrate how such an approach would function. In this initial prototype, the transient behavior specifications are defined as JSON objects which include an MTL formula, information regarding the logical variables in the formula, and information regarding the location and the retrieval of the event traces required for the verification of the specification. Regarding the verification of these specifications, 17 3 Elicitation of Requirements for the Extended Chaos Engineering Approach the initial prototype uses the already existing approach by Ulus [Ulu19b] and its implementation as a Python library [Ulua]. Additionally, the functionalities of the library are provided by a REST API in order to easily enable communication over the HTTP protocol. For the monitoring of the occurring behaviors and the creation of the event traces, a proprietary load generator and an InfluxDB database are used. These are selected since they are already a part of a test system provided by the supervisors of this thesis for the purpose of executing Chaos Experiments on it [FHW+21]. Furthermore, this test system is also used to demonstrate the capabilities of the initial prototype during the expert interviews. Introductory questions: The questions of this category are asked after the first part of the interview, namely after the presentation of the introduction and the motivation. The purpose of these questions is to learn more about the expertise of the interviewees, find out if they have already tried to analyze transient behaviors using Chaos Engineering and whether they have successfully achieved this. 1. How familiar are you with Chaos Engineering and how would you rate your expertise with it? 2. Do you agree with the presented problem statement? 2. Do you agree with the presented problem statement? 2.1. Have you ever experienced situations in which you were not able to define the behaviors that you wanted to verify? 2.2. Have you ever experienced situations in which you wanted to analyze behaviors that were not regarding the steady state and did you successfully achieve this? For example, knowing that a system would deviate from the steady state and analyzing how it actually deviates. 3.2.3 Requirements Elicitation Questions There were four question categories in total, which are listed below. For each category, a short description of the questions it contains and the motivation behind them is provided. There were four question categories in total, which are listed below. For each category, a short description of the questions it contains and the motivation behind them is provided. 18 3.2 Method Conceptual questions: The questions in this category are asked with the intention of finding out what the interviewees think about the concepts, such as MTL, the PSPs, and runtime verification, which are presented to them, how they would use the proposed approach, and whether they have any suggestions or recommendations regarding the presented concepts. This category contains the following questions: 3. Do you think that the proposed extension to the Chaos Engineering approach will enable a deeper insight into the resilience of software systems? Do you consider such an extension meaningful? 4. How do you practice Chaos Engineering and how could the proposed approach be integrated into your workflow? 5. In what kind of situations would you use the proposed approach? 6. How often would you use the proposed approach? 7. Would you use parts of the approach to verify transient behavior requirements outside the scope of Chaos Engineering, i.e., as a stand-alone analysis or coupled with other types of analysis? 8. Do you think that the concept regarding the transient behavior specification part is sufficient and complete? If not, what ideas would you include? 9. Do you think that the concept regarding the transient behavior verification part is sufficient and complete? If not, what ideas would you include? 10. Currently the approach considers only event traces with discrete timestamps, i.e. at certain time points, e.g. every second/minute/hour. Does this affect you and do you consider it a weakness of the approach? 19 3 Elicitation of Requirements for the Extended Chaos Engineering Approach Technical questions: The purpose of this category is to find out what the interviewees think about some of the technical aspects and properties of the prototype. The category contains the questions below. 11. Do you think that the outputs of the approach are informative enough? Would you like additional outputs to be included? If yes, what would they be? 12. Do you think that a figure, such as a line plot including the measurement data and the evaluation results, is a meaningful addition to the outputs of the proposed approach? How important would that be for you and would you use it frequently? 13. Which one of the following options would you prefer for the input of the formal behavior specification: 13.1. Use an MTL formula as the input. This would first require you to create the formula on your own or with the support of the PSPs and tools like PSP Wizard and then to transform it into the textual MTL syntax as shown in the presentation. This option would allow the coverage of a wider range of behaviors but it would also require you to get familiar with MTL. 13.2. Use a PSP defined as a simple text sentence as the input. This would only require you to select a pattern from a predefined pattern catalog and to provide information for the events in the pattern. This option does not require you to get familiar with MTL but the behavior coverage would be limited to the selection of supported patterns. 14. Do you think that the JSON inputs are expressive enough and contain all the necessary information? If not, what would you like to be added? 15. What do you think about the predicate inputs and functions? What predicate functions would you like to be included? 16. From what sources should the approach be able to retrieve monitoring data? 16. From what sources should the approach be able to retrieve monitoring data? 16.1. What monitoring techniques and tools do you use and how do they preserve the monitored data? 16.1. What monitoring techniques and tools do you use and how do they preserve the monitored data? 16.2. How would you like these to be integrated into the approach? 16.3. Do you have suggestions for additional monitoring techniques and tools that should be integrated? 17. Should the proposed approach and its proposed tooling be integrated with other existing approaches and tools? Non-functional requirements: The purpose of this last category of questions is to find out whether the interviewees have some non-functional requirements regarding the prototype presented to them. The category consisted of the following questions: 18. Do you think that Usability is an important non-functional requirement for the proposed approach and prototype? How could the Usability of the approach and the prototype be improved? 18. Do you think that Usability is an important non-functional requirement for the proposed approach and prototype? How could the Usability of the approach and the prototype be improved? 19. Do you think that Extensibility is an important non-functional requirement for the proposed approach and prototype? Would you like to be able to easily extend parts of the prototype, e.g. the predicate functions or the integration with additional monitoring approaches? 20. Do you think that Performance is an important non-functional requirement for the proposed approach and prototype? What would be an acceptable time until you get a result? What size of systems should the approach be able to handle? What monitoring duration, i.e. trace lengths should be supported? 3.3 Results The results of the three interviews for each category of questions are presented below in the respective section. Technical questions: 20 3.3 Results 3.3.1 Introductory questions All three of the interviewees stated that they have experience with Chaos Engineering which they rated on average as 7.2 out of 10, however, two of the interviewees said that they are lacking a bit of practical experience. Two out of the three interviewees have experienced situations in which they were not able to define the behaviors that they wanted to verify during Chaos Experiments. All of the interviewees have experienced situations in which they wanted to verify behaviors that were not concerning the steady state. Two of the interviewees were able to verify these behaviors, however, one of them had to do a lot of manual work to achieve this, and the third interviewee was able to achieve this only partially. 3.3.3 Technical questions When asked if they thought the results of the approach were informative enough, all interviewees added that additional information such as information about the time intervals and frames as well as the value deviations would be useful. Two of the interviewees said that a plot visualizing the monitoring data and the verification results would be a useful and important addition to the prototype, while the third participant said that this feature would be nice to have but not very important. When presented with the different input options for specifying transient behavior, namely an MTL formula or a PSP, all interviewees said they would like to have both for different reasons. Some of the main arguments in favor of PSPs are that it is generally quicker to create a specification with them and that they are more practical and readable, especially when a person is working with specifications that they did not create themselves. An argument in favor of MTL is that it could enable more precise and detailed specifications by using the full extent of the logical formalism. Regarding the JSON fields concerning the behavior specification, the interviewees only suggested additional logical predicates, such as increasing and decreasing trends and predicate functions supporting Boolean values occurring in the event traces. Regarding any monitoring tools that can be integrated into the approach, one of the interviewees could not think of any tools to be integrated, another interviewee said that he is currently using Comma-Separated Values (CSV) files to capture the events in his system, and the last one suggested supporting time series databases in general, and Prometheus in particular. 3.3.2 Conceptual questions All of the participants think that the proposed extension to the Chaos Engineering approach will provide a deeper insight into the resilience of software systems. In response to the question asking about how the extended approach could be adopted by the interviewees, one of them said that he was not actively practicing Chaos Engineering at the moment so he could not answer properly. Another interviewee answered that some interfaces or adapters would be required to integrate the extended Chaos Engineering approach into his workflow. The last interviewee answered that integration into his workflow would not be difficult since he uses similar technologies as the initial prototype. Two out of the three interviewees answered that they would regularly use the proposed approach if their 21 3 Elicitation of Requirements for the Extended Chaos Engineering Approach Chaos Experiments have some defined criteria regarding the transient behaviors of the system. The third interviewee answered that he would use the proposed approach in situations where he would like to know more about the system than just if a certain service is still running. Additionally, two of the interviewees said that they could imagine using the approach and the prototype for scenarios and use cases outside the scope of Chaos Engineering and Chaos Experiments. When asked whether they think the concept regarding the specification of the transient behaviors is complete and sufficient, all of them answered positively, however, two of them expressed the concern that extracting meaningful values for the specifications might be difficult. Additionally, two of the participants said that some form of assistance for the creation of the transient behavior specifications would be quite helpful. When the interviewees were asked whether they think that the concept regarding the verification of the specifications is complete and sufficient, they again answered mostly positively, only making remarks that including more information into the output of the verification such as which parts of the behavior specification failed, time intervals, and differences would be useful. Regarding the question enquiring whether the interviewees view the discrete nature of the event traces as a limitation of the approach, two of the participants said that this could be a limitation of the approach whereas the third interviewee said that in practice this is a strength of the approach since he is not aware of monitoring tools that capture measurement data in non-discrete ways. 3.3.5 Interpretation In this section, the results of the interviews are interpreted and analyzed with regard to the hypotheses defined in Section 3.2.2 which are subsequently confirmed or rejected. The results regarding the hypotheses will also decide what will be included in the final concept of the extended Chaos Engineering approach. The answers to the introductory questions confirm hypotheses 𝐻1 and 𝐻2, as the majority of participants had already experienced difficulties while defining the behaviors they wanted to verify by using Chaos Engineering and also while analyzing transient behaviors with it. These results further solidify the motivation of the thesis and the need for an extension of the Chaos Engineering approach that would allow an analysis of the transient behaviors. Furthermore, all of the interviewees consider such an extension meaningful which as a result confirms hypothesis 𝐻3. A majority of the participants indicated that they would use the approach and prototype for other scenarios outside of Chaos Engineering. This confirms hypothesis 𝐻4 and as a result, introduces the requirement that the approach and its implementation should be able to provide verification of transient behavior requirements to other tools and approaches outside of Chaos Engineering. The approach for specifying transient behavior requirements using MTL and PSPs was considered sufficient and complete by all interviewees. As a result, this confirms hypothesis 𝐻5 and leads to a requirement for the final concept of the extended Chaos Engineering approach, which is to support both MTL and PSPs for the specification of the behaviors. Furthermore, the interviewees also considered the concepts regarding the verification of the transient behavior requirements and in particular, the usage of runtime verification as sufficient and complete, which as a result confirms hypothesis 𝐻6 and results in the requirement to integrate runtime verification monitors into the final concept for the approach. Additionally, since a majority of the interviewees also stated that a visualization of the verification results and the monitoring data in the form of a plot is an important addition to the approach and the prototype, hypothesis 𝐻7 is also confirmed and the inclusion of a visualization is also considered as a requirement for the final concept. The interviewees also stated that the non-functional qualities of usability, extensibility, and performance are important for the final concept and the resulting prototype, and therefore also hypothesis 𝐻8 is confirmed. 3.3.4 Non-functional requirements One of the interviewees expressed the opinion that usability is the most important non-functional requirement for the prototype and that other features could even be neglected in favor of usability. Two of the interviewees also said that a user interface would increase the usability of the prototype. Additionally, two of the interviewees also said that having a validation of the JSON inputs against a 22 3.3 Results 3.3 Results JSON schema would also increase the usability. Furthermore, all interviewees said that extensibility is also an important feature of the prototype and would like to be able to extend parts of it if necessary. Regarding performance, two interviewees said that the response time of the verification is not of utmost importance and that the event trace lengths should not be very long as the transient behaviors normally occur for a shorter amount of time, e.g. couple of minutes. JSON schema would also increase the usability. Furthermore, all interviewees said that extensibility is also an important feature of the prototype and would like to be able to extend parts of it if necessary. Regarding performance, two interviewees said that the response time of the verification is not of utmost importance and that the event trace lengths should not be very long as the transient behaviors normally occur for a shorter amount of time, e.g. couple of minutes. 4 Extended Chaos Engineering Approach After identifying initial concepts that could be used to extend the Chaos Engineering approach to support the verification of transient behavioral specifications from literature, and after conducting the interviews with the experts, enough requirements have been collected so that they can now be combined into a final concept for the extended Chaos Engineering approach. As defined by Rosenthal and Jones [RJ20], the steps in the Chaos Experimentation process are as follows: (1) build a steady-state hypothesis, (2) vary real-world events, (3) run experiments in production, (4) automate experiments, (5) minimize blast radius. Since the Chaos Engineering approach is extended, also the above-mentioned steps have to be extended accordingly. A newly introduced step is the creation of the transient behavior hypothesis and its place in the extended Chaos Experimentation process is right after the creation of the steady-state hypothesis. As a result, the extended Chaos Experimentation process is defined as follows: (1) build a steady-state hypothesis, (2) build a transient behavior hypothesis, (3) vary real-world events, (4) run experiments in production, (5) automate experiments, (6) minimize blast radius. Furthermore, the characteristics of the transient behavior hypothesis and its differences from the steady-state hypothesis have to be defined. Traditionally, the steady-state hypothesis of an experiment is a simple statement regarding a single measurement of the system at a certain time point, which is verified against a certain tolerance [Mil19]. Even if the steady-state hypothesis is carried out continuously during an experiment, it is still a statement regarding only a single measurement at a single point in time and its acceptance criteria can be inherently different from those of the transient states. In contrast, the transient behavior hypothesis does not refer to a single point in time, but to a series of points in time or even to the entire experiment in which the transient states and behaviors of the system occur. Therefore, the transient behavior hypothesis must be able to define statements over longer time periods and not only a single point in time like the steady-state hypothesis. Furthermore, the transient behavior hypothesis should also be able to define statements regarding multiple measurements and metrics of the system since the transient behaviors and states of a system are rarely affecting a single component or service. These metrics must be able to cover properties that can be observed either from the outside of the system, e.g. response time, or the inside, e.g. 3.3.5 Interpretation Since a majority of the interviewed people said that the introduction of a graphical user interface would significantly increase the usability of the prototype, the introduction of such is included as a requirement for the final concept and the implementation of the prototype. 23 4 Extended Chaos Engineering Approach number of active instances, utilization of the instances, etc. As the transient behavior hypothesis can concern multiple measurements there must also be a way to connect and compare them in a formal way. Similarly to the steady-state hypothesis verification, also the verification of the transient behavior hypothesis should produce a single Boolean result. In addition to extending the Chaos Experimentation process, also the elements and concepts of the Chaos Engineering approach have to be extended. Traditionally these include a Chaos Experiment consisting of a steady-state hypothesis, an anomaly that is injected, and a test system on which the experiment is conducted. Even though the Chaos Experimentation steps that are listed above can be executed manually, the usage of tools that can support software engineers in executing experiments is particularly helpful also for the step which regards the automation of 25 4 Extended Chaos Engineering Approach the Chaos Experiments [RJ20]. Furthermore, various research has also successfully applied the principles of Chaos Engineering by using a Chaos Engineering tool to automate the execution of experiments [FHW+21; KHFH20]. Therefore the usage of a Chaos Engineering tool is considered a part of the traditional Chaos Engineering approach. Next, the traditional Chaos Engineering approach is extended similarly to how the steps of the Chaos Experimentation process are extended. The extended approach introduces several new elements to the already existing ones. These new elements are a transient behavior hypothesis, a transient behavior verification, a transient behavior monitoring of the system’s events, a visualization of the results, and a Graphical User Interface (GUI). The purpose of the transient behavior hypothesis is to make testable statements regarding the system under test as described above. The purpose of the transient behavior validation is to provide a concrete result regarding whether the behavior hypothesis holds or is rejected. The goal of the visualization of the result is to present the results in an intuitive and readable manner to the users of the approach as requested by the interviewees. As elicited in the expert interviews, the extended approach should also be usable as a stand-alone analysis without the need of a Chaos Engineering tool or Chaos Engineering itself, therefore the extended approach also includes a GUI providing this stand-alone analysis. Figure 4.1 depicts an overview of the extended Chaos Engineering approach, including the already existing and the newly introduced elements and the interactions between them. 4 Extended Chaos Engineering Approach Steady-state Hypothesis Anomaly Injection Transient Behavior Hypothesis Chaos Experiment Chaos Engineering Tool System Under Test Transient Behavior Verification Transient Behavior Monitoring Experiment Result <execute> <visualize result> <write results> <manipulate> <retrieve data> - existing element - new element - existing interaction - new interaction GUI Result Visualization <send specification> <monitor> <result> <result> Figure 4.1: Extended Chaos Engineering Approach - Concept Overview. <execute> Transient Behavior Verification Result Visualization Anomaly Injection <send specification> Transient Behavior Hypothesis Transient Behavior Monitoring GUI - existing element Figure 4.1: Extended Chaos Engineering Approach - Concept Overview. 4.1 Transient Behavior Hypothesis Specification The first element introduced by the concept is the transient behavior specification section. The specification section is introduced as an additional step into the Chaos Experiment since it is conceptually different from the steady-state hypothesis and a separate step is where it inherently 26 4.1 Transient Behavior Hypothesis Specification Steady-state Hypothesis Anomaly injection Transient Behavior Hypothesis Chaos Experiment MTL Formula/PSP Definition Predicate Information Event Trace - existing element - new element Figure 4.2: Transient Behavior Hypothesis Specification Concept. - new element Figure 4.2: Transient Behavior Hypothesis Specification Concept. belongs. The specification must contain a formal description of the transient behavior in order to give it a specific meaning and interpretation, and to make it compatible with approaches like runtime verification. As a result, the specification must use a formalism capable of describing the temporal properties of the transient behaviors, for which also runtime verification monitors have already been constructed. Such formalisms were already introduced in Chapter 2, namely MTL and the PSPs which can be seen as an abstraction layer, abstracting the complexity of the logical syntax of the different temporal logics. Additionally, including PSPs allows the users of the approach to also use tools like the PSP Wizard by Grunske et al. [AGL+15] in order to easily create PSP definitions, which as a result would also increase the usability of the proposed approach. Furthermore, the approach and the framework proposed by Grunske et al. [AGL+15] have already been successfully applied to formalize the requirements of an industrial case study as PSP definitions. As a result, the PSPs are also considered viable for the definition of the transient behavior hypotheses. Therefore, the transient behavior specification must contain a formal behavior description specified as either an MTL formula or a PSP definition. In addition, since the logical variables contained in an MTL formula or a PSP definition do not include information regarding how they should be interpreted, the behavior specification must also contain information concerning the logical variables, events, and predicates featured in the behavior description. To illustrate this, consider the following MTL formula: ■(𝐴→■(𝐵)). In order to enable an evaluation of the formula to either 𝑡𝑟𝑢𝑒or 𝑓𝑎𝑙𝑠𝑒, the logical variables 𝐴and 𝐵and their assignments have to be defined. 4.2 Transient Behavior Verification The second element required by the concept for the extended Chaos Engineering approach is the verification of the transient behavior hypothesis. Firstly, the verification must be able to be triggered and to receive transient behavior requirements from the Chaos Engineering tool and also from the GUI. Furthermore, the verification must be able to return its results to both the Chaos Engineering tool and the GUI. As elicited in the expert interviews, the verification must be able to return a Boolean result whether the transient behavior hypothesis holds. The purpose of such a concrete Boolean value regarding the result of the verification is the fact that it provides a simple means to understand whether the hypothesis holds or not. A further argument for the inclusion of such a result is the fact that also the result of the steady-state hypothesis is a concrete Boolean value in the traditional Chaos Engineering approach. Nevertheless, as requested by the interviewees, additional outputs regarding the result of the verification should also be included with the Boolean value. As elicited, this additional information includes more detailed information regarding the different time intervals in the verification, such as the time points in which the result changes from one value to another, e.g. from 𝑡𝑟𝑢𝑒to 𝑓𝑎𝑙𝑠𝑒, and the values that caused this change. The inclusion of these additional outputs provides the reasoning behind the final verification result and makes the outputs of the extended approach more informative altogether. In order to be able to create all of the above-discussed results and to correctly verify the transient behavior specifications, the verification must include the usage of an approach such as runtime verification [LS09]. Since this requires also the usage of monitoring data regarding the system’s states, the transient behavior verification must also be able to retrieve such from a transient behavior monitoring approach. 4.1 Transient Behavior Hypothesis Specification As already mentioned, these 27 4 Extended Chaos Engineering Approach assignments have to be comparisons concerning the event traces and the measurement data, and they should be easily mappable to Boolean values, i.e. 𝑡𝑟𝑢𝑒or 𝑓𝑎𝑙𝑠𝑒. Some example assignments are comparisons of the response time or the instance count against a certain threshold. Furthermore, the behavior specification requires the specification of an event trace of the behaviors occurring in the system, in order to enable the runtime verification of the specified behavior. Figure 4.2 depicts a Chaos Experiment containing a transient behavior specification including all newly introduced elements. 4.5 Transient Behavior Monitoring Even though transient behavior monitoring is considered in this concept, this thesis is focusing only on the specification and verification of transient behaviors. As a result, the concept simply introduces the usage of such an approach and assumes that it is capable of monitoring the events occurring in the system under test. Additionally, the concept assumes that the monitoring approach is monitoring and measuring the occurring events in a discrete-time manner and the monitored event traces can be easily retrieved by the transient behavior verification. Particularly suitable for this purpose are time series databases because of the way they store data. Furthermore, since there may be various monitoring approaches that provide the monitored data in different formats and ways, there should be a common format that is supported by the concept, in order to abstract the differences that the various monitoring approaches may have. Such a common format are CSV files. 4.4 Graphical User Interface A majority of the interviewees said that an inclusion of a GUI that supports the user in creating transient behavior hypothesis specifications and the insertion of these specifications into the Chaos Experiments would significantly increase the usability of the proposed approach. Furthermore, a majority of the interviewees also requested that the approach should also be usable outside the scope of Chaos Engineering and Chaos Experiments. Therefore, in order to provide these, the concept for the transient behavior verification also includes a GUI that enables and supports the users while specifying transient behavior hypothesis specifications, their insertion into Chaos Experiments, and the stand-alone execution of transient behavior verifications. 4.3 Verification Result Visualization As requested by the interviewees, the above-discussed results of the verification are also presented visually in the form of a plot which also includes the event traces used in the verification in order to make the interpretation of the results easier. The inclusion of a graphical presentation of the measurement data provides further reasoning behind the results of the verification without the need to manually retrieve and examine the measurement data which is often contained in unintuitive ways in the event traces. Additionally, the visualization of the measurement data and the verification results may enable further analyses or comparisons of the data as it contains more information than the Boolean results returned by the verification. The proposed plot for the visualization is a line chart that includes the event traces used during the verification presented as lines and the results of the verification are indicated by the background colors of the respective section in the plot. 28 4.4 Graphical User Interface 4.6 Interaction of the Proposed Concepts The purpose of this section is to describe how all elements interact with each other when a Chaos Experiment containing a transient behavior hypothesis specification is executed using the extended Chaos Engineering approach proposed in this chapter. First, the initial steady-state check and the anomaly injection get executed as in the traditional Chaos Engineering approach. After the anomaly has been injected, enough time has passed to let the transient behaviors occur, and the second steady-state check has been performed, the transient behavior check will be initiated. For this, the Chaos Engineering tool will trigger the verification and will also send the transient behavior specification to it. After the transient behavior specification has been received, the monitoring data required for the verification of the specification is retrieved, and afterward, the specification is prepared for verification. After the verification is complete, the results and the monitoring data are plotted and the results are returned from the verification to the Chaos Engineering tool. This whole process should take at most a couple of seconds and once it is completed, the Chaos Engineering tool should have a clear Boolean result regarding the verification of the transient behaviors, i.e. it should have received a clear 𝑡𝑟𝑢𝑒/ 𝑓𝑎𝑙𝑠𝑒result. Figure 4.3 depicts the above described interaction as a UML sequence diagram. 29 4 Extended Chaos Engineering Approach :Chaos Engineering Tool send transient behavior specification :Transient Behavior Monitoring check steady-state retrieve monitoring data :Transient Behavior Verification verify behavior prepare specification System under test experience transient behaviors start experiment execution inject anomaly throughout the whole experiment monitor behaviors plot data check steady-state return experiment result return verification result Figure 4.3: Transient Behavior Verification - Sequence Diagram. Figure 4.3: Transient Behavior Verification - Sequence Diagram. 30 5 Prototype Implementation In this chapter, an overview of how the concept presented in Chapter 4 is implemented into a final functioning prototype is presented. The development of the prototype includes the selection of a Chaos Engineering tool which is responsible for executing the Chaos Experiments, the definition of a format for the transient behavior specification, the development of the transient behavior verification tool, and the integration of several time series databases used as monitoring tools. In the final implementation of the prototype, Chaos Toolkit is used as the Chaos Engineering tool of choice. The behavior verification, result visualization, and GUI are implemented as the components of a Python [Pyt] application. Because of the discrete-time nature of the transient behavior verification, the prototype requires the integration of time series databases as monitoring tools. The monitoring tools that the final prototype supports are Prometheus [Pro] and InfluxDB [Inf]. Additionally, the prototype also supports the retrieval of monitoring data from CSV files. An overview of the prototype and the items it interacts with are presented in Figure 5.1. Chaos Toolkit Transient Behavior Verification Tool Transient Behavior Monitoring Tool Chaos Experiment CSV CSV System under test runs Transient Behavior Specifiation manipulates Figure 5.1: Prototype Components - Overview. System under test Chaos Toolkit Transient Behavior Specifiation Transient Behavior Monitoring Tool Figure 5.1: Prototype Components - Overview. 5.1 Chaos Engineering Tool The first thing required for the implementation of the concept is the selection of the Chaos Engineering tool that the prototype uses to run Chaos Experiments. For this purpose, the following are identified as possible candidates: Chaos Mesh [Chaa], Gremlin [Gre], ChaosBlade [Chad], Chaos Toolkit [Chac], Litmus [Lit], and Chaos Monkey [Chab]. Table 5.1 presents a comparison of the above-mentioned tools with regard to the different layers on which they support failure injections and whether they are open-source since those are important criteria in the selection of the tool in the final prototype implementation. 31 5 Prototype Implementation Chaos Mesh Chaos Monkey Gramlin Chaos Toolkit Litmus ChaosBlade Platform layer injections x x x x x x Network layer injections x x Application layer injections x x Open-source x x x x x Table 5.1: Chaos Engineering Tools - Comparison The first tool that is removed from the selection pool is ChaosBlade as it is lacking documentation written in English. The second tool that is removed from the selection pool is Gremlin, as it is not free-to-use. From the remaining tools, Chaos Toolkit supports the injection of failures and anomalies on more layers than the other tools. Another important factor also in favor of Chaos Toolkit is that it is the only tool that specifies Chaos Experiments in the format defined by Rosenthal and Jones [RJ20], namely containing a steady-state hypothesis and an anomaly injection. The rest of the Chaos Engineering tools either use proprietary Chaos Experiment formats that are not compatible with the one defined by Rosenthal and Jones [RJ20] or do not have any formal definition of what a Chaos Experiment contains. Furthermore, research has already shown that Chaos Toolkit is a tool capable of correctly executing Chaos Experiments [FHW+21; KHFH20]. In addition, Chaos Toolkit has already been successfully extended with an additional hypothesis validator [FHW+21], which is similar to the transient behavior verification proposed by the concept in Chapter 4. As a result, considering the above-discussed facts, Chaos Toolkit is chosen as the Chaos Engineering tool, which is used as part of the prototype. Additionally, a communication protocol has to be selected, with which the Chaos Engineering tool has to exchange data with the verification tool. A common protocol supported by at least two of the Chaos Engineering tools, namely Chaos Toolkit and Gremlin, is the HTTP protocol. 5.1 Chaos Engineering Tool As a result, the HTTP protocol is selected for the communication between the Chaos Engineering tool and the verification tool. 5.2 Transient Behavior Specification As already mentioned in the previous chapter, the specification should contain a transient behavior specified as an MTL formula or a PSP definition together with information regarding the behavior’s logical predicates and the measurement data. In addition, the format that is to be selected should be compatible with the HTTP protocol as this is how the specification will be sent to the verification tool. As a result, the technology and the format selected to contain the above-mentioned information is a JSON object. In the following, each field of the JSON object is listed and explained. behavior_description: An optional text field that allows the specification of a text description of the transient behavior. behavior_description: An optional text field that allows the specification of a text description o the transient behavior. specification: A required text field containing the formal description of the behavior, which is either an MTL formula, where the symbols for the temporal operators are replaced with their names or a textual PSP definition created using the PSP Wizard tool [AGL+15]. specification: A required text field containing the formal description of the behavior, which is either an MTL formula, where the symbols for the temporal operators are replaced with their names or a textual PSP definition created using the PSP Wizard tool [AGL+15]. 32 5.2 Transient Behavior Specification specification_type: A required text field specifying the type of the formal behavior specification, can be either “mtl“ or “psp“. specification_type: A required text field specifying the type of the formal behavior specification, can be either “mtl“ or “psp“. future-mtl: An optional field allowing to specify that a future-MTL formula has been defined in the “specification“ field. By default, the prototype assumes that the specified behaviors are in past-MTL. Additionally, the field should only be used when “specification_type“ is set to “mtl“. predicates_info: A required array containing the information regarding the logical predicates in the formal behavior specification. The array contains objects of the following format: predicate_description: An optional text field that allows the specification of a text descrip- tion of the logical predicate. predicate_name: A required text field containing the predicate name as it occurs in the formal specification above. predicate_logic: A required text field containing the name of a logical function from a predefined selection of Boolean operations provided by the prototype. predicate_comparison_value: A text field containing a value used as the comparison value for the respective logical operation. 5.2 Transient Behavior Specification Not all functions require the inclusion of this field, e.g., the “boolean“ function and the trend functions. measurement_source: A required text field defining the source of the measurement data, currently supporting “influx“ for InfluxDB, “prometheus“ for Prometheus,“csv“ for local CSV files, and “remote-csv“ for remote CSV files, for example hosted on a web server. remote-csv-address: Required when “measurement_source“ is set to “remote-csv“. This field contains the URL to the CSV table. contains the URL to the CSV table. measurement_points: A required array containing information regarding the measurement data. measurement_name: A required text field containing the name of the measurement as it occurs in the behavior specification. measurement_query: A text field specifying the query that will be used to retrieve the data. Only required when “measurement_source“ is specified as “influx“ or “prometheus“. measurement_column: A text field specifying the column from which the measurement data will be retrieved. Only required when “measurement_source“ is set to “csv“ or “remote-csv“. start_time: A text field for specifying the start time of the query interval. Only required when “measurement_source“ is set to “prometheus“. end_time: A text field for specifying the end time of the query interval. Only required when “measurement_source“ is set to “prometheus“. steps: A text field for specifying the steps of the query interval. Only required when “measurement_source“ is set to “prometheus“. easurement_points: A required array containing information regarding the measurement data. measurement_name: A required text field containing the name of the measurement as it occurs in the behavior specification. measurement_query: A text field specifying the query that will be used to retrieve the data. Only required when “measurement_source“ is specified as “influx“ or “prometheus“. measurement_column: A text field specifying the column from which the measurement data will be retrieved. Only required when “measurement_source“ is set to “csv“ or “remote-csv“. start_time: A text field for specifying the start time of the query interval. Only required when “measurement_source“ is set to “prometheus“. end_time: A text field for specifying the end time of the query interval. Only required when “measurement_source“ is set to “prometheus“. steps: A text field for specifying the steps of the query interval. Only required when “measurement_source“ is set to “prometheus“. 33 5 Prototype Implementation Listing 5.1 Transient Behavior Specification - MTL & CSV Example. Listing 5.1 Transient Behavior Specification - MTL & CSV Example. Listing 5.1 Transient Behavior Specification - MTL & CSV Example. 5.2 Transient Behavior Specification 1 { 2 "behavior_description": "When the response time of a service changes from normal to high, ↩→it can remain high for at most 5 time units.", 3 "specification": "( resp_time_high(resp_time) since[0,5] resp_time_stable(resp_time) )", 4 "specification_type": "mtl", 5 "predicates_info": [ 6 { 7 "predicate_name": "resp_time_stable", 8 "predicate_description": "Response time is stable below 0.5", 9 "predicate_comparison_value": "0.5", 10 "predicate_logic": "smallerEqual" 11 }, 12 { 13 "predicate_name": "resp_time_high", 14 "predicate_description": "Response time is unstable above 0.5", 15 "predicate_comparison_value": "0.5", 16 "predicate_logic": "bigger" 17 } 18 ], 19 "measurement_source": "csv", 20 "measurement_points": [ 21 { 22 "measurement_name": "resp_time", 23 "measurement_column": "resp_time" 24 } 25 ] 26 } Listing 5.1 presents an example transient behavior specification using an MTL formula as the formal specification and a CSV file as the source of the event traces. The formula contains two logical predicates, namely resp_time_high and resp_time_stable. The resp_time_high is 𝑡𝑟𝑢𝑒when the resp_time measurement is bigger than 0.5 and the resp_time_stable is 𝑡𝑟𝑢𝑒when the resp_time measurement is equal to or smaller than 0.5. The specification also defines that the resp_time measurement is retrieved from the table column with the same name. Listing 5.2 presents a transient behavior specification using a PSP definition. The pattern used in the specification contains the resp_time_high and instance_increase predicates. The resp_time_high predicate evaluates to 𝑡𝑟𝑢𝑒when the resp_time measurements are greater than or equal to 100 and the instance_increase predicate evaluates to 𝑡𝑟𝑢𝑒when the instance-count measurements are strictly increasing. The transient behavior specification also defines the source of the resp_time and instance_count measurements, which in this case is an InfluxDB instance, and the queries used to retrieve the individual measurements. 34 5.2 Transient Behavior Specification Listing 5.2 Transient Behavior Specification - PSP & InfluxDB Example. Listing 5.2 Transient Behavior Specification - PSP & InfluxDB Example. 5.2 Transient Behavior Specification 1 { 2 "behavior_description": "When the response time exceeds 100 ms, a new instance of the ↩→service should be spawned.", 3 "specification": "Globally, if {resp_time_high(resp_time)} [has occurred] then in ↩→response {instance_increase(instance_count)} [eventually holds] between 30 and 60 time ↩→units.", 4 "specification_type": "psp", 5 "predicates_info": [ 6 { 7 "predicate_name": "resp_time_high", 8 "predicate_description": "Response time is higher than 100ms", 9 "predicate_comparison_value": "100", 10 "predicate_logic": "biggerEqual" 11 }, 12 { 13 "predicate_name": "instance_increase", 14 "predicate_description": "Instance counts increase", 15 "predicate_logic": "trendUpwardStrict" 16 } 17 ], 18 "measurement_source": "influx", 19 "measurement_points": [ 20 { 21 "measurement_name":"resp_time", 22 "measurement_query":"SELECT \"AvgResponseTime\" FROM ↩→\"TimeBatchRuns\".\"autogen\".\"Batch_Time\"" 23 }, 24 { 25 "measurement_name":"instance_count", 26 "measurement_query":"SELECT \"counts\" FROM ↩→\"AliveInstances\".\"autogen\".\"InstanceCounts\"" 27 } 28 ] 29 } As already mentioned and shown in Listings 5.1 and 5.2, the prototype supports behavior definitions using MTL formulas and PSPs. When a PSP behavior specification is received by the verification tool, the PSP behavior definitions are translated into past-MTL formulas using already existing mappings for each pattern. At the time of writing, the supported patterns for which there are mappings into past-MTL are the following: Absence, Universality, Recurrence, and Response. For these patterns, a simple natural language definition as shown in Listing 5.2 can be used. For any other PSP that is not supported or that should be mapped into future-MTL, the MTL formula resulting from the mapping of the PSP to MTL should be used instead. One approach allowing an easy mapping of PSPs into future-MTL formulas is the PSP Wizard tool by Grunske et al. [AGL+15]. Additionally, the Boolean functions that can be assigned to the logical predicates in the behavior specifications are also limited to a certain selection. At the time of writing, the prototype supports simple numerical comparisons such as “bigger“, “biggerEqual‘, “smaller“, “smallerEqual“,“equal“, and trend functions such as “trendUpward“, “trendUpwardStrict“, “trendDownward“, “trendDownwardStrict“. Even though 35 5 Prototype Implementation the selection regarding the PSPs and the logical functions is limited, these are implemented in a way that allows easy extensions when required. The extension of the supported PSPs requires only the addition of a regular expression for the textual representation of the pattern in the file containing the remaining such and the MTL formula to which the pattern is mapped in another class containing the mappings. 5.2 Transient Behavior Specification The extension of the Boolean functions can be done by simply adding further Python functions that return a Boolean value in the respective class containing these. 5.2.1 Transient Behavior Specifications in Chaos Experiments The transient behavior specifications shown above now have to be incorporated into Chaos Experiments. As Chaos Toolkit is the Chaos Engineering tool of choice for the prototype, the Chaos Experiments are defined using Chaos Toolkit’s format and the transient behavior specification is adapted to be compatible with Chaos Toolkit. Furthermore, in order to send the behavior specification to the verification component using the HTTP protocol, the transient behavior specification in the Chaos Experiment is defined inside of a probe that uses the HTTP provider. This probe is placed in the steady-state hypothesis section of the experiment even though the transient behavior specification is conceptually different, because any probes outside of the steady-state hypothesis, for example in the method, do not check the predefined tolerances. Although this is a conceptual limitation of the Chaos Experiment and Chaos Toolkit, this does not affect the execution of the experiment and the verification of the transient behavior specification. As already mentioned, the probe uses the HTTP provider of Chaos Toolkit, which requires the specification of some additional fields, for example, the URL of the endpoint, the request method, and information regarding headers among others. The tolerance of the probe is set to the expected verification result. An example of a Chaos Experiment containing a transient behavior specification using Chaos Toolkit’s format can be seen in Listings 5.3 and 5.4. Listing 5.3 presents the steady-state hypothesis and the method of the experiment, however, without the probes of the steady-state hypothesis because of size restrictions. Thus, the probes are displayed in Listing 5.4. In the presented experiment, the steady-state hypothesis of the experiment is that the payment deployment is fully available. The failure that is injected is a termination of one of the payment pods. The failure is introduced at the 30th second of the experiment and the verification of the steady-state hypothesis and the transient behavior hypothesis are carried out 90 seconds after that, i.e. at the 120th second. The transient behavior hypothesis is specified using a PSP and its meaning is that after the response time has been stable, it is never the case that it becomes unstable. 36 5.2 Transient Behavior Specification Listing 5.3 Chaos Experiment Containing a Transient Behavior Specification - Steady-State Hypothesis (Without Probes) and Method. (Continued in Listing 5.4) Listing 5.3 Chaos Experiment Containing a Transient Behavior Specification - Steady-State Hypothesis (Without Probes) and Method. 5.2.1 Transient Behavior Specifications in Chaos Experiments (Continued in Listing 5.4) yp ( ) ( g ) 1 { 2 "title":"Terminate one payment pod", 3 "description":"Terminating one payment pod and observing steady-state hypothesis and ↩→transient behavior", 4 "tags":["kubernetes"], 5 "steady-state-hypothesis":{ 6 "title":"Verifying that deployment remains healthy", 7 "probes":[...] 8 }, 9 "method":[ 10 { 11 "type":"action", 12 "name":"terminate-payment-service-pod", 13 "provider":{ 14 "type":"python", 15 "module":"chaosk8s.pod.actions", 16 "func":"terminate_pods", 17 "arguments":{ 18 "name_pattern":"payment", 19 "rand":true 20 } 21 }, 22 "pauses":{ 23 "before":30, 24 "after":90 25 } 26 } 27 ] 28 } 37 37 5 Prototype Implementation Listing 5.4 Chaos Experiment Containing a Transient Behavior Specification - Probes in the Steady-State Hypothesis (Continuation of Listing 5.3). Listing 5.4 Chaos Experiment Containing a Transient Behavior Specification - Probes in the Steady-State Hypothesis (Continuation of Listing 5.3). Listing 5.4 Chaos Experiment Containing a Transient Behavior Specification - Probes in the Steady-State Hypothesis (Continuation of Listing 5.3). Listing 5.4 Chaos Experiment Containing a Transient Behavior Specification - Probes in th Steady-State Hypothesis (Continuation of Listing 5.3). Steady-State Hypothesis (Continuation of Listing 5.3). 1 { 2 ... 3 "probes":[ 4 { 5 "name":"deployment-is-fully-available", 6 "type":"probe", 7 "tolerance":true, 8 "provider":{ 9 "type":"python", 10 "module":"chaosk8s.probes", 11 "func":"deployment_fully_available", 12 "arguments":{ 13 "name":"payment" 14 } 15 } 16 }, 17 { 18 "name":"check-transient-behaviors", 19 "type":"probe", 20 "tolerance":{ 21 "type":"jsonpath", 22 "path":"$.result", 23 "target":"body", 24 "expect":[ 25 "True" 26 ] 27 }, 28 "provider":{ 29 "type":"http", 30 "url":"http://localhost:5000/monitor", 31 "method":"POST", 32 "headers":{ 33 "Content-Type":"application/json" 34 }, 35 "arguments":{ 36 "behavior_description":"description", 37 "specification":"After {resp_time_stable(resp_time)}, it is never the case ↩→that {resp_time_unstable(resp_time)} [holds].", 38 "specification_type":"psp", 39 "predicates_info":["..."], 40 "measurement_source":"influx", 41 "measurement_points":["..."] 42 } 43 } 44 } 45 ] 46 ... 47 } 38 5.3 Transient Behavior Verification Tool REST API To enable communication via the HTTP protocol, a REST API is implemented with the help of the web framework Flask [Fla]. There are several endpoints provided by the API, however, the most important one is the “/monitor“ endpoint. This endpoint is listening for HTTP requests containing transient behavior specifications as JSON objects and when such a request is received, the verification of the received specification is initiated. The REST API is also used to serve the web pages comprising the graphical user interface. 5.3 Transient Behavior Verification Tool The verification tool was implemented in Python since the approach for creating runtime verification monitors for MTL formulas by Ulus [Ulu19b] that was used as part of the prototype is implemented as a Python library [Ulua] and also since Chaos Toolkit is implemented in Python. An overview of the underlying components of the verification tool is presented in Figure 5.2. Transient Behavior Verification Tool REST API Evaluator Parser Plotter GUI Data Retriever Figure 5.2: Component Diagram of the Verification Tool. Transient Behavior Verification Tool Data Retriever Evaluator Plotter Figure 5.2: Component Diagram of the Verification Tool. Parsing Component The verification tool also contains a component that is responsible for the mapping of PSP specifications into MTL formulas and the preparation steps for the verification of the formula, which also includes the processing of the information regarding the logical predicates in the specification. 39 5 Prototype Implementation The Boolean functions corresponding to the “predicate_logic“ field from the behavior specification are implemented as the functions of a Python class located namely in this component. In Listing 5.5, a selection of some of the Boolean functions and the constructor of the class are presented. The value of the class attribute called “value“ which is assigned in the constructor is retrieved from the “predicate_comparison_value“ field in the specification. As shown in the listing, the implementation of the predicate functions is kept as simple as possible in order to allow easier extension and addition of further functions when required. Listing 5.5 Predicate Functions. 1 class Predicates: 2 def __init__(self, value=None): 3 self.value = value 4 self.trendLast = None 5 def equal(self, variable): 6 return variable == self.value 7 def notEqual(self, variable): 8 return variable != self.value 9 def bigger(self, variable): 10 return variable > self.value 11 def smaller(self, variable): 12 return variable < self.value 13 ... Listing 5.5 Predicate Functions. 1 class Predicates: 2 def __init__(self, value=None): 3 self.value = value 4 self.trendLast = None 5 def equal(self, variable): 6 return variable == self.value 7 def notEqual(self, variable): 8 return variable != self.value 9 def bigger(self, variable): 10 return variable > self.value 11 def smaller(self, variable): 12 return variable < self.value 13 ... Listing 5.5 Predicate Functions. 1 class Predicates: 2 def __init__(self, value=None): 3 self.value = value 4 self.trendLast = None 5 def equal(self, variable): 6 return variable == self.value 7 def notEqual(self, variable): 8 return variable != self.value 9 def bigger(self, variable): 10 return variable > self.value 11 def smaller(self, variable): 12 return variable < self.value 13 ... Listing 5.5 Predicate Functions. Data Retrieval Component Another component also a part of the verification tool is the data retriever component, which consists of one abstract class and three subclasses extending it, responsible for the retrieval of the monitoring data. At the time of writing, there are three data retriever subclasses, namely for InfluxDB, Prometheus, and CSV files. These three classes also have the same output format, namely a multidimensional array, where each column corresponds to a measurement point from the behavior specification. As a result, also the addition of further data retriever classes for different monitoring tools is kept as simple as possible. Evaluation Component The verification tool includes an additional evaluation component, responsible for the verification of the prepared MTL formula against the monitoring data. This is achieved by utilizing a Python library [Ulua] that creates the runtime verification monitors according to the approach of Ulus [Ulu19b]. Even though at the time of writing the used library is deprecated in favor of a newer implementation [Ulub] in C++, the old implementation is preferred due to several factors. First and most important, the input format of the MTL formulas in the Python implementation is more suitable for the needs of this thesis. This concerns in particular the definition of the predicates in the formulas. The Python library allows the binding of the predicates in the formula to Boolean functions as already described, while the newer implementation does not support this as it supports only a limited number of predefined functions. As a result, the choice of the Python 40 5.3 Transient Behavior Verification Tool implementation allows the definition of more complex logical predicates and also their extension and addition of further ones in the future. Another factor is that the newer implementation is realized in C++ and even though the monitors are accessible from Python via bindings, the original Python implementation is preferred as it is written in Python and ensures consistency with the remainder of the verification tool which is also implemented in Python. Furthermore, the claimed performance improvements in the newer implementation are not critical for the prototype since as elicited in the interviews, the response time of the verification is not of critical importance. Considering all of the above, the original Python implementation is deemed sufficient for the purpose of this thesis. The creation of the monitors and the verification of the behaviors is initiated by simply calling the methods provided by the imported library. During the verification, the monitor returns a Boolean value for each time point from the event trace. These results are collected in order to calculate the intervals in which the behavior specification was either satisfied or violated. The result regarding the last time point is accepted as the final result of the verification and is later returned as the final Boolean value. Additionally, the information regarding the intervals is passed to the plotting component in order to create the visualization of the verification results. Furthermore, the approach of Ulus [Ulu19b] supports only past-MTL formulas. Evaluation Component wrong as the results would not correspond to it since they are originating from the past verification. Nevertheless, when the evaluation component verifies a future-MTL formula, an additional output field is included which indicates that the results and the event traces have been reversed. ... ... ... ... □(𝐴→□[0,5]𝐵) A ... ↓ ... B ... B ... B ... B ↑ ... B ■(𝐴→■[0,5]𝐵) ... ... ... ... Table 5.3: Future and Past Example Trace. Table 5.3: Future and Past Example Trace. Plotting Component A further component of the verification tools is the plotting component. The sole purpose of this component is to create plots containing the monitoring data, the verification results, and the interval information. These plots are created with the help of the matplotlib Python library [Mat]. An example of a plot created by the prototype is presented in Figure 5.3. The plots created by this component depict the measurement data used in the verification as lines. In the example plot in Figure 5.3, these are the blue and black lines. Additionally, the background color of the plot indicates what the verification result is in the given time intervals, for example from Figure 5.3 it can be concluded that the behavior specification was satisfied from the start of the experiment at time 0 until around the 300th time unit, after which the specification was violated. Evaluation Component In order to also allow support for future-MTL formulas, some adaptations have to be introduced. To make a future-MTL formula verifiable by the past-MTL monitors, the future temporal operators in the formula have to be replaced with their past variants and the event trace has to be processed in reverse, i.e. starting from its end. To illustrate this, consider the following two PSPs and their mappings in respectively past and future-MTL. PSP After {𝐴}, it is always the case that {𝐵} [holds] for 5 time units. Past MTL Formula ■(♦[0,5] 𝐴→(𝐵S 𝐴)) Future MTL Formula □(𝐴→□[0,5]𝐵) PSP Before {𝐴}, it is always the case that {𝐵} [holds] for 5 time units. Past MTL Formula ■(𝐴→■[0,5]𝐵) Future MTL Formula □(♢[0,5] 𝐴→(𝐵U 𝐴)) Table 5.2: MTL Formula Conversion Example [AGL+15; Ulu19a] Table 5.2: MTL Formula Conversion Example [AGL+15; Ulu19a] As can be seen in Table 5.2, the future-MTL formula of the first PSP is almost the same as the past-MTL formula of the second PSP, the sole difference being the usage of different temporal operators, i.e. future 𝑎𝑙𝑤𝑎𝑦𝑠and past 𝑎𝑙𝑤𝑎𝑦𝑠. Additionally, when both PSPs are compared, it could be argued that they describe the same events, however, observed from different points. This is also depicted in Table 5.3. As implied in the table, by changing the future operators to their past equivalents and by reading the event traces in reverse, one could evaluate the future-MTL formulas by using only the past monitors. This is namely how the prototype handles future-MTL formulas. After a future-MTL formula is converted and its past variant is verified, its results are returned in the reversed manner in which the event traces were read, i.e. the first time point in the results is actually the last time point in the measurement data and the last time point in the results is the first time point in the measurement data. The reason for this is that the results correspond to the past verification of the formula and reversing the results and the event traces to their original future form would be 41 41 5 Prototype Implementation wrong as the results would not correspond to it since they are originating from the past verification. Nevertheless, when the evaluation component verifies a future-MTL formula, an additional output field is included which indicates that the results and the event traces have been reversed. Graphical User Interface Taking into account the requirement given by some of the interviewed people that the tool should be also usable as a stand-alone tool, meaning without the need for Chaos Experiments and Chaos Toolkit, the prototype also provides a GUI. Since the verification tool already implements a REST API, it is used to also serve the user interface to the users. This interface can be used to easily create transient behavior specifications either by using MTL or PSPs, insert these into existing Chaos Experiments, and verify such specifications straight from the user interface without the need and the execution of Chaos Experiments. The GUI is created using HTML, CSS, JavaScript, the Bootstrap framework [Boo], and Jinja [Jin], which is Flask’s template engine. The user interface is rather simple and minimalistic, and as a result, consists of only 7 web pages. Figure 5.4 presents the welcoming page of the user interface. On this page, the user is presented with the three options that can be selected, namely the creation of a transient behavior specification, the insertion of such a specification into a Chaos Experiment, and the stand-alone verification of transient behavior specifications. 42 5.3 Transient Behavior Verification Tool 0 100 200 300 400 500 600 time 0 200 400 600 800 resp_time 2.0 2.2 2.4 2.6 2.8 3.0 instance_count resp_time instance_count Figure 5.3: Result Plot Example. Figure 5.3: Result Plot Example. Figure 5.5 presents the page allowing the user to create a behavior specification using an MTL formula. The page consists of numerous text fields that extract information about the formula, predicates, and measurements, which are then inserted into the specification JSON object and provided to the user as a downloadable file. Figure 5.6 presents the page that can be used to create transient behavior specifications with a PSP. This page contains almost the same fields as those for the MTL behavior specifications, the only difference being the fields related to the formalism used in the specification. Both the MTL and the PSP page allow the user to select the logic of the predicates used in the specification and the measurement source by using drop-down menus with pre-configured information. The page for the creation of the PSP behavior specifications also allows a selection of the supported PSPs by using a drop-down menu. Figure 5.5 presents the page allowing the user to create a behavior specification using an MTL formula. Graphical User Interface The page consists of numerous text fields that extract information about the formula, predicates, and measurements, which are then inserted into the specification JSON object and provided to the user as a downloadable file. Figure 5.6 presents the page that can be used to create transient behavior specifications with a PSP. This page contains almost the same fields as those for the MTL behavior specifications, the only difference being the fields related to the formalism used in the specification. Both the MTL and the PSP page allow the user to select the logic of the predicates used in the specification and the measurement source by using drop-down menus with pre-configured information. The page for the creation of the PSP behavior specifications also allows a selection of the supported PSPs by using a drop-down menu. The page that allows the users to place a behavior specification already defined in the respective JSON format into a Chaos Experiment is presented in Figure 5.7. As shown in the figure, the page contains a simple text box where the JSON specification is to be pasted and a file input field for the Chaos Experiment. Similarly, the page that allows users to verify behavior specifications directly from the user interface also contains a simple text box for the JSON specification and an optional file input field for the cases in which the verification uses a CSV file. This page is presented in Figure 5.8. The page containing the results of the verification is shown in Figure 5.9. As shown in the figure, the page contains the Boolean result regarding the verification of the specification, the information regarding the intervals, and the resulting plot. 43 43 5 Prototype Implementation Figure 5.4: Graphical User Interface - Home Page. Figure 5.4: Graphical User Interface - Home Page. Figure 5.5: Graphical User Interface - Behavior Specification Using MTL. Figure 5.5: Graphical User Interface - Behavior Specification Using MTL. 44 44 5.3 Transient Behavior Verification Tool Figure 5.6: Graphical User Interface - Behavior Specification Using PSPs. Figure 5.6: Graphical User Interface - Behavior Specification Using PSPs. Figure 5.7: Graphical User Interface - Experiment Creation Page. Figure 5.7: Graphical User Interface - Experiment Creation Page. 45 45 5 Prototype Implementation Figure 5.8: Graphical User Interface - Stand-alone Verification Page. Figure 5.8: Graphical User Interface - Stand-alone Verification Page. Figure 5.9: Graphical User Interface - Verification Results Page. 6 Evaluation In this chapter, the proposed extended Chaos Engineering approach is evaluated through the evaluation of its implementation. There are three different types of evaluations conducted in total, namely an evaluation regarding the correctness of the verification of transient behavior specifications, an evaluation of the applicability with Chaos Experiments and their execution, and an evaluation of the integration with other approaches, based on the compatibility with the simulation approach MiSim [Wag21]. The different types of evaluation also utilize the Goal Question Metric (GQM) approach [BCR]. The GQM approach is particularly useful to operationalize project goals into measurable metrics [BRZ]. The GQM models defined by the GQM approach contain namely conceptual goals, questions derived from these goals the answers to which should indicate whether the goals are met, and metrics suited to measure the answers to the questions [BCR]. Figure 6.1 depicts the structural hierarchy of the GQM models used in the following evaluations. For the purpose of the evaluation of the approach proposed by this thesis, three GQM models were defined, one for each type of evaluation. In the following sections, each evaluation is thoroughly described. Verification of different specification types Prototype able to verify different types of specifications? Correctly verified specification types Specification & verification of trans. behavior requirements in Chaos Experiments Can trans. behavior hypotheses be defined in Chaos Experiments? Can trans. behavior hypotheses be verified in Chaos Experiments? Execute and verify multiple Chaos Experiments on the same system? Transient behavior hypotheses specified in Chaos Experiments Correctly executed & verified Chaos Experiments with a trans. behavior hypothesis Chaos Experiments containing a transient behavior hypothesis executed on the same test system Correctly verified transient behavior requirements using simulation data Use prototype with simulation data from a microservice simulation approach? Compatibility of the prototype with other methods and approaches Goals Questions Metrics Figure 6.1: GQM Models Overview. Specification & verification of trans. behavior requirements in Chaos Experiments Specification & verification of trans. behavior requirements in Chaos Experiments Can trans. behavior hypotheses be verified in Chaos Experiments? Can trans. behavior hypotheses be defined in Chaos Experiments? Correctly verified transient behavior requirements using simulation data Correctly verified specification types Metrics Figure 6.1: GQM Models Overview. Graphical User Interface Figure 5.9: Graphical User Interface - Verification Results Page. 46 6.1 Correctness Evaluation The goal of this first evaluation is to examine whether the approach and respectively the prototype implementing it accomplish their goal to enable a verification of transient behavior requirements. From this goal, a question and a respective metric were also derived. The complete GQM model for this evaluation is presented Table 6.1. 47 47 6 Evaluation Goal Enable verification of different types of transient behavior re- quirements specified using Metric Temporal Logic and Property Specification Patterns against monitoring data of software systems. Question Is the prototype able to verify different types of transient behavior specifications? Metric Number of correctly verified transient behavior specification types. Table 6 1: GQM Model for the Correctness Evaluation Table 6.1: GQM Model for the Correctness Evaluation. 6.1.1 Method To be able to answer the question defined by the GQM model and to successfully carry out the evaluation, several transient behavior specifications and respective event traces are required. To generate these, a benchmarking tool for MTL monitoring tools called Timescales [Ulu19a] is used. Even though Timescales is primarily focusing on the performance evaluation of monitoring tools, it can also be used to evaluate the correctness of the verification as it can generate the MTL formulas for 4 different PSPs with 4 different timing scopes and it is also able to generate event traces for each of the formulas as CSV files. The supported PSPs and scope combinations are the following: 1. Bounded Absence After Q 2. Bounded Absence Before R 3. Bounded Absence Between Q and R 4. Bounded Universality After Q 5. Bounded Universality Before R 6. Bounded Universality Between Q and R 7. Bounded Recurrence Globally 8. Bounded Recurrence Between Q and R 9. Bounded Response Globally 10. Bounded Response Between Q and R 6.1.2 Results and Discussion The results of the evaluation with each test suite are summarized in Table 6.2. As shown in the table, all of the 80 MTL formulas that are combinations of different PSPs and scopes contained in the test suites are correctly verified against 40 event traces that satisfy the respective behavior definitions, and another 40 event traces that violate them. As a result, the question in the GQM model from Table 6.1 is affirmatively answered. Thus, it is concluded that the goal to enable a verification of different types of transient behavior requirements is successfully achieved. Test Suite Number of initiated formula verifications Successfully verified formulas Past Small Suite - True 10 10 Past Small Suite - False 10 10 Future Small Suite - True 10 10 Future Small Suite - False 10 10 Past Large Suite - True 30 30 Past Large Suite - False 30 30 Future Large Suite - True 30 30 Future Large Suite - False 30 30 Table 6.2: Correctness Evaluation - Test Suite Results. Test Suite Number of initiated formula verifications Successfully verified formulas Past Small Suite - True 10 10 Past Small Suite - False 10 10 Future Small Suite - True 10 10 Future Small Suite - False 10 10 Past Large Suite - True 30 30 Past Large Suite - False 30 30 Future Large Suite - True 30 30 Future Large Suite - False 30 30 Table 6.2: Correctness Evaluation - Test Suite Results. Table 6.2: Correctness Evaluation - Test Suite Results. 10. Bounded Response Between Q and R Furthermore, the tool can generate two test suites, namely a small and a large suite. The small suite contains 10 MTL formulas, one for each supported PSP and scope combination, with short time bounds, in particular 10, and a trace length of ten thousand time units. The large suite contains 30 MTL formulas, 3 for each supported PSP and scope combination, with increasingly larger time bounds, from 10 up to 1000, and a trace length of a million time units. Furthermore, both suites can be configured to contain either past or future-MTL formulas. For the evaluation of the approach, both the small and large suites are generated two times, first containing past-MTL formulas, and then containing future-MTL formulas, together with the event traces for each behavior definition, i.e. past and future formula pair. By default Timescales generates traces in which the behaviors are satisfying the behavior definition, however, the tool allows to also generate traces that violate the behavior definition. As a result, both types of traces are generated for each suite. In total, for this evaluation, there are 80 MTL formulas, one half of them being in past-MTL and the other 48 6.2 Chaos Experiment Evaluation in future-MTL, with different timing constraints based on the 10 above-listed PSPs and scope combinations and 80 event traces in total, where for each formula pair there is a trace satisfying the behavior specification and another one that violates the specification. The MTL formulas and the respective event traces can be found in the supplementary material of this thesis [Zah22]. The evaluation is executed with the help of Postman [Pos], in which multiple collections containing the requests required for the evaluation are created. The collections containing the formulas and the traces that are satisfying the formulas are labeled as “Past Small Suite - True“, “Future Small Suite - True“, “Past Large Suite - True“ and “Future Large Suite - True“, whereas the collections in which the traces are violating the formulas are labeled as ‘Past Small Suite - False“, “Future Small Suite - False“, “Past Large Suite - False“ and “Future Large Suite - False“. Afterward, the evaluation is executed by executing each of the defined collections. 6.2 Chaos Experiment Evaluation Another goal of the approach that is evaluated is the ability to specify and verify transient behavior requirements using MTL and PSPs as part of Chaos Experiments. Therefore, the questions of whether the transient behavior requirements can be defined in Chaos Experiments, verified as part of these, and whether multiple Chaos Experiments containing different transient behavior hypotheses can be verified using the same test system are examined. The full GQM model defined for this evaluation is presented in Table 6.3. 49 49 6 Evaluation Goal Enable the specification and verification of transient behavior requirements using MTL and PSP as part of Chaos Experiments. Question 1 Can transient behavior hypotheses be defined as part of Chaos Experiments? Question 2 Can transient behavior hypotheses be verified as part of the execu- tion of Chaos Experiments? Question 3 Can multiple Chaos Experiments containing different types of transient behavior hypotheses be executed and correctly verified on the same continuously running test system instance? Metric 1 Number of transient behavior hypotheses specified as part of Chaos Experiments. Metric 2 Number of correctly executed and verified Chaos Experiments containing a transient behavior hypothesis. Metric 3 Number of Chaos Experiments containing a transient behavior hypothesis executed on the same test system instance. Table 6.3: GQM Model for the Chaos Experiment Evaluation. Table 6.3: GQM Model for the Chaos Experiment Evaluation. 6.2.2 Method In the following sections, three different Chaos Experiments are executed in order to provide answers to the questions defined in the GQM model regarding this evaluation. Two of the experiments include transient behavior hypotheses specified using MTL formulas and one using a PSP. Additionally, two of the experiments use the InfluxDB instance as the source of the measurement data, while the third one uses Prometheus as the source. A brief overview of the three Chaos Experiments and their characteristics is presented in Table 6.4. 6.2.1 Test System Table 6.4: Executed Chaos Experiments - Overview. API-Gateway Service2 Service1 Eureka Figure 6.2: Test System - Architecture. Eureka Figure 6.2: Test System - Architecture. 6.2.1 Test System In order to be able to execute Chaos Experiments, first, a test system is required. The test system used during this evaluation is a mocked version of a payment calculation system [FHW+21] developed by an industrial partner of the university, who however should remain confidential, as a result of which, also some details about the system have to be kept hidden and obscured. This system is selected for the conduction of this evaluation as it has already been used in Chaos Engineering research [FHW+21], in which Chaos Experiments were successfully executed onto the system. An additional argument for the selection of this system is that by using an experimental setup similar to the one in the already-mentioned research [FHW+21], it is possible to measure certain metrics, such as response time, for the whole duration of the experiment, which would imply that the setup is able to measure and capture the exhibited transient behaviors. The mock of the system uses the microservice architecture style and consists of four services. The purpose of the first service called API-Gateway is to distribute the incoming load to two other services, respectively called Service1 and Service2. Additionally, Service1 is also able to forward calls to Service2. There is also an Eureka service used for service discovery. A component diagram of the mentioned components is presented in Figure 6.2. Furthermore, the test system includes a load generator capable of generating a specific number of requests for the endpoints exposed by the system. The information regarding the generated load is additionally saved into an InfluxDB database instance. In order to enable the execution of Chaos Experiments on the test system, the API-Gateway, Eureka, Service1, and Service2 components were deployed on a local microk8s [Mic] Kubernetes installation. The Kubernetes deployment was configured in such a way that there were two instances of Service1 and two instances of Service2 hosted on different pods. In the following sections, three Chaos Experiments are described and executed on the presented system, after which their results are discussed. 50 6.2 Chaos Experiment Evaluation Chaos Experiment 1 Chaos Experiment 2 Chaos Experiment 3 Steady-state hypothesis Deployment is fully available Deployment is fully available Deployment is fully available Anomaly One Service1 instance is killed Sudden workload increase Sudden workload increase Transient behavior hypothesis type MTL formula PSP MTL formula Monitoring data source InfluxDB InfluxDB Prometheus Table 6.4: Executed Chaos Experiments - Overview. ■( ( 𝑖𝑛𝑠𝑡𝑎𝑛𝑐𝑒_𝑑𝑒𝑎𝑑(𝑖𝑛𝑠𝑡𝑎𝑛𝑐𝑒_𝑐𝑜𝑢𝑛𝑡) ∧𝑟𝑒𝑠𝑝_𝑡𝑖𝑚𝑒_𝑛𝑜𝑟𝑚𝑎𝑙(𝑟𝑒𝑠𝑝_𝑡𝑖𝑚𝑒) ) S[0,30] 𝑖𝑛𝑠𝑡𝑎𝑛𝑐𝑒𝑠_𝑛𝑜𝑟𝑚𝑎𝑙(𝑖𝑛𝑠𝑡𝑎𝑛𝑐𝑒_𝑐𝑜𝑢𝑛𝑡) ) In order to give some meaning to the MTL formula, the logical predicates and the measurements have to be defined. For the purpose of this Chaos Experiment, the 𝑖𝑛𝑠𝑡𝑎𝑛𝑐𝑒_𝑛𝑜𝑟𝑚𝑎𝑙predicate evaluates to 𝑡𝑟𝑢𝑒when the 𝑖𝑛𝑠𝑡𝑎𝑛𝑐𝑒_𝑐𝑜𝑢𝑛𝑡values are equal to 2, the 𝑖𝑛𝑠𝑡𝑎𝑛𝑐𝑒_𝑑𝑒𝑎𝑑predicate evaluates to 𝑡𝑟𝑢𝑒when the 𝑖𝑛𝑠𝑡𝑎𝑛𝑐𝑒_𝑐𝑜𝑢𝑛𝑡values are equal to 1 and 𝑟𝑒𝑠𝑝_𝑡𝑖𝑚𝑒_𝑛𝑜𝑟𝑚𝑎𝑙evaluates to 𝑡𝑟𝑢𝑒when the 𝑟𝑒𝑠𝑝_𝑡𝑖𝑚𝑒values are lower than or equal to 100. For the purpose of the Chaos Experiment, the 𝑖𝑛𝑠𝑡𝑎𝑛𝑐𝑒_𝑐𝑜𝑢𝑛𝑡measurements are corresponding to the instance count of Service1 and the 𝑟𝑒𝑠𝑝_𝑡𝑖𝑚𝑒measurements are corresponding to the response time of the requests sent to the API-Gateway by the load generator, and both of the measurements are preserved in the InfluxDB instance. The above-described Chaos Experiment expressed in Chaos Toolkit’s format can be found in Listings A.1 and A.2 as well as in the supplementary material of the thesis [Zah22]. The above-defined Chaos Experiment was executed on the test system while the load generator was generating 20 requests per second to the API-Gateway, after which they are distributed to the remaining services. The plot created by the transient behavior verification tool visualizing the results of transient behavior verification, the instance count of Service1, and the response time of the test system is presented in Figure 6.3. In the plot, the red and blue lines present the measurement data regarding the response time and the instance count, whereas the green and red background colors indicate the different verification results at the different time intervals during the experiment. For example, as can be seen in the plot, the behavior specification is evaluated as 𝑡𝑟𝑢𝑒until about the 35th second, which is indicated by the green background color, and then the evaluation turns to 𝑓𝑎𝑙𝑠𝑒, which is also indicated by the red background color. A deeper look into the measurement data which is also presented in the plot reveals that the behavior specification was indeed violated around the 35th second, in which the instance count of Service1 was 1 and the response time of the system under test was around 250 ms. As the transient behavior specification requires the response time to be at most 100 ms., the 𝑓𝑎𝑙𝑠𝑒verification result indicated by the red color starting from around the 35th second is indeed correct. Chaos Experiment With an MTL Transient Behavior Specification First, an anomaly that is injected as part of this Chaos Experiment has to be selected. One common anomaly that can occur in such microservice systems as the test system described above and used in the experiment, is the failure of single components or services. Also, it is assumed that such types of failures are inevitable in the long run [New15]. Thus, a failure of a single instance is the selected anomaly that will be injected as part of this Chaos Experiment. The failure is defined by stopping one of the Service1 pods while the system is in a normal mode of operation. The duration of the experiment is defined to be 120 seconds. The failure is introduced in the 30th second of the experiment. This is done in order to allow the system to adapt to the introduced workload and 51 6 Evaluation to get into a steady state in the first 30 seconds of the experiment. The steady-state hypothesis check is scheduled for the 120th second of the experiment as the system should recover until then and is checking whether the deployment is fully available. The hypothesis regarding the transient behaviors occurring in this experiment is informally defined as follows: Always when the instance count of a certain service drops from 2 to 1, a new instance has to be spawned in a maximum of 30 seconds and during this time the response time should be below 100 ms. The formal definition of the hypothesis specified using an MTL formula is as follows: ■( ( 𝑖𝑛𝑠𝑡𝑎𝑛𝑐𝑒_𝑑𝑒𝑎𝑑(𝑖𝑛𝑠𝑡𝑎𝑛𝑐𝑒_𝑐𝑜𝑢𝑛𝑡) ∧𝑟𝑒𝑠𝑝_𝑡𝑖𝑚𝑒_𝑛𝑜𝑟𝑚𝑎𝑙(𝑟𝑒𝑠𝑝_𝑡𝑖𝑚𝑒) ) S[0,30] 𝑖𝑛𝑠𝑡𝑎𝑛𝑐𝑒𝑠_𝑛𝑜𝑟𝑚𝑎𝑙(𝑖𝑛𝑠𝑡𝑎𝑛𝑐𝑒_𝑐𝑜𝑢𝑛𝑡) ) ■( ( 𝑖𝑛𝑠𝑡𝑎𝑛𝑐𝑒_𝑑𝑒𝑎𝑑(𝑖𝑛𝑠𝑡𝑎𝑛𝑐𝑒_𝑐𝑜𝑢𝑛𝑡) ∧𝑟𝑒𝑠𝑝_𝑡𝑖𝑚𝑒_𝑛𝑜𝑟𝑚𝑎𝑙(𝑟𝑒𝑠𝑝_𝑡𝑖𝑚𝑒) ) S[0,30] 𝑖𝑛𝑠𝑡𝑎𝑛𝑐𝑒𝑠_𝑛𝑜𝑟𝑚𝑎𝑙(𝑖𝑛𝑠𝑡𝑎𝑛𝑐𝑒_𝑐𝑜𝑢𝑛𝑡) ) In conclusion, it can be stated that the above described transient behavior specification including an MTL formula can be successfully incorporated into a Chaos Experiment and is correctly verified during its execution, meaning that the violation of the specification is correctly identified. 52 6.2 Chaos Experiment Evaluation 0 20 40 60 80 100 120 time 0 50 100 150 200 250 300 350 resp_time 1.0 1.2 1.4 1.6 1.8 2.0 instances resp_time instances Figure 6.3: Chaos Experiment 1 - Verification Results. Figure 6.3: Chaos Experiment 1 - Verification Results. Chaos Experiment With a PSP Transient Behavior Specification The Chaos Experiment defined in Chaos Toolkit’s format can be found in Listings A.3 and A.4, and also in the supplementary material of the thesis [Zah22]. When the experiment was executed, the transient behavior verification tool generated the plot presented in Figure 6.4. In the plot, the red line is depicting the response time of the system and the blue line depicts the instance count of Service1. Also in this plot, the green and red background colors indicate the verification result with regard to the different time frames during the experiment. 0 100 200 300 400 500 600 time 0 200 400 600 800 resp_time 2.0 2.2 2.4 2.6 2.8 3.0 instance_count resp_time instance_count Figure 6.4: Chaos Experiment 2 - Verification Results. Figure 6.4: Chaos Experiment 2 - Verification Results. As depicted in the plot above, the response time of the system exceeds 150 ms around the 110th second and the new instance is spawned around the 260th second, which is exceeding the constraint in the behavior specification by around 90 seconds. Therefore, the verification results indicating that the transient behavior requirement was violated are correct and the verification tool is capable of successfully verifying such behavior requirements specified as PSPs. Chaos Experiment With a PSP Transient Behavior Specification The anomaly that is injected in the second Chaos Experiment is a significant increase in the incoming workload to the API-Gateway. This is selected in order to examine the autoscaling capabilities of the test system and since a sudden increase in the workload has also been identified as a possible cause of transient behaviors [Bec20]. The duration of this experiment is 600 seconds which is 10 minutes. At the start of the experiment, the incoming load to the API-Gateway component is one request per second. Until the end of the experiment, the workload reaches 1000 requests per second. The steady-state hypothesis of this second experiment is checked at the 600th second and is the same as in the first one, namely that the deployment is fully available. The transient behavior hypothesis in this experiment can be informally defined as follows: When the response time of the system exceeds 150 ms. then a new instance of Service1 must be spawned in the next 30 to 60 seconds. The formal definition of the transient behavior hypothesis using a PSP is the following: Globally, if {𝑟𝑒𝑠𝑝_𝑡𝑖𝑚𝑒_ℎ𝑖𝑔ℎ(𝑟𝑒𝑠𝑝_𝑡𝑖𝑚𝑒)} [has occurred] then in response {𝑖𝑛𝑠𝑡𝑎𝑛𝑐𝑒_𝑖𝑛𝑐𝑟𝑒𝑎𝑠𝑒(𝑖𝑛𝑠𝑡𝑎𝑛𝑐𝑒_𝑐𝑜𝑢𝑛𝑡)} [eventually holds] between 30 and 60 time units. Globally, if {𝑟𝑒𝑠𝑝_𝑡𝑖𝑚𝑒_ℎ𝑖𝑔ℎ(𝑟𝑒𝑠𝑝_𝑡𝑖𝑚𝑒)} [has occurred] then in response {𝑖𝑛𝑠𝑡𝑎𝑛𝑐𝑒_𝑖𝑛𝑐𝑟𝑒𝑎𝑠𝑒(𝑖𝑛𝑠𝑡𝑎𝑛𝑐𝑒_𝑐𝑜𝑢𝑛𝑡)} [eventually holds] between 30 and 60 time units. Globally, if {𝑟𝑒𝑠𝑝_𝑡𝑖𝑚𝑒_ℎ𝑖𝑔ℎ(𝑟𝑒𝑠𝑝_𝑡𝑖𝑚𝑒)} [has occurred] then in response {𝑖𝑛𝑠𝑡𝑎𝑛𝑐𝑒_𝑖𝑛𝑐𝑟𝑒𝑎𝑠𝑒(𝑖𝑛𝑠𝑡𝑎𝑛𝑐𝑒_𝑐𝑜𝑢𝑛𝑡)} [eventually holds] between 30 and 60 time units. Additionally, the logical variables in the above-defined PSP have to be specified in order to give the PSP definition its actual meaning. The 𝑟𝑒𝑠𝑝_𝑡𝑖𝑚𝑒_ℎ𝑖𝑔ℎpredicate is 𝑡𝑟𝑢𝑒when the 𝑟𝑒𝑠𝑝_𝑡𝑖𝑚𝑒values are bigger than or equal to 150, which in this case corresponds to milliseconds. The 𝑖𝑛𝑠𝑡𝑎𝑛𝑐𝑒_𝑖𝑛𝑐𝑟𝑒𝑎𝑠𝑒predicate evaluates to 𝑡𝑟𝑢𝑒when the 𝑖𝑛𝑠𝑡𝑎𝑛𝑐𝑒_𝑐𝑜𝑢𝑛𝑡values concerning Service1 are bigger than 2. Similarly to the previous experiment, the measurement values are preserved in the InfluxDB instance. As next, the individual parts of the Chaos Experiment have to 53 6 Evaluation 6 Evaluation be put together and defined in Chaos Toolkit’s syntax. Since the increase of the load is configured in and initiated from the load generator component, the method section in the experiment is not required since Chaos Toolkit does not have to interact with the load generator. However, since the method field is required by Chaos Toolkit, a simple probe is specified in the method so that the lack of a method does not have any effect on the execution of the experiment. Chaos Experiment Evaluating the Integration of Prometheus as a Monitoring Tool The purpose of this Chaos Experiment is to evaluate the integration of Prometheus as a source of monitoring data since one of the interviewees requested its integration and the fact that Prometheus is an industry-standard with regard to monitoring approaches and tools. The anomaly that is introduced in this last Chaos Experiment is the same as in the previous one, namely an increase in the incoming workload. The steady-state hypothesis of the experiment is also the same, namely that the deployment is fully available. The only part differing in this experiment is the transient behavior 54 6.2 Chaos Experiment Evaluation requirement which requires the CPU usage of both Service1 pods should be below 0.8 cores. The same requirement is formally specified as the following MTL formula: ■( 𝑙𝑜𝑎𝑑_𝑛𝑜𝑟𝑚𝑎𝑙(𝑠𝑒𝑟𝑣𝑖𝑐𝑒1_1_𝑙𝑜𝑎𝑑) ∧𝑙𝑜𝑎𝑑_𝑛𝑜𝑟𝑚𝑎𝑙(𝑠𝑒𝑟𝑣𝑖𝑐𝑒1_2_𝑙𝑜𝑎𝑑) ) To complete the definition of the behavior, the logical predicates in the formula are defined as follows. The 𝑙𝑜𝑎𝑑_𝑛𝑜𝑟𝑚𝑎𝑙predicate is 𝑡𝑟𝑢𝑒when the values it receives are below 0.8, and the 𝑠𝑒𝑟𝑣𝑖𝑐𝑒1_1_𝑙𝑜𝑎𝑑and 𝑠𝑒𝑟𝑣𝑖𝑐𝑒1_2_𝑙𝑜𝑎𝑑measurements are regarding the number of cores used by each Service1 pod. The Chaos Experiment defined in Chaos Toolkit’s format can be found in Listings A.5 and A.6 and also in the supplementary material of the thesis [Zah22]. The plots generated from the transient behavior verification tool after executing the experiment are presented in Figure 6.5. The red and blue lines are depicting the utilization of the individual pods and similarly to the previous plots, the green and red background colors are indicating the verification results at the different time points. 0 100 200 300 400 500 600 time 0.0 0.2 0.4 0.6 0.8 1.0 service1_1_load 0.0 0.2 0.4 0.6 0.8 1.0 service1_2_load service1_1_load service1_2_load Figure 6.5: Chaos Experiment 3 - Verification Results. Figure 6.5: Chaos Experiment 3 - Verification Results. As presented in the plot above, the utilization of both pods exceeds 0.8 cores at around the 110th second. As a result, it is concluded that the verification tool correctly verified the violation of the behavior specification, while also using Prometheus as the source of the monitoring data. 6.2.3 Results and Discussion In the sections above, three Chaos Experiments were successfully specified and executed using the extended Chaos Engineering approach. Since each of these included a transient behavior hypothesis specified using either an MTL formula or a PSP, the first question of the GQM model in Table 6.3 55 6 Evaluation is positively answered. Furthermore, all of the transient behavior hypotheses were successfully and correctly verified as part of the execution of the experiments. Therefore, also the second question from the GQM model is answered positively. Lastly, all of the specified Chaos Experiments which contained different transient behavior hypotheses using different types of formalisms, i.e. MTL formulas and PSPs, were executed on the same test system and were correctly verified. Thus, also the third question in the GQM model is positively answered. As a result, the goal to enable the specification and verification of transient behavior requirements as part of Chaos Experiments has been achieved. 6.3 Evaluation of the Compatibility With Other Approaches As elicited in the expert interviews and included in the requirements for the extended Chaos Engineering approach, it must also be possible to use the approach as a stand-alone analysis without the need to run Chaos Experiments. Thus, a further goal of the approach proposed in this thesis is to be compatible with other approaches, by offering the verification of transient behavior requirements to these. One approach that can examine the resilience of software applications similarly to Chaos Engineering, however, by simulating experiments instead of actually running such, is MiSim [Wag21]. One important property of the usage of simulations is that they are particularly useful during design time, where the resilience of a system can be simulated and analyzed before the system is built. In comparison, the Chaos Engineering approach requires the existence of the analyzed systems. Introducing the ability to verify transient behavior requirements to the MiSim approach would further improve the resilience analysis that the tool is able to provide. Therefore, the ability to provide the verification of transient behavior requirements to further approaches such as MiSim is defined as an additional goal of the proposed approach and this goal is also used to derive additional questions and metrics which are summarized in the GQM model presented in Table 6.5. 6.3.1 Method In order to evaluate the compatibility with MiSim, first, the architecture on which Chaos Experiments will be simulated must be defined. The architecture that is defined as part of this evaluation is graphically presented in Figure 6.6. The architecture contains three components, namely an API-Gateway, Service1, and Service2. The API-Gateway is the entry point of the application, which routes the incoming requests to Service1, of which there are two instances that route their incoming requests to Service2. The simulated architecture is intentionally similar to the architecture of the test system used in Section 6.2 albeit with minor differences. The main differences are in that there is only a single instance of Service2 and that the API-Gateway cannot directly send requests to Service2. In addition, the MiSim architecture model does not include some of the properties of the test system from Section 6.2, such as its resilience mechanisms, but it does include an autoscaler for Service1 to simulate Kubernetes’ spawning of pods. The usage of an architecture model similar to the existing architecture of the test systems allows a comparison between the simulated data and the real measurements and a comparison of the results. The architecture model defined using MiSim’s JSON format can be found in Listings A.7 and A.8 as well as in the supplementary materials [Zah22]. 56 6.3 Evaluation of the Compatibility With Other Approaches After the architecture is defined, the next step is to define an experiment that is simulated. The experiment is specified to have a duration of 120 seconds and at the 30th second, one of the Service1 instances is killed. The experiment is also intentionally similar to the one executed in the Chaos Experiment evaluation in Section 6.2 in order to be able to detect any anomalies in the results. The experiments in MiSim are also defined as JSON objects and the above-described experiment is shown in Listing A.9. When the experiment gets simulated, the resulting simulation data is written into CSV tables, some of which are also provided in the supplementary materials of the thesis [Zah22]. Additionally, MiSim can create plots visualizing the data from the tables. Figure 6.7 presents the simulated response time of the example architecture whereas Figure 6.8 presents the instance counts throughout the experiment. 6.3.1 Method Goal Allow compatibility of the prototype with other methods and approaches that can capture transient behaviors in a discrete-time manner and offer the ability to specify and verify transient behavior requirements to these. Question Can the prototype be used with simulation data originating from approaches capable of simulating microservice architectures and their resilience? Metric Number of correctly verified transient behavior requirements using simulation data. Table 6.5: GQM Model for the Evaluation of the Compatibility With Other Approaches. Goal Allow compatibility of the prototype with other methods and approaches that can capture transient behaviors in a discrete-time manner and offer the ability to specify and verify transient behavior requirements to these. Question Can the prototype be used with simulation data originating from approaches capable of simulating microservice architectures and their resilience? Metric Number of correctly verified transient behavior requirements using simulation data. Table 6.5: GQM Model for the Evaluation of the Compatibility With Other Approaches. Table 6.5: GQM Model for the Evaluation of the Compatibility With Other Approaches. API-Gateway Service1 Service1 Service2 Figure 6.6: MiSim Simulated Architecture - Component Diagram. Figure 6.6: MiSim Simulated Architecture - Component Diagram. 57 57 6 Evaluation Figure 6.7: Simulated Response Time. Figure 6.8: Simulated Instance Counts. Figure 6.7: Simulated Response Time. Figure 6.7: Simulated Response Time. g p Figure 6.8: Simulated Instance Counts. Figure 6.8: Simulated Instance Counts. 58 6.3 Evaluation of the Compatibility With Other Approaches Next, the two transient behavior requirements that are verified against the simulation data are defined. The first transient behavior requirement is informally defined as follows: Whenever the number of instances of Service1 drops from 2 to 1, it should rise back to 2 within a maximum of 3 seconds. This requirement can be formally defined using an MTL formula as follows: ■( 𝑖𝑛𝑠𝑡𝑎𝑛𝑐𝑒_𝑑𝑒𝑎𝑑(𝑖𝑛𝑠𝑡𝑎𝑛𝑐𝑒_𝑐𝑜𝑢𝑛𝑡) S[0,3] 𝑖𝑛𝑠𝑡𝑎𝑛𝑐𝑒𝑠_𝑎𝑙𝑖𝑣𝑒(𝑖𝑛𝑠𝑡𝑎𝑛𝑐𝑒_𝑐𝑜𝑢𝑛𝑡) ) ■( 𝑖𝑛𝑠𝑡𝑎𝑛𝑐𝑒_𝑑𝑒𝑎𝑑(𝑖𝑛𝑠𝑡𝑎𝑛𝑐𝑒_𝑐𝑜𝑢𝑛𝑡) S[0,3] 𝑖𝑛𝑠𝑡𝑎𝑛𝑐𝑒𝑠_𝑎𝑙𝑖𝑣𝑒(𝑖𝑛𝑠𝑡𝑎𝑛𝑐𝑒_𝑐𝑜𝑢𝑛𝑡) ) The 𝑖𝑛𝑠𝑡𝑎𝑛𝑐𝑒_𝑑𝑒𝑎𝑑predicate is evaluating to 𝑡𝑟𝑢𝑒when the 𝑖𝑛𝑠𝑡𝑎𝑛𝑐𝑒_𝑐𝑜𝑢𝑛𝑡measurement is equal to 1 and the predicate 𝑖𝑛𝑠𝑡𝑎𝑛𝑐𝑒_𝑎𝑙𝑖𝑣𝑒is 𝑡𝑟𝑢𝑒when 𝑖𝑛𝑠𝑡𝑎𝑛𝑐𝑒_𝑐𝑜𝑢𝑛𝑡is equal to 2. Additionally, the 𝑖𝑛𝑠𝑡𝑎𝑛𝑐𝑒_𝑐𝑜𝑢𝑛𝑡measurement is referring to the number of instances of Service1. Since MiSim outputs the simulation results as CSV files, the stand-alone analysis functionality of the verification tool provided by the GUI is used to verify the above-defined behavior requirement. Furthermore, for the purpose of this behavior verification, the output table regarding the instance counts of the services has to be adjusted slightly. 6.3.1 Method The original table produced by MiSim contains information only about the moments when there are changes in the number of instances. For example, the table starts with timestamp 0 and instance count of 2 and then jumps straight to timestamp 30 and instance count of 1. Since the prototype requires measurements for each time unit, the data is manually adjusted to include the times in between the changes. The JSON object that contains the transient behavior specification and that is verified using the verification tool is presented in Listing A.10. The resulting plot created by the verification tool is presented in Figure 6.9. As shown in Figure 6.9, the time interval of 3 time units specified in the transient behavior requirement is violated, since the service needs 5 time units to respawn as depicted in Figure 6.8. As a result, it can be concluded that the first transient behavior requirement is successfully and correctly verified using the simulation data. 0 20 40 60 80 100 120 time 1.0 1.2 1.4 1.6 1.8 2.0 instance_count instance_count Figure 6.9: Compatibility Evaluation - Verification Result Plot 1. Figure 6.9: Compatibility Evaluation - Verification Result Plot 1. 59 6 Evaluation The second transient behavior requirement that is verified using the simulation data produced by MiSim can be informally defined as follows: Before the second instance of Service is respawned, the response time of the systems should be below 0.5. Formally defined using a PSP the requirement is the following: Before𝑖𝑛𝑠𝑡𝑎𝑛𝑐𝑒_𝑢𝑝(𝑖𝑛𝑠𝑡𝑎𝑛𝑐𝑒_𝑐𝑜𝑢𝑛𝑡), it is never the case that𝑟𝑒𝑠𝑝_𝑡𝑖𝑚𝑒_ℎ𝑖𝑔ℎ(𝑟𝑒𝑠𝑝_𝑡𝑖𝑚𝑒) [holds]. The 𝑖𝑛𝑠𝑡𝑎𝑛𝑐𝑒_𝑢𝑝predicate in the behavior specification is defined as a strictly increasing function, i.e. it is evaluating to 𝑡𝑟𝑢𝑒when the 𝑖𝑛𝑠𝑡𝑎𝑛𝑐𝑒_𝑐𝑜𝑢𝑛𝑡measurement values increase from 1 to 2. The 𝑟𝑒𝑠𝑝_𝑡𝑖𝑚𝑒_ℎ𝑖𝑔ℎpredicate is evaluating to 𝑡𝑟𝑢𝑒when the 𝑟𝑒𝑠𝑝_𝑡𝑖𝑚𝑒measurement for the response time of the system is higher than 0.5. The transient behavior requirement expressed as a JSON object is presented in Listing A.11. Similarly to the previous behavior specification, also this one has to be verified by using the GUI of the verification tool. Furthermore, the output CSV tables of MiSim had to be adjusted once again, since the information regarding the instance counts and the response times is located in two separate tables. As a result, the two tables had to be combined into a new one. After the verification, the resulting plot presented in 6.10 was created. 6.3.2 Results and Discussion In this evaluation, two transient behavior hypotheses are defined using both an MTL formula and a PSP. All of these hypotheses are then verified against simulation data produced by MiSim. Each of the hypotheses is also correctly verified against the simulation data required for its validation. As a result, the question from the GQM model in Table 6.5, whether the proposed approach can be used with simulation data, is positively answered. Thus, it can be concluded that the goal to allow compatibility with other approaches and to offer the ability to verify transient behavior hypotheses to these is achieved. 6.3.1 Method Looking at the data in Figures 6.8, 6.7, and 6.10, it is concluded that the verification component was able to successfully verify also the second transient behavior requirement against the simulation data produced by MiSim. 0 20 40 60 80 100 120 time 1.0 1.2 1.4 1.6 1.8 2.0 instance_count 0.3 0.4 0.5 0.6 0.7 0.8 resp_time instance_count resp_time Figure 6.10: Compatibility Evaluation - Verification Result Plot 2. 0 20 40 60 80 100 120 ti 1.0 1.2 1.4 1.6 1.8 2.0 instance_count 0.3 0.4 0.5 0.6 0.7 0.8 resp_time Figure 6.10: Compatibility Evaluation - Verification Result Plot 2. 60 6.4 Evaluation of Research Questions 6.4 Evaluation of Research Questions In this section, the results of the thesis are discussed with regard to the research questions formulated in Chapter 1, presented once again below. RQ1 How can a transient behavior requirement be specified in the context of Chaos Engineering and Chaos Experiments? RQ1 How can a transient behavior requirement be specified in the context of Chaos Engineering and Chaos Experiments? RQ2 How can the Chaos Experimentation process incorporate a verification of transient behavior requirement specifications? RQ2 How can the Chaos Experimentation process incorporate a verification of transient behavior requirement specifications? RQ3 What does a concept for an extended Chaos Engineering approach contain and how can it be implemented into a prototype? RQ4 Is the extended Chaos Engineering approach able to correctly verify different transient behavior requirements? RQ5 Is the extended Chaos Engineering approach capable of correctly executing Chaos Experiments and verifying transient behavior requirements as part of the experiments? RQ6 Is the extended Chaos Engineering approach compatible with other approaches and tools? Regarding the first research question, the results of the evaluation show that MTL and the PSPs are two formalisms capable of describing transient behavior requirements as part of the Chaos Engineering approach. This is further solidified by the multiple successfully executed Chaos Experiments containing namely MTL and PSP transient behavior specifications. Concerning the second research question, the concept for the extended Chaos Engineering approach proposes the integration of runtime verification as an additional step in the Chaos Experimentation process. The evaluation of the approach and its implementation shows that this is a viable extension to the experimentation process. With regard to the third research question, the thesis proposes a concept for an extended Chaos Engineering approach in Chapter 4, consisting of a transient behavior hypothesis, a transient behavior verification, a result visualization, a GUI, and a transient behavior monitoring approach. Furthermore, the concept is successfully implemented into a prototype, which is described in Chapter 5. After conducting the correctness evaluation in Section 6.1 and showing that the approach and its implementation are capable of verifying various different transient behavior specifications, the fourth research question is answered positively. Additionally, the Chaos Experiment evaluation carried out in Section 6.2, particularly the three Chaos Experiments containing transient behavior hypotheses which are correctly executed and verified by the extended 61 6 Evaluation approach, demonstrate that the approach is able to execute and verify Chaos Experiments containing such hypotheses. 6.4 Evaluation of Research Questions Thus, the fifth research question is also answered positively. Moreover, the evaluation conducted in Section 6.3 shows that the extended approach and its implementation are able to provide the verification of transient behavior requirements to other tools and approaches. Therefore, the last research question is also answered positively. 6.5 Threats to Validity One potential threat to the validity of the research and in particular to the evaluation and the results might be the selection of the experimental setup used to run Chaos Experiments during the evaluation. In particular, the test system and its behaviors have been already thoroughly studied by other research projects [FHW+21]. This previous knowledge about the test system might have made the specification of the transient behavior hypotheses and the verification of these easier than the same with a different system, which was not previously studied. Thus, the results and findings gathered with the selected test system might not be applicable or transferable to other systems. To mitigate this threat, the selection of the test system was discussed with the supervisors of the thesis. Nevertheless, the microservice architecture style and the technologies utilized by the test system and also by an increasing number of modern systems would imply that the results and the exhibited behaviors would also be observable with other systems utilizing the same properties. Additionally, the selection of the experts that were interviewed in order to elicit requirements regarding the extended Chaos Engineering approach could be a further threat to the validity of the results of the thesis. A different selection of interviewees with different experiences and backgrounds could have resulted in different answers to the questions and as a result, different requirements for the approach. To mitigate this threat, the selection pool and the inclusion of candidates in it were thoroughly discussed with the supervisors of the thesis. However, the number of interviewed people could also be a threat, since the selection of experts was rather limited, and interviewing additional people could have resulted in different answers and requirements. Another potential threat to the validity of the thesis and its results are the expert interviews and in particular, the questions that the interviewees were asked. It might have been the case that the questions were not selected correctly in order to elicit the necessary and correct requirements for the extended Chaos Engineering approach. As a result, a different method for conducting the expert interviews and a different selection of questions might have resulted in a different extension to the Chaos Engineering approach. Furthermore, the questions asked during the interviews might have been biased by the author of the thesis. To mitigate these, the interview questions were reviewed multiple times by the supervisors of the thesis. 6.6 Limitations One limitation of the implementation of the approach is that the transient behavior hypothesis has to be specified in the probes of the steady-state hypothesis. This is a limitation since the hypothesis regarding the transient behavior is conceptually different from the steady-state hypothesis. Since these two hypotheses are inherently different, the transient behavior hypothesis should not be a part of the steady-state hypothesis. This limitation is originating from the Chaos Engineering tool 62 6.6 Limitations selected as part of the prototype, namely Chaos Toolkit. In Chaos Toolkit, only the tolerances of the probes in the steady-state hypothesis are checked and all tolerances that are specified outside these do not get checked. As a result, in order for the tolerances of the transient behavior hypothesis to be checked, this hypothesis has to be included in the probes of the steady-state hypothesis. If at a later point in time a check of the tolerances specified in the method is introduced, the transient behavior hypothesis could be easily moved to a probe in the method. selected as part of the prototype, namely Chaos Toolkit. In Chaos Toolkit, only the tolerances of the probes in the steady-state hypothesis are checked and all tolerances that are specified outside these do not get checked. As a result, in order for the tolerances of the transient behavior hypothesis to be checked, this hypothesis has to be included in the probes of the steady-state hypothesis. If at a later point in time a check of the tolerances specified in the method is introduced, the transient behavior hypothesis could be easily moved to a probe in the method. Another limitation of the implementation of the approach is the visualization of the verification results of future-MTL formulas. Since the future temporal operators have to be converted into their past equivalents, the event trace has to be read in the reverse direction and the results are also gathered in a reverse manner. As a result of this, the visualization of the results and the intervals in it are also reversed in order to comply with the past variant of the formula. Since this reversal of all of the results is contradicting the original direction of the event traces, the results of the verification of future-MTL formulas might be confusing for the users of the approach and are therefore counted as a limitation of the approach. 7.1 Summary This work examines how the traditional Chaos Engineering approach and in particular the Chaos Experimentation process can be extended in order to enable the specification of requirements regarding the transient behaviors of software systems and to verify them as part of Chaos Experiments. To achieve this, first formalisms, namely temporal logic and the PSPs, capable of expressing transient behaviors and their temporal properties, are researched. In addition, the Chaos Engineering approach and its tools are also researched. After some initial ideas are gathered, they are implemented into an initial prototype with the purpose of presenting this initial prototype to people with Chaos Engineering expertise with the goal of studying the feasibility of the prototype and the concepts included in it and also gathering their requirements for an extended Chaos Engineering approach, that includes the verification of transient behavior requirements. After conducting multiple interviews and eliciting requirements for the extended approach, a concept for the approach is proposed. The concept contains various new elements, namely a transient behavior hypothesis specification, a transient behavior verification, a result visualization, a graphical user interface, and a transient behavior monitoring approach. Afterward, a prototype implementing the concept is developed with the goal of evaluating the extended Chaos Engineering approach. The approach is evaluated in three different ways, first, its correctness is evaluated using benchmark data, then a series of Chaos Experiments containing transient behavior hypotheses are executed using the approach, and afterward, the compatibility with other approaches and tools is evaluated. All of the above-listed evaluations are successfully completed, proving that the extended approach fulfills its goals and purpose. Additionally, the research questions formulated at the beginning of the thesis are addressed and answered. Lastly, the threats to the validity of the research and the limitations of the proposed approach are discussed. 7 Conclusion and Future Work This chapter concludes the thesis by providing a summary of the conducted research and discusses possible future work. 6.6 Limitations Nevertheless, the final result of the verification is unaffected by this limitation. A further limitation of the prototype is that there is no consistent way to define the beginning and the end of the event traces that should be used in the verification. For example, when InfluxDB is used as the monitoring tool, the beginning and the end of the trace have to be manually specified as part of the query string. However, when Prometheus is used as the measurement data source, the definition of the measurement points in the behavior specification allows the definition of the start and end times of the event trace that is used during the verification in separate fields. The definition of those however still has to be done manually. On the contrary, when a CSV file is used as the monitoring data source, the definition of a start and end point of the trace is not supported and the whole CSV file is used. 63 63 7.2 Future Work One aspect that can be further studied is the optimization of transient behavior specifications. The approach presented in this thesis showed that it is capable of verifying whether the specifications hold or are violated, however, an approach for optimizing the specification based on the verification results would be particularly useful. For example, assuming there is a service that actually takes 15 seconds to be revived, after a specification that requires the service to be revived in 30 seconds is 65 7 Conclusion and Future Work verified, the optimization approach would be able to adjust and update the specification to reflect the actual time the service takes to be revived, i.e. update the specification from 30 seconds to 15 seconds. verified, the optimization approach would be able to adjust and update the specification to reflect the actual time the service takes to be revived, i.e. update the specification from 30 seconds to 15 seconds. Another aspect that can be studied is the introduction of further runtime verification monitors for other temporal logics or even other types of formalisms. Additionally, the existing catalog of PSPs that is supported by the prototype can be further expanded, and if additional temporal logics are introduced into the specification, also the mappings of the PSPs into these newly introduced temporal logics can be added. In addition, further monitoring tools and approaches can also be studied. Since at the time of writing the approach only supports discrete-time event traces, it would be beneficial to examine how other types of event traces with different time interpretations such as continuous can be included in the extended Chaos Engineering approach. Additionally, the introduction of additional discrete-time monitoring tools is also a thing that can be considered. Moreover, the limitation of Chaos Toolkit which is part of the prototype, which requires the specification of the transient behavior hypothesis as part of the probes of the steady-state hypothesis can be addressed. Research can be conducted in order to find possible ways to extend the Chaos Experiments of Chaos Toolkit in order to enable the specification of the transient behavior hypothesis outside the probes of the steady-state hypothesis. This may include the addition of different types of hypotheses or the introduction of a tolerance check for probes outside the steady-state hypothesis. 7.2 Future Work An additional aspect that can be further studied is the usability and the user experience of the approach and its prototype implementation introduced in this thesis. A user study with Chaos Engineering practitioners as subjects, evaluating the usability of the prototype and how practitioners would use it to run Chaos Experiments, could also provide further insights into how the tool can be extended to better support the needs of the practitioners. 66 Bibliography [AGL+15] M. Autili, L. Grunske, M. Lumpe, P. Pelliccione, A. Tang. “Aligning Qualitative, Real-Time, and Probabilistic Property Specification Patterns Using a Structured English Grammar”. In: IEEE Transactions on Software Engineering 41.7 (2015), pp. 620–638. doi: 10.1109/TSE.2015.2398877 (cit. on pp. 1, 10–12, 27, 32, 35, 41). [AH92] R. Alur, T. A. Henzinger. “Logics and models of real time: A survey”. In: Real- Time: Theory in Practice. Ed. by J. W. de Bakker, C. Huizing, W. P. de Roever, G. Rozenberg. Berlin, Heidelberg: Springer Berlin Heidelberg, 1992, pp. 74–106. isbn: 978-3-540-47218-6 (cit. on pp. 8, 10). [BCE+03] M. Bruneau, S. E. Chang, R. T. Eguchi, G. C. Lee, T. D. O’Rourke, A. M. Reinhorn, M. Shinozuka, K. J. Tierney, W. A. Wallace, D. von Winterfeldt. “A Framework to Quantitatively Assess and Enhance the Seismic Resilience of Communities”. In: Earthquake Spectra 19 (2003), pp. 733–752 (cit. on p. 14). V. R. Basili, G. Caldiera, H. D. Rombach. The Goal Question Metric Approach (cit. on p. 47). [Bec20] S. Beck. “Evaluating human-computer interfaces for specification and comprehension of transient behavior in microservice-based software systems”. Master’s Thesis. University of Stuttgart, 2020. doi: http://dx.doi.org/10.18419/opus-11846 (cit. on pp. 2, 8, 53). [BFH+21] S. Beck, S. Frank, M. A. Hakamian, L. Merino, A. van Hoorn. “TransVis: Using Visualizations and Chatbots for Supporting Transient Behavior in Microservice Sys- tems”. In: Proceedings of the 9th IEEE Working Conference on Software Visualization (VISSOFT 2021). Accepted. 2021 (cit. on p. 14). [BK08] C. Baier, J.-P. Katoen. Principles of model checking. The MIT Press. London, England: MIT Press, Apr. 2008 (cit. on p. 10). A. Bauer, M. Leucker, C. Schallhart. Runtime Verification for LTL and TLTL. Tech. rep. 2007 (cit. on p. 13). [Boo] Bootstrap. Official Website. https://getbootstrap.com/. (Accessed on 07/28/2022) (cit. on p. 42). B. Boehm, H. D. Rombach, M. V. Zelkowitz. Foundations of Empirical Software Engineering (cit. on p. 47). [CCR13] I. M. Copi, C. Cohen, V. Rodych. Introduction to Logic. en. 14th ed. Pearson custom library. London, England: Pearson Education, June 2013 (cit. on p. 12). Chaos Mesh. https://chaos-mesh.org/. (Accessed on 07/28/2022) (cit. on p. 31). [Chab] Chaos Monkey. Chaos Monkey - Documentation. https://netflix.github.io/ chaosmonkey/. (Accessed on 07/28/2022) (cit. on p. 31). 67 Bibliography [Chac] Chaos Toolkit. Official Website. https : / / chaostoolkit . org/. (Accessed on 07/28/2022) (cit. on pp. 8, 9, 14, 31). [Chad] ChaosBlade. https://chaosblade.io/. (Accessed on 07/28/2022) (cit. on p. 31). J. Cámara, R. Matplotlib. https://matplotlib.org/. (Accessed on 07/28/2022) (cit. on p. 42). [Chac] Chaos Toolkit. Official Website. https : / / chaostoolkit . org/. (Accessed on 07/28/2022) (cit. on pp. 8, 9, 14, 31). Bibliography de Lemos. “Evaluation of resilience in self-adaptive systems using probabilistic model-checking”. In: 2012 7th International Symposium on Software Engineering for Adaptive and Self-Managing Systems (SEAMS). 2012, pp. 53–62. doi: 10.1109/SEAMS.2012.6224391 (cit. on pp. 1, 13). [CZ18] C. Czepa, U. Zdun. “On the Understandability of Temporal Properties Formalized in Linear Temporal Logic, Property Specification Patterns and Event Processing Language”. In: IEEE Transactions on Software Engineering PP (July 2018), pp. 1–1. doi: 10.1109/TSE.2018.2859926 (cit. on p. 11). [DAC99] M. Dwyer, G. Avrunin, J. Corbett. “Patterns in property specifications for finite-state verification”. In: Proceedings of the 1999 International Conference on Software Engineering (IEEE Cat. No.99CB37002). 1999, pp. 411–420. doi: 10.1145/302405. 302672 (cit. on pp. 1, 10, 11). [FHW+21] S. Frank, A. Hakamia, L. Wagner, D. Kesim, J. von Kistowski, A. van Hoorn. Scenario- based Resilience Evaluation and Improvement of Microservice Architectures: An Experience Report. 2021 (cit. on pp. 14, 18, 26, 32, 50, 62). [Fla] Flask. Flask Documentation. https://flask.palletsprojects.com/en/2.1.x/. (Accessed on 07/28/2022) (cit. on p. 39). GD18] Y. Gan, C. Delimitrou. “The Architectural Implications of Cloud Microservices”. In: IEEE Computer Architecture Letters 17.2 (2018), pp. 155–158. doi: 10.1109/LCA. 2018.2839189 (cit. on p. 1). Gre] Gremlin. https://www.gremlin.com/. (Accessed on 07/28/2022) (cit. on p. 31).l [Inf] InfluxDB. Open Source Time Series Database. https://www.influxdata.com/ products/influxdb-overview/. (Accessed on 07/28/2022) (cit. on p. 31). [Jin] [Jin] Jinja. Jinja Documentation (3.1.x). https://jinja.palletsprojects.com/en/3.1.x/. (Accessed on 07/28/2022) (cit. on p. 42). [KHFH20] D. Kesim, A. van Hoorn, S. Frank, M. Häussler. “Identifying and Prioritizing Chaos Experiments by Using Established Risk Analysis Techniques”. In: 2020 IEEE 31st International Symposium on Software Reliability Engineering (ISSRE). 2020, pp. 229– 240. doi: 10.1109/ISSRE5003.2020.00030 (cit. on pp. 26, 32). [Lit] [Lit] Litmus. Litmus Documentation. https://docs.litmuschaos.io/. (Accessed on 07/28/2022) (cit. on p. 31). [LS09] M. Leucker, C. Schallhart. “A brief account of runtime verification”. In: The Journal of Logic and Algebraic Programming 78.5 (2009). The 1st Workshop on Formal Languages and Analysis of Contract-Oriented Software (FLACOS’07), pp. 293–303. issn: 1567-8326. doi: https://doi.org/10.1016/j.jlap.2008.08.004. url: https://www.sciencedirect.com/science/article/pii/S1567832608000775 (cit. on pp. 12, 13, 28). [Mat] Matplotlib. https://matplotlib.org/. (Accessed on 07/28/2022) (cit. on p. 42). 68 Bibliography [Mic] MicroK8s. MicroK8s Documentation. https://microk8s.io/docs. (Accessed on 07/28/2022) (cit. on p. 50). R. Miles. Learning chaos engineering : discovering and overcoming system weaknesses through experimentation / Russ Miles. eng. First edition. Beijing: O’Reilly, 2019. isbn: 1-4920-5099-7 (cit. on pp. 7, 8, 25). S. Newman. Building microservices : designing fine-grained systems. Sebastopol, CA: O’Reilly Media, 2015. All links were last followed on July 28, 2022. Bibliography isbn: 9781491950357 (cit. on pp. 1, 51). [Pos] Postman. Postman API Platform. https : / / www . postman . com/. (Accessed on 07/28/2022) (cit. on p. 49). [Pri] Principles of chaos engineering. https://principlesofchaos.org/. (Accessed on 07/28/2022) (cit. on p. 7). Prometheus. Monitoring system & time series database. https://prometheus.io/. (Accessed on 07/28/2022) (cit. on p. 31). Python. Homepage. https://www.python.org/. (Accessed on 07/28/2022) (cit. on p. 31). C. Rosenthal, N. Jones. Chaos Engineering - System Resiliency in Practice. Sebastopol: O’Reilly Media, 2020. isbn: 978-1-492-04386-7 (cit. on pp. 1, 7, 25, 26, 32). D. Ulus. A pure Python package to monitor formal specifications over temporal se- quences. https://github.com/doganulus/python-monitors. (Accessed on 07/28/2022) (cit. on pp. 18, 39, 40). D. Ulus. Reelay Monitors. https://doganulus.github.io/reelay/. (Accessed on 07/28/2022) (cit. on p. 40). [Ulu19a] D. Ulus. “Timescales: A Benchmark Generator for MTL Monitoring Tools”. In: Oct. 2019, pp. 402–412. isbn: 978-3-030-32078-2. doi: 10.1007/978-3-030-32079-9_25 (cit. on pp. 10, 41, 48). [Ulu19b] D. Ulus. “Online Monitoring of Metric Temporal Logic using Sequential Networks”. In: CoRR abs/1901.00175 (2019). arXiv: 1901.00175. url: http://arxiv.org/abs/ 1901.00175 (cit. on pp. 10, 13, 18, 39–41). [VFH14] V. Venikov, D. Fry, W. Higinbotham. Transient Phenomena in Electrical Power Systems: International Series of Monographs on Electronics and Instrumentation, Vol. 24. v. 24. Elsevier Science, 2014. isbn: 9781483222714. url: https://books.google. de/books?id=czWoBQAAQBAJ (cit. on pp. 1, 8). M. Villamizar, O. Garcés, H. Castro, M. Verano, L. Salamanca, R. Casallas, S. Gil. “Evaluating the monolithic and the microservice architecture pattern to deploy web applications in the cloud”. In: 2015 10th Computing Colombian Conference (10CCC). IEEE. 2015, pp. 583–590 (cit. on p. 1). [Wag21] L. Wagner. “Simulating Scenario-based Chaos Experiments for Microservice Archi- tectures”. Bachelor’s Thesis. University of Stuttgart, 2021 (cit. on pp. 47, 56). [Zah22] D. Zahariev. Supporting and Verifying Transient Behavior Specifications in Chaos Engineering - Supplementary Materials. July 2022. doi: 10.5281/zenodo.6845089 (cit. on pp. 5, 49, 52, 54–57). All links were last followed on July 28, 2022. 69 69 71 71 A Evaluation Resources A Evaluation Resources A.1 Chaos Experiment Evaluation - Chaos Experiments Listing A.1 Evaluation - Chaos Experiment 1 (Without Transient Behavior Hypothesis Probe). A.1 Chaos Experiment Evaluation - Chaos Experiments Listing A.1 Evaluation - Chaos Experiment 1 (Without Transient Behavior Hypothesis Probe). 1 { 2 "title":"Terminate one service1 pod", 3 "description":"Terminating one service1 pod and observing steady-state hypothesis and ↩→transient behavior", 4 "tags":["kubernetes"], 5 "steady-state-hypothesis":{ 6 "title":"Verifying service remains healthy", 7 "probes":[ 8 { 9 "name":"deployment-is-fully-available", 10 "type":"probe", 11 "tolerance":true, 12 "provider":{ 13 "type":"python", 14 "module":"chaosk8s.probes", 15 "func":"deployment_fully_available", 16 "arguments":{ 17 "name":"Service1" 18 } 19 } 20 }, 21 .... 22 ] 23 }, 24 "method":[ 25 { 26 "type":"action", 27 "name":"terminate-service1-pod", 28 "provider":{ 29 "type":"python", 30 "module":"chaosk8s.pod.actions", 31 "func":"terminate_pods", 32 "arguments":{ 33 "name_pattern":"Service1", 34 "rand":true 35 } 36 }, 37 "pauses":{ 38 "before":30, 39 "after":90 40 } 41 } 42 ] 43 } 72 1 { 2 "title":"Terminate one service1 pod", 3 "description":"Terminating one service1 pod and ↩→transient behavior", 4 "tags":["kubernetes"], 5 "steady-state-hypothesis":{ 6 "title":"Verifying service remains healthy", 7 "probes":[ 8 { 9 "name":"deployment-is-fully-available" 10 "type":"probe", 11 "tolerance":true, 12 "provider":{ 13 "type":"python", 14 "module":"chaosk8s.probes", 15 "func":"deployment_fully_available" 16 "arguments":{ 17 "name":"Service1" 18 } 19 } 20 }, 21 .... 22 ] 23 }, 24 "method":[ 25 { 26 "type":"action", 27 "name":"terminate-service1-pod", 28 "provider":{ 29 "type":"python", 30 "module":"chaosk8s.pod.actions", 31 "func":"terminate_pods", 32 "arguments":{ 33 "name_pattern":"Service1", 34 "rand":true 35 } 36 }, 37 "pauses":{ 38 "before":30, 39 "after":90 40 } 41 } 42 ] 43 } 72 A.1 Chaos Experiment Evaluation - Chaos Experiments A.1 Chaos Experiment Evaluation - Chaos Experiments Listing A.2 Evaluation - Chaos Experiment 1 (Transient Behavior Hypothesis Probe). Listing A.2 Evaluation - Chaos Experiment 1 (Transient Behavior Hypothesis Probe). A Evaluation Resources 1 { 2 "name":"check-transient-behaviors", 3 "type":"probe", 4 "tolerance":{ 5 "type":"jsonpath", 6 "path":"$.result", 7 "target":"body", 8 "expect":[ 9 "True" 10 ] 11 }, 12 "provider":{ 13 "type":"http", 14 "url":"http://localhost:5000/monitor", 15 "method":"POST", 16 "headers":{ 17 "Content-Type":"application/json" 18 }, 19 "arguments":{ 20 "behavior_description":"Always when the instance count of a Service1 drops from 2 to ↩→1, a new instance has to be spawned in max 30 sec", 21 "specification":"always(((instances_dead(instances) and ↩→(resp_time_normal(resp_time)) ) since[0,30] instances_normal(instances)))", 22 "specification_type":"mtl", 23 "predicates_info":[{ 24 "predicate_name":"resp_time_normal", 25 "predicate_logic":"smallerEqual", 26 "predicate_comparison_value":"100" 27 },{ 28 "predicate_name":"instances_normal", 29 "predicate_logic":"equal", 30 "predicate_comparison_value":"2" 31 },{ 32 "predicate_name":"instances_dead", 33 "predicate_logic":"smaller", 34 "predicate_comparison_value":"2" 35 }], 36 "measurement_source":"influx", 37 "measurement_points":[{ 38 "measurement_name":"resp_time", 39 "measurement_query":"SELECT \"AvgResponseTime\" FROM ↩→\"TimeBatchRuns\".\"autogen\".\"Batch_Time\"" 40 },{ 41 "measurement_name":"instances", 42 "measurement_query":"SELECT \"counts\" FROM ↩→\"AliveInstances\".\"autogen\".\"InstanceCounts\"" 43 }] 44 } 45 } 46 } 1 { 2 "name":"check-transient-behaviors", 3 "type":"probe", 4 "tolerance":{ 5 "type":"jsonpath", 6 "path":"$.result", 7 "target":"body", 8 "expect":[ 9 "True" 10 ] 11 }, 12 "provider":{ 13 "type":"http", 14 "url":"http://localhost:5000/monitor", 15 "method":"POST", 16 "headers":{ 17 "Content-Type":"application/json" 18 }, 19 "arguments":{ 20 "behavior_description":"Always when the instance count of a Service1 drops from 2 to ↩→1, a new instance has to be spawned in max 30 sec", 21 "specification":"always(((instances_dead(instances) and ↩→(resp_time_normal(resp_time)) ) since[0,30] instances_normal(instances)))", 22 "specification_type":"mtl", 23 "predicates_info":[{ 24 "predicate_name":"resp_time_normal", 25 "predicate_logic":"smallerEqual", 26 "predicate_comparison_value":"100" 27 },{ 28 "predicate_name":"instances_normal", 29 "predicate_logic":"equal", 30 "predicate_comparison_value":"2" 31 },{ 32 "predicate_name":"instances_dead", 33 "predicate_logic":"smaller", 34 "predicate_comparison_value":"2" 35 }], 36 "measurement_source":"influx", 37 "measurement_points":[{ 38 "measurement_name":"resp_time", 39 "measurement_query":"SELECT \"AvgResponseTime\" FROM ↩→\"TimeBatchRuns\".\"autogen\".\"Batch_Time\"" 40 },{ 41 "measurement_name":"instances", 42 "measurement_query":"SELECT \"counts\" FROM ↩→\"AliveInstances\".\"autogen\".\"InstanceCounts\"" 43 }] 44 } 45 } 46 } 73 A Evaluation Resources Listing A.3 Evaluation - Chaos Experiment 2 (Without Transient Behavior Hypothesis P isting A.3 Evaluation - Chaos Experiment 2 (Without Transient Behavior Hypothesis Probe). Listing A.3 Evaluation - Chaos Experiment 2 (Without Transient Behavior Hypothesis Probe). 1 { 2 "title":"Increasing workload scenario", 3 "description":"Increasing load and observing steady-state hypothesis and transient ↩→behavior", 4 "tags":[ 5 "kubernetes" 6 ], 7 "steady-state-hypothesis":{ 8 "title":"Verifying service remains healthy", 9 "probes":[ 10 { 11 "name":"deployment-is-fully-available", 12 "type":"probe", 13 "tolerance":true, 14 "provider":{ 15 "type":"python", 16 "module":"chaosk8s.probes", 17 "func":"deployment_fully_available", 18 "arguments":{ 19 "name":"Service1" 20 } 21 } 22 }, 23 ..... A Evaluation Resources 24 ] 25 }, 26 "method":[ 27 { 28 "type":"action", 29 "name":"do-nothing", 30 "provider":{ 31 "type":"python", 32 "module":"chaosk8s.probes", 33 "func":"deployment_fully_available", 34 "arguments":{ 35 "name":"Service1" 36 } 37 }, 38 "pauses":{ 39 "after":600 40 } 41 } 42 ] 43 } 1 { 2 "title":"Increasing workload scenario", 3 "description":"Increasing load and observing steady ↩→behavior", 4 "tags":[ 5 "kubernetes" 6 ], 7 "steady-state-hypothesis":{ 8 "title":"Verifying service remains healthy", 9 "probes":[ 10 { 11 "name":"deployment-is-fully-available", 12 "type":"probe", 13 "tolerance":true, 14 "provider":{ 15 "type":"python", 16 "module":"chaosk8s.probes", 17 "func":"deployment_fully_available", 18 "arguments":{ 19 "name":"Service1" 20 } 21 } 22 }, 23 ..... 24 ] 25 }, 26 "method":[ 27 { 28 "type":"action", 29 "name":"do-nothing", 30 "provider":{ 31 "type":"python", 32 "module":"chaosk8s.probes", 33 "func":"deployment_fully_available", 34 "arguments":{ 35 "name":"Service1" 36 } 37 }, 38 "pauses":{ 39 "after":600 40 } 41 } 42 ] 43 } 74 A.1 Chaos Experiment Evaluation - Chaos Experiments A.1 Chaos Experiment Evaluation - Chaos Experiments Listing A.4 Evaluation - Chaos Experiment 2 (Transient Behavior Hypothesis Probe). Listing A.4 Evaluation - Chaos Experiment 2 (Transient Behavior Hypothesis Probe). 1 { 2 "name":"check-transient-behaviors", 3 "type":"probe", 4 "tolerance":{ 5 "type":"jsonpath", 6 "path":"$.result", 7 "target":"body", 8 "expect":["True"] 9 }, 10 "provider":{ 11 "type":"http", 12 "url":"http://localhost:5000/monitor", 13 "method":"POST", 14 "headers":{ 15 "Content-Type":"application/json" 16 }, 17 "arguments":{ 18 "specification":"Globally, if {resp_time_high(resp_time)} [has occurred] then in ↩→response {instance_increase(instance_count)} [eventually holds] between 30 and 60 time ↩→units.", 19 "specification_type":"psp", 20 "predicates_info":[ 21 { 22 "predicate_name":"resp_time_high", 23 "predicate_description":"Response time is higher than 100ms", 24 "predicate_comparison_value":"100", 25 "predicate_logic":"biggerEqual" 26 }, 27 { 28 "predicate_name":"instance_increase", 29 "predicate_description":"Instance counts increase", 30 "predicate_comparison_value":"2", 31 "predicate_logic":"bigger" 32 }], 33 "measurement_source":"influx", 34 "measurement_points":[ 35 { 36 "measurement_name":"resp_time", 37 "measurement_query":"SELECT \"AvgResponseTime\" FROM ↩→\"TimeBatchRuns\".\"autogen\".\"Batch_Time\"" 38 }, 39 { 40 "measurement_name":"instance_count", 41 "measurement_query":"SELECT \"counts\" FROM ↩→\"AliveInstances\".\"autogen\".\"InstanceCounts\"" 42 }] 43 } 44 } 45 } 75 75 A Evaluation Resources Listing A.5 Evaluation - Chaos Experiment 3 (Without Transient Behavior Hypothesis P isting A.5 Evaluation - Chaos Experiment 3 (Without Transient Behavior Hypothesis Probe). Listing A.5 Evaluation - Chaos Experiment 3 (Without Transient Behavior Hypothesis Probe). 1 { 2 "title":"Increasing workload scenario", 3 "description":"Increasing load and observing steady-state hypothesis and transient ↩→behavior", 4 "tags":[ 5 "kubernetes" 6 ], 7 "steady-state-hypothesis":{ 8 "title":"Verifying service remains healthy", 9 "probes":[ 10 { 11 "name":"deployment-is-fully-available", 12 "type":"probe", 13 "tolerance":true, 14 "provider":{ 15 "type":"python", 16 "module":"chaosk8s.probes", 17 "func":"deployment_fully_available", 18 "arguments":{ 19 "name":"Service1" 20 } 21 } 22 }, 23 ..... A Evaluation Resources 24 ] 25 }, 26 "method":[ 27 { 28 "type":"action", 29 "name":"do-nothing", 30 "provider":{ 31 "type":"python", 32 "module":"chaosk8s.probes", 33 "func":"deployment_fully_available", 34 "arguments":{ 35 "name":"Service1" 36 } 37 }, 38 "pauses":{ 39 "after":600 40 } 41 } 42 ] 43 } 1 { 2 "title":"Increasing workload scenario", 3 "description":"Increasing load and observing steady ↩→behavior", 4 "tags":[ 5 "kubernetes" 6 ], 7 "steady-state-hypothesis":{ 8 "title":"Verifying service remains healthy", 9 "probes":[ 10 { 11 "name":"deployment-is-fully-available", 12 "type":"probe", 13 "tolerance":true, 14 "provider":{ 15 "type":"python", 16 "module":"chaosk8s.probes", 17 "func":"deployment_fully_available", 18 "arguments":{ 19 "name":"Service1" 20 } 21 } 22 }, 23 ..... 24 ] 25 }, 26 "method":[ 27 { 28 "type":"action", 29 "name":"do-nothing", 30 "provider":{ 31 "type":"python", 32 "module":"chaosk8s.probes", 33 "func":"deployment_fully_available", 34 "arguments":{ 35 "name":"Service1" 36 } 37 }, 38 "pauses":{ 39 "after":600 40 } 41 } 42 ] 43 } 76 A.1 Chaos Experiment Evaluation - Chaos Experiments Listing A.6 Evaluation - Chaos Experiment 3 (Transient Behavior Hypothesis Probe). Listing A.6 Evaluation - Chaos Experiment 3 (Transient Behavior Hypothesis Probe). 1 { 2 "name":"check-transient-behaviors", 3 "type":"probe", 4 "tolerance":{ 5 "type":"jsonpath", 6 "path":"$.result", 7 "target":"body", 8 "expect":["True"] 9 }, 10 "provider":{ 11 "type":"http", 12 "url":"http://localhost:5000/monitor", 13 "method":"POST", 14 "headers":{ 15 "Content-Type":"application/json" 16 }, 17 "arguments":{ 18 "behavior_description":"description", 19 "specification":"always(load_normal(service1_1_load) and ↩→load_normal(service1_2_load))", 20 "specification_type":"mtl", 21 "predicates_info":[ 22 { 23 "predicate_name":"load_normal", 24 "predicate_description":"description", 25 "predicate_comparison_value":"0.8", 26 "predicate_logic":"smaller" 27 } 28 ], 29 "measurement_source":"prometheus", 30 "measurement_points":[{ 31 "measurement_name":"service1_1_load", 32 "measurement_query":" (rate (container_cpu_usage_seconds_total{ ↩→pod='Service1-588bdc5d88-tkclz'}[1m])) ", 33 "start_time":"10m", 34 "end_time":"now", 35 "steps":"1" 36 },{ 37 "measurement_name":"service1_2_load", 38 "measurement_query":" (rate ↩→(container_cpu_usage_seconds_total{pod='Service1-588bdc5d88-bvfn5'}[1m])) ", 39 "start_time":"10m", 40 "end_time":"now", 41 "steps":"1" 42 } 43 ] 44 } 45 } 46 } 77 77 A Evaluation Resources A.2 Compatibility Evaluation Resources Listing A.7 MiSim Architecture Definition. 1 { 2 "network_latency":".02+.002-0.001", 3 "microservices":[ 4 { 5 "name":"gateway", 6 "instances":1, 7 "capacity":10000, 8 "loadbalancer_strategy":"even", 9 "operations":[ 10 { 11 "name":"API_Endpoint", 12 "demand":1, 13 "dependencies":[ 14 { 15 "service":"service1", 16 "operation":"dependentCalculation" 17 } 18 ] 19 } 20 ] 21 }, 22 { 23 "name":"service1", 24 "instances":2, 25 "patterns":[], 26 "capacity":200, 27 "operations":[ 28 { 29 "name":"dependentCalculation", 30 "demand":10, 31 "dependencies":[ 32 { 33 "operation":"service2.doWork", 34 "probability":1 35 } 36 ] 37 } 38 ], A.2 Compatibility Evaluation Resources A.2 Compatibility Evaluation Resources Listing A.7 MiSim Architecture Definition. 1 { 2 "network_latency":".02+.002-0.001", 3 "microservices":[ 4 { 5 "name":"gateway", 6 "instances":1, 7 "capacity":10000, 8 "loadbalancer_strategy":"even", 9 "operations":[ 10 { 11 "name":"API_Endpoint", 12 "demand":1, 13 "dependencies":[ 14 { 15 "service":"service1", 16 "operation":"dependentCalculation" 17 } 18 ] 19 } 20 ] 21 }, 22 { 23 "name":"service1", 24 "instances":2, 25 "patterns":[], 26 "capacity":200, 27 "operations":[ 28 { 29 "name":"dependentCalculation", 30 "demand":10, 31 "dependencies":[ 32 { 33 "operation":"service2.doWork", 34 "probability":1 35 } 36 ] 37 } 38 ], 78 Listing A.8 MiSim Architecture Definition (Continuation). 39 "s_patterns":[ 40 { 41 "type":"autoscaling", 42 "config":{ 43 "period":0.5 44 }, 45 "strategy":{ 46 "type":"reactive", 47 "config":{ 48 "lower_bound":0.2, 49 "upper_bound":0.8 50 } 51 } 52 } 53 ] 54 }, 55 { 56 "name":"service2", 57 "instances":1, 58 "capacity":32000, 59 "operations":[ 60 { 61 "name":"doWork", 62 "demand":1 63 } 64 ] 65 } 66 ] 67 } Listing A.9 MiSim Experiment Definition. 1 { 2 "name": "Minimal Scenario", 3 "duration": 120, 4 "seed": 42, 5 "artifact": "gateway", 6 "component": "ALL ENDPOINTS", 7 "stimulus": "LOAD~ ./Examples/PaperExample/constant_loadArrivalRates_120sec.csv AND KILL ↩→service1 1@30" 8 } Listing A.8 MiSim Architecture Definition (Continuation). Listing A.8 MiSim Architecture Definition (Continuation). 39 "s_patterns":[ 40 { 41 "type":"autoscaling", 42 "config":{ 43 "period":0.5 44 }, 45 "strategy":{ 46 "type":"reactive", 47 "config":{ 48 "lower_bound":0.2, 49 "upper_bound":0.8 50 } 51 } 52 } 53 ] 54 }, 55 { 56 "name":"service2", 57 "instances":1, 58 "capacity":32000, 59 "operations":[ 60 { 61 "name":"doWork", 62 "demand":1 63 } 64 ] 65 } 66 ] 67 } Listing A.8 MiSim Architecture Definition 39 "s_patterns":[ 40 { 41 "type":"autoscaling", 42 "config":{ 43 "period":0.5 44 }, 45 "strategy":{ 46 "type":"reactive", 47 "config":{ 48 "lower_bound":0.2, 49 "upper_bound":0.8 50 } 51 } 52 } 53 ] 54 }, 55 { 56 "name":"service2", 57 "instances":1, 58 "capacity":32000, 59 "operations":[ 60 { 61 "name":"doWork", 62 "demand":1 63 } 64 ] 65 } 66 ] 67 } "s_patterns":[ Listing A.9 MiSim Experiment Definition. Listing A.9 MiSim Experiment Definition. 1 { 2 "name": "Minimal Scenario", 3 "duration": 120, 4 "seed": 42, 5 "artifact": "gateway", 6 "component": "ALL ENDPOINTS", 7 "stimulus": "LOAD~ ./Examples/PaperExample/constant_loadArrivalRates_120sec.csv AND KILL ↩→service1 1@30" 8 } Listing A.10 Compatibility Evaluation - Transient Behavior Specification 1. A.2 Compatibility Evaluation Resources 1 { 2 "specification": " always(instance_dead(instance_count) since[0,3] ↩→instance_alive(instance_count)) ", 3 "specification_type": "mtl", 4 "predicates_info": [ 5 { 6 "predicate_name": "instance_alive", 7 "predicate_comparison_value": "2", 8 "predicate_logic": "equal" 9 }, 10 { 11 "predicate_name": "instance_dead", 12 "predicate_comparison_value": "1", 13 "predicate_logic": "equal" 14 } 15 ], 16 "measurement_source": "csv", 17 "measurement_points": [ 18 { 19 "measurement_name": "instance_count", 20 "measurement_column": "instance_count" 21 } 22 ] 23 } Listing A.11 Compatibility Evaluation - Transient Behavior Specification 2. i 1 { 2 "specification":"Before {instance_up(instance_count)}, it is never the case that ↩→{resp_time_high(resp_time)} [holds].", 3 "specification_type":"psp", 4 "predicates_info":[ 5 { 6 "predicate_name":"instance_up", 7 "predicate_description":"A new instance has been spawned.", 8 "predicate_logic":"trendUpwardStrict" 9 }, 10 { 11 "predicate_name":"resp_time_high", 12 "predicate_description":"Response time is too high when higher than 0.5.", 13 "predicate_comparison_value":"0.5", 14 "predicate_logic":"bigger" 15 } 16 ], 17 "measurement_source":"csv", 18 "measurement_points":[ 19 { 20 "measurement_name":"instance_count", 21 "measurement_column":"instance_count" 22 }, 23 { 24 "measurement_name":"resp_time", 25 "measurement_column":"resp_time" 26 } 27 ] 28 } place, date, signature Declaration I hereby declare that the work presented in this thesis is entirely my own and that I did not use any other sources and references than the listed ones. I have marked all direct or indirect statements from other sources contained therein as quotations. Neither this work nor significant parts of it were part of another examination procedure. I have not published this work in whole or in part before. The electronic copy is consistent with all submitted copies.
https://openalex.org/W2883650593	https://europepmc.org/articles/pmc6080792?pdf=render	English	null	Chronic schistosomiasis suppresses HIV-specific responses to DNA-MVA and MVA-gp140 Env vaccine regimens despite antihelminthic treatment and increases helminth-associated pathology in a mouse model	PLOS pathogens	2,018	cc-by	13,647	RESEARCH ARTICLE OPEN ACCESS Citation: Dzhivhuho GA, Rehrl SA, Ndlovu H, Horsnell WGC, Brombacher F, Williamson A-L, et al. (2018) Chronic schistosomiasis suppresses HIV-specific responses to DNA-MVA and MVA- gp140 Env vaccine regimens despite antihelminthic treatment and increases helminth-associated pathology in a mouse model. PLoS Pathog 14(7): e1007182. https://doi.org/10.1371/journal. ppat.1007182 ¤ Current address: Department of Microbiology, Immunology and Cancer Biology, Myles H. Thaler Center for AIDS and Human Retrovirus Research, University of Virginia, Charlottesville, Virginia, United States of America. Godfrey A. Dzhivhuho1¤, Samantha A. Rehrl1, Hlumani Ndlovu2, William G. C. Horsnell1,3,4,5, Frank Brombacher1,3,6,7, Anna-Lise Williamson1,3, Gerald K. Chege1,7* Godfrey A. Dzhivhuho1¤, Samantha A. Rehrl1, Hlumani Ndlovu2, William G. C. Horsnell1,3,4,5, Frank Brombacher1,3,6,7, Anna-Lise Williamson1,3, Gerald K. Chege1,7* 1 Department of Pathology, Faculty of Health Sciences, University of Cape Town, Cape Town, Western Cape, South Africa, 2 Department of Integrative Biomedical Sciences, Faculty of Health Sciences, University of Cape Town, Cape Town, Western Cape, South Africa, 3 Institute of Infectious Disease and Molecular Medicine, Faculty of Health Sciences, University of Cape Town, Cape Town, Western Cape, South Africa, 4 Institute of Microbiology and Infection, University of Birmingham, Birmingham, United Kingdom, 5 Laboratory of Molecular and Experimental Immunology and Neurogenetics, Centre Nationnal de la Recherche Scientifique, University of Orleans and Le Studium Institute for Advanced Studies, Rue Dupanloup, Orle´ans, France, 6 International Centre for Genomic Engineering and Biotechnology–Cape Town Component, Cape Town, Western Cape, South Africa, 7 South African Medical Research Council, Cape Town, Western Cape, South Africa Abstract Editor: Guido Silvestri, Emory University, UNITED STATES Chronic schistosomiasis suppresses HIV- specific responses to DNA-MVA and MVA- gp140 Env vaccine regimens despite antihelminthic treatment and increases helminth-associated pathology in a mouse model Godfrey A. Dzhivhuho1¤, Samantha A. Rehrl1, Hlumani Ndlovu2, William G. C. Horsnell1,3,4,5, Frank Brombacher1,3,6,7, Anna-Lise Williamson1,3, Gerald K. Chege1,7* Author summary Chronic parasitic worm infections are thought to reduce the efficacy of vaccines. Given that HIV and worm infections are common in sub-Saharan Africa (SSA) and their geo- graphical distribution vastly overlaps, it is likely that future HIV vaccines in SSA will be administered to a large proportion of people with chronic worm infections. This study examined the impact of S. mansoni worm infections on the immunogenicity of candidate HIV vaccines in a mouse model. S. mansoni worm-infected animals had lower magnitudes of HIV vaccine responses compared with uninfected animals and elimination of worms by praziquantel treatment prior to vaccination conferred only partial restoration of nor- mal immune responses to vaccination. The presence of S. mansoni eggs trapped in the tis- sues in the absence of live infection was associated with poor vaccine responses. In addition, this study found that effective immunization with some HIV vaccine regimens could potentially worsen worm-associated pathology when given to infected individuals. These novel findings suggest further research in HIV vaccines and future vaccination poli- cies regarding the current clinical vaccines and future HIV vaccination with respect to parasitic worm infections especially in SSA. Impact of chronic schistosomiasis on HIV vaccination +gp140 compared to uninfected control mice. Similarly, gp140 Env-specific antibody responses were significantly lower in vaccinated Sm-infected mice compared to controls. Treatment with PZQ partially restored cellular but not humoral immune responses in vacci- nated Sm-infected mice. Gp140 Env-specific antibody responses were attenuated in mice that were inoculated with SmE compared to controls. Lastly, Sm-infected mice that were vaccinated with DNA+MVA displayed exacerbated liver pathology as indicated by larger granulomas and increased hepatosplenomegaly when compared with unvaccinated Sm- infected mice. This study shows that chronic schistosomiasis attenuates both HIV-specific T-cell and antibody responses and parasite elimination by chemotherapy may partially restore cellular but not antibody immunity, with additional data suggesting that the presence of SmE retained in the tissues after antihelminthic therapy contributes to lack of full immune restoration. Our data further suggest that helminthiasis may compromise HIV vaccine safety. Overall, these findings suggested a potential negative impact on future HIV vaccina- tions by helminthiasis in endemic areas. Research Chairs Initiative of the DST and NRF. The SAAVI vaccine production was funded by the South African Medical Research Council and NIH NIAID Contract NOI-AI-95371. The Poliomyelitis Research Foundation (PRF) of South Africa provided additional student bursaries (to GAD and SAR). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript. Competing interests: The authors have declared that no competing interests exist. Editor: Guido Silvestri, Emory University, UNITED STATES Future HIV vaccines are expected to induce effective Th1 cell-mediated and Env-specific antibody responses that are necessary to offer protective immunity to HIV infection. How- ever, HIV infections are highly prevalent in helminth endemic areas. Helminth infections induce polarised Th2 responses that may impair HIV vaccine-generated Th1 responses. In this study, we tested if Schistosoma mansoni (Sm) infection altered immune responses to SAAVI candidate HIV vaccines (DNA and MVA) and an HIV-1 gp140 Env protein vaccine (gp140) and whether parasite elimination by chemotherapy or the presence of Sm eggs (SmE) in the absence of active infection influenced the immunogenicity of these vaccines. In addition, we evaluated helminth-associated pathology in DNA and MVA vaccination groups. Mice were chronically infected with Sm and vaccinated with DNA+MVA in a prime+boost combination or MVA+gp140 in concurrent combination regimens. Some Sm-infected mice were treated with praziquantel (PZQ) prior to vaccinations. Other mice were inoculated with SmE before receiving vaccinations. Unvaccinated mice without Sm infection or SmE inocu- lation served as controls. HIV responses were evaluated in the blood and spleen while Sm- associated pathology was evaluated in the livers. Sm-infected mice had significantly lower magnitudes of HIV-specific cellular responses after vaccination with DNA+MVA or MVA Received: April 5, 2018 Accepted: June 27, 2018 Published: July 26, 2018 Copyright: © 2018 Dzhivhuho et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited. Data Availability Statement: All relevant data are within the paper and its Supporting Information files. Funding: The main funding was provided by the South African Department of Science and Technology through a National Research Foundation (NRF) grant to GC. Some work is based upon research supported by the South African PLOS Pathogens \| https://doi.org/10.1371/journal.ppat.1007182 July 26, 2018 1 / 23 PLOS Pathogens \| https://doi.org/10.1371/journal.ppat.1007182 July 26, 2018 Impact of chronic schistosomiasis on HIV vaccination protection would be conferred through a T helper 1 (Th1) immune pathway [8, 9]. Induction of potent T cell mediated immune responses has previously been demonstrated using heterolo- gous prime-boost vaccination strategies that utilise DNA and viral vaccine vectors such as modified Vaccinia Ankara (MVA) [10–14], while induction of durable antibody immune responses may require immunisation with HIV envelope protein-based vaccines [15–18]. It is widely accepted that an ideal HIV vaccine should induce both anti-HIV cellular responses and HIV Env-specific antibodies to destroy virus-infected cells and neutralize viruses at portals of entry respectively in order to clear the virus before dissemination into the tissues or block viral entry at the mucosal sites [8, 9, 19, 20]. During chronic schistosomiasis, parasite eggs are lodged in the liver and intestinal tissue [21, 22] resulting in predominantly T-helper 2 (Th2) immune responses [23–27] and the induction of anti-inflammatory regulatory T-cells (Treg) which suppress the innate and adap- tive T- and B-cell responses [24, 28, 29]. This has been shown to lead to general hyporespon- siveness which may adversely impact standard immunizations, by suppressing immune responses to Th1-type vaccine and impairing the expansion of pathogen-specific cytotoxic T lymphocyte (CTL) responses [30–37]. Parasitic helminth infections are currently treated with chemotherapeutic drugs such as praziquantel (PZQ) for schistosomiasis [38–40] and ivermectin or mebendazole for geohel- minths [41–43], which are cost-effective interventions. However, re-infection after effective treatment is common and frequent in populations in endemic areas [44]. Several animal and clinical studies have reported that helminth infections impair the outcome of a variety of vac- cines, including Salmonella [45]; BCG [30, 46–48], tetanus [46, 49–51], diphtheria toxoid [52], HBV [53], pneumococcal [54] and live attenuated oral cholera vaccines [55]. However, elimi- nation of helminth infection has also been shown to at least partially restore this abrogation [56]. Furthermore, individuals treated with antihelminthics show higher frequencies of BCG- specific IFN-γ and IL-12 producing cells than untreated helminth infected individuals [30]. Previous HIV vaccines studies reported reduced vaccine-induced immunity in schisto- some-infected mice [57] and partial restoration after elimination of helminths [58, 59]. How- ever, it is not clear if antibody responses are attenuated as these studies evaluated only cellular responses to Gag as they were monovalent candidate vaccines. PLOS Pathogens \| https://doi.org/10.1371/journal.ppat.1007182 July 26, 2018 Introduction Human immunodeficiency virus (HIV) and parasitic helminthic worm infections are highly prevalent and geographically overlap each other in sub Saharan Africa (SSA) [1, 2]. A majority of inhabitants harbor at least one or more species of parasitic helminth infection [3–6] and an estimated 50% of the chronically infected individuals living in high-risk rural communities are co-infected with HIV [7]. Furthermore, re-infections after successful treatments are also very common in endemic areas. Therefore, it is very likely that successful future HIV vaccines will be administered to people who already have ongoing helminthiasis or have been previously infected and treated. Current HIV-1 vaccine research suggests that a successful HIV vaccine will need to induce effective T cell and functional antibody responses, where a key component of immune PLOS Pathogens \| https://doi.org/10.1371/journal.ppat.1007182 July 26, 2018 2 / 23 Sm-infection impairs Th1-like cellular and humoral immunity in HIV vaccinated mice To assess the impact of Sm-infection on systemic immune responses, we quantified systemic Th1 and Th2 immune responses in uninfected and Sm-infected mice following MVA+gp140 and DNA+MVA vaccination. Con A stimulation of splenocytes from Sm-infected mice resulted in a significantly reduced IFN-γ:IL-4 ratio in DNA+MVA (p<0.05) and MVA+gp140 (p<0.01) vaccine regimens compared to Sm-uninfected mice (Fig 1A). Similarly, Con A-stimulated splenocytes from Sm-infected mice produced significantly lower (p<0.05) levels of IFN-γ and IL-2 ELISpot responses compared to splenocytes from uninfected mice (Fig 1B). Furthermore, Sm-infected mice had a reduced frequency (p<0.05) of cytokine-producing CD4+ and CD8+ T cells after re-stimulation of splenocytes with PMA/ Ionomycin compared to splenocytes from uninfected control mice (Fig 1C). Moreover, vacci- nated Sm-infected mice displayed an impaired type 1 antibody response, indicated by reduced amount of type 1 total antibody isotypes [(IgG2a (p<0.001), IgG2b (p<0.001)] and increased type 2 antibody isotype [(total IgG1 (p<0.001) (Type 2-associated antibodies), IgM (p<0.001)] compared to uninfected but vaccinated mice and uninfected control mice (Fig 1D). Sm-infec- tion was accompanied with increased levels of the regulatory cytokine IL-10 (S1C and S1I Fig). Impact of chronic schistosomiasis on HIV vaccination investigated whether the S. mansoni eggs (SmE) in the absence of active infection, which mim- ics the state whereby SmE remain trapped in the tissues shortly after antihelminthic treatment, has an adverse effect on vaccine immunogenicity. Lastly, we evaluated helminth-induced pathology to predict HIV vaccine safety in helminth endemic areas. Our findings show that mice infected with S. mansoni displayed reduced magnitudes of vaccine-specific cellular and humoral responses and anthelminthic treatment with PZQ failed to restore levels of anti- gp140 antibodies while partially reversing the adverse impact on cellular responses. Unexpect- edly, vaccination with a T-cell based vaccine regimen was observed to worsen helminth-associ- ated pathology suggesting potential safety concerns in future mass HIV vaccination in helminth endemic areas. PLOS Pathogens \| https://doi.org/10.1371/journal.ppat.1007182 July 26, 2018 Current vaccine candidates and future successful vaccines will likely include multiple immunogens, including Env, in order to broaden the vaccine targets and the capacity to cross-neutralise the majority of transmitted viruses. [60, 61]. It is well-accepted that helminth-induced Th2 responses play an important role in host pro- tection [62, 63]. Since Th1 and Th2 display reciprocal antagonist, it would be anticipated that HIV vaccine-generated Th1 responses may reduce host immunity against helminth-associated pathology, thereby compromising the safety of an otherwise effective HIV T-cell vaccine. Pox- virus-vectored HIV vaccines are promising candidates for induction of T cell responses and therefore this is a relevant safety issue which remains under-investigated in the helminthic infection background. We have previously described the development of two multigene candidate vaccines, the SAAVI DNA-C2 and SAAVI MVA-C, which express matched HIV-1 subtype C proteins (Gag, RT, Tat, Nef and Env) [11, 64–66]. These vaccine candidates have been evaluated further in nonhuman primates [18, 67] and Phase 1 clinical trials [15, 16]. Also, we have evaluated these vaccines in combination with an HIV-1C gp140ΔV2 Env protein [11, 15, 16, 18, 64, 65, 67]. In the current study, we investigated the impact of chronic schistosomiasis on the immuno- genicity of these vaccines in a mouse model and whether the elimination of worms by antihel- minthic chemotherapy prior to immunization benefits vaccination outcome. We further PLOS Pathogens \| https://doi.org/10.1371/journal.ppat.1007182 July 26, 2018 3 / 23 Impact of chronic schistosomiasis on HIV vaccination SFU/106 splenocytes) ELISpot responses in uninfected mice compared to Sm-infected mice Fig 1. Th1/Th2 profile: S. mansoni infection induces a strong Th2 biased cellular and antibody responses. Spleens and blood were obtained from vaccinated Sm-free (blue) and Sm-infected (red) mice 16 weeks post infection. Splenocytes were prepared and stimulated with 1 μg/ml Con A for 48 (A) and 23 (B) hours and cytokines were measured in cytometric bead array and ELISpot assays respectively. Other splenocytes were stimulated with PMA/Ionomycin for 6hrs (C) and the induced intracellular cytokines were measured by flow cytometry. Total Th1 and Th2 antibodies were also measured in sera using an antibody ELISA (D). Results represent 3 independent experiments and plotted as the mean + SEM. Statistical analysis was performed using unpaired, two-tailed t-test analysis followed by FDR for multiple comparisons. (: p<0.05; : p<0.01; : p<0.001). https://doi.org/10.1371/journal.ppat.1007182.g001 Fig 1. Th1/Th2 profile: S. mansoni infection induces a strong Th2 biased cellular and antibody responses. Spleens and blood were obtained from vaccinated Sm-free (blue) and Sm-infected (red) mice 16 weeks post infection. Splenocytes were prepared and stimulated with 1 μg/ml Con A for 48 (A) and 23 (B) hours and cytokines were measured in cytometric bead array and ELISpot assays respectively Other splenocytes were stimulated with PMA/Ionomycin for 6hrs (C) and Fig 1. Th1/Th2 profile: S. mansoni infection induces a strong Th2 biased cellular and antibody responses. Spleen Fig 1. Th1/Th2 profile: S. mansoni infection induces a strong Th2 biased cellular and antibody responses. Spleens and blood were obtained from vaccinated Sm-free (blue) and Sm-infected (red) mice 16 weeks post infection. Splenocytes were prepared and stimulated with 1 μg/ml Con A for 48 (A) and 23 (B) hours and cytokines were measured in cytometric bead array and ELISpot assays respectively. Other splenocytes were stimulated with PMA/Ionomycin for 6hrs (C) and the induced intracellular cytokines were measured by flow cytometry. Total Th1 and Th2 antibodies were also measured in sera using an antibody ELISA (D). Results represent 3 independent experiments and plotted as the mean + SEM. Statistical analysis was performed using unpaired, two-tailed t-test analysis followed by FDR for multiple comparisons. (: p<0.05; : p<0.01; : p<0.001). Fig 1. Th1/Th2 profile: S. mansoni infection induces a strong Th2 biased cellular and antibody responses. Spleens and blood were obtained from vaccinated Sm-free (blue) and Sm-infected (red) mice 16 weeks post infection. T cell vaccine-specific responses to SAAVI HIV-1 vaccines are attenuated by chronic helminth infection while antihelminthic treatment only partially restores the magnitudes to normal responses To determine the effect of chronic Sm infection on the HIV-1 vaccine-specific T cell immu- nity, Sm-infected and uninfected mice were vaccinated with either DNA+MVA or MVA +gp140 vaccine regimens and vaccine-specific T cell responses were determined using ELI- Spot, CBA and flow cytometry. To determine if elimination of schistosome infection prior to vaccination could reverse the effect on those responses, groups of mice were treated with PZQ before vaccinations. Vaccination with DNA+MVA induced significantly higher cumulative HIV-1 specific IFN-γ (2014 ± 177.4 SFU/106 splenocytes) and IL-2 (174.1 ± 71.13 SFU/106 splenocytes) ELISpot responses in uninfected mice compared to Sm-infected mice (IFN-γ: 1420 ± 61.54 SFU/106 splenocytes and IL-2: 0 SFU/106 splenocytes) (Fig 2A and 2B). Responses to the RT (CD8) peptide induced the highest number of IFN-γ secreting CD8+ and CD4+ T cells (1019 ± 217.3 SFU/106 splenocytes) compared to other peptides in uninfected mice vaccinated with DNA+MVA. However, IL-2 SFU/106 cells were similar among different peptides stimulations. Similarly, vaccination with MVA+gp140 induced significantly higher cumulative HIV-1 specific IFN-γ (1838 ± 173.3 SFU/106 splenocytes) and IL-2 (197.7 ± 20.12 To determine the effect of chronic Sm infection on the HIV-1 vaccine-specific T cell immu- nity, Sm-infected and uninfected mice were vaccinated with either DNA+MVA or MVA +gp140 vaccine regimens and vaccine-specific T cell responses were determined using ELI- Spot, CBA and flow cytometry. To determine if elimination of schistosome infection prior to vaccination could reverse the effect on those responses, groups of mice were treated with PZQ before vaccinations. Vaccination with DNA+MVA induced significantly higher cumulative HIV-1 specific IFN-γ (2014 ± 177.4 SFU/106 splenocytes) and IL-2 (174.1 ± 71.13 SFU/106 splenocytes) ELISpot responses in uninfected mice compared to Sm-infected mice (IFN-γ: 1420 ± 61.54 SFU/106 splenocytes and IL-2: 0 SFU/106 splenocytes) (Fig 2A and 2B). Responses to the RT (CD8) peptide induced the highest number of IFN-γ secreting CD8+ and CD4+ T cells (1019 ± 217.3 SFU/106 splenocytes) compared to other peptides in uninfected mice vaccinated with DNA+MVA. However, IL-2 SFU/106 cells were similar among different peptides stimulations. Similarly, vaccination with MVA+gp140 induced significantly higher cumulative HIV-1 specific IFN-γ (1838 ± 173.3 SFU/106 splenocytes) and IL-2 (197.7 ± 20.12 PLOS Pathogens \| https://doi.org/10.1371/journal.ppat.1007182 July 26, 2018 4 / 23 PLOS Pathogens \| https://doi.org/10.1371/journal.ppat.1007182 July 26, 2018 Impact of chronic schistosomiasis on HIV vaccination Fig 2. Sm infection alters vaccine-specific cellular responses and treatment with PZQ only partially restores these responses close to normal. Mice were chronically infected with Sm before vaccination with a DNA-vectored and MVA-vectored HIV-1 (DNA+MVA) vaccine or an MVA-vectored HIV-1 and a HIV-1 gp140 Env protein (MVA+gp140) vaccine regimen with or without prior anti-helminthic treatment with PZQ. Spleens were collected 12 days after the last vaccination. The splenocytes were stimulated and the induced cytokines measured in an IFN-γ (A) and IL-2 (B) ELISpot, cytometric bead array (C, D, and E), and intracellular cytokine staining (F, G, and H) assays. Results represent 3 independent experiments and plotted as the mean + SEM. Statistical analysis was performed using unpaired, two-tailed t-test analysis followed by FDR for multiple comparisons. (: p<0.05; : p<0.01; : p<0.001). https://doi.org/10.1371/journal.ppat.1007182.g002 Fig 2. Sm infection alters vaccine-specific cellular responses and treatment with PZQ only partially restores these responses close to normal. Mice were chronically infected with Sm before vaccination with a DNA-vectored and MVA-vectored HIV-1 (DNA+MVA) vaccine or an MVA-vectored HIV-1 and a HIV-1 gp140 Env protein (MVA+gp140) vaccine regimen with or without prior anti-helminthic treatment with PZQ. Spleens were collected 12 days after the last vaccination. The splenocytes were stimulated and the induced cytokines measured in an IFN-γ (A) and IL-2 (B) ELISpot, cytometric bead array (C, D, and E), and intracellular cytokine staining (F, G, and H) assays. Results represent 3 independent experiments and plotted as the mean + SEM. Statistical analysis was performed using unpaired, two-tailed t-test analysis followed by FDR for multiple comparisons. (: p<0.05; : p<0.01; : p<0.001). https://doi org/10 1371/journal ppat 1007182 g002 https://doi.org/10.1371/journal.ppat.1007182.g002 pg/ml) and TNF-α (251.2 ± 30.33 pg/ml) (Fig 2C–2E) were released by splenocytes from unin- fected mice in the DNA+MVA vaccine regimen compared to lower levels of IFN-γ (1899 ± 244.6 pg/ml); IL-2 (23.55 ± 4.094 pg/ml) and TNF-α (122.9 ± 17.45 pg/ml) released in Sm- infected vaccinated mice (Fig 2C–2E). Similarly, for the MVA+gp140 vaccine regimen, signifi- cantly higher levels of net cumulative Th1 cytokines: IFN-γ (2416 pg/ml); IL-2 (63.79 pg/ml) and lower TNF-α (112.48 pg/ml) were released from splenocytes of uninfected vaccinated mice compared to lower levels of IFN-γ (789 pg/ml); IL-2 (4.0 pg/ml) and higher TNF-α (123.87 pg/ ml) released in Sm-infected vaccinated mice (Fig 2C–2E). Splenocytes were prepared and stimulated with 1 μg/ml Con A for 48 (A) and 23 (B) hours and cytokines were measured in cytometric bead array and ELISpot assays respectively. Other splenocytes were stimulated with PMA/Ionomycin for 6hrs (C) and the induced intracellular cytokines were measured by flow cytometry. Total Th1 and Th2 antibodies were also measured in sera using an antibody ELISA (D). Results represent 3 independent experiments and plotted as the mean + SEM. Statistical analysis was performed using unpaired, two-tailed t-test analysis followed by FDR for multiple comparisons. (: p<0.05; : p<0.01; : p<0.001). g p g y p p Sm-free (blue) and Sm-infected (red) mice 16 weeks post infection. Splenocytes were prepared and stimulated with 1 μg/ml C and cytokines were measured in cytometric bead array and ELISpot assays respectively. Other splenocytes were stimulated w the induced intracellular cytokines were measured by flow cytometry. Total Th1 and Th2 antibodies were also measured in se Results represent 3 independent experiments and plotted as the mean + SEM. Statistical analysis was performed using unpair by FDR for multiple comparisons. (: p<0.05; : p<0.01; : p<0.001). https://doi.org/10.1371/journal.ppat.1007182.g001 https://doi.org/10.1371/journal.ppat.1007182.g001 SFU/106 splenocytes) ELISpot responses in uninfected mice compared to Sm-infected mice (IFN-γ: 1166 ± 132.2 SFU/106 splenocytes) and IL-2: 11.89 ± 5.951 SFU/106 splenocytes) (Fig 2A and 2B). Responses to the Env (CD8) peptide induced the highest number of IFN-γ secret- ing CD8+ and CD4+ T cells (553.3 ± 55.86 SFU/106 splenocytes) compared to other peptides in uninfected mice vaccinated with MVA+gp140. However, IL-2 SFU/106 cells were similar among different peptides stimulations. Vaccination after PZQ treatment had varying effects on the magnitudes of ELISpot responses. For DNA+MVA vaccine regimen, the cumulative magnitude of IFN-γ but not IL-2 SFU/106 cells was still significantly lower in treated mice compared with uninfected mice indi- cating partial restoration of responses to normal magnitudes (Fig 2A and 2B). In contrast, the cumulative magnitudes of both IFN-γ and IL-2 SFU/106 cells between PZQ-treated and vacci- nated mice and uninfected mice were similar for MVA+gp140 vaccine regimen, indicating res- toration to near normal SFU/106 cells (Fig 2A and 2B). Th1 cytokine levels were significantly reduced in Sm-infected mice. As shown in Fig 2C–2E, significantly higher levels of net cumulative IFN-γ (6523 ± 282.0 pg/ml); IL-2 (84.86 ± 0.3147 5 / 23 PLOS Pathogens \| https://doi.org/10.1371/journal.ppat.1007182 July 26, 2018 PLOS Pathogens \| https://doi.org/10.1371/journal.ppat.1007182 July 26, 2018 Impact of chronic schistosomiasis on HIV vaccination infected vaccinated mice whilst they were readily detected at similar levels in both uninfected and PZQ-treated vaccinated mice indicating restoration of cytokine responses by PZQ treat- ment (Fig 2F). However, Pol- and Env- specific CD8+ T cells were detected in Sm-infected mice at similar levels as the uninfected and PZQ-treated vaccine group except in the MVA+ gp140 vaccine regimen where a significantly higher percentage of cumulative cytokine-produc- ing CD8+ T cells in response to the Pol and Env CD8 peptides stimulation was observed in uninfected vaccinated mice (1.79 ± 0.04%) compared to Sm-infected vaccinated (1.49 ± 0.06%) mice. (Fig 2G). Most of the cytokine producing CD8+ and CD4+ T cells belonged to the effector memory phenotype (Fig 2H) and the profiles of the memory phenotypes were similar in both uninfected and PZQ-treated vaccinated groups. Env-specific antibody responses to SAAVI MVA-C+gp140 Env vaccine regimen are attenuated by chronic helminth infection despite prior antihelminthic treatment To determine the effect of helminth infection of the development of Env-specific antibody responses, mice were infected with Sm and vaccinated with MVA+gp140 and humoral responses were determined by ELISA. Uninfected and vaccinated mice produced higher amounts of gp140-specific IgG antibodies compared to Sm-infected vaccinated mice across all IgG isotypes (IgG1 [1681 ± 373.9 vs 140.8 ± 29.42 AUs]; IgG2a [4746 ± 1154 vs 71.14 ± 15.98 AUs]; IgG2b [2247 ± 553.9 vs 45.23 ± 12.65 AUs]). Treatment of infected mice with PZQ did not restore vaccine-specific antibody responses in infected mice as indicate by significantly lower titers of gp140-specific IgG antibodies across all IgG isotypes (IgG1 [287.0 ± 79.96 AUs], IgG2a [866.1 ± 514.3 AUs], IgG2b [126.3 ± 28.41 AUs]) compared to uninfected vaccinated control mice (Fig 3). PLOS Pathogens \| https://doi.org/10.1371/journal.ppat.1007182 July 26, 2018 After treatment with PZQ, the levels of IFN-γ and IL-2 were observed to be significantly higher compared to Sm-infected mice for both DNA+MVA and MVA+gp140 vaccine regimens but noticeably lower than those of unin- fected vaccinated mice, indicating only partial restoration to normal magnitudes. Furthermore, the frequencies of vaccine-specific cytokine (IFN-γ, IL-2 and TNF-α) producing T cells as deter- mined by flow cytometry showed a similar general trend whereby lower levels of HIV-specific T cells were detected in Sm-infected mice compared with uninfected animals (Fig 2F and 2G). For both vaccine regimens, Pol- and Env- specific CD4+ T cells were undetectable in Sm- pg/ml) and TNF-α (251.2 ± 30.33 pg/ml) (Fig 2C–2E) were released by splenocytes from unin- fected mice in the DNA+MVA vaccine regimen compared to lower levels of IFN-γ (1899 ± 244.6 pg/ml); IL-2 (23.55 ± 4.094 pg/ml) and TNF-α (122.9 ± 17.45 pg/ml) released in Sm- infected vaccinated mice (Fig 2C–2E). Similarly, for the MVA+gp140 vaccine regimen, signifi- cantly higher levels of net cumulative Th1 cytokines: IFN-γ (2416 pg/ml); IL-2 (63.79 pg/ml) and lower TNF-α (112.48 pg/ml) were released from splenocytes of uninfected vaccinated mice compared to lower levels of IFN-γ (789 pg/ml); IL-2 (4.0 pg/ml) and higher TNF-α (123.87 pg/ ml) released in Sm-infected vaccinated mice (Fig 2C–2E). After treatment with PZQ, the levels of IFN-γ and IL-2 were observed to be significantly higher compared to Sm-infected mice for both DNA+MVA and MVA+gp140 vaccine regimens but noticeably lower than those of unin- fected vaccinated mice, indicating only partial restoration to normal magnitudes. Furthermore, the frequencies of vaccine-specific cytokine (IFN-γ, IL-2 and TNF-α) producing T cells as deter- mined by flow cytometry showed a similar general trend whereby lower levels of HIV-specific T cells were detected in Sm-infected mice compared with uninfected animals (Fig 2F and 2G). For both vaccine regimens, Pol- and Env- specific CD4+ T cells were undetectable in Sm- PLOS Pathogens \| https://doi.org/10.1371/journal.ppat.1007182 July 26, 2018 6 / 23 Schistosoma mansoni eggs (SmE) attenuates Env-specific responses and partially suppresses T cell responses to SAAVI MVA-C+gp140 Env vaccine regimen Values are plotted and expressed as mean antibody units (AUs) ± SEM for the 8–12 animals in each group. Statistical differences between the groups were calculated by unpaired t-test (two- tailed). (: p<0.05; : p<0.01; : p<0.001). Fig 3. Schistosomamansoni infection significantly reduces HIV-1 Env gp140-specific antibodies. Blood for preparation of serum was collected two weeks after the last vaccination. HIV-1 Env gp140-specific IgG1, IgG2a and IgG2b antibodies were analysed by ELISA. Values are plotted and expressed as mean antibody units (AUs) ± SEM for the 8–12 animals in each group. Statistical differences between the groups were calculated by unpaired t-test (two- tailed). (: p<0.05; : p<0.01; : p<0.001). https://doi.org/10.1371/journal.ppat.1007182.g003 SEA resulted in a trend towards reduced IFN-γ:IL-4 ratio for SmE-inoculated mice compared to uninfected but vaccinated mice (S2 Fig). SEA resulted in a trend towards reduced IFN-γ:IL-4 ratio for SmE-inoculated mice compared to uninfected but vaccinated mice (S2 Fig). Schistosoma mansoni eggs (SmE) attenuates Env-specific responses and partially suppresses T cell responses to SAAVI MVA-C+gp140 Env vaccine regimen We next sought to investigate whether Sm eggs (SmE) alone are capable of attenuating HIV vaccine-specific responses in the absence of an active Sm infection. To achieve this, we sensi- tized mice with 2 500 SmE intraperitoneally, challenged them with 2 500 SmE intravenously 14 days later and vaccinated them with the MVA+gp140 vaccine regimen. Cumulative cellular responses to the HIV peptides were measured in the spleens using CBA and ELISpot (Fig 4A–4E) and gp140 Env-specific antibodies in the sera (Fig 4F). Cumulative HIV-1 IFN-γ SFU/106 (Fig 4A), and IL-2 SFU/106 (Fig 4B) in SmE-inoculated vaccinated mice were noticeably lower, but not significantly when compared to SmE-free vaccinated mice. Sim- ilarly, levels of cumulative IFN-γ; TNF-α and IL-2 secreted by splenocytes were noticeably lower in SmE-inoculated vaccinated mice compared to those secreted by splenocytes from SmE-free vaccinated mice (Fig 4C–4E respectively). However, SmE-inoculated mice had sig- nificantly lower amounts of gp140-specific IgG1 (656.8 ± 177.1 versus 1203 ± 152.0 AUs), IgG2a (71.14 ± 15.98 versus 238.1 ± 34.33 AUs), and IgG2b antibodies (82.73 ± 18.20 versus 218.3 ± 41.86 AUs) compared to SmE-free mice (Fig 4F), indicating broad attenuation of gp140 Env-specific antibody responses. Furthermore, we confirmed that SmE alone, in the absence of active infection, is capable of skewing the Th1/Th2 profile towards a Th2 response. As shown in Fig 4G, at 9 weeks post inoculation, the IFN-γ:IL-4 ratio was significantly lower in vaccinated SmE-inoculated (367.0 ± 38.34 pg/ml) compared to SmE-free (597.2 ± 72.93 pg/ ml) vaccinated mice after stimulation with Con A. Similarly, stimulation of splenocytes with PLOS Pathogens \| https://doi.org/10.1371/journal.ppat.1007182 July 26, 2018 7 / 23 Impact of chronic schistosomiasis on HIV vaccination Fig 3. Schistosomamansoni infection significantly reduces HIV-1 Env gp140-specific antibodies. Blood for preparation of serum was collected two weeks after the last vaccination. HIV-1 Env gp140-specific IgG1, IgG2a and IgG2b antibodies were analysed by ELISA. Values are plotted and expressed as mean antibody units (AUs) ± SEM for the 8–12 animals in each group. Statistical differences between the groups were calculated by unpaired t-test (two- tailed). (: p<0.05; : p<0.01; : p<0.001). https://doi.org/10.1371/journal.ppat.1007182.g003 Fig 3. Schistosomamansoni infection significantly reduces HIV-1 Env gp140-specific antibodies. Blood for preparation of serum was collected two weeks after the last vaccination. HIV-1 Env gp140-specific IgG1, IgG2a and IgG2b antibodies were analysed by ELISA. PLOS Pathogens \| https://doi.org/10.1371/journal.ppat.1007182 July 26, 2018 Vaccination with DNA+MVA vaccine regimen worsens the Sm-associated pathology in mice We investigated whether vaccination with DNA+MVA exacerbates helminth associated pathology in chronically infected mice by determining granuloma sizes and hydroxyproline content in mouse livers and assessing hepatosplenomegaly. We also investigated whether treat- ment with PZQ prior to vaccination with the DNA+MVA regimen ameliorates tissue pathol- ogy in infected mice. Sm-infected and vaccinated mice developed significantly larger (60.27 ± 2.37 mm2) granulomas when compared to all the other groups (vaccinated Sm- infected-PZQ treated [41.29 ± 1.66 mm2]; Sm-infected alone [40.79 ± 2.38 mm2] and Sm- infected-PZQ treated vaccinated [42.36 ± 2.26 mm2]) (Fig 5A and 5B). Sm-infected mice that were either vaccinated or unvaccinated developed hepatosplenomegaly as indicated by larger spleens and livers compared to vaccinated infected mice, naïve mice and vaccinated Sm- infected-PZQ treated mice (Fig 5C and 5D). No difference in hydroxyproline content was observed between unvaccinated Sm-infected and vaccinated Sm-infected mice (Fig 5E). Sur- prisingly, high levels of hydroxyproline content were observed in unvaccinated Sm-infected- PZQ treated mice compared to Sm-infected (Fig 5E). However, Sm-infected mice had signifi- cantly higher number of eggs per gram of liver compared to unvaccinated and vaccinated Sm- infected mice that were treated with PZQ (Fig 5F). PLOS Pathogens \| https://doi.org/10.1371/journal.ppat.1007182 July 26, 2018 8 / 23 Impact of chronic schistosomiasis on HIV vaccination Fig 4. Cellular and antibody responses to HIV vaccines in SmE inoculated mice. Spleens and blood were collected 12 days after the last vaccination. Splenocytes were stimulated the induced cytokines were analysed in an IFN-γ (A) and IL-2 (B) ELISpot, cytokine bead array (CBA) (C, D, and E) assays. The individual bars represent the magnitudes of the cumulative cytokine levels. HIV-1 Env gp140-specific IgG1, IgG2a and IgG2b antibodies were analysed in the sera by an ELISA (F). Values are expressed as antibody units and the mean (AUs) for the 8–12 animals in each group shown as a horizontal bar. An IFN-γ:IL-4 ratio (G) was Fig 4 Cellular and antibody responses to HIV vaccines in SmE inoculated mice Spleens and blood were collected 12 days after Fig 4. Cellular and antibody responses to HIV vaccines in SmE inoculated mice. Spleens and blood were collected 12 days after the last vaccination. Splenocytes were stimulated the induced cytokines were analysed in an IFN-γ (A) and IL-2 (B) ELISpot, cytokine bead array (CBA) (C, D, and E) assays. The individual bars represent the magnitudes of the cumulative cytokine levels. Impact of chronic schistosomiasis on HIV vaccination calculated from the CBA data. Results represent 3 independent experiments. Statistical analysis was performed using unpaired, two-tailed t-test analysis. (: p<0.05; : p<0.01). calculated from the CBA data. Results represent 3 independent experiments. Statistical analysis was performed using unpaired, two-tailed t-test analysis. (: p<0.05; : p<0.01). https://doi.org/10.1371/journal.ppat.1007182.g004 Discussion The present study investigated the impact of chronic schistosomiasis on the induction of T cell-mediated and antibody responses to candidate HIV vaccines in a mouse model, whether attenuation of vaccine responses can be reversed by pre-vaccination anti-helminthic treat- ment, and if vaccination with a T cell-based candidate vaccine has an adverse effect on hel- minth-associated pathology. Firstly, we sought to establish that the mouse model of chronic schistosomiasis worked well as to be expected on our hands. As we expected, our data confirmed that prior to vaccination, Sm-infected mice elicited predominantly Th2 responses and a decreased Th1 cytokine profile (Fig 1A and 1B; S1 Fig), had impaired Th1 cytokine-producing CD8+ and CD4+ T cells (Fig 1C) and an increase in Th2 total antibodies in serum (Fig 1D) as well as enlarged spleens and livers (Fig 5C and 5D) compared to uninfected mice. These findings are consistent with Fig 5. Analysis of livers and spleens of mice after vaccination with a DNA+MVA vaccine regimen. Groups of mice were infected with 30–35 Sm cercariae. Some groups were treated twice with PZQ after 8.5 weeks post infection. Vaccination with a DNA+MVA vaccine regimen was performed as shown in Table 1. Spleens and livers were harvested 12 days after last vaccination, weighed and prepared for various analyses. (A): Representative histological micrographs showing liver granulomas after staining with either H&E or CAB; (B): Granuloma sizes; (C) Liver weights; (D): Spleen weights; (E): hydroxyproline content and (F): Number of eggs per gram of liver. For Figs B-D, the results represent 3 independent experiments. For Figs E and F, the data is representative of 2 independent experiments. Means are shown as horizontal bars. Statistical analysis was performed using unpaired, two-tailed t-test analysis followed by FDR for multiple comparisons. (: p>0.05; : p<0.01; : p<0.001). https://doi.org/10.1371/journal.ppat.1007182.g005 Fig 5. Analysis of livers and spleens of mice after vaccination with a DNA+MVA vaccine regimen. Groups of mice were infected with 30–35 Sm cercariae. Some groups were treated twice with PZQ after 8.5 weeks post infection. Vaccination with a DNA+MVA vaccine regimen was performed as shown in Table 1. Spleens and livers were harvested 12 days after last vaccination, weighed and prepared for various analyses. (A): Representative histological micrographs showing liver granulomas after staining with either H&E or CAB; (B): Granuloma sizes; (C) Liver weights; (D): Spleen weights; (E): hydroxyproline content and (F): Number of eggs per gram of liver. Vaccination with DNA+MVA vaccine regimen worsens the Sm-associated pathology in mice HIV-1 Env gp140-specific IgG1, IgG2a and IgG2b antibodies were analysed in the sera by an ELISA (F). Values are expressed as antibody units and the mean (AUs) for the 8–12 animals in each group shown as a horizontal bar. An IFN-γ:IL-4 ratio (G) was PLOS Pathogens \| https://doi.org/10.1371/journal.ppat.1007182 July 26, 2018 9 / 23 Impact of chronic schistosomiasis on HIV vaccination previous reports that demonstrate that Sm infection and a host of other helminths skews the host’s immune responses from a Th1 towards a Th2 type with egg deposition and increased production of IL-4 as key driving forces [23, 26, 27, 63, 68–72]. Our data also agrees with previ- ous studies that have demonstrated that IL-10 is also responsible for down-regulation of Th1 responses that is observed in schistosome infections [73, 74]. IL-10 has been shown to mediate this down-regulation via an activation-induced cell death process resulting in apoptosis of CD4+ and CD8+ T cells which is also linked to the onset of egg-laying by the helminth parasite and formation of granulomas [75–77]. Our data also suggested a correlation between the decrease of cytokine-producing CD8+ T cells and levels of IgG2a and IgG2b antibodies observed in unvaccinated Sm and vaccinated Sm-infected mice. Previous studies have reported that cytokines produced by CD4+ and CD8+ T cells play an important role in the regulation of the humoral immune response and isotype switching [78–80]. The immunological conse- quence of a predominant Th2 biasing demonstrated in Sm-infected mice agrees with the con- cept of reciprocal antagonism between Th1 and Th2 as previously suggested [62, 81]. Secretion of IFN-γ and IL-2 by T cells has been associated with suppression of viral replica- tion in HIV-infected individuals and better proliferation of HIV-1-specific CD4+ and CD8+ T cells, suggesting that production of these Th1 cytokines by candidate HIV vaccines is a good indicator of vaccine-mediated immune protection [82] and good indications of polyfunctional CD4+ T cell responses [83]. In this study, we observed that the presence of Sm infection pre- vented optimal generation of vaccine-specific T cell responses following immunization with SAAVI vaccine candidates (Fig 2). As shown in Fig 2C–2E, vaccine-specific Th1 cumulative cytokine levels (IFN-γ, IL-2 and TNF-α) in recall responses to HIV peptides were significantly reduced in vaccinated Sm-infected compared to Sm-free mice vaccinated with DNA+MVA and MVA+gp140, with exception to TNF-α in MVA+gp140 vaccinated groups which were similar in both vaccinated Sm-infected and Sm-uninfected but vaccinated mice. Similarly, the magnitudes of vaccine-specific cumulative IFN-γ and IL-2 T cell responses measured by ELI- Spot were significantly lower in Sm-infected vaccinated mice compared to Sm-uninfected mice vaccinated with DNA+MVA and MVA+gp140 (Fig 2A and 2B). PLOS Pathogens \| https://doi.org/10.1371/journal.ppat.1007182 July 26, 2018 Discussion For Figs B-D, the results represent 3 independent experiments. For Figs E and F, the data is representative of 2 independent experiments. Means are shown as horizontal bars. Statistical analysis was performed using unpaired, two-tailed t-test analysis followed by FDR for multiple comparisons. (: p>0.05; : p<0.01; : p<0.001). https://doi.org/10.1371/journal.ppat.1007182.g005 https://doi.org/10.1371/journal.ppat.1007182.g005 PLOS Pathogens \| https://doi.org/10.1371/journal.ppat.1007182 July 26, 2018 10 / 23 Impact of chronic schistosomiasis on HIV vaccination reports [89, 90]. Treatment with PZQ has been shown to eliminate adult worms with no direct impact on the SmE already trapped in the tissues other than preventing continued egg deposi- tion in treated subjects [38–40] and potentially restoring normal T cell immune responsive- ness. However, our immunological data showed that treating mice prior to vaccination only partially restored the hosts’ vaccine-specific T cells responses (Fig 2). Surprisingly, the partial recovery of these responses did not translate in reduction of the magnitudes of Th2 cytokine responses (S1 Fig). Anti-inflammatory cytokines such as IL-10 remained elevated despite treat- ment with antihelminth (S1C and S1I Fig) whilst previous studies in which PZQ was used reported similar findings [58, 59]. However, it was unclear if the antibody responses were affected. In our study, Th1-type gp140-Env-specific antibody responses in Sm-infected mice were significantly lower despite treatment with PZQ (Fig 3). To our knowledge, no study has evaluated the ability of antihelminthic treatment in the restoration of antibody responses to HIV vaccines. However, it is has been suggested that the duration of infection prior and post treatment is an important factor which determines subsequent restoration of normal responses to vaccination [53, 89, 90]. A study by Chen et al., showed a recovery of immune balance 16 weeks post-treatment [53]. Also, findings from the studies conducted by Da’dara’s group and Shollenberger’s groups, demonstrated that normal immune responses can be achieved 2–10 weeks post-treatment [58, 59]. In contrast to our study, only a 1.5-week post-treatment period was allowed prior to commencing the vaccinations. Future studies should investigate varying post-treatment periods including multiple vaccinations to establish the optimal recovery dura- tions to start vaccinations after antihelminthic treatment. This is particularly relevant if there will arise a need to integrate future HIV vaccinations with helminthic worms control pro- grammes to improve the vaccination outcomes in helminth-endemic areas. This study went further to demonstrate that Th1 cellular responses elicited by DNA- and MVA- vectored HIV-1 vaccines exacerbated helminth-induced pathology. Sm-infected mice vaccinated with a DNA+MVA regimen had significantly larger granulomas as well as enlarged spleens and livers compared to Sm-infected unvaccinated groups (Fig 5B and 5C). Treatment significantly reduced the pathology; however, a considerable number of eggs were still present in the liver tissues of treated mice (Fig 5F). PLOS Pathogens \| https://doi.org/10.1371/journal.ppat.1007182 July 26, 2018 A similar decline has been reported in Sm-infected mice compared to uninfected controls following immunization with a DNA-vectored HIV-1 vaccine [57]. Also, the frequencies of vaccine-specific cytokine- producing CD4+ and CD8+ T cells in Sm-infected mice were observed in the current study (Fig 2F and 2G), which may translate to a decrease in antibody population [84]. However, it was noted that the cytokine-producing T cells were predominantly of the effector memory phenotype in both Sm-infected and uninfected vaccinated mice (Fig 2H). Vaccine-induced effector memory T cells have been associated with protection against mucosal SIV challenge in vaccinated rhesus monkeys [85]. In this study, the downregulation of these vaccine-specific cellular responses demonstrates the ability of Sm infection to negatively affect protective potential of candidate HIV-1 vaccines as have suggested by others [57, 59]. The antibody responses to the gp140-Env protein were significantly impaired in Sm- infected vaccinated mice (Fig 3). The mean concentration of anti-gp140 antibodies in Sm- infected mice vaccinated with MVA+gp140 was significantly lower than that in Sm-free vacci- nated control for all IgG isotypes (IgG1; IgG2 and IgG2b). Antibody responses to specific HIV antigens have been proposed to correlate with protection [82, 86]; thus, this was an interesting finding with far-reaching implications for future vaccine development. Elimination of helminth parasites with an antihelminth drug prior to immunization was expected to restore normal vaccine T cell responsiveness as previously demonstrated [59, 87, 88]. Our results show that treating mice with PZQ reversed tissue pathology as indicated by reduced spleen and liver weights sizes of granulomas and SmE deposition in the liver tissue in PZQ-treated mice compared with untreated mice (Fig 5B–5D and 5F) is consistent with earlier PLOS Pathogens \| https://doi.org/10.1371/journal.ppat.1007182 July 26, 2018 11 / 23 Impact of chronic schistosomiasis on HIV vaccination specific cellular responses observed in SmE-inoculated mice was not significant. Surprisingly as with Sm live infection, antibody responses to HIV Env-gp140 were significantly reduced in the presence of SmE (Fig 4F). As suggested previously [72], this finding confirms that SmE trapped in the tissues play a critical role in attenuating the host’s vaccine-specific responses in Sm-infected individual and may explain why both cellular and antibody responses are still sup- pressed despite treatment in PZQ-treated groups. Thus, the possible mechanism by which Sm infection suppresses these HIV-specific cellular and humoral responses appear to involve the deposition of SmE in the tissues, which stimulates increased production of IL-4 and IL-10 with concomitant polarization of Th2 immune responses. This in turn may promote activation- induced apoptosis of HIV-specific CD4+ and CD8+ T cells resulting in attenuated induction of Th1 immune responses which are key components of HIV vaccine-specific responses. This finding further highlights another challenge that even after antihelminth treatment with PZQ, generation of optimal vaccine responses may not be achieved as helminth eggs left trapped in the tissues could still attenuate HIV vaccine-induced immune responses. In light of these find- ings, this study suggests that, whilst elimination of worms can offer an affordable and a simple means of antihelminthic treatment, only partially restoration of immune responsiveness to T cell-based vaccines for HIV-1 and other infectious diseases in helminth endemic settings may be achieved. Thus, it would be important to evaluate vaccine delivery systems that can poten- tially overcome the negative impact of concurrent helminthiasis as previously suggested [93]. An alternative avenue would be the discovery of antihelminthic drugs which are effective in elimination of SmE from the host’s tissues in addition to the elimination of the parasitic worms. Although, this study gives further information on the impact of helminth infection on the immunogenicity of HIV vaccines, not all immunological aspects could be elucidated. Thus, this study justifies further investigations with use of a nonhuman primate model such as baboons (immune system is highly similar to humans) to obtain a better understanding of these immune responses. The present study demonstrated that chronic helminth infection is associated with Th2-dri- ven attenuation of both T cell and antibody response to HIV vaccines, and elimination of worm by chemotherapy partially restored T cell responses but not necessarily antibody responses. This study further demonstrated that vaccinating helminth-infected individuals with HIV vaccines that induce strong cellular responses may increase the pathology induced by the parasites, rendering the vaccine unsafe in helminth endemic areas. Lastly, this study suggests that the often-suggested integration of antihelminthic treatment programme with a successful future HIV vaccine might not result in improved vaccination outcome unless alter- native antihelminthic drugs with a capacity to eliminate schistosome eggs from tissues are developed. In addition, we recommend that HIV vaccine development programs should con- sider designing vaccines that can overcome the adverse effects of helminth-induced immunity. Surprisingly, the amount of hydroxyproline con- tent, which is a measure of collagen content, was significantly higher in PZQ-treated unin- fected mice compared with unvaccinated Sm-infected groups (Fig 5E), suggesting that PZQ treatment may contribute to increased fibrosis of the liver (Fig 5E) as an adverse side effect. A recent study showed that a novel experimental drug (Paeoniflorin) used for treating schistoso- miasis managed to control sclerosis better than PZQ [91], pointing to a possible future replace- ment of PZQ as the antihelminthic drug of choice. Nevertheless, PZQ treatment resulted in reduced number of eggs per gram of liver tissue when compared with Sm-infected untreated mice (Fig 5F). These findings highlighted the scientific challenges in the development of HIV vaccines for SSA, where parasitic helminthiasis is endemic. As discussed above, this study found lack of restoration of vaccine-specific responses upon PZQ-treatment prior to vaccinations while a substantial level of SmE burden was observed in PZQ-treated mice several weeks post-treatment. We, therefore investigated if Sm eggs in the absence of a live infection could result in downregulation of HIV-specific responses. Following an established Sm-egg model [92], mice were inoculated with S. mansoni eggs and then vacci- nated with candidate HIV vaccines to evaluate how these eggs affect vaccination outcomes. The IFN-γ:IL-4 ratio for the SmE-sensitized vaccinated mice was significantly smaller than the unsensitized vaccinated mice (Fig 4G), indicating a considerable elevation of Th2 cytokines and down regulation of Th1 in SmE-inoculated mice comparable to SmE-unsensitized mice. However, this polarized Th2 immune responses appear to have had only partial effects on the vaccine-specific T cell responses (Fig 4A–4E). Although reduced, the decrease in vaccine- PLOS Pathogens \| https://doi.org/10.1371/journal.ppat.1007182 July 26, 2018 12 / 23 PLOS Pathogens \| https://doi.org/10.1371/journal.ppat.1007182 July 26, 2018 Impact of chronic schistosomiasis on HIV vaccination Vaccines I. SAAVI DNA-C2 (DNA): Composed of two DNA plasmids, expressing a human immuno- deficiency virus subtype C (HIV-1C) polyprotein comprising gag, reverse transcriptase, tat and nef (grttnC) and an HIV-1C truncated env (gp150CT) as previously described [65]. II. SAAVI MVA-C (MVA): Recombinant MVA expressing the same immunogens as the DNA vaccine as previously described [64]. Both DNA- and MVA-vectored HIV-1 vaccines have been shown to elicit strong T cell responses in mice [11], nonhuman primates [18, 67] and clinical trials [16]. III. HIV-1 gp140 Env protein (gp140 Env): The gp140 (TV-1) (HIV-1/Clade C) was pur- chased from Immune Technology, USA. The Env amino acid consequences were derived from a South African HIV-1 subtype C primary isolate, TV1 [94]. III. HIV-1 gp140 Env protein (gp140 Env): The gp140 (TV-1) (HIV-1/Clade C) was pur- chased from Immune Technology, USA. The Env amino acid consequences were derived from a South African HIV-1 subtype C primary isolate, TV1 [94]. Ethics statement Female BALB/c mice (6–8 weeks old) were purchased from South African Vaccine Producers (SAVP) (Johannesburg, South Africa), housed in an Animal Biosafety Level 2 facility at the Uni- versity of Cape Town and maintained in accordance with the South African National Guidelines for Use of Animals for Scientific Purposes (SANS Code 10386: 2008) which is also in line with EU Directive 2010/63/EU. Experimental protocols performed in this study were reviewed and approved by the Animal Ethics Committee of the University of Cape Town (UCT AEC: protocol number: 014/026) and performed by qualified personnel in compliance with the South African Veterinary Council regulations. A mixture of ketamine hydrochloride and xylazine was used to anaesthesise mice for all procedures that involved intramuscular or intravenous injections, infec- tion with live Schistosoma mansoni cercariae, collection of blood by cardiac puncture and prepa- ration for euthanasia. Euthanasia was done by cervical dislocation while the animals were under anaesthesia (induced with a mixture of ketamine and xylazine as describe above). Parasites and vaccines Biomphalaria glabrata snails (Strain NMRI, NR-21962), infected with Schistosoma mansoni (Strain NMRI) were provided by the Schistosome Research Reagent Resource Center (NIAID, NIH, USA) and maintained in our laboratory for generation of live S. mansoni (Sm) cercariae that were used in this study. S. mansoni eggs (SmE) were purchased from the Theodor Bilharz Research Institute (Schistosome Biological Supply Center, Egypt) and stored at -80˚C until use. The integrity and viability of the eggs were evaluated using a light microscope prior to use. PLOS Pathogens \| https://doi.org/10.1371/journal.ppat.1007182 July 26, 2018 13 / 23 Impact of chronic schistosomiasis on HIV vaccination to Sm (4 groups) were infected percutaneously via the abdomen with 35 live S. mansoni cercar- iae at the beginning of the experimentation. Those receiving antihelminthic treatment were given two doses of PZQ (Sigma Aldrich, USA) by oral gavage (500 mg/kg; diluted in water containing 2% Kolliphor EL [Sigma Aldrich, USA]) three days apart, between 8 and 8.5 weeks post infection. Animals were vaccinated twice, 4 weeks apart, starting at 10 weeks post infec- tion (Table 1). In Experiment 3 (Table 2), mice were allocated to 4 groups (5 mice per group). Two groups were inoculated twice with SmE (2500 eggs per mouse), 14 days apart, initially by intraperitoneal route, and subsequently by intravenous route. As in Experiments 1 and 2, mice were vaccinated twice, 4 weeks apart, starting at 1 week after the second inoculation with SmE (Table 2). Vaccinations with DNA-vectored (100μg DNA per mouse) and MVA-vectored (106 plaque forming units per mouse) HIV-1 vaccines were given intramuscularly. Vaccination with HIV- 1 gp140 Env (10μg protein per mouse formulated in Imject Alum adjuvant [Thermo Scientific, USA]) was administered subcutaneously. DNA+MVA vaccine regimens were given as DNA prime and MVA boost vaccine regimens whilst MVA+gp140 Env were given concurrently. Twelve days following the last vaccination, blood was collected by cardiac punctured, mice were euthanised and spleens and livers were harvested for evaluation of HIV immune responses and helminth-induced pathology. Immunogenicity assays to evaluate host immune response and SAAVI vaccines specific responses cellular immune response to HIV vaccines Splenocytes were prepared using a standard protocol [95] and stimulated with HIV peptides or mitogen stimulant at 2μg/ml (S1 Table). IFN-γ and IL-2 ELISpot assays were carried out as previously described [11]. Cytometric bead array (CBA) assays were carried out as previously described [96]. Intracellular cytokine staining (ICS) and flow cytometry analysis was per- formed as previously described [12, 14] with minor modifications. Briefly, cells were stained with a viability dye, violet amine reactive dye (ViViD; Invitrogen, USA), at a pre-determined optimal concentration before staining for cell surface molecules with the following fluoro- chrome-conjugated antibodies: anti-CD3-Alexa 700, anti-CD4-PE-Cy7, anti-αCD8-APC-Cy7, anti-CD62L-APC, and anti-CD44-FITC diluted to 0.2μg in staining buffer (BD Biosciences, USA). Further intracellular cytokine staining was done with pooled PE-conjugated anti-TNF (0.2μg) anti-IL-2 (0.06μg) and anti-IFN-γ (0.06μg) antibodies diluted in Perm/Wash buffer (BD Biosciences, USA). Infection, inoculation with SmE, PZQ treatment and HIV immunization The study comprised of three experiments. In Experiment 1 and 2 (Table 1), mice were ran- domly allocated to six groups (5–8 mice per group) per experiment. Mice receiving exposure Table 1. Immunasation schedule for live S. mansoni cercaria infection groups. Groups: n = 5–8 Infection PZQ 1 PZQ 2 Vaccine 1 Vaccine 2 Sampling Week 0 8 8.5 10 14 16 Experiment 1 MVA+Gp140 - - - ✓ ✓ ✓ Sm+MVA+Gp140 ✓ - - ✓ ✓ ✓ Sm+PZQ+MVA+Gp140 ✓ ✓ ✓ ✓ ✓ ✓ Sm+PZQ ✓ ✓ ✓ - - ✓ Sm ✓ - - - - ✓ Naïve - - - - - ✓ Experiment 2 DNA+MVA - - - ✓ ✓ ✓ Sm+DNA+MVA ✓ - - ✓ ✓ ✓ Sm+PZQ+DNA+MVA ✓ ✓ ✓ ✓ ✓ ✓ Sm+PZQ ✓ ✓ ✓ - - ✓ Sm ✓ - - - - ✓ Naïve - - - - - ✓ htt //d i /10 1371/j l t 1007182 t001 Table 1. Immunasation schedule for live S. mansoni cercaria infection groups. Groups: n = 5–8 Infection PZQ 1 PZQ 2 Vaccine 1 Vaccine 2 Sampling Week 0 8 8.5 10 14 16 Experiment 1 MVA+Gp140 - - - ✓ ✓ ✓ Sm+MVA+Gp140 ✓ - - ✓ ✓ ✓ Sm+PZQ+MVA+Gp140 ✓ ✓ ✓ ✓ ✓ ✓ Sm+PZQ ✓ ✓ ✓ - - ✓ Sm ✓ - - - - ✓ Naïve - - - - - ✓ Experiment 2 DNA+MVA - - - ✓ ✓ ✓ Sm+DNA+MVA ✓ - - ✓ ✓ ✓ Sm+PZQ+DNA+MVA ✓ ✓ ✓ ✓ ✓ ✓ Sm+PZQ ✓ ✓ ✓ - - ✓ Sm ✓ - - - - ✓ Naïve - - - - - ✓ htt //d i /10 1371/j l t 1007182 t001 Table 1. Immunasation schedule for live S. mansoni cercaria infection groups. PLOS Pathogens \| https://doi.org/10.1371/journal.ppat.1007182 July 26, 2018 PLOS Pathogens \| https://doi.org/10.1371/journal.ppat.1007182 July 26, 2018 1 14 / 23 Histology Spleens and livers were weighed prior to processing for immunological evaluation in the labo- ratory to determine if HIV vaccination worsens helminth-associated pathology. Livers were then fixed in 4% (v/v) buffered formalin solution. The fixed samples were then embedded in wax and processed. Sections (5–7μm) were stained with hematoxylin and eosin (H&E) (Sigma Aldrich, USA) to show aggregation of white blood cells around the Sm eggs and Chromotrop- aniline blue solution (CAB) (Sigma Aldrich, USA) and counterstained with Weigert’s hema- toxylin (Sigma Aldrich, USA) to stain for collagen. Micrographs of liver granuloma were cap- tured using a Nikon 90i wide-field microscope using a 5.0 megapixel colour digital camera running Nikon’s NIS-Elements v. 4.30 software (Nikon Instruments Inc., USA). The area of each granuloma containing a single egg was measured with the ImageJ 1.34 software (National Institutes of Health, USA). A total of 25–30 granulomas per slide per mouse were included in the analyses. Data was presented as a mean area of each granuloma containing a single egg. The number of eggs per gram of liver was determined by counting individual eggs from hydro- lysed liver under a microscope. Hydroxyproline assay Hydroxyproline content, which is a direct measure of collagen content in liver was determined using a modified hydroxyproline protocol by Bergman and Loxley [98]. Briefly, liver samples were weighed, hydrolyzed and added to a 40mg Dowex/Norit mixture. The supernatants were neutralised with 1% phenolphthalein and titrated against 10 M NaOH. An aliquot was mixed with isopropanol and added to chloramine-T/citrate buffer solution (pH 6.5). Erlich’s reagent (95% ethanol containing dimethylaminobenzaldehyde (DMAB) and concentrated hydrochlo- ric acid) was added and absorbance was read at 570 nm. Hydroxyproline levels were calculated using 4-hydroxy-L-proline (Sigma Aldrich, USA) as a standard, and results were expressed as μmoles hydroxyproline per weight of tissue that contained 104 eggs. Impact of chronic schistosomiasis on HIV vaccination was used as a negative control while a reference serum sample prepared from mice previously vaccinated with HIV gp140 protein was used in 12 two-fold dilutions, starting at 1:100, to gen- erate a standard curve. For detection, appropriate secondary anti-mouse antibodies conjugated with horseradish peroxidase were used including the three anti-mouse IgG isotypes (IgG1; IgG2a and IgG2b; Southern Biotechnology). After colour development using tetramethyl-benzidine substrate (TMB; KPL, USA), the optical density (OD) was measured at 450nm (with a reference filter set at 540 nm) using a microplate reader (Molecular Devices Corporation, USA). Based on the constants of the standard curve generated from the serially diluted reference sample, the recip- rocal dilution giving an OD value of 1 (against gp140) was assigned a value of 1000 antibody units (AUs). The negative control (unvaccinated mouse serum) was assigned a reciprocal dilu- tion of 0 and zero AUs. A reference sample was used on each ELISA plate to generate a stan- dard curve from which the assigned AUs were used to extrapolate for test samples at a fixed dilution of 1:1000. A cut-off value for positive antibody responses was set at 2 x the OD value of the negative control serum (unvaccinated) and those below the cut of value were assigned an antibody unit of zero. Antibody responses to HIV vaccines To measure the level of HIV gp140 Env-specific antibodies in mouse sera, a standardised ELISA assay was established as previously described [97]. Briefly, ELISA plates were coated with 0.5μg/ml of gp140 protein diluted in PBS and incubated overnight at 4˚C. Test mouse sera (diluted 1:1000) were tested in duplicates. Mouse sera obtained from unvaccinated mice Table 2. Immunisations schedule for S. mansoni egg challenge groups. Groups: n = 5 i.p. inoculation with 2500 SmE i.v. inoculation with 2500 SmE Vaccine 1 Vaccine 2 Sampling Week 0 2 3 7 9 Experiment 3 MVA+gp140 - - ✓ ✓ ✓ SmE+MVA+gp140 ✓ ✓ ✓ ✓ ✓ SmE ✓ ✓ - - ✓ Naïve - - - - ✓ https://doi.org/10.1371/journal.ppat.1007182.t002 PLOS Pathogens \| https://doi.org/10.1371/journal.ppat.1007182 July 26, 2018 15 / 23 PLOS Pathogens \| https://doi.org/10.1371/journal.ppat.1007182 July 26, 2018 Acknowledgments The authors thank Dr Justin Nono and Mr Rodney Lucas for technical assistance with the maintenance of S. mansoni parasites and animal protocols respectively. Impact of chronic schistosomiasis on HIV vaccination tests were two-tailed. p values <0.05 were considered as significant. The false discovery rate (FDR) with Benjamini-Hochberg-adjusted p<0.05 was performed as previously described [99]. Supporting information S1 Fig. Cytokine production after Con A and SEA stimulation in Sm-free and Sm-infected mice vaccinated with MVA+gp140 or DNA+MVA regimen. Splenocytes were harvested from mice vaccinated with the indicated regimen described in Fig 1. They were then stimu- lated with an irrelevant peptide (negative control), Con A or with SEA for 48 hours. Culture supernatants were collected and the level of Th1 and Th2 cytokines released into the medium for MVA+gp140 (A-C and D-F respectively) and DNA+MVA (G-I and J-L respectively) vacci- nated mice was measured using a cytokine bead array assay. The individual bars represent the magnitude of the net cytokine levels for vaccinated Sm-free (blue); vaccinated Sm-infected (red) and Sm-infected-PZQ treated (green) vaccinated mice. Results represent 3 independent experiments and plotted as the mean + SEM, and cytokine levels were expressed as pg/ml. Sta- tistical analysis was performed using unpaired, two-tailed t-test analysis followed by FDR for multiple comparisons. (: p>0.05; : p<0.01; : p<0.001). (TIF) S2 Fig. Th1/Th2 profile: The presence of S. mansoni Eggs (SmE) in the tissues tends to polarize the Th1/Th2 balance towards a Th2 profile. Splenocytes were obtained from SmE- sensitized and non-sensitized mice after two vaccinations with MVA-vectored HIV-1 and HIV-1 gp140 Env protein vaccines as shown in Table 2. They were then stimulated with an irrelevant peptide (negative control) or with SEA for 48 hours. Culture supernatants were col- lected and the level of Th1 and Th2 cytokines released into the medium was measured using a cytokine bead array assay. The individual bars represent the IFN-γ/IL-4 ratio for vaccinated non-sensitized (blue) and vaccinated SmE-sensitized (red) mice. Results represent 3 indepen- dent experiments and plotted as the mean + SEM. (TIF) PLOS Pathogens \| https://doi.org/10.1371/journal.ppat.1007182 July 26, 2018 Statistical analysis Statistical analysis was performed using Prism version 5.0 (GraphPad Software, USA). The t-test for independent unpaired non-parametric comparisons was applied to assess the level of significance between means ±SEM. Three independent experiments were conducted and all PLOS Pathogens \| https://doi.org/10.1371/journal.ppat.1007182 July 26, 2018 16 / 23 References 1. Means AR, Burns P, Sinclair D, Walson JL. Antihelminthics in helminth-endemic areas: effects on HIV disease progression. The Cochrane Library. 2016. 2. WHO. Schistosomiasis Fact Sheet. 2016 [updated January 2017; cited 2017 16 Decemebr 2016]. Avail- able from: http://www.who.int/mediacentre/factsheets/fs366/en/. 3. WHO UNAIDS. AIDS epidemic update: December 2009. WHO Regional Office Europe, 2009 9291738328. 4. De Silva NR, Brooker S, Hotez PJ, Montresor A, Engels D, Savioli L. Soil-transmitted helminth infec- tions: updating the global picture. Trends Parasitol. 2003; 19(12):547–51. PMID: 14642761 5. UNICEF U, WHO. Global HIV/AIDS response: epidemic update and health sector progress towards uni- versal access: progress report 2011: World Health Organization; 2011. 6. Walson JL, Stewart BT, Sangare L, Mbogo LW, Otieno PA, Piper BK, et al. Prevalence and correlates of helminth co-infection in Kenyan HIV-1 infected adults. PLOS Neglected Tropical Disease. 2010; 4(3): e644. https://doi.org/10.1371/journal.pntd.0000644 PMID: 20361031; PubMed Central PMCID: PMCPMC2846937. 7. Bustinduy A, King C, Scott J, Appleton S, Sousa-Figueiredo JC, Betson M, et al. HIV and schistosomia- sis co-infection in African children. The Lancet Infectious Diseases. 2014; 14(7):640–9. https://doi.org/ 10.1016/S1473-3099(14)70001-5 PMID: 24747019 8. Haynes BF, McElrath MJ. Progress in HIV-1 vaccine development. Current opinion in HIV and AIDS. 2013; 8(4):326. https://doi.org/10.1097/COH.0b013e328361d178 PMID: 23743722 9. McMichael AJ, Koff WC. Vaccines that stimulate T cell immunity to HIV-1: the next step. Nat Immunol. 2014; 15(4):319–22. https://doi.org/10.1038/ni.2844 PMID: 24646598; PubMed Central PMCID: PMCPMC4324504. 10. Chege GK, Burgers W, Muller T, Shephard EG, Williamson C, Williamson A. A single dose of SAAVI MVA-C reboosts rhesus macaques after more than 3 years post DNA-MVA prime-boost vaccination. Retrovirology. 2012; 9(Suppl 2):P32. https://doi.org/10.1186/1742-4690-9-s2-p32 11. Shephard E, Burgers WA, Van Harmelen JH, Monroe JE, Greenhalgh T, Williamson C, et al. A multi- gene HIV type 1 subtype C modified vaccinia Ankara (MVA) vaccine efficiently boosts immune responses to a DNA vaccine in mice. AIDS Res Hum Retroviruses. 2008; 24(2):207–17. https://doi.org/ 10.1089/aid.2007.0206 PMID: 18240963. 12. Jongwe TI, Chapman R, Douglass N, Chetty S, Chege G, Williamson A-L. HIV-1 Subtype C Mosaic Gag Expressed by BCG and MVA Elicits Persistent Effector T Cell Responses in a Prime-Boost Regi- men in Mice. PloS one. 2016; 11(7):e0159141. https://doi.org/10.1371/journal.pone.0159141 PMID: 27427967 13. Chapman R, Stutz H, Jacobs W Jr., Shephard E, Williamson AL. Priming with recombinant auxotrophic BCG expressing HIV-1 Gag, RT and Gp120 and boosting with recombinant MVA induces a robust T cell response in mice. PLoS One. 2013; 8(8):e71601. Impact of chronic schistosomiasis on HIV vaccination Supervision: Hlumani Ndlovu, William G. C. Horsnell, Gerald K. Chege. Supervision: Hlumani Ndlovu, William G. C. Horsnell, Gerald K. Chege. Supervision: Hlumani Ndlovu, William G. C. Horsnell, Gerald K. Chege. Writing – original draft: Godfrey A Dzhivhuho Supervision: Hlumani Ndlovu, William G. C. Horsnell, Gerald K. Chege. Writing – review & editing: Hlumani Ndlovu, William G. C. Horsnell, Frank Brombacher, Anna-Lise Williamson, Gerald K. Chege. PLOS Pathogens \| https://doi.org/10.1371/journal.ppat.1007182 July 26, 2018 16. Churchyard G, Mlisana K, Karuna S, Williamson A-L, Williamson C, Morris L, et al. Sequential Immuni- zation with gp140 Boosts Immune Responses Primed by Modified Vaccinia Ankara or DNA in HIV- Author Contributions Conceptualization: Gerald K. Chege. Data curation: Godfrey A. Dzhivhuho, Samantha A. Rehrl. Investigation: Godfrey A. Dzhivhuho, Samantha A. Rehrl. Methodology: Godfrey A. Dzhivhuho, Samantha A. Rehrl. Project administration: Gerald K. Chege. Project administration: Gerald K. Chege. Resources: Frank Brombacher, Anna-Lise Williamson, Gerald K. Chege. Resources: Frank Brombacher, Anna-Lise Williamson, Gerald K. Chege. 17 / 23 PLOS Pathogens \| https://doi.org/10.1371/journal.ppat.1007182 July 26, 2018 Impact of chronic schistosomiasis on HIV vaccination Uninfected South African Participants. PloS one. 2016; 11(9):e0161753. https://doi.org/10.1371/ journal.pone.0161753 PMID: 27583368 Uninfected South African Participants. PloS one. 2016; 11(9):e0161753. https://doi.org/10.1371/ journal.pone.0161753 PMID: 27583368 17. Chege GK, Montefiori DC, Barnett S, Gray CM, Burgers WA, Muller TL, et al. A Combined Saavi MVA- C and Novartis Gp140 Env Protein Vaccine Boost Elicits Robust HIV Cellular and Humoral Responses in a Nonhuman Primate Model. AIDS Res Hum Retroviruses. 2013; 29(11):A39–A. PubMed PMID: WOS:000326037500091. 18. Chege GK, Burgers WA, Mu¨ller TL, Gray CM, Shephard EG, Barnett SW, et al. DNA-MVA-protein vac- cination of rhesus macaques induces HIV-specific immunity in mucosal-associated lymph nodes and functional antibodies. Vaccine. 2017. 19. Excler J-L, Robb ML, Kim JH. Prospects for a globally effective HIV-1 vaccine. Vaccine. 2015; 33:D4– D12. https://doi.org/10.1016/j.vaccine.2015.03.059 PMID: 26100921 20. Mascola JR, Montefiori DC. The role of antibodies in HIV vaccines. Annu Rev Immunol. 2009; 28:413– 44. 21. Hams E, Aviello G, Fallon PG. The schistosoma granuloma: friend or foe? Front Immunol. 2013; 4:89. https://doi.org/10.3389/fimmu.2013.00089 PMID: 23596444; PubMed Central PMCID: PMCPMC3625856. 22. Lundy SK, Lukacs NW. Chronic schistosome infection leads to modulation of granuloma formation and systemic immune suppression. Frontiers in immunology. 2013;4. https://doi.org/10.3389/fimmu.2013. 00004 23. Maizels RM, Balic A, Gomez-Escobar N, Nair M, Taylor MD, Allen JE. Helminth parasites–masters of regulation. Immunol Rev. 2004; 201(1):89–116. 24. Maizels RM, Yazdanbakhsh M. Immune regulation by helminth parasites: cellular and molecular mech- anisms. Nat Rev Immunol. 2003; 3(9):733–44. https://doi.org/10.1038/nri1183 PMID: 12949497. 25. Dauby N, Goetghebuer T, Kollmann TR, Levy J, Marchant A. Uninfected but not unaffected: chronic maternal infections during pregnancy, fetal immunity, and susceptibility to postnatal infections. The Lan- cet infectious diseases. 2012; 12(4):330–40. https://doi.org/10.1016/S1473-3099(11)70341-3 PMID: 22364680 26. Moreau E, Chauvin A. Immunity against helminths: interactions with the host and the intercurrent infec- tions. BioMed Research International. 2010; 2010. 27. Yin J, Dai A, Arango T, Kasinathan RS, Greenberg RM, Boyer JD. IL-4 and IFN-gamma induced by human immunodeficiency virus vaccine in a schistosome infection model. Human vaccines & immu- notherapeutics. 2012; 8(11):1555–63. https://doi.org/10.4161/hv.22142 PMID: 23151453; PubMed Central PMCID: PMCPMC3601131. 28. Bourke CD, Maizels RM, Mutapi F. Acquired immune heterogeneity and its sources in human helminth infection. Parasitology. 2011; 138(2):139–59. https://doi.org/10.1017/S0031182010001216 PMID: 20946693; PubMed Central PMCID: PMCPMC3021922. 29. van Riet E, Hartgers FC, Yazdanbakhsh M. Chronic helminth infections induce immunomodulation: consequences and mechanisms. Immunobiology. 2007; 212(6):475–90. https://doi.org/10.1016/j. imbio.2007.03.009 PMID: 17544832. 30. PLOS Pathogens \| https://doi.org/10.1371/journal.ppat.1007182 July 26, 2018 References https://doi.org/10.1371/journal.pone.0071601 PMID: 23977084; PubMed Central PMCID: PMCPMC3748047. 14. Chapman R, Jongwe TI, Douglass N, Chege G, Williamson A-L. Heterologous prime-boost vaccination with DNA and MVA vaccines, expressing HIV-1 subtype C mosaic Gag virus-like particles, is highly immunogenic in mice. PloS one. 2017; 12(3):e0173352. https://doi.org/10.1371/journal.pone.0173352 PMID: 28278263 15. Gray GE, Mayer KH, Elizaga ML, Bekker LG, Allen M, Morris L, et al. Subtype C gp140 Vaccine Boosts Immune Responses Primed by the South African AIDS Vaccine Initiative DNA-C2 and MVA-C HIV Vac- cines after More than a 2-Year Gap. Clinical and vaccine immunology: CVI. 2016; 23(6):496–506. Epub 2016/04/22. https://doi.org/10.1128/CVI.00717-15 PMID: 27098021; PubMed Central PMCID: PMCPMC4895009. 16. Churchyard G, Mlisana K, Karuna S, Williamson A-L, Williamson C, Morris L, et al. Sequential Immuni- zation with gp140 Boosts Immune Responses Primed by Modified Vaccinia Ankara or DNA in HIV- PLOS Pathogens \| https://doi.org/10.1371/journal.ppat.1007182 July 26, 2018 18 / 23 Impact of chronic schistosomiasis on HIV vaccination 36. Obuku AE, Asiki G, Abaasa A, Ssonko I, Harari A, van Dam GJ, et al. Effect of Schistosoma mansoni Infection on Innate and HIV-1-Specific T-Cell Immune Responses in HIV-1-Infected Ugandan Fisher Folk. AIDS Res Hum Retroviruses. 2016; 32(7):668–75. https://doi.org/10.1089/AID.2015.0274 PMID: 26864743 37. Elias D, Britton S, Kassu A, Akuffo H. Chronic helminth infections may negatively influence immunity against tuberculosis and other diseases of public health importance. Expert review of anti-infective ther- apy. 2007; 5(3):475–84. Epub 2007/06/06. https://doi.org/10.1586/14787210.5.3.475 PMID: 17547511. 38. Shaheen AA, Ebeid FA, Fahim A. Effect of praziquantel on some aspects of carbohydrate metabolism in mice infected with Schistosoma mansoni. Pharmacol Res. 1989; 21(3):263–70. PMID: 2546139 39. Tallima H, El Ridi R. Praziquantel binds Schistosoma mansoni adult worm actin. Int J Antimicrob Agents. 2007; 29(5):570–5. https://doi.org/10.1016/j.ijantimicag.2006.12.018 PMID: 17341443 40. Utzinger J, Keiser J. Schistosomiasis and soil-transmitted helminthiasis: common drugs for treatment and control. Expert opinion on pharmacotherapy. 2004; 5(2):263–85. Epub 2004/03/05. https://doi.org/ 10.1517/14656566.5.2.263 PMID: 14996624. 41. Cañete R, Escobedo A, Almirall P, Gonza´lez M, Brito K, Cimerman S. Mebendazole in parasitic infec- tions other than those caused by soil-transmitted helminths. Trans R Soc Trop Med Hyg. 2009; 103 (5):437–42. https://doi.org/10.1016/j.trstmh.2008.11.029 PMID: 19195670 42. Churcher TS, Pion SD, Osei-Atweneboana MY, Prichard RK, Awadzi K, Boussinesq M, et al. Identifying sub-optimal responses to ivermectin in the treatment of River Blindness. Proceedings of the National Academy of Sciences. 2009; 106(39):16716–21. 43. Geary TG. Are new anthelmintics needed to eliminate human helminthiases? Curr Opin Infect Dis. 2012; 25(6):709–17. https://doi.org/10.1097/QCO.0b013e328359f04a PMID: 23041774 44. N’Goran E, Utzinger J, N’guessan A, Mu¨ller I, Zamble´ K, Lohourignon K, et al. Reinfection with Schisto- soma haematobium following school-based chemotherapy with praziquantel in four highly endemic vil- lages in Coˆte d’Ivoire. Tropical medicine & international health. 2001; 6(10):817–25. 45. Bobat S, Darby M, Mrdjen D, Cook C, Logan E, Auret J, et al. Natural and vaccine-mediated immunity to Salmonella Typhimurium is impaired by the helminth Nippostrongylus brasiliensis. PLOS Neglected Tropical Diseases. 2014; 8(12):e3341. https://doi.org/10.1371/journal.pntd.0003341 PMID: 25474738; PubMed Central PMCID: PMCPMC4256288. 46. Kilian H, Nielsen G. Cell-mediated and humoral immune responses to BCG and rubella vaccinations and to recall antigens in onchocerciasis patients. Tropical medicine and parasitology: official organ of Deutsche Tropenmedizinische Gesellschaft and of Deutsche Gesellschaft fur Technische Zusamme- narbeit (GTZ). 1989; 40(4):445–53. 47. Stewart G, Boussinesq M, Coulson T, Elson L, Nutman T, Bradley J. PLOS Pathogens \| https://doi.org/10.1371/journal.ppat.1007182 July 26, 2018 Elias D, Britton S, Aseffa A, Engers H, Akuffo H. Poor immunogenicity of BCG in helminth infected pop- ulation is associated with increased in vitro TGF-beta production. Vaccine. 2008; 26(31):3897–902. https://doi.org/10.1016/j.vaccine.2008.04.083 PMID: 18554755. 31. Elias D, Akuffo H, Pawlowski A, Haile M, Schon T, Britton S. Schistosoma mansoni infection reduces the protective efficacy of BCG vaccination against virulent Mycobacterium tuberculosis. Vaccine. 2005; 23(11):1326–34. Epub 2005/01/22. https://doi.org/10.1016/j.vaccine.2004.09.038 PMID: 15661380. 32. Kamal S, El Sayed Khalifa K. Immune modulation by helminthic infections: worms and viral infections. Parasite Immunol. 2006; 28(10):483–96. https://doi.org/10.1111/j.1365-3024.2006.00909.x PMID: 16965284 33. Wammes LJ, Hamid F, Wiria AE, de Gier B, Sartono E, Maizels RM, et al. Regulatory T cells in human geohelminth infection suppress immune responses to BCG and Plasmodium falciparum. Eur J Immu- nol. 2010; 40(2):437–42. https://doi.org/10.1002/eji.200939699 PMID: 20063313 34. Labeaud AD, Malhotra I, King MJ, King CL, King CH. Do antenatal parasite infections devalue childhood vaccination? PLoS neglected tropical diseases. 2009; 3(5):e442. Epub 2009/05/30. https://doi.org/10. 1371/journal.pntd.0000442 PMID: 19478847; PubMed Central PMCID: PMCPMC2682196. 35. Borkow G, Bentwich Z. Chronic immune activation associated with chronic helminthic and human immunodeficiency virus infections: role of hyporesponsiveness and anergy. Clin Microbiol Rev. 2004; 17(4):1012–30. https://doi.org/10.1128/CMR.17.4.1012-1030.2004 PMID: 15489359; PubMed Central PMCID: PMCPMC523563. PLOS Pathogens \| https://doi.org/10.1371/journal.ppat.1007182 July 26, 2018 19 / 23 Impact of chronic schistosomiasis on HIV vaccination vaccine CVD 103-HgR. J Infect Dis. 2000; 182(4):1199–206. https://doi.org/10.1086/315837 PMID: 10979918 vaccine CVD 103-HgR. J Infect Dis. 2000; 182(4):1199–206. https://doi.org/10.1086/315837 PMID: 10979918 56. Cooper PJ, Chico M, Sandoval C, Espinel I, Guevara A, Levine MM, et al. Human infection with Ascaris lumbricoides is associated with suppression of the interleukin-2 response to recombinant cholera toxin B subunit following vaccination with the live oral cholera vaccine CVD 103-HgR. Infect Immun. 2001; 69 (3):1574–80. Epub 2001/02/17. https://doi.org/10.1128/IAI.69.3.1574-1580.2001 PMID: 11179329; PubMed Central PMCID: PMCPMC98058. 57. Da’Dara AA, Lautsch N, Dudek T, Novitsky V, Lee TH, Essex M, et al. Helminth infection suppresses T- cell immune response to HIV-DNA-based vaccine in mice. Vaccine. 2006; 24(24):5211–9. https://doi. org/10.1016/j.vaccine.2006.03.078 PMID: 16675073. 58. Shollenberger LM, Bui CT, Paterson Y, Nyhoff L, Harn DA. HIV-1 vaccine-specific responses induced by Listeria vector vaccines are maintained in mice subsequently infected with a model helminth para- site, Schistosoma mansoni. Vaccine. 2013; 31(48):5651–8. https://doi.org/10.1016/j.vaccine.2013.09. 067 PMID: 24120546. 59. Da’dara AA, Harn DA. Elimination of helminth infection restores HIV-1C vaccine-specific T cell responses independent of helminth-induced IL-10. Vaccine. 2010; 28(5):1310–7. https://doi.org/10. 1016/j.vaccine.2009.11.019 PMID: 19941988; PubMed Central PMCID: PMCPMC2831209. 60. Wee EG, Ondondo B, Berglund P, Archer J, McMichael AJ, Baltimore D, et al. HIV-1 Conserved Mosa- ics Delivered by Regimens with Integration-Deficient DC-Targeting Lentiviral Vector Induce Robust T Cells. Mol Ther. 2017; 25(2). 61. Korber B, Hraber P, Wagh K, Hahn BH. Polyvalent vaccine approaches to combat HIV-1 diversity. Immunol Rev. 2017; 275(1):230–44. https://doi.org/10.1111/imr.12516 PMID: 28133800 62. Finkelman FD, Urban JF Jr. The other side of the coin: the protective role of the TH2 cytokines. The Journal of allergy and clinical immunology. 2001; 107(5):772–80. Epub 2001/05/10. https://doi.org/10. 1067/mai.2001.114989 PMID: 11344341. 63. Maizels RM, Hewitson JP, Smith KA. Susceptibility and immunity to helminth parasites. Curr Opin Immunol. 2012; 24(4):459–66. https://doi.org/10.1016/j.coi.2012.06.003 PMID: 22795966 64. Burgers WA, Shephard E, Monroe JE, Greenhalgh T, Binder A, Hurter E, et al. Construction, characteri- zation, and immunogenicity of a multigene modified vaccinia Ankara (MVA) vaccine based on HIV type 1 subtype C. AIDS Res Hum Retroviruses. 2008; 24(2):195–206. Epub 2008/02/05. https://doi.org/10. 1089/aid.2007.0205 PMID: 18240957. 65. Burgers WA, van Harmelen JH, Shephard E, Adams C, Mgwebi T, Bourn W, et al. Design and preclini- cal evaluation of a multigene human immunodeficiency virus type 1 subtype C DNA vaccine for clinical trial. J Gen Virol. 2006; 87(Pt 2):399–410. https://doi.org/10.1099/vir.0.81379-0 PMID: 16432028. 66. PLOS Pathogens \| https://doi.org/10.1371/journal.ppat.1007182 July 26, 2018 Onchocerciasis modulates the immune response to mycobacterial antigens. Clin Exp Immunol. 1999; 117(3):517. https://doi.org/10. 1046/j.1365-2249.1999.01015.x PMID: 10469056 48. Wajja A, Kizito D, Nassanga B, Nalwoga A, Kabagenyi J, Kimuda S, et al. The effect of current Schisto- soma mansoni infection on the immunogenicity of a candidate TB vaccine, MVA85A, in BCG-vacci- nated adolescents: An open-label trial. PLOS Neglected Tropical Disease. 2017; 11(5):e0005440. 49. Prost A, Schlumberger M, Fayet M. Response to tetanus immunization in onchocerciasis patients. Annals of Tropical Medicine & Parasitology. 1983; 77(1):83–5. 50. Cooper P, Espinel I, Paredes W, Guderian R, Nutman T. Impaired tetanus-specific cellular and humoral responses following tetanus vaccination in human onchocerciasis: a possible role for interleukin-10. J Infect Dis. 1998; 178(4):1133–8. PMID: 9806045 51. Sabin EA, Araujo MI, Carvalho EM, Pearce EJ. Impairment of tetanus toxoid-specific Thl-like immune responses in humans infected with Schistosoma mansoni. J Infect Dis. 1996; 173(1):269–72. PMID: 8537675 52. Haseeb M, Craig JP. Suppression of the immune response to diphtheria toxoid in murine schistosomia- sis. Vaccine. 1997; 15(1):45–50. PMID: 9041665 53. Chen L, Liu WQ, Lei JH, Guan F, Li MJ, Song WJ, et al. Chronic Schistosoma japonicum infection reduces immune response to vaccine against hepatitis B in mice. PLoS One. 2012; 7(12):e51512. Epub 2012/12/29. https://doi.org/10.1371/journal.pone.0051512 PMID: 23272112; PubMed Central PMCID: PMCPMC3522692. 54. Apiwattanakul N, Thomas PG, Iverson AR, McCullers JA. Chronic helminth infections impair pneumo- coccal vaccine responses. Vaccine. 2014; 32(42):5405–10. https://doi.org/10.1016/j.vaccine.2014.07. 107 PMID: 25131738 55. Cooper PJ, Chico ME, Losonsky G, Sandoval C, Espinel I, Sridhara R, et al. Albendazole treatment of children with ascariasis enhances the vibriocidal antibody response to the live attenuated oral cholera 20 / 23 PLOS Pathogens \| https://doi.org/10.1371/journal.ppat.1007182 July 26, 2018 Impact of chronic schistosomiasis on HIV vaccination 1-and T helper cell 2-type cytokine responses in vivo. The Journal of Immunology. 1997; 159(10):5014– 23. PMID: 9366429 1-and T helper cell 2-type cytokine responses in vivo. The Journal of Immunology. 1997; 159(10):5014– 23. PMID: 9366429 75. Estaquier J, Marguerite M, Sahuc F, Bessis N, Auriault C, Ameisen J. Interleukin-10-mediated T cell apoptosis during the T helper type 2 cytokine response in murine Schistosoma mansoni parasite infec- tion. European cytokine network. 1997; 8(2):153–60. PMID: 9262964 76. Fallon PG, Smith P, Dunne DW. Type 1 and type 2 cytokine-producing mouse CD4+ and CD8+ T cells in acute Schistosoma mansoni infection. Eur J Immunol. 1998; 28(4):1408–16. https://doi.org/10.1002/ (SICI)1521-4141(199804)28:04<1408::AID-IMMU1408>3.0.CO;2-H PMID: 9565381 77. Lundy SK, Lerman SP, Boros DL. Soluble egg antigen-stimulated T helper lymphocyte apoptosis and evidence for cell death mediated by FasL+ T and B cells during murine Schistosoma mansoni infection. Infect Immun. 2001; 69(1):271–80. https://doi.org/10.1128/IAI.69.1.271-280.2001 PMID: 11119515 78. Kemeny DM, Noble A, Holmes BJ, Diaz-Sanchez D. Immune regulation: a new role for the CD8+ T cell. Immunol Today. 1994; 15(3):107–10. https://doi.org/10.1016/0167-5699(94)90152-X PMID: 8172642 79. Burger E, Vaz CC, Sano A, Calich VL, Singer-Vermes LM, Xidieh CF, et al. Paracoccidioides brasilien- sis infection in nude mice: studies with isolates differing in virulence and definition of their T cell-depen- dent and T cell-independent components. The American journal of tropical medicine and hygiene. 1996; 55(4):391–8. PMID: 8916794 80. Snapper CM, Marcu KB, Zelazowski P. The immunoglobulin class switch: beyond “accessibility”. Immu- nity. 1997; 6(3):217–23. PMID: 9075922 81. Thomas PG, Harn DA. Immune biasing by helminth glycans. Cell Microbiol. 2004; 6(1):13–22. PMID: 14678327 82. Tomaras GD, Plotkin SA. Complex immune correlates of protection in HIV-1 vaccine efficacy trials. Immunol Rev. 2017; 275(1):245–61. https://doi.org/10.1111/imr.12514 PMID: 28133811 83. Lin L, Finak G, Ushey K, Seshadri C, Hawn TR, Frahm N, et al. COMPASS identifies T-cell subsets cor- related with clinical outcomes. Nat Biotechnol. 2015; 33(6):610–6. https://doi.org/10.1038/nbt.3187 PMID: 26006008 84. Swain SL, McKinstry KK, Strutt TM. Expanding roles for CD4+ T cells in immunity to viruses. Nature Reviews Immunology. 2012; 12(2):136–48. https://doi.org/10.1038/nri3152 PMID: 22266691 85. Hansen SG, Ford JC, Lewis MS, Ventura AB, Hughes CM, Coyne-Johnson L, et al. Profound early con- trol of highly pathogenic SIV by an effector memory T-cell vaccine. Nature. 2011; 473(7348):523–7. https://doi.org/10.1038/nature10003 PMID: 21562493 86. Haynes BF, Gilbert PB, McElrath MJ, Zolla-Pazner S, Tomaras GD, Alam SM, et al. PLOS Pathogens \| https://doi.org/10.1371/journal.ppat.1007182 July 26, 2018 Williamson AL, Rybiki E, Shephard E, Gray G, Bekker LG, Downing K, et al. South African HIV-1 vac- cine candidates—the journey from the bench to clinical trials. South African medical journal = Suid-Afri- kaanse tydskrif vir geneeskunde. 2012; 102(6):452–5. Epub 2012/06/07. PMID: 22668934. 67. Burgers WA, Chege GK, Muller TL, van Harmelen JH, Khoury G, Shephard EG, et al. Broad, high-mag- nitude and multifunctional CD4+ and CD8+ T-cell responses elicited by a DNA and modified vaccinia Ankara vaccine containing human immunodeficiency virus type 1 subtype C genes in baboons. J Gen Virol. 2009; 90(Pt 2):468–80. https://doi.org/10.1099/vir.0.004614-0 PMID: 19141458. 68. Maizels RM, McSorley HJ. Regulation of the host immune system by helminth parasites. J Allergy Clin Immunol. 2016; 138(3):666–75. https://doi.org/10.1016/j.jaci.2016.07.007 PMID: 27476889 69. Henderson G, Conary J, Summar M, McCurley T, Colley D. In vivo molecular analysis of lymphokines involved in the murine immune response during Schistosoma mansoni infection. I. IL-4 mRNA, not IL-2 mRNA, is abundant in the granulomatous livers, mesenteric lymph nodes, and spleens of infected mice. The Journal of Immunology. 1991; 147(3):992–7. PMID: 1907310 70. Pearce EJ, Caspar P, Grzych JM, Lewis FA, Sher A. Downregulation of Th1 cytokine production accompanies induction of Th2 responses by a parasitic helminth, Schistosoma mansoni. J Exp Med. 1991; 173(1):159–66. PMID: 1824635; PubMed Central PMCID: PMCPMC2118762. 71. Scott P, Pearce E, Cheever AW, Coffman RL, Sher A. Role of Cytokines and CD4+ T-Cell Subsets in the Regulation of Parasite Immunity and Disease. Immunol Rev. 1989; 112(1):161–82. 72. Kullberg MC, Pearce E, Hieny S, Sher A, Berzofsky J. Infection with Schistosoma mansoni alters Th1/ Th2 cytokine responses to a non-parasite antigen. J Immunol. 1992; 148(10):3264–70. PMID: 1533656 73. Sher A, Fiorentino D, Caspar P, Pearce E, Mosmann T. Production of IL-10 by CD4+ T lymphocytes correlates with down-regulation of Th1 cytokine synthesis in helminth infection. J Immunol. 1991; 147 (8):2713–6. PMID: 1680917. 74. Wynn TA, Morawetz R, Scharton-Kersten T, Hieny S, Morse H, Ku¨hn R, et al. Analysis of granuloma formation in double cytokine-deficient mice reveals a central role for IL-10 in polarizing both T helper cell PLOS Pathogens \| https://doi.org/10.1371/journal.ppat.1007182 July 26, 2018 21 / 23 PLOS Pathogens \| https://doi.org/10.1371/journal.ppat.1007182 July 26, 2018 Immune-correlates analysis of an HIV-1 vaccine efficacy trial. The New England journal of medicine. 2012; 366(14):1275– 86. Epub 2012/04/06. https://doi.org/10.1056/NEJMoa1113425 PMID: 22475592; PubMed Central PMCID: PMCPMC3371689. 87. Elias D, Wolday D, Akuffo H, Petros B, Bronner U, Britton S. Effect of deworming on human T cell responses to mycobacterial antigens in helminth-exposed individuals before and after bacille calmette– gue´rin (BCG) vaccination. Clinical & Experimental Immunology. 2001; 123(2):219–25. 88. Ghosh K, Wu W, Antoine AD, Bottazzi ME, Valenzuela JG, Hotez PJ, et al. The impact of concurrent and treated Ancylostoma ceylanicum hookworm infections on the immunogenicity of a recombinant hookworm vaccine in hamsters. J Infect Dis. 2006; 193(1):155–62. https://doi.org/10.1086/498528 PMID: 16323144 89. el-Fakahany AF, Abdalla KF, el-Hady HM, Abd el-Aziz SM, Afifi LM. The effect of praziquantel treatment on the liver functions, worm burden, and granuloma size using two drug regimen in murine Schistosoma mansoni infection. J Egypt Soc Parasitol. 1993; 23(3):877–86. PMID: 8308363. 90. Tawfik AF, Carter CE, Colley DG. Effects of anti-schistosomal chemotherapy on immune responses, protection and immunity. The American journal of tropical medicine and hygiene. 1986; 35(1):100–9. PMID: 3080912 91. Abd El-Aal NF, Hamza RS, Harb O. Paeoniflorin targets apoptosis and ameliorates fibrosis in murine schistosomiasis mansoni: A novel insight. Exp Parasitol. 2017; 183:23–32. Epub 2017/10/19. https:// doi.org/10.1016/j.exppara.2017.10.005 PMID: 29037782. 92. Joyce KL, Morgan W, Greenberg R, Nair MG. Using eggs from Schistosoma mansoni as an in vivo model of helminth-induced lung inflammation. JoVE (Journal of Visualized Experiments). 2012;(64): e3905–e. 93. Shollenberger LM, Bui C, Paterson Y, Allen K, Harn D. Successful vaccination of immune suppressed recipients using Listeria vector HIV-1 vaccines in helminth infected mice. Vaccine. 2013; 31(16):2050– 6. https://doi.org/10.1016/j.vaccine.2013.02.037 PMID: 23470236. 22 / 23 PLOS Pathogens \| https://doi.org/10.1371/journal.ppat.1007182 July 26, 2018 Impact of chronic schistosomiasis on HIV vaccination 94. Lian Y, Srivastava I, Go´mez-Roma´n VR, Zur Megede J, Sun Y, Kan E, et al. Evaluation of envelope vac- cines derived from the South African subtype C human immunodeficiency virus type 1 TV1 strain. J Virol. 2005; 79(21):13338–49. https://doi.org/10.1128/JVI.79.21.13338-13349.2005 PMID: 16227256 94. Lian Y, Srivastava I, Go´mez-Roma´n VR, Zur Megede J, Sun Y, Kan E, et al. Evaluation of envelope vac- cines derived from the South African subtype C human immunodeficiency virus type 1 TV1 strain. J Virol. 2005; 79(21):13338–49. https://doi.org/10.1128/JVI.79.21.13338-13349.2005 PMID: 16227256 95. Shen Y-J. An investigation into the Use of Lumpy Skin Disease Virus as a Vaccine Vector for a Potential HIV-1 vaccine 2010. 96. Chapman R, Shephard E, Stutz H, Douglass N, Sambandamurthy V, Garcia I, et al. Priming with a recombinant pantothenate auxotroph of Mycobacterium bovis BCG and boosting with MVA elicits HIV-1 Gag specific CD8+ T cells. PLoS One. 2012; 7(3):e32769. https://doi.org/10.1371/journal.pone. 0032769 PMID: 22479338; PubMed Central PMCID: PMCPMC3315557. 97. Cheung YB, Xu Y, Remarque EJ, Milligan P. Statistical estimation of antibody concentration using multi- ple dilutions. J Immunol Methods. 2015; 417:115–23. https://doi.org/10.1016/j.jim.2015.01.001 PMID: 25585264 98. Bergman I, Loxley R. Two improved and simplified methods for the spectrophotometric determination of hydroxyproline. Analytical Chemistry. 1963; 35(12):1961–5. 9. McDonald JH. Handbook of biological statistics: Sparky House Publishing Baltimore, MD; 2009. 99. McDonald JH. Handbook of biological statistics: Sparky House Publishing Baltimore, MD; 2009. 23 / 23
https://openalex.org/W4392732059	https://ejournal.itn.ac.id/index.php/jati/article/download/7756/4769	Indonesian	null	PENGEMBANGAN MEDIA PEMBELAJARAN INTERAKTIF PADA MATA PELAJARAN ADMINISTRASI INFRASTRUKTUR JARINGAN SMK NEGERI 2 PADANG	JATI	2,024	cc-by-sa	2,076	JATI (Jurnal Mahasiswa Teknik Informatika) JATI (Jurnal Mahasiswa Teknik Informatika) Vol. 7 No. 6, Desember 2023 PENGEMBANGAN MEDIA PEMBELAJARAN INTERAKTIF PADA MATA PELAJARAN ADMINISTRASI INFRASTRUKTUR JARINGAN SMK NEGERI 2 PADANG Widi Juliandes Putra, Thomson Mary, Anggri Yulio Pernanda Universitas PGRI Sumatera Barat Jl. Gn. Pangilun, Gn. Pangilun, Padang Utara Thomsonmary1980@gmail.com ABSTRAK Permasalahan yang ada pada mata pelajaran administrasi infrastruktur jaringan, materi mengevaluasi dan konfigurasi routing dinamis, media pembelajaran administrasi infrastruktur jaringan masih bersifat konfensional, akibat dari media pembelajaran yang kurang menarik maka berdampak pada hasil belajar siswa yang kurang memuaskan, belum adanya media pembelajaran interaktif yang dijadikan sebagai media pendukung proses pembelajaran. Penelitian ini bertujuan untuk menghasilkan media pembelajaran interaktif yang valid dan praktis. Penelitian ini menggunakan metode Research and Development (R&D) dengan model pengembangan ADDIE. Subjek penelitian ini berjumlah 32 orang siswa sebagai sampel. Teknik pengumpulan data menggunakan angket dengan skala likert. Hasil penelitian ini menunjukkan nilai rata-rata uji validitas ahli media sebesar 0,7 dengan kategori valid dan ahli materi sebesar 0,718 dengan kategori valid. Untuk uji praktikalitas terhadap respon guru mencapai 92,08% dengan kategori sangat praktis dan untuk respon siswa mencapai 89,62% dengan kategori sangat praktis. 1. PENDAHULUAN menggunakan media pembelajaran yang menampilkan teks, gambar, video, dan evaluasi pembelajaran, siswa dapat lebih mudah memahami materi yang disampaikan. Namun, masih banyak siswa yang belum dapat memahami materi yang disampaikan oleh guru karena kurangnya penggunaan media pembelajaran dalam proses pembelajaran. Selain itu, siswa yang tertinggal dalam pembelajaran juga disebabkan oleh kurangnya media pendukung belajar mandiri di rumah dan variasi media pembelajaran yang terbatas. Akibat dari masalah ini adalah rendahnya hasil belajar siswa pada mata pelajaran Administrasi Infrastruktur Jaringan. Oleh karena itu, perlu dilakukan perbaikan dalam metode pembelajaran dan penggunaan media pendukung agar siswa dapat lebih tertarik dan mudah memahami materi pelajaran. Media pendidikan sangatlah berperan penting dalam Pembelajaran, inovasi dalam perkembangan media sangatlah dibutuhkan karena media pembelajaran dengan teknologi yang berkembang sangatlah berhubungan erat. Media Pembelajaran harus menyamai dengan keberadaan teknologi pada saatsekarang kemudian mengadopsinya dan menyertakannya dalam kegiatan pembelajaran. Oleh karena itu, peneliti mengembangkan sebuah platfrom desain grafis untuk menjadikan media pembelajaran yang menarik. Kemudian media tersebut disosialisasikan kepada siswa SMK Negeri 2 Padang. Berharap memanfaatkan media tersebut akan meningkatkan minat belajar kejuruan khususnya siswa TKJ. Setelah melakukan observasi di SMK Negeri 2 Padang pada jurusan Teknik Komputer dan Jaringan dikelas XI, terlihat bahwa penggunaan smartphone oleh siswa saat proses pembelajaran belum optimal. Meskipun siswa diizinkan menggunakan smartphone untuk mencari materi pembelajaran, kenyataannya banyak siswa yang hanya menggunakan smartphone untuk hal- hal yang tidak bermanfaat seperti melihat media sosial atau bermain game. Selain itu, metode pembelajaran yang dilakukan oleh guru pada mata pelajaran Administrasi Infrastruktur Jaringan terkesan monoton dengan menggunakan metode ceramah dan hanya menggunakan media yang tergolong konvensional, sehingga siswa merasa bosan dan belum tertarik dalam proses pembelajaran. 2. TINJAUAN PUSTAKA Penelitian ini mengenai pengembangan Media Pembelajaran Interaktif Menggunakan canva Pada mata Pelajaran Administrasi Infrastruktur Jaringan di SMK Negeri 2 Padang. Pada obsevasi awal terdapat beberapa kendala yaitu media yang digunakan dalam proses pembelajaran disekolah tersebut masih monoton, guru masih menggunakan buku sebagai media dalam proses pembelajaran dimana materi pada buku tersebut difoto oleh guru dan di bagikan melalui whatshapp, untuk itu diperlukan media pembelajaran yang valid dan praktis untuk menunjang proses pembelajaran lebih baik. Berdasarkan kajian teoristis yang telah dilakukan, penelitian yang relevan dengan penelitian ini adalah penelitian yang dilakukan oleh: Media pembelajaran yang digunakan dalam proses belajar dapat memudahkan siswa dalam memahami dan mening katkan hasil belajar. Dengan 3919 JATI (Jurnal Mahasiswa Teknik Informatika) Vol. 7 No. 6, Desember 2023 Penelitian (Adi & Kusumadewi, 2011) “Pengembangan Media Pembelajaran Berbasis Android Pada Mata Pelajaran Pemograman Dasar Kelas X Di Smk Negeri 3 Surabaya” berdasarkan dari hasil penelitian ini menunjukan bahwa : (1) perbedaan hasil belajar siswa yang menggunakan media pembelajaran berbasis Android dengan yang tidak menggunakan media pembelajaran berbasis Android pada mata pelajaran pemrograman dasar kelas X di SMK Negeri 3 Surabaya. (2) Mengetahui respon siswa terhadap media pembelajaran berbasis Android pada mata pelajaran pemrograman dasar Gambar 1. Rancangan tampilan awal media Gambar 1. Rancangan tampilan awal media Pada Tampilan menu utama berisikan beberapa menu seperti profil, Materi, Quiz, Petunjuk, Video, Home. Penelitian (Jufri, 2022) “Pengembangan Media Pembelajaran Berbasis Android Menggunakan MIT App Inventor pada Materi Barisan dan Deret Aritmatika Kelas X SMKN 1 Kinali” berdasarkan dari hasil penelitian : (1) Agar dapat menghasilkan media pembelajaran berbasis android menggunakan MIT App Inventor pada materi barisan dan deret aritmatika kelas X SMK Negeri 1 Kinali yang valid, praktis dan efektif ditinjau dari kemandirian penggunaan oleh siswa. (2) Penelitian ini menggunakan model Plomp yang terdiri dari 3 fase yaitu preliminary research, prototyping phase dan assessment phase. (3) Instrument penelitian adalah angket validasi, angktet uji praktikalitas dan angket kemandirian belajar siswa menggunakan media pembelajaran. (4) Berdasarkan hasil penilaian validator yaitu ahli media dan materi diperoleh persentase penilaian akhir kevalidan 91,04% dengan kategori sangat valid. (5) Gambar 2. Tampilan menu Gambar 2. Tampilan menu Adapun defenisi operasional penelitian ini adala Media adalah sarana atau prasarana yang digunakan untuk menyampaikan sesuatu dalam proses pembelajaran dikelas. Media pembelajaran interaktif adalah suatu suatu sistem penyampaian materi dengan menyajikan materi video rekaman dengan pengendalian komputer salah satu mata pelajaran wajib pada program keahlian Teknik Komputer dan Jaringan (TKJ). 2. TINJAUAN PUSTAKA Canva adalah sebuah situs sekaligus aplikasi yang menyediakan tools untuk membuat desain grafis dan publikasi online. Diluncurkan pada tahun 2013, Canva kini menjadi salah satu aplikasi yang digemari karena terbilang mudah digunakan untuk pemula. 3. METODE PENELITIAN Pada bagian ini memuat metode yang digunakan pada pembuatan skripsi. Adapun metode yang akan digunakan olePenelitian ini menggunakan metode research and development (R&D) dengan model pengembangan. Penelitian dan pengembangan (research and Development-R&D) berasal dari dua kata yaitu penelitian (research) dan pengembangan (development). Frase ini merupakan gabungan 2 (dua) kata kerja yang memiliki tujuan aktivitas. Penelitian (research) merupakan suatu mekanisme atau kegiatan ilmiah dengan mengikuti aturanaturan atau norma- norma penelitian yang sudah standar dan diakui secara universal, sedangkan pengembangan (development) berarti suatu aktivitas yang merujuk pada penambahan, peningkatan, baik dari segi kuantitas maupun kualitas dari suatu kegiatan atau objek yang menjadi kegiatan. Pada dasarnya definisi operasional untuk memudahkan dalam pengambilan data. Dengan adanya defenisi operasional, maka akan memperjelas ruang lingkup variabel penelitian. Penelitian ini dilakukan sebagai pengembangan media pembelajaran untuk meningkatkan ketertarikan siswa terhadap pembelajaran pada kelas XI TKJ di SMK Negeri 2 Padang. DAFTAR PUSTAKA [1] Amalia, B., Irma, A., & Dikananda, A. R. (2023). Pengembangan Game Edukasi Untuk Mengenalkan Jenis Profesi Pada Anak TK Dengan Menggunakan Metode ADDIE di Paud Kayuwalang Yayasan Al-Mutaqin. 7(1), 615– 618. [2] Arsyad, M. N., & Fatmawati, F. (2018). Penerapan Media Pembelajaran Berbasis Multimedia Interaktif Terhadap Mahasiswa IKIP Budi Utomo Malang. Agastya: Jurnal Sejarah Dan Pembelajarannya, 8(2), 188. https://doi.org/10.25273/ajsp.v8i2.2702 4. HASIL DAN PEMBAHASAN Penelitian ini menggunakan metode Research and Development (R&D) adalah penelitian yang digunakan untuk menghasilkan produk tertentu, dan menguji keefektifan produk tersebut, dapat disimpulkan bahwa: Telah dihasilkan media pembelajaran interaktif pada mata pelajaran administrasi infrastruktur jaringan dengan menggunakan platform desain komunikasi visual online canva untuk membuat media pembelajaran ineraktif. Berdasarkan hasil analisis validasi ahli media pada media pembelajaran interaktif maka diketahui aspek visual dan aspek penggunaan dengan rumus pencariannya yaitu Diketahui Aspek Visual V= ( ∑s / N ( 5-1) ) ( ) ) = 8,1 / 3 ( 3 )= 0,9 P d b Pada gambar diatas terdapat tampilan KD 3.6 dan 4.6 yang akan ditampilkan dalam media agar bisa dibahas oleh siswa kelas XI TKJ pada mata pelajaran administrasi Pada tampilan video ini akan disediakan 2 buah bagian Tampilan video yang akan menampilkan video yang berisikan pembelajaran tentang konfigurasi routing dinamis. Validasi media pembelajaran yang dilakukan ahli media bertujuan untuk mengetahui kelayakan produk sebagai media pembelajaran serta sebagai dasar dalam perbaikan dan meningkatkan kualitas dari media pembelajaran. Validasi dilakukan dengan cara memberikan lembar validasi yang berisikan 14 butir pertanyaan ke dalam 2 aspek yaitu aspek visual dan aspek penggunaan. Uji validitas media pembelajaran interaktif dilakukan oleh 3 orang validator yaitu validator media dan validator materi, untuk validator ahli media dilakukan oleh 3 orang dosen, sedangkan validator ahli materi dilakukan oleh 3 orang guru yang mengajar mata pelajaran administrasi infrastruktur jaringan di SMK Negeri 2 Padang. Hasil validasi untuk ahli media didapatkan angka sebesar 0,94 dengan kategori valid, validitas materi sebesar 0,96 dengan kategori valid. Dengan demikian dapat disimpulkan bahwa dari hasil validasi ahli media dan ahli materi media pembelajaran interaktif dinyatakan valid. Maka hasil pencarian aspek visual mendapatkan nilai sebesar 0.9 dengan kategori valid dan aspek penggunaan mendapatkan nilai sebesar 0.97 dengan kategori valid, dinyatakan valid karena kategori kevalidan pada media pembelajaran >0.667 sedangkan 0-0.666 dinyatakan tidak valid. Rumus untuk mendapatkan nilai total dan nilai rata-rata yaitu : Diketahui nilai total dengan rumus =sum (nilai validitas) T = sum (0.9, 0.97) Maka hasil dari nilai total sebesar 1.87 dan nilai rata-rata dari keseluruhan sebesar 0.94, dapat disimpulkan masuk dalam kategori valid. 4. HASIL DAN PEMBAHASAN Memuat hasil, Pengujian dan pembahasan tentang skripsi yang telah dilakukan Didalam penelitian ini dapat menghasilkan sebuah produk yaitu media pembelajaran interaktif, dimana media pembelajaran tersebut dibuat dan dirancang oleh peneliti. Media pembelajaran tersebut bertujuan untuk menjadi alat bantu bagi guru dalam proses pembelajaran dan sebagai sumber belajar mandiri bagi peserta didik dimanapun mereka berada. merupakan halaman menu media pembelajaran interaktif, menu ini memiliki 5 tombol untuk menghubungkan menu ke menu berikutnya yaitu profile, materi, quiz, petunjuk, video tampilan menu utama dapat dilihat dibawah ini. 3920 Vol. 7 No. 6, Desember 2023 JATI (Jurnal Mahasiswa Teknik Informatika) Gambar 4. Tampilan Menu Utama Media siswa dalam menggunakan media dan memahami pelajaran. Intruksi tampilan Tahap terakhir yang dilakukan pada pengembangan media pembelajaran adalah tahap evaluasi, dimana perancangan media pembelajaran interaktif dilakukan perbaikan ataupun revisi terhadap media oleh validator ahli media berdasarkan hasil dari evaluasi. Validator ahli media pada media pembelajaran interaktif dilakukan oleh dosen program studi Pendidikan Informatika Universitas PGRI Sumatera Barat yaitu validator 1 (AP), Validator 2 (BN) dan validator 3 (MD). Gambar 4. Tampilan Menu Utama Media Pada gambar ini menjelaskan menu-menu yang pada media pembelajaran interaktif dalam media menggunakan media pembelajaran dan membantu Tabel 1. Hasil Validasi Media No Aspek Penilaian Validator Skala Reter ∑s V Kategori VI V2 V3 S1 S2 S3 1 Visual 3,7 3,7 3,7 2,7 2,7 2,7 8,1 0,9 Valid 2 Penggunaan 3,7 4 4 2,7 3 3 8,7 0,97 Valid Jumlah 1,87 - Rata-rata 0,94 Valid Sumber : hasil penelitian, 2023 (data diolah) Tabel 1. Hasil Validasi Media No Aspek Penilaian Validator Skala Reter ∑s V Kategori VI V2 V3 S1 S2 S3 1 Visual 3,7 3,7 3,7 2,7 2,7 2,7 8,1 0,9 Valid 2 Penggunaan 3,7 4 4 2,7 3 3 8,7 0,97 Valid Jumlah 1,87 - Rata-rata 0,94 Valid Sumber : hasil penelitian, 2023 (data diolah) Sumber : hasil penelitian, 2023 (data diolah) SMK Negeri 2 Padang. Penelitian ini menggunakan metode Research and Development (R&D) adalah penelitian yang digunakan untuk menghasilkan produk tertentu, dan menguji keefektifan produk tersebut, dapat disimpulkan bahwa: Telah dihasilkan media pembelajaran interaktif pada mata pelajaran administrasi infrastruktur jaringan dengan menggunakan platform desain komunikasi visual online canva untuk membuat media pembelajaran ineraktif. SMK Negeri 2 Padang. 5. KESIMPULAN DAN SARA Berdasarkan hasil penelitian dan pembahasan pengembangan media pembelajaran interaktif pada mata pelajaran administrasi infrastruktur jaringan 3921 Vol. 7 No. 6, Desember 2023 JATI (Jurnal Mahasiswa Teknik Informatika) https://jbasic.org/index.php/basicedu/article/view/ 1237 [3] Mulyono, H., Irsyadunas, & Oktaliardi, R. (2017). Pengembangan Media Pembelajaran Berbasis Mobile Pada Mata Pelajaran Jaringan Dasar di SMKN 5 Padang. Seminar Nasional: Jambore Konseling 3, 00(00), XX–XX. https://doi.org/10.1007/XXXXXX-XX-0000-00 [7] Hidayat, F., & Nizar, M. (2021). Model Addie (Analysis, Design, Development, Implementation and Evaluation) Dalam Pembelajaran Pendidikan Agama Islam. Jurnal Inovasi Pendidikan Agama Islam (JIPAI), 1(1), 28–38. https://doi.org/10.15575/jipai.v1i1.11042 [4] Budiman, A., Arifin, A., Marlianto, F., Putra, P., Ikip, B., & Pontianak, P. (2019). Pengembangan Media Pembelajaran Berbasis E-Learning Pada SMK di Pontianak. 2(2), 133–139. [8] Journal, B., & Kunci, K. (2022). Jurnal BaJET Pengembangan Media Pembelajaran Berbasis Pada Mata Pelajaran Biologi di Kelas XI. 6(1), 358– 363. [5] Dela, A., Prawesti, D., Mary, T., & Kurniawan, H. (2023). Pengembangan Multimedia Interaktif pada Materi Pembelajaran Pemrograman Dasar. Jurnal Pendidikan Vokasi Dan Seni, 1(2), 66– 76. [9] Jufri, L. H. (2022). Pengembangan Media Pembelajaran Berbasis Android Menggunakan MIP App Inventor pada Materi Barisan dan Deret Aritmatika Kelas X SMKN 1 Kinali. 06(02), 1475–1485. [6] Hapsari, G. P. P., & Zulherman. (2021). Pengembangan Media Video Animasi Berbasis Aplikasi Canva untuk Meningkatkan Motivasi dan Prestasi Belajar Siswa. Jurnal Basicedu, 5(4), 2384–2394. [10] Kurniawan, R., & Azwar, S. (2017). Konstruksi Skala Kepedulian terhadap Penggunaan Energi. Jurnal Ilmu Perilaku, 1(1), 22. https://doi.org/10.25077/jip.1.1.22-32.201 3922
https://openalex.org/W4396715645	https://journals.kabarak.ac.ke/index.php/ajcl/article/download/139/138	English	null	Backlash Against ISDS	null	2,021	cc-by	12,378	BACKLASH AGAINST ISDS A CASE FOR A SUSTAINABLE DEVELOPMENT APPROACH TO INVESTMENT ARBITRATION FOR DEVELOPING COUNTRIES Nilfat Ali* * LLM (Chinese University of Hong Kong), LLB (Moi University). African Journal of Commercial Law 1 (2019/2020) 49-76 * LLM (Chinese University of Hong Kong), LLB (Moi University). Af i J l f C i l L 1 (2019/2020) 49 76 1 UNCTAD, ‘The Entry into Force of Bilateral Investment Treaties (BITS)’ IIA Monitor No. 3 (2006) International Investment Agreements, available at https://unctad.org/en/Docs/webite- iia20069_en.pdf, accessed 11 Dec 2018. 4 Equador has terminated twelve investment treaties, withdrawal from ICSID Convention and notification of revision of 24 BITs. See also UNCTAD Report (2017), Phase 2 of IIA Reform: Modernizing The Existing Stock of Old-Generation Treaties, “of 212 BITs terminated as of March 2017, 19 treaties (9 per cent) were jointly terminated, without any replacement or consolidation; another 59 (28 per cent) were unilaterally terminated, while 134 (63 per cent) were replaced by a new treaty”, available at https://unctad.org/en/PublicationsLibrary/diaepcb2017d3_en.pdf. 5 Ibid 6 See for instance, South Africa. 7 UNCTAD Report 2017, supra n 4. 8 Vera Korzun, ‘The Right to Regulate in Investor-State Arbitration: Slicing and Dicing Reg- ulatory Carve-Outs’ (2016) 50 Vanderbilt Journal of Transnational Law 355. Abstract Over the years, Investor-State Dispute Settlement (ISDS) has spurred a lot of controversies on its legitimacy and integrity. The criticisms mainly stem from the obscuring effect of ISDS on legitimate government regulation and the adventurist manner in which arbitral tribunals interpret substantive rights favouring foreign investors. This, in turn, has resulted in a backlash by States which is illustrated through various forms ranging from withdrawal from traditional Bilateral Investment Treaties (BITs), the development of new era BITs, to extremities of eliminating ISDS all-together. Although there are plausible reforms that have been adopted, the system is still a work in progress. Mechanisms must be incorporated to strike a balance between promotion and protection of investment and the sovereign right of the State to regulate investment in the public interest. Innovation and the adoption of regulatory tools that meet individual States developmental needs will play a focal role in attaining a comprehensive International Investment Agree- ments (IIAs) reform. This paper argues that a paradigm shift from the competing interests of States and investors to a sustainable development approach will achieve a necessary balance in ISDS taking into account the interests of all stake- holders involved in investment arbitration. Keywords: Investor-State Dispute Settlement, Promotion and Protection of Investment, International Investment Agreements, Bilateral Investment Agreements, Sustainable Development 2 Stephan W. Schill, ‘Reforming Investor-State Dispute Settlement: A Comparative and Inter- national Constitutional Law Framework’ (2017) 20 Journal of International Economic Law 649. 1 UNCTAD, ‘The Entry into Force of Bilateral Investment Treaties (BITS)’ IIA Monitor No. 3 (2006) International Investment Agreements, available at https://unctad.org/en/Docs/webite- iia20069_en.pdf, accessed 11 Dec 2018. 2 Stephan W. Schill, ‘Reforming Investor-State Dispute Settlement: A Comparative and Inter- national Constitutional Law Framework’ (2017) 20 Journal of International Economic Law 649. 3 See generally, M Sornarajah, Resistance and Change in the International Law on Foreign Investment, (Cambridge University Press, 2015). 4 Equador has terminated twelve investment treaties, withdrawal from ICSID Convention and notification of revision of 24 BITs. See also UNCTAD Report (2017), Phase 2 of IIA Reform: Modernizing The Existing Stock of Old-Generation Treaties, “of 212 BITs terminated as of March 2017, 19 treaties (9 per cent) were jointly terminated, without any replacement or consolidation; another 59 (28 per cent) were unilaterally terminated, while 134 (63 per cent) were replaced by a new treaty”, available at https://unctad.org/en/PublicationsLibrary/diaepcb2017d3_en.pdf. 5 Ibid. ( ) 3 See generally, M Sornarajah, Resistance and Change in the International Law on Foreign estment, (Cambridge University Press, 2015). 9 United Nations Conference on Trade and Development has published numerous literature on the current investment regime. 10 UNCTAD, World Investment Report 2016, Investor Nationality: Policy Challenges (United Nations, New York and Geneva 2016) 108, available at https://unctad.org/en/PublicationsLibrary/ wir2016_en.pdf, accessed 18 April 2019. 11 Ibid. 1.0 Introduction The period between 1990 – 2007 saw a proliferation of Bilateral Investment Treaties (BITs): over 2500 BITs were signed between States1 with capital exporting countries signing such BITs mainly for the protection of their investors abroad while capital importing countries doing so to attract foreign investments. However, with the boom in BITs, disputes unavoidably arose out of the agreements. The number of claims brought against States has continued to rise and being mostly on the losing end, States have had to pay numerously in compensation. As a result, the investment arbitration regime has been perceived by States as being pro-investor.2 The lack of predictability and coherence in decisions on significant political and economic matters and the lack of sustainability of the system as a whole in terms of promotion and protection of investments has, inevitably, led to the legitimacy deficit or what is referred to as a ‘legitimacy crisis’ in international investment arbitration.3 This has resulted in a backlash by States against the system. The backlash is illustrated through various forms ranging from withdrawal from traditional BITs,4 the development of new era BITs,5 to extremities of eliminating ISDS all-together.6 In addition, a number of Latin American States have recently withdrawn from the International Centre for Settlement of Investment Disputes (ICSID) Convention7 while others, for instance, Australia, have refrained from including ISDS in new International Investment Agreements (IIAs).8 ~ 50 ~ 12 Schill, supra n 2. 13 M Sornarajah, supra n 3 at 451. Backlash Against ISDS In addition to the backlash by States, international organisations, in particular United Nations Conference on Trade and Development (UNCTAD), have been very vocal regarding the current IIA regime.9 UNCTAD has identified key areas for reform, in particular, the policy objectives of safeguarding the right to regulate, reforming ISDS, promoting and facilitating investment and ensuring responsible investment. It is worth noting that the backlash by States and other international organisations against the current IIA regime and the reforms underway are geared towards not only protection and promotion of foreign investments, but also compatibility with sustainable development. In fact, it is safe to say that, investment protection is enmeshed with the achievement of sustainable development goals. As noted in UNCTAD’s 2016 World Investment Report, ‘[r]eform to bring the IIA regime in line with today’s sustainable development imperative is well underway. Today, the question is not about whether to reform, but about the what, how and extent of such reform’.10 There are plausible reforms that have been adopted so far. However, the system is still a work in progress.11 Innovation and the adoption of regulatory tools that meet individual State’s developmental needs will play a focal role in attaining a comprehensive IIA reform. This paper argues that a paradigm shift from the competing interests of States and investors to a sustainable development approach will help achieve the necessary balance in ISDS taking into account the interests of all stakeholders involved. This paper is divided into four main parts: the first part provides a general introduction to the issues under discussion. The second part will problematise the regime of ISDS from the perspective of developing countries in order to lay the basis for core arguments that will be developed later. The third part will do an appraisal of some approaches that countries have proposed or have employed as alternatives to the current ISDS system. Mainly, this paper will assess the case of South Africa, Brazil and India. Finally, the fourth part will make a case for a sustainable development approach in ISDS. ~ 51 ~ 12 Schill, supra n 2. 2.0 ISDS: A Troubled Regime The contemporary ISDS regime has generated a lot of debate among scholars of international investment law. For the proponents of the system, it is argued that ISDS is the most effective mechanism for the protection of foreign investments and to ensure that States are held accountable for government con- duct affecting such investments either directly or indirectly.12 Among the critics, the expansionary and adventurist manner in which arbitral tribunals interpret treaty provisions is pro-investor and this forms the underlying basis of the le- gitimacy crisis in ISDS.13 In this context, the inclination towards pro-investor interpretations is based upon the ideology that such standards of investment protection are necessary for the promotion of investment inflows and ultimately economic development of the host State. Even though it has been argued that this is the root cause of the ISDS legitimacy crisis, this paper seeks to explore a nuanced approach to the root cause in the context of developing countries. Upon attaining independence, most developing countries were left economically crippled and in desperation for investment in-flows so as to boost and place their economies into the global commons. During this time, most countries were seeking ways to attract all forms of investments. Coupled with the perception that developing countries had weak justice systems and were prone to political instability, most IIAs arising from this epoch were drafted with the protection of investors and their investments in mind while significantly undermining States’ rights on regulation and overall protection of their citizens on matters pertaining human rights, sustainable development and environmental standards. In this race to the bottom, viewed from the perspective of public interest, very weak agreements ensued. In addition, factors such as corruption, a glaring menace in developing countries, lack of adequate expertise in the negotiation of investment treaties and ignorance of the ultimate rights being given up, further weakened the position of developing countries in investment relations. This was the ‘original sin.’ Over the years, the ISDS regime has spurred a lot of controversies on its legitimacy and integrity which have revolved around five main issues. 14 M Sornarajah, supra n 3; Aaron Cosby, ‘The Road to Hell? Investor Protections in Nafta’s Chapter 11’ in Lyuba Zarsky (ed), International Investment for Sustainable Development: Balancing Rights and Rewards, (Monetary Institute for International Studies 2005), Frank Emmert and Begaiym Esenkulova, ‘Balancing Investor Protection and Sustainable Development in Investment Arbitration – Trying to Square the Circle?’ available https://www.researchgate.net/publication/327729018_Bal- ancing_Investor_Protection_and_Sustainable_Development_in_Investment_Arbitration_-Trying_ to_Square_the_Circle, accessed 15 December 2018. 19 For instance, the broad manner in which arbitral tribunals have interpreted legitimate ex- pectation, FET, FPS. See, for example, in Railroad Development Corporation (RDC) v Republic of Guatemala, ICSID Case No. ARB/07/23, Award, where the arbitral tribunal rejected the argument to link the FET standard to state practice and opinion juris, which is a more restrictive standard, this broad interpretation of FET is also evidenced in Waste Management v United Mexican States (II), ICSID Case No. ARB (AF)/00/3. 18 See, however, Giovanni Zarra, ‘Right to Regulate, Margin of Appreciation and Proportion- ality: Current Status in Investment Arbitration in Light of Philip Morris v. Uruguay’ (2017) 14 (2) Brazilian Journal of International Law, 95 who argues that, on a positive note, tribunals are seen to react to this criticism and embracing the notion of putting due weight on the purpose of the govern- ment measure rather than just the negative effects of the regulation on investments. 16 Lisa Diependaele, Ferdi De Ville & Sigrid Sterckx ‘Assessing the Normative Legitimacy of Investment Arbitration: The EU’s Investment Court System’ (2019) 24 (1) New Political Economy, 37. 17 Ibid. 15 Ibid, Emmert and Esenkulova. 20 See for instance the Lauder cases, CME Czech Republic B.V. v Czech Republic, Ad hoc – UNCITRAL Arbitration Rules, Partial Award of 13 September 2001, Final Award of 14 March 2003; and Lauder v Czech Republic, Final Award, 3 September 2001, Ad hoc - UNCITRAL Arbitration Rules. 2.0 ISDS: A Troubled Regime First, it has been contended that the system exhibits bias in favour of investors by granting them expansive rights while failing to adequately ~ 52 ~ Backlash Against ISDS address the rights of host States, mainly in terms of government regulation for the public interest.14 This issue concerns both the obscuring effect of ISDS on government regulation and the adventurist nature of interpretations of substantive rights by arbitral tribunals favouring foreign investors. On the effect of ISDS on government regulation, it has been argued that this stems from the fact that States have broad obligations under investment agreements, such that, any legitimate government regulation in the interest of the public affecting investments either directly or indirectly, can be subject to investment arbitration.15 This observation explains why most cases before arbitral tribunals concern government regulation.16 Moreover, tribunals are in the habit of looking into the extent to which a regulation affects an investor rather than looking at the overall purpose of a regulation or a government measure.17 This kind of reasoning has a direct effect on States in terms of regulation for the public interest and can potentially result in a regulatory chill where States opt to avoid regulating or simply not regulate to the extent they should in matters of public welfare such as health, safety and environmental standards.18 In addition, arbitral tribunals have also been criticised on the innovative and broad manner in which they interpret substantive rights solely for the benefit of foreign investors.19 This further curtails a host State’s roles of exercising its regulatory function in the public interest. ~ 53 ~ ~ 53 ~ 21 SD Meyers Inc. v Canada, 2000. 22 Metalclad Corporation v United Mexican States (2000). ICSID Case No. ARB(AF)/97/ 1 (2000). 26 Chin Leng Lim, Jean Ho & Martins Paparinskis, International Investment Law and Arbitra- tion: Commentary, Awards and other Materials, (Cambridge: Cambridge University Press, 2018) 73. , 24 Juan Pablo Charris Benedetti, ‘The Proposed Investment Court System: Does It Really Solve the Problems’, (2019) 42 Review Derecho del Estado 83. 23 Mann, H. et al. Comments on ICSID Discussion Paper, Possible Improvements of the Framework for ICSID Arbitration’ International Institute for Sustainable Development (IND), 2004, 6. , ( ) 25 M. Sornarajah supra n 3. 21 SD Meyers Inc. v Canada, 2000. 22 Metalclad Corporation v United Mexican States (2000). ICSID Case No. ARB(AF)/97/ 1 (2000). 23 Mann, H. et al. Comments on ICSID Discussion Paper, ‘Possible Improvements of the Framework for ICSID Arbitration’ International Institute for Sustainable Development (IND), 2004, 6. 24 Juan Pablo Charris Benedetti, ‘The Proposed Investment Court System: Does It Really Solve the Problems’, (2019) 42 Review Derecho del Estado 83. 25 M. Sornarajah supra n 3. 27 See Susan D Frank, ‘Rationalizing Costs in Investment Treaty Arbitration’ (2011) 88(4) Washington University Law Review 769. 28 See, for instance, Yukos Universal Limited (Isle of Man) v The Russian Federation, UNCIT- RAL, PCA Case No. AA 227, Yukos tribunal ordered Russia to pay $60 million in legal fees to Yukos lawyers. Nilfat Ali Second, there is unpredictability and lack of coherence in decisions. Consistency and coherence in decisions are the cornerstone of any legal system. Arbitral tribunals are increasingly rendering conflicting decisions on cases involving similar issues20 and at times even arising out of the same IIA. A notable example of this situation is the SD Meyers v Canada21 case and the Metalclad v Mexico22 case where the respective tribunals adopted different interpretation for fair and equitable treatment arising out of the same treaty, i.e. NAFTA. The inconsistency in decisions, especially seeing that such decisions are mainly on public issues that have significant financial and political implications, is a major blow to the ISDS regime. As aptly argued, ‘…the mantra of one case not being binding on any other, each one being an individual, one-off, ad-hoc process, has no place in a legal system that passes judgment on a vast range of government measures affecting international investments’.23 It has been suggested that a well-established appellate court system could be the way forward to cure such inconsistencies and to enhance predictability and coherence in decision making. 24 Sornarajah, on the other hand, argues that an appellate facility would merely be, ‘a superstructure built onto the rotten foundations of the existing system’.25 He suggests ‘wiping the slate clean’ as the ultimate solution in ISDS. This argument, although theoretically appealing, as Lim et al. note, the global political will required for starting on a clean slate is lacking especially in light of the reforms that have been introduced in the system.26 Given that under the current ISDS regime, there is no system of precedence and each case is decided as a stand-alone, ad- hoc process, an appellate court may present a good avenue to cure the system of the criticisms of inconsistency and unpredictability. 26 Chin Leng Lim, Jean Ho & Martins Paparinskis, International Investment Law and Arbitra- tion: Commentary, Awards and other Materials, (Cambridge: Cambridge University Press, 2018) 73. ~ 54 ~ ~ 54 ~ 27 See Susan D Frank, ‘Rationalizing Costs in Investment Treaty Arbitration’ (2011) 88(4) Washington University Law Review 769. 28 See, for instance, Yukos Universal Limited (Isle of Man) v The Russian Federation, UNCIT- RAL, PCA Case No. AA 227, Yukos tribunal ordered Russia to pay $60 million in legal fees to Yukos lawyers. 29 Ibid. 30 Emmert and Esenkulova supra n 14. 31 Lisa Diependaele et al, supra n 16. 32 ICSID Convention, Article 14. 33 UNCITRAL, Article 11. 34 See for instance, Transatlantic Trade and Investment Partnership, EU-Canada Comprehen- sive Economic and Trade Agreement. 32 ICSID Convention, Article 14. 33 UNCITRAL, Article 11. 31 Lisa Diependaele et al, supra n 16. 34 See for instance, Transatlantic Trade and Investment Partnership, EU-Canada Comprehen- sive Economic and Trade Agreement. 30 Emmert and Esenkulova supra n 14. 35 Jandhyala, Srividya, ‘Why Do Countries Commit to ISDS for Disputes with Foreign Inves- tors?’ (2016) 16 (1) A I B Insights; East Lansing, 7. 36 Emmert and Esenkulova supra n 14. 37 Philip Morris Brands Sàrl, Philip Morris Products S.A. and Abal Hermanos S.A. v Oriental Republic of Uruguay, ICSID Case No. ARB/10/7. 38 Vattenfall AB and Others v Federal Republic of Germany, ICSID Case No. ARB/12/12. 39 Methanex Corporation v United States of America,NAFTA/UCITRAL Award,03/08/2005. 40 Cosby, supra n 14. Backlash Against ISDS Third, the cost of investment arbitration has been perceived as being high.27 Costs incurred in investment arbitration are quite substantial.28 Empirical studies show that even partial costs could represent up to more than 10% of the total arbitration award.29 This coupled with the inconsistencies and unpredictability in awarding costs,30 further escalates the controversies surrounding ISDS. Fourth, the credibility of arbitrators is a rising concern among States, especially, on the so-called role-issue conflict where an arbitrator would also be an investment lawyer. As such, they may have to decide on a particular issue where they may also potentially have to advise and represent clients concerning the same issue. A conflict of interest is inevitable and the perception of impartiality and neutrality is, thus, compromised. In addition, it has been argued that there is insufficient attention on arbitrators’ qualification. Procedural frameworks and IIAs are normally silent on substantive requirements regarding qualifications of arbitrators.31 For instance, the ICSID Convention merely provides that arbitrators should be ‘of high moral character and [have] competence in the field of law, commerce, industry, or finance and may be relied upon to exercise independent judgement.’32 United Nations Commission on International Trade Law (UNCITRAL) Rules only require the arbitrators to disclose circumstances that would raise doubts regarding their impartiality.33 However, some new era treaties address this issue and prescribe precise qualifications to be held by arbitrators should a dispute ensue.34 Finally, the lack of transparency and the lack of sufficient representation of the public in matters that generate great public interest further exacerbate the problems of ISDS. ISDS still remains a closed system of dispute settlement with disputes mainly decided behind closed doors. Non-disputing ~ 55 ~ ~ 55 ~ Nilfat Ali parties, under the current ISDS regime, have no legal rights to participate or influence the outcomes of the disputes.35 Amicus briefs still remain a rarity in ISDS.36 Investments and even investment disputes have social, economic and political impacts on the local communities of a host State and exclusion of these communities, especially from dispute settlement, effectively shuts out their experiences. As noted, the bulk of ISDS cases arise out of governmental regulations and measures on public policy challenged by investors. In fact, even legitimate measures of the host State on environmental and health regulation, affecting investments either directly or indirectly can be challenged before an arbitral tribunal. Backlash Against ISDS For instance, in the Philip Morris case,37 Philip Morris sued Uruguay and Australia for their requirements for health warnings on cigarette packets. In Vattenfall case,38 a Swedish energy company, sued Germany for regulations phasing out nuclear energy and similarly in Methanex case,39 the extent to which a government may exercise its sovereign right to ban a potentially harmful substance to both health and the environment was challenged. As rightly noted, in these circumstances, the decisions of the tribunals will have ‘implications that go far beyond commercial impacts to such public policy objectives as the protection of the environment and public health and safety.’40 Under all these circumstances, the resulting effect of the process is felt by more than just the parties involved. 3.0 Whither alternatives to ISDS? The legitimacy crisis in the ISDS regime has resulted in a backlash by States against the system as noted in the introduction. Under this section, the paper will critically assess some approaches employed by States, in particular, States in the Global South, as alternatives to the current ISDS system. Mainly, the research will assess the case of South Africa, Brazil and India, especially because of the recent developments in their investment regimes and the possible ~ 56 ~ Backlash Against ISDS impact that they could have in reshaping the current investment treaty system. In each of these cases, as will be discussed below, there are circumstances that have triggered and galvanised the policies adopted by the States. 41 D. Games, ‘The Experience of South African Firms in Doing Business in Africa: A Prelimi- nary Survey and Analysis’ (2004) South African Institute for International Affairs, Business in Africa Research Project. 42 Sibanda Omphemetse S., ‘The Promotion and Protection of Foreign Investment Law Bill: Denunciation of BITs, and the De-internationalisation of Investor-State Arbitration in South Africa’ (2014) 4(4) London: The Academy of Business and Retail Management. 43 Piero Foresti, Laura de Carli & Others v The Republic of South Africa, ICSID Case No. ARB (AF)/07/1 43 Piero Foresti, Laura de Carli & Others v The Republic of South Africa, ICSID Case No. ARB (AF)/07/1 j 42 Sibanda Omphemetse S., ‘The Promotion and Protection of Foreign Investment Law B Denunciation of BITs, and the De-internationalisation of Investor-State Arbitration in South Afr (2014) 4(4) London: The Academy of Business and Retail Management. 3.1.0 South Africa Like many other post-colonial States and after the end of its social, economic and political global isolation due to apartheid, South Africa got into a state of euphoria signing as many investment agreements as possible so as to attract Foreign Direct Investments (FDI) into its economy. This was confirmed by a study that concluded that, ‘…[i]n less than a decade, South Africa ha[d] become one of the top 10 investors in, and trading partner of, many African countries, displacing those companies from Europe (particularly in countries that were former colonial powers) and America, which ha[d] traditionally retained their economic links in Africa.’41 However, as with the conclusion of the BITs, disputes inevitably arose between the host State and the foreign investors. With each of these disputes, South Africa became more sceptical of the rights being given up by the State in the name of attracting FDI.42 This, especially, in terms of the sovereign rights of the State to regulate in the public interest. More so, this has to be seen in light of South Africa’s history of apartheid, the economic inequalities suffered during that period and its constitutional-based transformation agenda. Of significance, is the case of Piero Foresti, Laura de Carli & Others v The Republic of South Africa of 200743 which was the ultimate eye opener for South Africa. In this case, the Black Economic Empowerment policies (BEE), a crucial feature of South Africa’s National Development Plan, guaranteed under the Constitution as affirmative action measures whose mandate is to ~ 57 ~ ~ 57 ~ Nilfat Ali cure the lingering effects of the apartheid era,44 was challenged as being contrary to the obligations of South Africa under the Italy-South Africa BIT. Against this backdrop, the Government of South Africa began the process of reviewing its BITs and their impact on economic development and the sovereign right of the State to regulate.45 Following this review, it was concluded that, ‘the current system [had] open[ed] the door for narrow commercial interests to subject matters of vital national interest to unpredictable international arbitration that may constitute direct challenges to legitimate, constitutional and democratic policy-making.’46 Further, that the ‘Cabinet understood that the relationship between BITs and FDI was ambiguous at best, and that BITs pose risks and limitations on the ability of the Government to pursue its Constitutional-based transformation agenda. 44 Elizabeth Whitsitt, An ICSID Tribunal Introduces Innovative Steps into Non-Disputing Party Procedure, Investment Treaty News, (10 October, 2009), available at https://www.iisd.org/ itn/2009/10/10/an-icsid-tribunal-introduces-innovative-steps-into-non-disputing-party-procedure/. 45 Tarcisio Gazzini, ‘Travelling the National Route: South Africa’s Protection of Investment Act 2015’ (2018) 26 (2) African Journal of International and Comparative Law 242. 46 See Dr. Rob Davis, Minister of Trade and Industry, in his speech at the launch of UNC- TAD’s Investment Policy Framework for Sustainable Development in Geneva in September 2012, available at http://www.dti.gov.zadelegationspeechdetail.jsp?id=2506. 47 Minister of Trade and Industry, Dr. Rob Davis, in his speech at the launch of UNCTAD’s Investment Policy Framework for Sustainable Development in Geneva in September 2012. 48 Gazini, supra n 48. 49 South Africa, Protection of Investment Act 2015 (Act 22 of 2015), Government Gazette, 606, 39514, 15 December 2015, available at http://www.gov.za/sites/www.gov.za/files/39514_Ac- t22of2015ProtectionOfInvestmentAct.pdf, accessed on 28 December 2018. 49 South Africa, Protection of Investment Act 2015 (Act 22 of 2015), Government Gazette, 606, 39514, 15 December 2015, available at http://www.gov.za/sites/www.gov.za/files/39514_Ac- t22of2015ProtectionOfInvestmentAct.pdf, accessed on 28 December 2018. y 48 Gazini, supra n 48. 50 Richard Jansen van Vuuren, ‘Investment Laws in South Africa – Foreign Investors take Note’ [2018] Mining Review Africa, available at https://www.miningreview.com/international/ changes-investment-laws-south-africa-foreign-investors/. 51 Section 8, South Africa Protection of Investment Act, 2015. 3.1.0 South Africa Cabinet concluded that South Africa should refrain from entering into BITs in future, except in cases of compelling economic and political circumstances.’47 Following this, South Africa terminated several Bilateral Investment Treaties (BITs) to which it was a party.48 Subsequently South Africa opted for a legislative approach in its investment relations and in 2013, South Africa published for public comment the Promotion and Protection of Investment Bill which was consequently enacted in 2015 as the Promotion of Investment Act 22 of 2015.49 Of its striking features, the Act explicitly grants the South African Government the right to take regulatory measures in order to: redress historical, social and economic inequalities and injustices; uphold the rights, values and principles contained in the Constitution of the Republic of South Africa; and to promote and preserve cultural heritage, foster economic development, protect ~ 58 ~ Backlash Against ISDS the environment; and achieve the progressive realisation of socio-economic rights.50 The Act also provides for investor protections: it provides inter alia, for national treatment,51 just and equitable compensation on expropriation,52 a physical security of property standard53 and a right to transfer funds. It is interesting to note that on compensation for expropriation, the Act has shifted from the traditional rule of compensation of full market value to what is deemed to be just and equitable. On dispute resolution, the Act is argued to be reminiscent of the Calvo doctrine,54 and it provides for the settlement of investment disputes through the domestic justice system.55 The Act further provides for international arbitration. However, it requires that the ‘arbitration will be conducted between the Republic and the home state of the applicable investor.’56 The 52 Section 10, Protection of Investment Act, 2015, reference is made to section 25 of the South n Constitution which in relation to expropriation provides: 52 Section 10, Protection of Investment Act, 2015, reference is made to section 25 of the South African Constitution which in relation to expropriation provides: 25(2) Property may be expropriated only in terms of law of general application— (a) for a public purpose or in the public interest; and (b) subject to compensation, the amount of which and the time and manner of payment of which have either been agreed to by those affected or decided or approved by a court. 60 Tarcisio Gazzini, ‘Travelling the National Route: South Africa’s Protection of Investment Act 2015’ (2018) 26(2) African Journal of International and Comparative Law 242. See also UNC- TAD Series on International Investment Agreements II, Investor-State Dispute Settlement: A Sequel, available at https://unctad.org/en/PublicationsLibrary/diaeia2013d2_en.pdf, accessed 5 January 2019. 58 Roberto Echandi, ‘What do Developing Countries Expect from the International Investment Regime?’ in Jose E. Alvarez, Karl P. Sauvant, Kamil Girard Ahmed, and Gabriela P. Vizcamno, (eds), The Evolving International Investment Regime: Expectations, Realities, Options, (Oxford: Oxford University Press, 2011). y ) 59 Mmiselo Freedom Qumba, ‘Safeguarding Foreign Direct Investment in South Africa: Does the Protection of Investment Act Live up to its Name?’ (2018) 25 (3) South African Journal of Inter- national Affairs 341. 57 Won Kidane, ‘Contemporary International Investment Law Trends and Africa’s Dilemmas in the Draft Pan-African Investment Code’ (2018) 50 The George Washington International Law Review Journal 523. 3.1.0 South Africa 25(2) Property may be expropriated only in terms of law of general application— (a) for a public purpose or in the public interest; and (b) subject to compensation, the amount of which and the time and manner of payment of which have either been agreed to by those affected or decided or approved by a court. (b) subject to compensation, the amount of which and the time and manner of payment of which have either been agreed to by those affected or decided or approved by a court. 25(3) The amount of the compensation and the time and manner of payment must be just and equitable, reflecting an equitable balance between the public interest and the inter- ests of those affected, having regard to all relevant circumstances, including— (a) the current use of the property; ( ) p p y (b) the history of the acquisition and use of the property; (c) the market value of the property; (d) the extent of direct state investment and subsidy in the  acquisition and beneficial (d) the extent of direct state investment and su capital improvement of the  property; and capital improvement of the  property; and (e) the purpose of the expropriation. 54 Tarcisio Gazzini, ‘Travelling the National Route: South Africa’s Protection of Investment Act 2015’ (2018) 26(2) African Journal of International and Comparative Law 242, making reference to ‘calvo grandchildren’. 55 Section 13(4): Subject to applicable legislation, an investor, upon becoming aware of a dispute as referred to in subsection (1), is not precluded from approaching any competent court, independent tribunal or statutory body within the Republic for the resolution of a dispute relating to an investment. 56 Ibid, Section 13(5): The government may consent to international arbitration in respect of investments covered by this Act, subject to the exhaustion of domestic remedies. The consideration of a request for international arbitration will be subject to the administrative processes set out in sec- tion 6. Such arbitration will be conducted between the Republic and the home state of the applicable investor. ~ 59 ~ ~ 59 ~ Nilfat Ali State-State arbitration can only be resorted to after exhaustion of domestic remedies. South Africa takes investment disputes back to the era of diplomatic protection where the home state would be invited to espouse a claim on behalf of its investor. This, as Kidane observes, is ‘reminiscent of the pre-BIT days of self-help and gunboat diplomacy.’57 ISDS, as Echandi rightly observes, being rule-oriented, governed by principles and rules established by law, is instrumental in the elimination of politics and power-oriented diplomacy in international investment relations.58 In addition, critics have argued that the Act seems to be leaning more towards the protection of public interest and significantly curtails investors’ protections as found in traditional BITs.59 South Africa has taken a rather radical measure by eliminating ISDS. Even though ISDS has its weaknesses, it is the one innovation that has successfully governed international investment relations excluding interference through States politics, which the South African approach is prone to. In addition, the idea of espousal of claims by home States means that the investor will have limited access in terms of control and direction of the dispute. As Gazzini observes, ‘the entire proceedings and all decisions related to the dispute (including a possibly friendly settlement or an agreement on compensation) will be firmly in the hand of the home state with all the associated shortcomings and implications.’60 ~ 60 ~ Backlash Against ISDS 64 For a detailed discussion on Brazil’s resistance to traditional BITs, see Michelle Ratton San- chez Badin and Fabio Morosini, ‘Navigating between Resistance and Conformity with the Interna- tional Investment Regime: The Brazilian Agreements on Cooperation and Facilitation of Investments (ACFIs)’in Fabio Morosini and Michelle Ratton Sanchez Badin (Eds), Reconceptualizing Interna- tional Investment Law from the Global South (Cambridge: Cambridge University Press, 2018). 61 United Nations Conference on Trade and Development (UNCTAD), Investment Policy Hub, International Investment Agreements Navigator, available at https://investmentpolicy.unctad. org/international-investment-agreements, accessed 27 June 2019 . g g 62 United Nations Conference on Trade and Development (UNCTAD). (2017). International investment agreements navigator available at http://investmentpolicyhub.unctad.org/IIA, accessed 23 April 2019. 63 Ibid 66 José Henrique Vieira Martins, ‘Brazil’s Cooperation and Facilitation Investment Agree- ments (CFIA) and Recent Developments’, (2017), available at https://www.iisd.org/itn/2017/06/12/ brazils-cooperation-facilitation-investment-agreements-cfia-recent-developments-jose-henrique- vieira-martins/, accessed 20 April 2019. 61 United Nations Conference on Trade and Development (UNCTAD), Investment Policy Hub, International Investment Agreements Navigator, available at https://investmentpolicy.unctad. org/international-investment-agreements, accessed 27 June 2019 . 62 United Nations Conference on Trade and Development (UNCTAD). (2017). International investment agreements navigator available at http://investmentpolicyhub.unctad.org/IIA, accessed 23 April 2019. 63 Ibid. 64 For a detailed discussion on Brazil’s resistance to traditional BITs, see Michelle Ratton San- chez Badin and Fabio Morosini, ‘Navigating between Resistance and Conformity with the Interna- tional Investment Regime: The Brazilian Agreements on Cooperation and Facilitation of Investments (ACFIs)’in Fabio Morosini and Michelle Ratton Sanchez Badin (Eds), Reconceptualizing Interna- tional Investment Law from the Global South (Cambridge: Cambridge University Press, 2018). 65 UNCTAD World Investment Report (2016), supra n 10. 66 José Henrique Vieira Martins, ‘Brazil’s Cooperation and Facilitation Investment Agree- ments (CFIA) and Recent Developments’, (2017), available at https://www.iisd.org/itn/2017/06/12/ brazils-cooperation-facilitation-investment-agreements-cfia-recent-developments-jose-henrique- vieira-martins/, accessed 20 April 2019. 67 Brazilian Central Bank. (2014). Brazilian Capital Abroad, available at http://www4.bcb. gov.br/rex/CBE/Ingl/CBE2014Results.pdf, accessed 20 April 2019. 68 Ibid. 69 Ibid at 65. f ( g 65 UNCTAD World Investment Report (2016), supra n 10. 67 Brazilian Central Bank. (2014). Brazilian Capital Abroad, available at http://www4.bcb. gov.br/rex/CBE/Ingl/CBE2014Results.pdf, accessed 20 April 2019. 68 Ibid. 69 Ibid at 65. 73 Stephanie Grace Hawes,’ Trends in Investment Treaty Arbitration: A Perspective on Brazil’ (Arbitration Blog, 30 October 2017), available at http://arbitrationblog.practicallaw.com/trends-in- investment-treaty-arbitration-a-perspective-on-brazil/, accessed 20 December 2018; see also Vivian Gabriel, ‘The New Brazilian Cooperation and Facilitation Investment Agreement: An Analysis of the Conflict Resolution Mechanism in Light of the Theory of the Shadow of the Law’ (2016) 34(2) Conflict Resolution Quarterly, 141. 70 The Cooperation and Facilitation Investment Agreement, vailable at http://www.mdic.gov. br/arquivos/CFIA-Presentation-EN.pdf, accessed 5 January 2019. 71 Ibid. 3.2.0 Brazil Out of the 2352 BITs in force signed between 1957 and 2019,61 none involves Brazil.62 It is worth noting that between 1994 and 1999, Brazil had actually signed 14 traditional BITs. However, none of them was approved by Congress63 for granting excessive protection of private interests at the expense of government’s regulatory flexibility. In addition, there was a general lack of trust on the capacity of BITs to attract investments in Brazil.64 Despite having no BIT in force, Brazil continued to receive significant FDI inflows.65 This reinforced the comprehension that investment inflows are not necessarily influenced by BITs in force.66 However, having undergone political and economic transformations, Brazil’s status, as mainly a capital importer, altered and it became an exporter of capital as well.67 In fact, by the end of 2014, the stock of Brazilian FDI abroad corresponded to almost half of the amount of FDI in Brazil.68 This coupled ‘with the interest of partner countries in negotiating investment agreements, the several problems perceived in traditional BITs and the growing number of investor–state arbitration cases raised the debate of investment agreements again in Brazil.’69 66 José Henrique Vieira Martins, ‘Brazil’s Cooperation and Facilitation Investment Agree- ments (CFIA) and Recent Developments’, (2017), available at https://www.iisd.org/itn/2017/06/12/ brazils-cooperation-facilitation-investment-agreements-cfia-recent-developments-jose-henrique- vieira-martins/, accessed 20 April 2019. 67 Brazilian Central Bank. (2014). Brazilian Capital Abroad, available at http://www4.bcb. gov.br/rex/CBE/Ingl/CBE2014Results.pdf, accessed 20 April 2019. 68 Ibid. 69 Ibid at 65. ~ 61 ~ Nilfat Ali 70 The Cooperation and Facilitation Investment Agreement, vailable at http://www.mdic.gov. br/arquivos/CFIA-Presentation-EN.pdf, accessed 5 January 2019. 71 Ibid. 72 Ibid. 73 Stephanie Grace Hawes,’ Trends in Investment Treaty Arbitration: A Perspective on Brazil’ (Arbitration Blog, 30 October 2017), available at http://arbitrationblog.practicallaw.com/trends-in- investment-treaty-arbitration-a-perspective-on-brazil/, accessed 20 December 2018; see also Vivian Gabriel, ‘The New Brazilian Cooperation and Facilitation Investment Agreement: An Analysis of the Conflict Resolution Mechanism in Light of the Theory of the Shadow of the Law’ (2016) 34(2) Conflict Resolution Quarterly, 141. Nilfat Ali These developments warranted a new plan of action for Brazil and acted as a platform for Brazil to exercise innovation in coming up with a model that would not only protect investments but also promote and facilitate productive investments. Brazil in 2015, thus, adopted a new approach: the Cooperation and Facilitation Investment Agreement (CFIA). The model CFIA seeks to protect and promote investments as well as preserve Brazil’s right to regulate in the public interest. Institutional governance, risk mitigation and thematic agendas for investment cooperation and facilitation form the substratum of the CFIA.70 Undoubtedly impressive and innovative, risk mitigation and institutional governance aim at prevention of disputes and improvement of investment relations between the parties through the establishment of focal points such as ombudsmen to enhance dialogue between the parties.71 Like the South African Act, investor-state arbitration is eliminated and instead it provides for State-State arbitration. Other interesting innovations in the model include incorporation of corporate and social responsibility principles and clauses concerning corruption. In addition, CFIAs require investors to partake in the sustainable development of the host State and adherence to environmental protection standards and respect for human rights.72 Among the strong points on the Brazil CFIAs, as has been observed by Hawes, is the promotion of amicable settlement of disputes before opting for a rather adversarial arbitration approach.73 In addition, CFIAs expressly foster sustainable development for the Host State. Although traditional BITs are intended to promote this aspect as well, they are normally drafted with the protection of investment as a priority. On the other hand, the negatives include the lack of mention of substantive provisions on fair and equitable treatment and indirect expropriation and the ~ 62 ~ Backlash Against ISDS limitation of access to dispute resolution.74 In addition, the proposed system of focal points has been criticised for limiting lodging of complaints to only the investors and the State since investments have social and economic impacts on the local community. As observed by Bernasconi-Osterwalder and Brauch,75 the focal points should be expanded to accommodate concerns raised by all stakeholders involved including the members of the public. In addition, they argue that the system does not provide an avenue for public participation and accountability and, as such, is prone to corruption opportunities. On state- state arbitration, the system risks the same challenges as South Africa, as discussed above. 75 Bernasconi-Osterwalder, Nathalie, and Martin Dietrich Brauch, ‘Comparative Commentary to Brazil’s Cooperation and Investment Facilitation Agreements (CIFAs) with Mozambique, Angola, Mexico, and Malawi’ (2015) International Institute for Sustainable Development, available at https:// www.iisd.org/sites/default/files/publications/commentary-brazil-cifas-acfis-mozambique-angola- mexico-malawi.pdf, accessed 29 December 2018. p 78 See Vivian Gabriel, ‘The New Brazilian Cooperation and Facilitation Investment Agree- ment: An Analysis of the Conflict Resolution Mechanism in Light of the Theory of the Shadow of the Law’ (2016) 34(2) Conflict Resolution Quarterly, 141. 74 See Ely Caetano Xavier Jr, José Augusto Fontoura Costa, ‘Expropriation in Brazil’s Coop- eration and Facilitation Investment Agreements: A Failed Attempt to Think Outside the Box’ (2017) Yearbook on International Investment Law and Policy, 444. 77 Supra n 65. 76 Ibid. 77 Supra n 65. 76 Ibid. 79 Ibid. 80 Prabhash Ranjan and P. Anand, ‘The 2016 Model Indian Bilateral Investment Treaty: A Critical Deconstruction’ (2017) 38 Northwestern Journal of International Law & Business 1. 81 White Industries Australia Limited v. Republic of India, UNCITRAL, Final Award (Nov. 30, 2011). 82 Centre for Public Interest Litigation and Others v Union of India and Others, 2012 (3) Supreme Court cases 104. 83 Forum against FDAs, We Call Upon the Government to Review and Rescind Its Decision to Sign BIT! BIPA with the USA-Open letter to the Indian Prime Minister, Dr ManhomanSingh (Sept. 26, 2013), available at http://www.bilaterals.org/IMG/pdf/lettter-forum-agftas us indiabits_26 sept. pdf. 84 Statement by P. Rajeeve, Member of Parliament (India), Transcript of the Proceedings of the Rajyasabha (22 May 2012) 52-4, available at http://l64.100.47.5/newdebate/225/22052012/ Full- day.pdf. Nilfat Ali The system is further criticised for not having a model text of the CFIA.76 So far, CFIAs have varied depending on the counterparty of the agreement. This in turn will, no doubt, result in inconsistencies in the rather ‘young’ system. However, it has been observed that the lack of uniformity of the CFIAs is indicative of Brazil attempting to cater for the different needs of each partner and the possibility to continually improve the model.77 Brazil’s CFIA system is greatly inspired by the South Korea’s success story of preventive dispute settlement. However, Brazil made a significant modification on the South Korea system and, instead of ISDS, it opted for State-State arbitration. This has serious consequences on the bargaining power of the parties. Under State-State arbitration, an investor is not assured that its home State will take up and pursue their claim. Therefore, investors have no leverage in the negotiations and hence lack incentive to partake in the preventive negotiations.78 Nevertheless, it is yet to be seen how this regime will fair in the international investment regime. The incorporation of promotion of sustainable development and ‘regulatory space’ for the Host State are in no doubt a major plus for the Brazil CFIA. In addition, as observed, ‘the positive ~ 63 ~ Nilfat Ali repercussion of the model among relevant economic agents and partners, as well as in the international academic and cooperation circles, show that the model seems to be heading in the right way.’79 p 84 Statement by P. Rajeeve, Member of Parliament (India), Transcript of the Proceedings of the Rajyasabha (22 May 2012) 52-4, available at http://l64.100.47.5/newdebate/225/22052012/ Full- day.pdf. ) 82 Centre for Public Interest Litigation and Others v Union of India and Others, 2012 (3) preme Court cases 104. 83 Forum against FDAs, We Call Upon the Government to Review and Rescind Its Decision to IT! BIPA with the USA-Open letter to the Indian Prime Minister, Dr ManhomanSingh (Sept. 13), available at http://www.bilaterals.org/IMG/pdf/lettter-forum-agftas us indiabits_26 sept. 3.3.0 India Until very recently, there was hardly any domestic discussion on India’s approach to negotiating investment treaties. However, there are a series of developments that woke up India to the realities of investment treaty arbitration. There were three such developments.80 First, up to 2010, there were barely any claims against India. This position changed and India increasingly found itself as a respondent in ISDS claims. In fact, it wasn’t until 2011, that the first publicly known BIT arbitral award was issued against India in White Industries v India.81 After this award, the Government of India has received a series of notices of disputes from a number of foreign corporations (investors) challenging a wide array of regulatory measures taken by India, mainly, in the telecom industry following the cancellation of telecom licenses by the Indian Supreme Court in 2012.82 The second development relates to the domestic pressures against the investment regime in India by different actors in academia, civil society organisations83 and parliamentarians.84 Following the White Industries case and the swelling number of ISDS notices against India, the demands for a crucial assessment of BITs in India only escalated. Finally, there were internal discussions within the Indian Government on the urgent need to review its BITs. This is evidenced by the Ministry of ~ 64 ~ Backlash Against ISDS Commerce’s discussion paper ‘International Investment Agreements Between India and Other Countries.’85 The paper noted, inter alia, that it would crucial for developing countries to strike a balance between investors’ rights and domestic policy when entering BITs86 and further that ‘legitimate public concerns must not be subordinated to investment protection issues.’87 The paper identified areas of concern in Indian BITs, such as the broad definition of investment, an expansive provision on expropriation without any reservations and undefined terms, in particular, fair and equitable treatment. The paper concluded by recommending a review of India’s existing BITs and the striking of a balance between investor rights and the government’s sovereign power to regulate in the public interest on health, safety and the environment.88 Consequently, the above-mentioned developments paved way for India to review its BITs and critically rethink its strategy on BITs. 85 Ministry of Commerce, Government of India, International Investment Agreements between India and Other Countries [hereinafter, Commerce Ministry Paper]. As quoted in Prabhash supra n 81. It is worth noting that the paper was prepared in 2011 after the White Industries case. It also appears to be greatly inspired by UNCTAD’s work on BITs and IIAs generally as this is quoted severally to explain various arguments. The developments taking place elsewhere on BITs such as South Africa and some Latin American countries further played a focal role in the development of the paper. 86 Ibid Commerce Ministry Paper, supra 74. 89 Government of India, Ministry of Commerce and Industry, Rajya Sabha, Question No. 1122, Answered on Jul. 26, 2017, available at http://164.100.47.4/newrsquestion/ShowQn.aspx. 90 Ibid. 88 Commerce Ministry Paper, supra 74. 3.3.0 India The review revealed that India’s BITs were inadequately drafted and contained vague and ambiguous provisions that could be subject to broad interpretations by arbitral tribunals.89 India further admitted that such vague provisions were susceptible to interpretations that would encroach upon government’s regulatory powers.90 It was, therefore, imperative to adopt a new model BIT for India to reengage with its trade and investment partners. The adoption of a new model BIT was inevitable. From the above background and the developments leading to the adoption of the model BIT, it is evident that the model BIT is an avenue for India to claw back the balance between investor rights and state’s regulatory powers. For instance, on the definition of investment, the elements of indirect expropriation and the coverage of national treatment, tighter language has been introduced. Also, for the avoidance of a repeat of White Industries case, ~ 65 ~ Nilfat Ali where the tribunal expansively interpreted the Most Favored Nation (MFN) clause, the MFN clause has been excluded in the model BIT. The model BIT, also, excluded the umbrella clause and the Fair and Equitable Treatment (FET) clause which are considered as the catch-all phrase which investors pry on and arbitral tribunals have a field day in interpreting. Instead, the model BIT contains a provision entitled “Treatment of Investments,” thereby limiting the scope of its application. In addition, the model BIT, quite interestingly, incorporates corporate social responsibility which requires investors to voluntarily incorporate internationally recognised standards of corporate social responsibility in their practices and internal policies. This is a plus for India and is geared towards the sustainable and equitable development of the host state. On ISDS, the model BIT still maintains ISDS but the same is subject to the exhaustion of local remedies requirement before initiation of arbitral proceedings. However, the model BIT allows room for non-applicability of the conditions precedent if the claimant can show that there are no domestic legal remedies reasonably capable of providing any relief in respect of the same measure or similar factual matters for which a breach of treaty is claimed by the investor. The jurisdiction of an arbitral tribunal has been significantly minimised to matters arising from only alleged breach of Chapter II of the model BIT. This chapter provides for Treatment of Investments that includes full protection and security clause, national treatment, expropriation, money transfer provisions and compensation for loses. 91 See Government of India, Ministry of Commerce & Industry, Department of Indus- trial Policy & Promotion, Lok Sabha Unstarred Question July 25, 2016), No. 1290, available at http://164.100.47.190/loksabhaquestions/annex/9/AU1290.pdf, accessed 24 December 2018. 92 James J. Nedumpara, ‘India’s Trade and Investment Agreements: Striking a Balance be- tween Investor Protection Rights and Development Concerns’ in Fabio Morosini and Michelle Rat- ton Sanchez Badin (Eds), Reconceptualizing International Investment Law from the Global South (Cambridge: Cambridge University Press, 2018). 3.3.0 India Other limitations include, lack of jurisdiction of an arbitral tribunal to review the merits of a decision rendered by a domestic court and that a tribunal cannot accept jurisdiction over any claim that is or has been subject to arbitration under Chapter V of the treaty which provides for State-State investor settlement. According to the Indian government, the model BIT is meant to strike a balance between protection of investments with host State’s legitimate right to regulate in the public interest and to remove ambiguity and vagueness in treaty provisions so as to limit arbitration discretion.91 A cursory look at the model ~ 66 ~ Backlash Against ISDS BIT is sufficient to note that for India, carving out regulatory space for the exercise of the State’s sovereign power to legitimately regulate in the public interest took center stage in the model BIT. This is quite understandable given the challenges of ISDS and the limitations host States have had in regulation for public interest. In addition, as noted, ‘India has been long echoing the need to preserve policy autonomy in several spheres of trade policy. India has played a pivotal role in retaining policy flexibility in various areas of economic diplomacy…’92 Critics have argued that the model BIT is far from attaining a balance between investor rights and the regulatory powers of the host State. They argue, if anything, the model BIT tilts the scale towards regulatory rights of the host State and significantly limits investor protections.93 Further, that the model still fails considerably to delineate clear and precise treaty provisions hence giving too much leeway to arbitrators in interpretation of treaty provisions.94 However, this observation is not accurate. Although the model BIT ex- cludes some substantive provisions granted in traditional BITs, comparatively, it has significantly attempted to balance investor rights and the State’s right to regulate in the public interest. Indeed, as further discussed below, to achieve the Sustainable Development Goals, States have to consider both sufficient guarantee of investor rights and the need for regulatory flexibility. The Indian alternative is quite bold. While the role of BITs to influence investment inflows is debatable, for India, no doubt, they played a major role in attracting foreign investments.95 This was mainly because foreign investors felt assured that their rights were adequately protected. 95 See Niti Bhasin and Rinku Manocha, ‘Do Bilateral Investment Treaties Promote FDI In- flows? Evidence from India’ (2016) 41(4) Vikalpa: J. Decision Makers 275. 96 Ibid. ( g g y ) 93 Prabhash Ranjan & Pushkar Anand, supra n 81. 93 Prabhash Ranjan & Pushkar Anand, supra n 81. 94 Ibid. g g y , ) 93 Prabhash Ranjan & Pushkar Anand, supra n 81. 94 Ibid 93 Prabhash Ranjan & Pushkar Anand, supra n 81. 94 Ibid. 98 Fabio Morosini and Mitchell Ratton Sanchez Badin, ‘Reconceptualizing International Investment Law from the Global South: An Introduction’ in Fabio Morosini and Michelle Ratton Sanchez Badin(Eds), Reconceptualizing International Investment Law from the Global South (Cam- bridge: Cambridge University Press, 2018). 3.3.0 India However, with the new model, investment inflows have significantly reduced.96 In conclusion, the model BIT is quite innovative and generally high- lights some boundaries that India is not willing to overstep. In addition, it ~ 67 ~ Nilfat Ali expressly limits the powers that an arbitral tribunal would normally exercise in traditional BITs, in relation to State’s regulatory functions. Undoubtedly, there are criticisms against the model BIT and allegations of it being ‘investor- unfriendly’. However, as observed, India seems to be ready to face such al- legations head-on.97 97 James J. Nedumpara, supra n 89. 4.0 Conclusion From the above discussions, it is clear that countries from the Global South are no longer willing to sit back as their sovereign rights are encroached upon by way of expansive and innovative BIT interpretations by arbitral tribunals. The legitimacy crisis, coupled with the lack of sustainability of the current ISDS regime was obviously the main crystalising factor that developing countries needed to take action and be proactive in the negotiations of investment agreements either between South-South or North-South fostering investment relations. To take back what was given up, by way of advancing the neoliberal ideological stance which was seen as the only viable option at that time, de- veloping countries are hands-on terminating traditional BITs and renegoti- ating more reasonable and balanced investment agreements. As seen, others like South Africa have opted for the domestic route in governing investment relations. Brazil was even more innovative in coming up with not only a pro- motion and protection of investments agreement but also one that facilitates productive investment for both the host State and the foreign investors. Whereas, from the above, it is only India that maintains ISDS in its new model BIT, it is evident that ISDS was cautiously incorporated with significant limits on arbitral tribunals’ powers in a bid to carve out regulatory space and maintain flexibility in legislation. The neoliberal narrative on economic order that has prevailed for so many years is now facing uncertainties.98 The Global South seems to be ~ 68 ~ Backlash Against ISDS writing a new economic order which has the capability of reshaping the global economic order at large. A paradigm shift from the competing interests of States and foreign investors to a sustainable development approach will achieve the necessary balance in ISDS. This model proposes incorporation of mechanisms that will focus on both the promotion and protection of investment without compromising on the health, security and environmental standards of the public, who in this case form a third party in investment relations. The proposed mechanisms will seek to take into account the interests of all stakeholders involved in an investment arbitration substantively and procedurally. 99 For instance, in the realm of international environmental law, it relates to the protection of the environment for the enjoyment of the present and the future generations, under human rights law it encompasses environmental sustainability, equitable development for the alienation of poverty, improved health standards, promote peace, protect human rights and pursue gender equality. 100 United Nations Declaration on the Right to Development, Article 1: The right to develop- ment is an inalienable human right by virtue of which every human person and all peoples are entitled to participate in, contribute to, and enjoy economic, social, cultural and political development, in which all human rights and fundamental freedoms can be fully realized, available at http://www. un.org/documents/ga/res/41/a41r128.htm, accessed December 20 2018. 101 J Anthony VanDuzer, Penelope Simons and Graham Mayeda, Integrating Sustainable De- velopment into International Investment Agreements: A Guide for Developing Country Negotiators (Commonwealth Secretariat, 2012). 102 M. Sornarajah, supra n 3. g g , 101 J Anthony VanDuzer, Penelope Simons and Graham Mayeda, Integrating Sustainable De- velopment into International Investment Agreements: A Guide for Developing Country Negotiators (Commonwealth Secretariat, 2012). 102 M. Sornarajah, supra n 3. 99 For instance, in the realm of international environmental law, it relates to the protection of the environment for the enjoyment of the present and the future generations, under human rights law it encompasses environmental sustainability, equitable development for the alienation of poverty, improved health standards, promote peace, protect human rights and pursue gender equality. 100 United Nations Declaration on the Right to Development, Article 1: The right to develop- ment is an inalienable human right by virtue of which every human person and all peoples are entitled to participate in, contribute to, and enjoy economic, social, cultural and political development, in which all human rights and fundamental freedoms can be fully realized, available at http://www. un.org/documents/ga/res/41/a41r128.htm, accessed December 20 2018. improved health standards, promote peace, protect human rights and pursue gender equality. 100 United Nations Declaration on the Right to Development, Article 1: The right to develop- ment is an inalienable human right by virtue of which every human person and all peoples are entitled to participate in, contribute to, and enjoy economic, social, cultural and political development, in which all human rights and fundamental freedoms can be fully realized, available at http://www. un.org/documents/ga/res/41/a41r128.htm, accessed December 20 2018. 101 J Anthony VanDuzer, Penelope Simons and Graham Mayeda, Integrating Sustainable De- velopment into International Investment Agreements: A Guide for Developing Country Negotiators 4.0 Conclusion Depending on the context, sustainable development can mean several things.99 This paper will adopt the United Nations’ conception of right to de- velopment, which encompasses facets of the definitions of sustainable devel- opment under the environmental, human rights and economics realms.100 Van- Duzer et al, advance this approach and in their formulations of what constitutes sustainable development, they observe that ‘economic growth is regarded as compatible with the preservation of the environment and positive social de- velopment including the alleviation of poverty in developing countries.’101 In this notion, development incorporates environmental protections, human health and welfare, human rights and the rights of indigenous people. On the role of IIAs to enhance foreign investments in developing countries, scholars have argued that the signing of an IIA does not necessarily mean increased foreign investments. In fact, according to Sonarajah, studies on this issue suggest that there is little correlation between the signing of IIAs and increase in investment in-flows.102 For instance, Brazil for many years did not have any IIA in place but was extremely successful in attracting ~ 69 ~ Nilfat Ali investments.103 However, according to investor surveys, the existence of a BIT has a positive impact on FDI flows into developing countries as this will offer investors protection, stability and predictability for FDI projects.104 Nonetheless, it is acknowledged that this is only one variable in the attraction of FDI. Other factors include domestic policy environment in a State, openness to investment and trade as well as transparency.105 Undoubtedly, IIAs have a significant role in the attraction of FDI in de- veloping countries. However, the main issue is how such IIAs are drafted regarding the reciprocal obligation of investors to enhance development in host States. Therefore, the increase in FDI does not necessarily result in the automatic development of the host State, even though this was the whole point for developing countries in the race to the bottom for FDI. As Sornarajah observes, ‘liberalization of investment flows was seen as the path to develop- ment, while competing models were regarded as unsuccessful.’106 There also exist competing models, being those that were based on strict regulation of the economy by the host State. He further rightly observes that states, and in particular, developing countries barely understood the repercussions of the treaties they adopted.107 This position is illustrated by the boom of IIAs and the consequential investor-state disputes. 103 Gabriel, supra n 58. 104 UCTAD, The Role of International Investment Agreements in Attracting Foreign Direct Investment to Developing Countries, UNCTAD Series on International Investment Policies for De- velopment, 2009, available at https://unctad.org/en/Docs/diaeia20095_en.pdf. 105 Ibid. 106 M.Sornarajah, supra n 3. 107 Ibid. 108 EU-Canada Comprehensive Economic and Trade Agreement (CETA), EU-Singapore, Can- ada-China 2012 BIT, Indian Model BIT, US 2012 Model. b d. 108 EU-Canada Comprehensive Economic and Trade Agreement (CETA), EU-Singapore, Can- ada-China 2012 BIT, Indian Model BIT, US 2012 Model. 4.0 Conclusion The attraction of investment is among the main agenda in the develop- ment strategies of developing countries. However, under traditional BITs, this need is seen as secondary to the protection of investors and their investments. The development of the host state is barely addressed. There is no doubt that there is need for a paradigm shift in the current status in investment treaties to address the sustainable development issues under BITs. The very recent or new era BITs recognise this deficit in the traditional BITs and incorporate certain clauses on regulation for public interest, provisions on corporate social responsibility, limitation on issues to be forwarded for arbitration and compo- nents for promotion of sustainable development for host States.108 In addition, ~ 70 ~ Backlash Against ISDS the vagueness and ambiguity of provisions is being cured by delineating clear and precise definitions of terms. The notion of sustainable development in ISDS encompasses elements and mechanisms that can be incorporated into ISDS which will not only achieve a fundamental balance between investor rights and the sovereign powers of the State to regulate in the public interest, but also take into account the interests of the crucial 3rd party: the public, in terms of sufficient representation and participation in ISDS. To achieve this goal, efforts should be made on several dimensions. The first dimension, which should also cover the substantive aspect of ISDS is on the initial stage of treaty negotiating and drafting. This stage also includes involvement of the public through active consultation: ensuring transparency and consistency. The other dimension to look at are the procedural aspect of ISDS, which will include the role of the tribunal in enhancing legitimacy in ISDS and the overall balance between investor protection and regulation in the public interest. This aspect will also include mechanisms of inclusion and representation of the public in ISDS and avenues of instilling confidence in arbitrators in ISDS. These aspects are discussed below in detail. 4.1.0 The Initial Stage and the Substantive Aspects of ISDS This stage mainly focuses on the incorporation of a sustainable development approach in IIAs and consequently in ISDS. At this stage, it is imperative that the host State is proactive in both treaty negotiation and drafting. In negotiations, the host State has to understand the nature of its obligations and at the same time carve out space for regulatory flexibility. In order to achieve this delicate balance in ISDS, IIA negotiations and drafting should be clear on the rights and the obligations of both parties. In particular, in order to achieve a sustainable development approach in ISDS, IIAs should be drafted with both protection of investment and promotion of sustainable development of the host State in mind. This can be achieved through, delineating clear and precise definitions of terms to minimise the expansive interpretations by arbitral tribunals, express provision on matters an arbitral tribunal can exercise jurisdiction and most importantly carve out space for legislative flexibility which can be done through clarification of scope and meaning of treaty provisions and by using specific flexibility mechanisms ~ 71 ~ Nilfat Ali such as exceptions and reservations.109 Further, States can shield themselves against unjustified liability and high procedural costs through treaty design by looking into options on how dispute settlement is conducted and in the scope and application of treaty clauses.110 On dispute settlement, an IIA can set out preconditions on ISDS. For instance, exhaustion of domestic judicial systems within the host State, the use of other Alternative Dispute Resolution (ADR) mechanisms that can be utilised prior to institution of ISDS as a last resort and exploit other innovative mechanisms, such as, opting for State-State arbitrations for matters arising out of pre-establishments rights provisions, qualifying consent to ISDS and providing for the possibility of counter-claims by States.111 On the other hand, to ensure protection of investors, IIAs should provide express provisions on compensation for expropriation. As Giovanni notes, in the presence of an express provision on compensation for an expropriation, States are bound by such clauses and thus cannot disregard an obligation voluntarily assumed.112 4.2.0 Procedural Aspects of ISDS To fully enmesh a sustainable development approach in ISDS, the procedural aspects of ISDS too have to be reformed. This goal can be achieved through several avenues as discussed below. y p , , p tad.org/en/PublicationsLibrary/diaepcb2015d5_en.pdf, accessed 6 May 2019. 110 Ibid. 111 Ibid. 112 Giovanni, supra n 18. 113 See generally, Annalisa M. Leibold , ‘The Friction Between Investor Protection and Human Rights: Lessons from Foresti v South Africa’ [2015] Houston Journal of International Law, Forth- coming, available at https://ssrn.com/abstract=2586252, accessed 20 April 2019. 114 Philip Morris Brands Sàrl, Philip Morris Products S.A. and Abal Hermanos S.A. v Oriental Republic of Uruguay, ICSID Case No. ARB/10/7. 109 UNCTAD Investment Policy for Sustainable Development, 2015, available at ht 109 UNCTAD Investment Policy for Sustainable Development, 2015, available at https://unc- org/en/PublicationsLibrary/diaepcb2015d5_en.pdf, accessed 6 May 2019. 110 Ibid. 111 Ibid 109 UNCTAD Investment Policy for Sustainable Development, 2015, available at https://unc- org/en/PublicationsLibrary/diaepcb2015d5_en.pdf, accessed 6 May 2019. 115 Giovanni supra n 18, non-commercial values as those “not pertaining to the protection of property but relating to the safeguard of other essential interests such as environment and human health” also see Mary E. Footer, ‘Bits And Pieces: Social And Environmental Protection In The Regulation of Foreign Investment’ (2009) 18(1) Michigan State Journal of International Law, 33. 116 Ibid at para 305. 117 Methanex Corporation v United States of America,NAFTA/UCITRAL Award,03/08/2005, Chemtura Corporation v Government of Canada, NAFTA/UNCITRAL Award,02/08/2010, Philip Morris Brands Sàrl, Philip Morris Products S.A. and Abal Hermanos S.A. v Oriental Republic of Uruguay, ICSID Case No. ARB/10/7. 118 Foresti case, supra n 43 119 Ibid. 120 UNCTAD, supra n 105. 121 Vanduzer et al., supra n 98. 4.2.1.0 Legitimacy Enhancements Essentially, these are mechanisms adopted by a tribunal to alleviate pressure from critics and at the same time strengthen the legitimacy of ISDS.113 Arbitral tribunals are seen to be embracing this nuanced approach in a bid to enhance their legitimacy. For instance, in the recent Philip Morris decision,114 ~ 72 ~ Backlash Against ISDS the tribunal took into consideration the so-called non-commercial values115 in an attempt to strike a balance between State obligations and investor rights. In this case, the tribunal found that the government measures in issue, had been adopted bona fides for the purpose of protecting public welfare, were non- discriminatory and were proportionate. That they were not to be considered as an expropriation and, therefore, the Claimants did not have any right to compensation despite the negative effect on their business.116 This tendency of legitimacy enhancement by tribunals is further observed in various other awards117 including, the Foresti case118 where the tribunal allowed for further ‘Non-Disputing Party’ participation and shifted costs to the Claimant due to the friction generated from a case in which an investor’s rights directly conflicted with a state’s right to enact positive human rights legislation.119 Another way an arbitral tribunal can enhance its legitimacy can be by taking into account investor behaviour when settling investor-state disputes.120 This inclination by tribunals is commendable and a significant step towards the revival of legitimacy in ISDS. 115 Giovanni supra n 18, non-commercial values as those “not pertaining to the protection of property but relating to the safeguard of other essential interests such as environment and human health” also see Mary E. Footer, ‘Bits And Pieces: Social And Environmental Protection In The Regulation of Foreign Investment’ (2009) 18(1) Michigan State Journal of International Law, 33. 4.2.2.0 Public Participation Participation of the members of the public in matters that have significant social and economic impact on them is another limb of sustainable development. As VanDuzer et al observe, to ensure sustainable development through international investment rules, ‘they should be developed through wide consultation with people in the host country and decisions about the negotiation, application and interpretation of agreements should be transparent and consistent.’121 This participation extrapolates to dispute settlement too. Members of the public are a significant stakeholder in ISDS as taxpayers. ~ 73 ~ Nilfat Ali This participation can be achieved through amicus submissions. Amicus participation has several advantages: aids in the protection of public interests, enhances transparency, improves the quality of the award and enhances public scrutiny of the process.122 In the current ISDS regime, amicus briefs are a rarity.123 It has been argued that there is an inherent conflict in the ICSID Rules on the admissibility of amicus briefs.124 However, as observed, amicus briefs are now becoming a significant player in matters involving important public policy considerations.125 In fact, in the recent case of Achmea v Slovak,126 which is deemed as a significant milestone in ISDS history on amici briefs, the amicus submissions were made at the invitation of the tribunal.127 122 Manjiao Chi, Integrating Sustainable Development in International Investment Law: Nor- mative Incompatibility, System Integration and Governance Implications, (Routledge, 2017) 123. 123 Emmert and Esenkulova supra n 14. 124 Ibid, “On the one hand, Rule 37(2)(a) requires that “the non-disputing party submission would assist the Tribunal in the determination of a factual or legal issue related to the proceeding by bringing a perspective, particular knowledge or insight that is different from that of the disputing par- ties”. On the other hand, unless the disputing parties give broad consent, the non-disputing party will have very limited rights and even more limited access. But how are the amici supposed to know what they might be able to add beyond what is already presented to the tribunal by the disputing parties, if they do not have access to the files?”. 125 Ibid. 126 Manjiao Chi, Ibid at 122. 127 Ibid. 128 These can include pleadings and submissions filed by the parties, submissions of non-dis- puting parties, procedural decisions, arbitral award and records of the hearing. 129 2012 U.S. Model BIT. 122 Manjiao Chi, Integrating Sustainable Development in International Investment Law: Nor- Incompatibility, System Integration and Governance Implications, (Routledge, 2017) 123. 123 128 These can include pleadings and submissions filed by the parties, submissions of non-dis- puting parties, procedural decisions, arbitral award and records of the hearing. 1292012 U S M d l BIT 129 2012 U.S. Model BIT. 130 EU-Canada Comprehensive Economic and Trade Agreement. 131 Canada-China BIT as quoted in Manjiao Chi, supra n 117 at 123. 132 Nigel Blackaby, Public Interest and Investment Treaty Arbitration, in International Com- mercial Arbitration: Important Contemporary Questions (Albert Jan van den Berg ed., 2002) as quoted in Susan D. Frank, ‘The Legitimacy Crisis in Investment Arbitration: Privatizing Public In- ternational Law Through Inconsistent Decisions’ (2005) 73 Fordham Law Review 1521. 133 Susan D. Frank, supra n 127. 134 Ibid. 135 Ibid. 4.2.3.0 Transparency It is acknowledged that arbitration as an alternative dispute settlement mechanism is a private process between a claimant and a respondent. However, in the context of ISDS, the respondent is a State and the implications of the process are inevitably extended to the public. Transparency in this respect can be enhanced through making available/public documents128 related to the dispute, public access to the hearing and as mentioned above admission of amicus submissions. New era BITs seem keen on incorporating transparency-related clauses. For instance, the 2012 U.S BIT model contains provisions titled “Transparency of Arbitral Proceedings”;129 Comprehensive Economic and Trade Agreement 128 These can include pleadings and submissions filed by the parties, submissions of non-dis- puting parties, procedural decisions, arbitral award and records of the hearing. 129 2012 U.S. Model BIT. ~ 74 ~ ~ 74 ~ Backlash Against ISDS (CETA) adopts UNCITRAL Transparency Rules;130 and in the Canada-China BIT, an arbitral tribunal has no power to decide whether hearings should be made public and that decision is left to the disputing State subject to consultation with the disputing investor.131 136 Zeng Huaqun, ‘Balance, Sustainable Development, Integration: Innovative Path for BIT Practice’ (2014) 17 Journal of International Economic Law, 299. 4.2.4.0 Reform of Arbitration Rules These reforms should be aimed at revision of institutional rules governing international arbitrations to curb the problem of inconsistency in decisions arising out of ISDS. Such revisions are such that will allow more transparency of arbitral proceedings and can include an establishment of review bodies to permit review of investment treaty awards.132 It has been suggested that this option is more practical and has a chance of being established efficiently compared to seeking amendment of treaties which can be cumbersome and less effective.133 The reforms suggested include establishment of working groups from different backgrounds drawn from arbitration practitioners, international lawyers, former government officials, among others to establish a uniform code of best practices for ISDS that would take into account the inherent uniqueness of ISDS and public implications of investment treaty arbitrations. Best practices can include, heightening or allowing transparency of arbitration proceedings through publication of awards and access to documentation in ISDS especially in matters that generate public interest and revision of cost-shifting rules to give tribunals clearer guidance on when it would be appropriate to shift costs to discourage institution of unmeritorious claims.134 However, the establishment of these protocols would require instigation of political will to revise or supplement the current rules.135 For instance, for ICSID Rules, there could be constraints arising from the ICSID Convention which may prove to be a real barrier to reform in this area. ~ 75 ~ Nilfat Ali ISDS is the main form of dispute resolution platform relied upon by parties in investment treaties as it is perceived to offer a neutral and impartial forum for resolving disputes between States and foreign investors. However, the regime is currently facing a legitimacy crisis due to the mounting criticisms ranging from inconsistencies in decisions, costs of arbitration and even credibility of arbitrators. In addition, the inherent structural imbalance of BITs heavily undermines ISDS from playing a more constructive role in investment governance. Mechanisms must be incorporated to strike a balance between promotion and protection of investment and the sovereign right of the State to regulate in the public interest. A paradigm shift to a sustainable development approach in ISDS taking into account the rights of all stakeholders involved in investment arbitration is the most promising avenue to revive its legitimacy. As noted, ‘[d] eveloping states have to change their traditional passive stance and participate positively in international investment policy making, and make their great efforts and contributions as pioneers for the initiative and development of BIT practice with their developing and developed counterparts.’136 ~ 76 ~ ~ 76 ~
https://openalex.org/W4232147445	https://dwes.copernicus.org/preprints/dwes-2020-37/dwes-2020-37.pdf	English	null	Comment on dwes-2020-37	null	2,021	cc-by	15,876	Drinking Water Engineering and Science Discussions Open Access Drinking Water Engineering and Science Discussions Open Access https://doi.org/10.5194/dwes-2020-37 Preprint. Discussion started: 5 February 2021 c⃝Author(s) 2021. CC BY 4.0 License. Abstract In this research, reliability indicators of water distribution networks are evaluated under pipe failure conditions. The case studies include two benchmark and one real-life water distribution networks in Iran with more hydraulic constraints. Some important reliability indicators are presented such as resilience index, network resilience, modified resilience index and minimum surplus In this research, reliability indicators of water distribution networks are evaluated under pipe failure conditions. The case studies include two benchmark and one real-life water distribution networks in Iran with more hydraulic constraints. Some important reliability indicators are presented such as resilience index, network resilience, modified resilience index and minimum surplus head index. GANetXL is used to do one-objective and two-objective optimization of the previously mentioned water distribution 20 networks in order to not only minimize the cost, but also maximize the reliability indicators. Moreover, the results of a statistical analysis for each pipe is used to determine the sensitive pipes that are of the most failure probability. GANetXL is an optimization tool in Excel environment and works based on Genetic Algorithm. GANetXL has the capability of being linked to EPANET (Hydraulic simulation software). The results obtained clearly show that network resilience index is of poor performance when head index. GANetXL is used to do one-objective and two-objective optimization of the previously mentioned water distribution 20 networks in order to not only minimize the cost, but also maximize the reliability indicators. Moreover, the results of a statistical analysis for each pipe is used to determine the sensitive pipes that are of the most failure probability. GANetXL is an optimization tool in Excel environment and works based on Genetic Algorithm. GANetXL has the capability of being linked to EPANET (Hydraulic simulation software). The results obtained clearly show that network resilience index is of poor performance when head index. GANetXL is used to do one-objective and two-objective optimization of the previously mentioned water distribution 20 networks in order to not only minimize the cost, but also maximize the reliability indicators. Moreover, the results of a statistical analysis for each pipe is used to determine the sensitive pipes that are of the most failure probability. GANetXL is an optimization tool in Excel environment and works based on Genetic Algorithm. GANetXL has the capability of being linked to EPANET (Hydraulic simulation software). The Evaluation of Reliability Indices in Water Distribution Networks under Pipe Failure Condition Alireza Moghaddam1, Roya Peirovi-Minaee2, Hossein Rezaei 3, Alireza Faridhosseini 4,5, Ali Naghi Ziaei 4 1 PhD in Water Resources Engineering, Department of Civil Engineering, University of Gonabad, Gonabad, Iran 5 2Assistant Professor, Department of Environmental Health Engineering, Faculty of Public Health, Social Determinants of Health Research Center, Gonabad University of Medical Sciences, Gonabad, Iran 3 Professor, Department of Water Engineering, College of Agriculture, University of Urmia, Urmia, Iran 4Associate Professor, Department of Water Engineering, College of Agriculture, Ferdowsi University of Masshad, Mashhad, Iran 10 5Visiting Professor, College of Engineering, Mathematics & Physical Sciences, University of Exeter, Exeter, UK Alireza Moghaddam1, Roya Peirovi-Minaee2, Hossein Rezaei 3, Alireza Faridhosseini 4,5, Ali Naghi Ziaei 4 1 PhD in Water Resources Engineering, Department of Civil Engineering, University of Gonabad, Gonabad, Iran 5 2Assistant Professor, Department of Environmental Health Engineering, Faculty of Public Health, Social Determinants of Health Research Center, Gonabad University of Medical Sciences, Gonabad, Iran 3 Professor, Department of Water Engineering, College of Agriculture, University of Urmia, Urmia, Iran 4Associate Professor, Department of Water Engineering, College of Agriculture, Ferdowsi University of Masshad, Mashhad, Iran 10 5Visiting Professor, College of Engineering, Mathematics & Physical Sciences, University of Exeter, Exeter, UK Visiting Professor, College of Engineering, Mathematics & Physical Sciences, University of Exeter, Exeter, UK Abstract The results obtained clearly show that network resilience index is of poor performance when compared with the other indexes under pipe failure conditions, especially in real-life networks that include small pipe diameters. 25 It was also showed that if a water distribution network was optimized only in terms of cost, there would be an unacceptable pressure drop at some nodes in case of pipe failure. compared with the other indexes under pipe failure conditions, especially in real-life networks that include small pipe diameters. 25 It was also showed that if a water distribution network was optimized only in terms of cost, there would be an unacceptable pressure drop at some nodes in case of pipe failure. Keywords: GANetXL, Optimization, Pipe Reliability, Resiliency, Water distribution Network Corresponding Author(fist): E-mail addresses: Peirovi.r@gmu.ac.ir Corresponding Author(fist): E-mail addresses: Peirovi.r@gmu.ac.ir Corresponding Author(second): E-mail addresses: h.rezaie@urmia.ac.ir Corresponding Author(second): E-mail addresses: h.rezaie@urmia.ac.ir 15 1. Introduction 30 1. Introduction 30 Water distribution networks (WDNs) are designed to provide users with a minimum acceptable level of supply, in terms of pressure, availability, and water quality at all times under a range of operating conditions. Nowadays, WDNs have become . t oduct o 30 Water distribution networks (WDNs) are designed to provide users with a minimum acceptable level of supply, in terms of pressure, availability, and water quality at all times under a range of operating conditions. Nowadays, WDNs have become 1 1 Drinking Water Engineering and Science Discussions Open Access https://doi.org/10.5194/dwes-2020-37 Preprint. Discussion started: 5 February 2021 c⃝Author(s) 2021. CC BY 4.0 License. https://doi.org/10.5194/dwes-2020-37 Preprint. Discussion started: 5 February 2021 c⃝Author(s) 2021. CC BY 4.0 License. complex and need huge investments in construction and maintenance. As a result, there is an avid desire to improve their efficiency through minimizing their cost and maximizing their benefit. complex and need huge investments in construction and maintenance. As a result, there is an avid desire to improve their efficiency through minimizing their cost and maximizing their benefit. Optimal WDN design is a computationally complex problem because of its non-linear nature and the constraints involved. 35 Therefore, finding the globally optimal solution is difficult if we use optimization methods as the non-linearity is significant. A majority of the previous models have utilized different optimization algorithms for ordinary benchmark networks meaning that optimal design problems have turned out to be more of mathematical or computational challenges rather than civil engineering ones. Optimal WDN design is a computationally complex problem because of its non-linear nature and the constraints involved. 35 Therefore, finding the globally optimal solution is difficult if we use optimization methods as the non-linearity is significant. A majority of the previous models have utilized different optimization algorithms for ordinary benchmark networks meaning that optimal design problems have turned out to be more of mathematical or computational challenges rather than civil engineering ones. In the last three decades, several researchers have broadly studied the design optimization problem of WDNs. The problems 40 have been solved using linear, non-linear and various meta-heuristic methods. Linear and non-linear methods were predominantly used in the period 1960–1990(Jacoby 1968, Watanatada 1973, Alperovits and Shamir 1977, Quindry, Liebman et al. 1981, Lansey and Mays 1989, Fujiwara and Khang 1990). Linear methods applied to nonlinear problems have not resulted in optimal solutions. 1. Introduction 30 The non-linear methods did not necessarily yield a global optimum, and the final solution depended on the In the last three decades, several researchers have broadly studied the design optimization problem of WDNs. The problems 40 have been solved using linear, non-linear and various meta-heuristic methods. Linear and non-linear methods were predominantly used in the period 1960–1990(Jacoby 1968, Watanatada 1973, Alperovits and Shamir 1977, Quindry, Liebman et al. 1981, Lansey and Mays 1989, Fujiwara and Khang 1990). Linear methods applied to nonlinear problems have not resulted in optimal solutions. The non-linear methods did not necessarily yield a global optimum, and the final solution depended on the initial solution used as a starting point for the search procedure (Piratla 2016). In addition, the use of discrete variables, specific- 45 size pipe diameters, limits the quality of the optimal solution obtained. These limitations led to the employment of meta- heuristics that use stochastic optimization methods. Murphy and Simpson were the first researchers who used a simple Genetic Algorithm (GA) to optimally design water (Murphy and Simpson 1992). GA has been integrated with hydraulics simulator to optimize the solutions by many researchers 50 (Simpson, Dandy et al. 1994, Simpson and Goldberg 1994, Savic and Walters 1997, Lippai, Heaney et al. 1999, Neelakantan, Suribabu et al. 2008). (Vasan and Simonovic 2010) recently applied a differential evolutionary algorithm (DE), an improved GA. The major difference between GA and DE is that GA relies on crossover, a mechanism of probabilistic exchange of information among solutions to create better solutions, while DE uses mutation as the primary search mechanism. DE uses a uniform crossover that can take child vector parameters from one parent more often than from the other one. It is said that GA 55 most of the times succeed in finding the global optimum or at least arriving at somewhere very close to it. More importantly, GA is capable of handling discrete optimization (as pipe diameters are discrete) (Savic and Walters 1997). Many other optimization algorithms have been used in the optimal design of water distribution systems (Tayfur 2017). (Loganathan, Greene et al. 1995) and (Cunha and Sousa 1999) applied simulated annealing for optimal design of water distribution systems. (Geem, Kim et al. 2002) developed a harmony search optimization approach to solve network design 60 problems while (Eusuff and Lansey 2003) developed the shuffled frog leaping algorithm. (Maier, Simpson et al. 1. Introduction 30 2003) applied the ant colony optimization approach and improved GA both in terms of computational efficiency and its ability to find nearly optimal solutions. (Baños, Gil et al. 2007) analyzed the performance of memetic algorithms for optimal design of looped water distribution systems and demonstrated that it works well for problems of large scale. (Mohan and Babu 2009) proposed to use a heuristic based approach called heuristics-based algorithm (HBA) to identify the least cost combination of pipe diameters. They 65 demonstrated that the HBA is capable of identifying the least cost combination of pipe diameters with fewer numbers of evaluations. (Moghaddam, Alizadeh et al. 2018) applied a Simple Modified Particle Swarm Optimization (SMPSO) to minimize the cost of water distribution networks. SMPSO then used a novel factor to decrease the inertia weight of the algorithm in proportion with simulation time to facilitate both global and local search. heuristic based approach called heuristics-based algorithm (HBA) to identify the least cost combination of pipe diameters. They 65 demonstrated that the HBA is capable of identifying the least cost combination of pipe diameters with fewer numbers of evaluations. (Moghaddam, Alizadeh et al. 2018) applied a Simple Modified Particle Swarm Optimization (SMPSO) to minimize the cost of water distribution networks. SMPSO then used a novel factor to decrease the inertia weight of the algorithm in proportion with simulation time to facilitate both global and local search. 2 2 Drinking Water Engineering and Science Discussions Open Access https://doi.org/10.5194/dwes-2020-37 Preprint. Discussion started: 5 February 2021 c⃝Author(s) 2021. CC BY 4.0 License. https://doi.org/10.5194/dwes-2020-37 Preprint. Discussion started: 5 February 2021 c⃝Author(s) 2021. CC BY 4.0 License. Objective function is important in optimizing the design of distribution systems. The main negative aspect of the single-objective 70 constrained formulation is that it does not effectively set up a trade-off between cost and reliability/robustness of a design (Todini 2000). Reliability can be considered as the ability of providing an adequate supply under both usual and unusual conditions (Farmani, Savic et al. 2005), including demand uncertainty, pipe failure, etc. One of the most used reliability criteria is the concept of resilience index suggested by (Todini 2000),which is a measure of the ability of the network to handle failures and is Objective function is important in optimizing the design of distribution systems. 1. Introduction 30 There are a few applications of 85 GANetXL in water systems, which include the development of a model for optimal management of groundwater contamination (Farmani, Savic et al. 2005, Farmani, Henriksen et al. 2009) and multi-objective optimization of water distribution systems (Piratla and Ariaratnam 2012, Mala-Jetmarova, Barton et al. 2015, Piratla 2016). in Iran in which there are more hydraulic constraints compared with the benchmark networks. There are a few applications of 85 GANetXL in water systems, which include the development of a model for optimal management of groundwater contamination (Farmani, Savic et al. 2005, Farmani, Henriksen et al. 2009) and multi-objective optimization of water distribution systems (Piratla and Ariaratnam 2012, Mala-Jetmarova, Barton et al. 2015, Piratla 2016). 2. Material and Methods 90 2.1 Optimization Model for WDN Design In this paper, WDNs are optimized with pipe diameters as decision variables. Cost is considered as the objective function that must be minimized [Eq. (1)] and the reliability criteria are modeled in the form of a two-objective function [Eq. (2)]. 𝑀𝑖 𝑓 ∑ 𝐷× 𝑙 𝑁 (1) 𝑀𝑖𝑛 𝑓1 = ∑ 𝑐𝑖𝐷𝑖× 𝑙𝑖 𝑁 𝑖=1 𝑀𝑖𝑛 𝑓1 = ∑ 𝑐𝑖𝐷𝑖× 𝑙𝑖 𝑁 𝑖=1 𝑀𝑎𝑥 𝑓2 = 𝑅𝑒𝑙𝑖𝑎𝑏𝑖𝑙𝑖𝑡𝑦 𝑓𝑎𝑐𝑡𝑜𝑟 95 𝑀𝑎𝑥 𝑓2 = 𝑅𝑒𝑙𝑖𝑎𝑏𝑖𝑙𝑖𝑡𝑦 𝑓𝑎𝑐𝑡𝑜𝑟 (2) 95 Where 𝑓1 is network cost, 𝑓2 is network reliability, 𝑐𝑖 is cost for unit length of pipe with diameter, 𝐷𝑖 length 𝑙𝑖 and 𝑁 is pipe numbers in the network. Where 𝑓1 is network cost, 𝑓2 is network reliability, 𝑐𝑖 is cost for unit length of pipe with diameter, 𝐷𝑖 length 𝑙𝑖 and 𝑁 is pipe numbers in the network. Where 𝑓1 is network cost, 𝑓2 is network reliability, 𝑐𝑖 is cost for unit length of pipe with diameter, 𝐷𝑖 length 𝑙𝑖 and 𝑁 is pipe numbers in the network. 1. Introduction 30 The main negative aspect of the single-objective 70 constrained formulation is that it does not effectively set up a trade-off between cost and reliability/robustness of a design (Todini 2000). Reliability can be considered as the ability of providing an adequate supply under both usual and unusual conditions (Farmani, Savic et al. 2005), including demand uncertainty, pipe failure, etc. One of the most used reliability criteria is the concept of resilience index suggested by (Todini 2000),which is a measure of the ability of the network to handle failures and is related indirectly to system reliability. Several suggestions were made to modify the resilience index introduced by Todini 75 (Prasad and Park 2004, Farmani, Savic et al. 2005, Jayaram and Srinivasan 2008, Reca, Martinez et al. 2008, Raad, Sinske et al. 2010, Baños, Reca et al. 2011, Greco, Di Nardo et al. 2012, Pandit and Crittenden 2012). Literature review shows that stochastic models, particularly the GA types, give better results than linear and non-linear optimization models (Pandit and Crittenden 2012). Subsequently, a genetic algorithm technique is used in this research as a part of GANetXL(Savić, Bicik et al. 2011) , an add-in to Microsoft Excel. GANetXL is used to optimize two benchmark networks 80 from literature (Two-loop and Hanoi water networks) in two different conditions including single-objective (cost) and two- objective (cost and reliability criteria) optimizations. Afterwards, the solutions obtained, as well as the performance of the proposed Resilience Index, Network Resilience, Modified Resilience Index and Minimum Surplus Head Index are discussed. Finally, as the results obtained for the benchmark networks are satisfactory, GANetXL is used to design a real-life water network of GANetXL(Savić, Bicik et al. 2011) , an add-in to Microsoft Excel. GANetXL is used to optimize two benchmark networks 80 from literature (Two-loop and Hanoi water networks) in two different conditions including single-objective (cost) and two- objective (cost and reliability criteria) optimizations. Afterwards, the solutions obtained, as well as the performance of the proposed Resilience Index, Network Resilience, Modified Resilience Index and Minimum Surplus Head Index are discussed. Finally, as the results obtained for the benchmark networks are satisfactory, GANetXL is used to design a real-life water network in Iran in which there are more hydraulic constraints compared with the benchmark networks. 2.2 Constraints head available at node 𝑗; 𝐻𝑗 𝑚𝑖𝑛= minimum hydraulic-head required at node 𝑗; 𝐻𝑗 𝑚𝑎𝑥=maximum hydraulic-head at node 𝑗; 𝑛𝑑 = 110 number of demand nodes; 𝑉𝑖 𝑚𝑖𝑛= minimum velocity required at pipe 𝑖 and 𝑉𝑖 𝑚𝑎𝑥=maximum velocity at pipe 𝑖; 𝑛𝑝 = number of pipes. 2.3.1 Resilience Index (𝐼𝑟) 120 Todini’s resilience index is a popular surrogate measure within the WDN research field (Todini 2000, Reca, Martinez et al. 2008, Atkinson, Farmani et al. 2014). It considers surplus hydraulic power as a proportion of available hydraulic power. The resilience index,𝐼𝑟, is measured in the continuous range of [0-1] (for feasible solutions of 𝐻𝑗 𝑚𝑖𝑛≤𝐻𝑗) and is formulated as below (Todini 2000): Todini’s resilience index is a popular surrogate measure within the WDN research field (Todini 2000, Reca, Martinez et al. 2008, Atkinson, Farmani et al. 2014). It considers surplus hydraulic power as a proportion of available hydraulic power. The resilience index,𝐼𝑟, is measured in the continuous range of [0-1] (for feasible solutions of 𝐻𝑗 𝑚𝑖𝑛≤𝐻𝑗) and is formulated as below (Todini 2000): 𝐼𝑟= ∑ 𝑞𝑗(𝐻𝑗−𝐻𝑗 𝑚𝑖𝑛 𝑛𝑛 𝑗=1 ) ∑ 𝑄𝑘 𝑛𝑟 𝑘=1 𝐻𝑘+∑ 𝑃𝑖𝛾 ⁄ −∑ 𝑞𝑗𝐻𝑗 𝑚𝑖𝑛 𝑛𝑛 𝑗=1 𝑛𝑝 𝑖=1 (6) 125 𝐼𝑟= ∑ 𝑞𝑗(𝐻𝑗−𝐻𝑗 𝑚𝑖𝑛 𝑛𝑛 𝑗=1 ) ∑ 𝑄𝑘 𝑛𝑟 𝑘=1 𝐻𝑘+∑ 𝑃𝑖𝛾 ⁄ −∑ 𝑞𝑗𝐻𝑗 𝑚𝑖𝑛 𝑛𝑛 𝑗=1 𝑛𝑝 𝑖=1 (6) 125 Where 𝑛𝑛 is the number of supply and demand nodes; 𝑛𝑟 is the set of supply nodes (reservoir/emptying tanks); 𝑛𝑝 denotes the number of pumps; 𝐻𝑗is the available head at supply node 𝑗; 𝐻𝑗 𝑚𝑖𝑛 represents the required head at supply node 𝑗; 𝑞𝑗 is the demand at node 𝑗; 𝑄𝑘is the supply at input node 𝑘; 𝐻𝑘is representative of head associated with the input node 𝑘; 𝑃𝑖 is the power of pump 𝑖; and finally 𝛾 is the specific weight of water. Maximization of the resilience index improves the ability of a pipeline (6) Where 𝑛𝑛 is the number of supply and demand nodes; 𝑛𝑟 is the set of supply nodes (reservoir/emptying tanks); 𝑛𝑝 denotes the number of pumps; 𝐻𝑗is the available head at supply node 𝑗; 𝐻𝑗 𝑚𝑖𝑛 represents the required head at supply node 𝑗; 𝑞𝑗 is the demand at node 𝑗; 𝑄𝑘is the supply at input node 𝑘; 𝐻𝑘is representative of head associated with the input node 𝑘; 𝑃𝑖 is the power of pump 𝑖; and finally 𝛾 is the specific weight of water. Maximization of the resilience index improves the ability of a pipeline network in encountering failure conditions 130 2.3 Reliability Indicators A range of reliability criteria has been introduced to different degrees of complexity. Usually, these criteria give some suggestion 115 of the ability of a WDN to handle changing conditions and are straightforward to analyze so are practical for optimization studies that compare the performance of network design. This section presents the definition of the key criteria and their derivatives as well as the advantages and disadvantages of them. 2.2 Constraints The constraints to the optimization problem are as follows: 100 The constraints to the optimization problem are as follows: 100 1) Explicit system constraints such as conservation of mass of flow, conservation of energy and conservation of mass of constituent, which all are controlled by water network simulator software, EPANET (Rossman 2000, Mala-Jetmarova, Barton et al. 2015). 2) Implicit bound constraints, which include choosing pipe diameters from a commercially available set of discrete pipe sizes [Eq. (3)], minimum and maximum pressure at load nodes [Eq. (4)], minimum and maximum velocity in pipes [Eq. (5)]. 105 𝐷𝑖∈{𝐶𝐷𝑘} ∀𝑖 𝑘= 1,2,3, … , 𝑛𝑐 (3) The constraints to the optimization problem are as follows: 100 1) Explicit system constraints such as conservation of mass of flow, conservation of energy and conservation of mass of constituent, which all are controlled by water network simulator software, EPANET (Rossman 2000, Mala-Jetmarova, Barton et al. 2015). 2) Implicit bound constraints, which include choosing pipe diameters from a commercially available set of discrete pipe sizes [Eq. (3)], minimum and maximum pressure at load nodes [Eq. (4)], minimum and maximum velocity in pipes [Eq. (5)]. 105 2) Implicit bound constraints, which include choosing pipe diameters from a commercially available set of discrete pipe sizes [Eq. (3)], minimum and maximum pressure at load nodes [Eq. (4)], minimum and maximum velocity in pipes [Eq. (5)]. 105 𝐷𝑖∈{𝐶𝐷𝑘} ∀𝑖 𝑘= 1,2,3, … , 𝑛𝑐 (3) 3 3 Drinking Water Engineering and Science Discussions Open Access https://doi.org/10.5194/dwes-2020-37 Preprint. Discussion started: 5 February 2021 c⃝Author(s) 2021. CC BY 4.0 License. 𝐻𝑗 𝑚𝑖𝑛≤𝐻𝑗 ≤𝐻𝑗 𝑚𝑎𝑥, 𝑗= 1,2, … , 𝑛𝑑 (4) 𝑉𝑖 𝑚𝑖𝑛≤𝑉𝑖 ≤𝑉𝑖 𝑚𝑎𝑥, 𝑖= 1,2, … , 𝑛𝑝 (5) where 𝐷𝑖 = diameter of pipe 𝑖; 𝐶𝐷𝑘= kth commercially available pipe size; 𝑛𝑐 = number of available pipe sizes; 𝐻𝑗= hydraulic- (5) ere 𝐷𝑖 = diameter of pipe 𝑖; 𝐶𝐷𝑘= kth commercially available pipe size; 𝑛𝑐 = number of available pipe sizes; 𝐻𝑗= h head available at node 𝑗; 𝐻𝑗 𝑚𝑖𝑛= minimum hydraulic-head required at node 𝑗; 𝐻𝑗 𝑚𝑎𝑥=maximum hydraulic-head at node 𝑗; 𝑛𝑑 = 110 number of demand nodes; 𝑉𝑖 𝑚𝑖𝑛= minimum velocity required at pipe 𝑖 and 𝑉𝑖 𝑚𝑎𝑥=maximum velocity at pipe 𝑖; 𝑛𝑝 = number of pipes. 2.3.2 Network Resilience (𝐼𝑛) Prasad and Park (2004) introduced another reliability measure called network resilience (𝐼𝑛), which incorporates the effects of both surplus power and reliable loops. Reliable loops can be ensured if the pipes connected to the same node do not vary greatly in diameter. If 𝐷1𝑗, 𝐷2𝑗, ..., 𝐷𝑛𝑝𝑗 (where 𝐷1𝑗≥𝐷2𝑗≥⋯≥𝐷𝑛𝑝𝑗) are the diameters of the np pipes connected to node j, then 135 uniformity of that node is given by Eq. 7, Prasad and Park (2004) introduced another reliability measure called network resilience (𝐼𝑛), which incorporates the effects of both surplus power and reliable loops. Reliable loops can be ensured if the pipes connected to the same node do not vary greatly in diameter. If 𝐷1𝑗, 𝐷2𝑗, ..., 𝐷𝑛𝑝𝑗 (where 𝐷1𝑗≥𝐷2𝑗≥⋯≥𝐷𝑛𝑝𝑗) are the diameters of the np pipes connected to node j, then 135 uniformity of that node is given by Eq. 7, Prasad and Park (2004) introduced another reliability measure called network resilience (𝐼𝑛), which incorporates the effects of both surplus power and reliable loops. Reliable loops can be ensured if the pipes connected to the same node do not vary greatly both surplus power and reliable loops. Reliable loops can be ensured if the pipes connected to the same node do not vary greatly in diameter. If 𝐷1𝑗, 𝐷2𝑗, ..., 𝐷𝑛𝑝𝑗 (where 𝐷1𝑗≥𝐷2𝑗≥⋯≥𝐷𝑛𝑝𝑗) are the diameters of the np pipes connected to node j, then 135 uniformity of that node is given by Eq. 7, in diameter. If 𝐷1𝑗, 𝐷2𝑗, ..., 𝐷𝑛𝑝𝑗 (where 𝐷1𝑗≥𝐷2𝑗≥⋯≥𝐷𝑛𝑝𝑗) are the diameters of the np pipes connected to node j, then 135 uniformity of that node is given by Eq. 7, 𝐶𝑗= ∑ 𝐷𝑖𝑗 𝑛𝑝 𝑖=1 𝑛𝑝×max 𝐷𝑖𝑗 𝐶𝑗= ∑ 𝐷𝑖𝑗 𝑛𝑝 𝑖=1 𝑛𝑝×max 𝐷𝑖𝑗 (7) 𝐶𝑗= ∑ 𝐷𝑖𝑗 𝑛𝑝 𝑖=1 𝑛𝑝×max 𝐷𝑖𝑗 (7) where 𝑛𝑝𝑗 is the number of pipes connected to node 𝑗. The value of 𝐶𝑗 = 1 if the diameter of the pipes connected to the same node are the same; and 𝐶𝑗<1 if the pipes connected to a node have different diameters. For nodes connected to only one pipe, where 𝑛𝑝𝑗 is the number of pipes connected to node 𝑗. The value of 𝐶𝑗 = 1 if the diameter of the pipes connected to the same node are the same; and 𝐶𝑗<1 if the pipes connected to a node have different diameters. For nodes connected to only one pipe, the value of 𝐶𝑗 is taken to be one. 2.3.2 Network Resilience (𝐼𝑛) 140 the value of 𝐶𝑗 is taken to be one. 140 4 Drinking Water Engineering and Science Discussions Open Access https://doi.org/10.5194/dwes-2020-37 Preprint. Discussion started: 5 February 2021 c⃝Author(s) 2021. CC BY 4.0 License. (8) Theoretically, the value of network resilience may vary between 0 and 1. However, for real-world systems it never attains a value of 1, since imposing the same diameter to all pipes in a network need not always provide a Pareto-optimal solution in Cost-𝐼𝑛 space, as 𝐼𝑛 is a measure of the combined effect of surplus power and nodal uniformity. 145 145 2.3.4 Minimum Surplus Head Index (𝐼𝑚) In a WDN, minimum surplus head, 𝐼𝑚, is defined as the lowest nodal pressure difference between the minimum required and 155 observed pressure, formulated as In a WDN, minimum surplus head, 𝐼𝑚, is defined as the lowest nodal pressure difference between the minimum required and 155 observed pressure, formulated as 𝐼𝑚= 𝑚𝑖𝑛{𝐻𝑗−𝐻𝑗 𝑚𝑖𝑛} 𝑗= 1,2, … , 𝑛𝑛 (10) Maximization of the available surplus head at the most depressed node to some extent improves the reliability of a network (Prasad and Park 2004) 𝑚𝑖𝑛{𝐻𝑗−𝐻𝑗 𝑚𝑖𝑛} 𝑗= 1,2, … , 𝑛𝑛 (10) 𝐼𝑚= 𝑚𝑖𝑛{𝐻𝑗−𝐻𝑗 𝑚𝑖𝑛} 𝑗= 1,2, … , 𝑛𝑛 (10) (10) Maximization of the available surplus head at the most depressed node to some extent improves the reliability of a network (Prasad and Park 2004). 160 2.4 GANetXL GANetXL is used as the optimization tool in this research. GANetXL has been developed by the Center for Water System of University of Exeter as an add-on in Microsoft Excel (Miri and Afshar 2014). It is a common optimization tool with spreadsheet- based interface for solving both single-objective and multi-objective optimization problems (Savić, Bicik et al. 2011). The 2.4 GANetXL GANetXL is used as the optimization tool in this research. GANetXL has been developed by the Center for Water System of University of Exeter as an add-on in Microsoft Excel (Miri and Afshar 2014). It is a common optimization tool with spreadsheet- based interface for solving both single-objective and multi-objective optimization problems (Savić, Bicik et al. 2011). The primary advantage of GANetXL is its capability of easy integration with EPANET via Visual Basic. GANetXL incorporates 165 GA for single-objective and NSGA-II for multi-objective optimizations (Deb, Pratap et al. 2002). In addition, it has the capability to apply penalty functions. GANetXL is well suited for solving multi-objective optimization problems (Mala-Jetmarova, Barton et al. 2014). In this paper GANetXL is employed in two steps: in the first step for single-objective optimization based on GA and the second primary advantage of GANetXL is its capability of easy integration with EPANET via Visual Basic. GANetXL incorporates 165 GA for single-objective and NSGA-II for multi-objective optimizations (Deb, Pratap et al. 2002). In addition, it has the capability to apply penalty functions. GANetXL is well suited for solving multi-objective optimization problems (Mala-Jetmarova, Barton et al. 2014). 2.3.3 Modified Resilience Index (𝑀𝑅𝐼) Jayaram and Srinivasan(Jayaram and Srinivasan 2008) proposed a modified resilience index (𝑀𝑅𝐼), which theoretically overcomes the drawback of Todini’s resilience index when evaluating networks with multiple sources. In contrast to Todini’s resilience index, the value of the modified resilience index is directly proportional to the total surplus power at the demand nodes. Eq. (9) describes 𝑀𝑅𝐼, which only considers the solutions with pressures equal to or higher than that required in all nodes. 150 While Todini’s 𝐼𝑟 and Prasad’s 𝐼𝑛 take values up to a maximum of 1, Jayaram’s 𝑀𝑅𝐼 can be greater than 1(Baños, Reca et al. 2011). ∑ 𝑞𝑗𝐻𝑗−𝐻𝑗 𝑚𝑖𝑛 𝑛𝑛 𝑗=1 ∑ 𝑞𝑗𝐻𝑗 𝑚𝑖𝑛 𝑛𝑛 𝑗=1 (9) um Surplus Head Index (𝐼𝑚) 𝑀𝑅𝐼= ∑ 𝑞𝑗𝐻𝑗−𝐻𝑗 𝑚𝑖𝑛 𝑛𝑛 𝑗=1 ∑ 𝑞𝑗𝐻𝑗 𝑚𝑖𝑛 𝑛𝑛 𝑗=1 (9) 4 Minimum Surplus Head Index (𝐼𝑚) 2.3.4 Minimum Surplus Head Index (𝐼𝑚) 2.3.4 Minimum Surplus Head Index (𝐼𝑚) In this paper GANetXL is employed in two steps: in the first step for single-objective optimization based on GA and the second step for two-objective optimization based on NSGA-II. GA and NSGA-II parameters such as population size, the number of 170 generations, selection method, crossover and mutation operators, crossover and mutation probability and the type of algorithm were tested and reasonably well-performing parameters selected for final optimization runs. These parameters are presented in Table 1 for three example networks, which are described in the following sections. The crossover and mutation types are described in details in CWS (2011). step for two-objective optimization based on NSGA-II. GA and NSGA-II parameters such as population size, the number of 170 generations, selection method, crossover and mutation operators, crossover and mutation probability and the type of algorithm were tested and reasonably well-performing parameters selected for final optimization runs. These parameters are presented in Table 1 for three example networks, which are described in the following sections. The crossover and mutation types are described in details in CWS (2011). Table 1. Optimum GA and NSGA-II values for the three case studies in this paper Table 1. Optimum GA and NSGA-II values for the three case studies in this paper 175 175 5 5 Drinking Water Engineering and Science Discussions Open Access https://doi.org/10.5194/dwes-2020-37 Preprint. Discussion started: 5 February 2021 c⃝Author(s) 2021. CC BY 4.0 License. https://doi.org/10.5194/dwes-2020-37 Preprint. Discussion started: 5 February 2021 c⃝Author(s) 2021. CC BY 4.0 License. Parameter values or method selected Parameter Hanoi and Real- life Network Two-loop Network Generational Elitist Generalation Algorithm (Only in single-objective mode) 100 70 Population size 1000 1000 Number of generations Tourenmate Roulette Selection Method Simple one point Simple one point Crossover operator Simple Simple by gene Mutation operator 0.95 0.8 Crossover probability 0.7 0.01 Mutation rate Yes Yes Adaptive mutation Parameter values or method selected Parameter values or method selected 3. Results and Discussion Three example applications are studied: the Two-loop (Alperovits and Shamir 1977), Hanoi (Fujiwara and Khang 1990) ,which are the benchmark networks, as well as a real-life case study in Iran. 180 3.1 Example 1: The Two - loop network The Two-loop network, shown in figure. 1, was originally presented by Alperovits and Shamir (Alperovits and Shamir 1977). The network consists of 7 nodes and 8 pipes with two loops, and is fed by gravity from a reservoir with a 210 m ﬁxed head. Nodal demands and elevations are given in Table 2. The pipes are all 1000 m long with the assumed Hazen–Williams coefficient of 130. The minimum pressure head requirement of the other nodes is 30 m above the nodal elevations. There are 14 commercial 185 diameters to be selected whose costs and diameters are given in Table 3. The Two-loop network, shown in figure. 1, was originally presented by Alperovits and Shamir (Alperovits and Shamir 1977). The network consists of 7 nodes and 8 pipes with two loops, and is fed by gravity from a reservoir with a 210 m ﬁxed head. Nodal demands and elevations are given in Table 2. The pipes are all 1000 m long with the assumed Hazen–Williams coefficient The Two-loop network, shown in figure. 1, was originally presented by Alperovits and Shamir (Alperovits and Shamir 1977). The network consists of 7 nodes and 8 pipes with two loops, and is fed by gravity from a reservoir with a 210 m ﬁxed head. Nodal demands and elevations are given in Table 2. The pipes are all 1000 m long with the assumed Hazen–Williams coefficient of 130. The minimum pressure head requirement of the other nodes is 30 m above the nodal elevations. There are 14 commercial 185 diameters to be selected whose costs and diameters are given in Table 3. Fig. 1 The layout of Two-loop network. Fig. 1 The layout of Two-loop network. 190 Table 2. Node demands and elevations for Two-loop network 190 Node Elevation (m) Demand (m3/h) Table 2. Node demands and elevations for Two-loop network Node Elevation (m) Demand (m3/h) 6 https://doi.org/10.5194/dwes-2020-37 Drinking Water Engineering and Science Discussions Open Access Preprint. Discussion started: 5 February 2021 c⃝Author(s) 2021. CC BY 4.0 License. Drinking Water Engineering and Science Discussions Open Access https://doi.org/10.5194/dwes-2020-37 Preprint. Discussion started: 5 February 2021 c⃝Author(s) 2021. CC BY 4.0 License. Table 3. 3.1 Example 1: The Two - loop network Pipe sizes and costs for Two-loop network Pipe Diameter (mm) Cost ($/m) 1 25.4 2 2 50.8 5 3 76.2 8 4 101.6 11 5 152.4 16 6 203.2 23 7 254.0 32 8 304.8 50 9 355.6 60 10 406.4 90 11 457.2 130 12 508.0 170 13 558.8 300 14 609.6 550 In the first step, as a result of single-objective optimization of the Two-loop network using GA technique in GANetXL, the In the first step, as a result of single-objective optimization of the Two-loop network using GA techniq minimum cost obtained 419000$ with 35000 number of function evaluations (NFEs) which is the same to minimum costs 195 obtained by GA (Savic and Walters 1997), Simulated Annealing (SA) (Cunha and Sousa 1999), Shuffled frog leaping Algorithm (SFLA) (Eusuff and Lansey 2003), Harmony Search (HS) (Geem 2009)and Scatter search (SS)(Lin, Liu et al. 2007) with 250000, 25000, 11323, 5000 and 3215 NFEs, respectively. minimum cost obtained 419000$ with 35000 number of function evaluations (NFEs) which is the same to minimum costs 195 obtained by GA (Savic and Walters 1997), Simulated Annealing (SA) (Cunha and Sousa 1999), Shuffled frog leaping Algorithm (SFLA) (Eusuff and Lansey 2003), Harmony Search (HS) (Geem 2009)and Scatter search (SS)(Lin, Liu et al. 2007) with 250000, 25000, 11323, 5000 and 3215 NFEs, respectively. As a result, minimum cost is 419000$ for one-objective optimization of this network using GANetXL after 1000 generations that is equal with minimum costs obtained by GA, Simulated Annealing (SA), Shuffled frog leaping Algorithm (SFLA) Harmony 200 Search (HS) and Scatter search (SS)(Savic and Walters 1997, Cunha and Sousa 1999, Geem, Kim et al. 2002, Eusuff and Lansey 2003, Geem 2009). that is equal with minimum costs obtained by GA, Simulated Annealing (SA), Shuffled frog leaping Algorithm (SFLA) Harmony 200 Search (HS) and Scatter search (SS)(Savic and Walters 1997, Cunha and Sousa 1999, Geem, Kim et al. 2002, Eusuff and Lansey 2003, Geem 2009). In the second step, figure 2 (a-d) shows the obtained Pareto front for two-objective optimization of two-loop network using NSGA-II in GANetXL considering rI , nI ,MRI and m I as the second objective function, respectively. All of the solutions In the second step, figure 2 (a d) shows the obtained Pareto front for two objective optimization of two loop network using NSGA-II in GANetXL considering rI , nI ,MRI and m I as the second objective function, respectively. 3.1 Example 1: The Two - loop network All of the solutions 7 7 Drinking Water Engineering and Science Discussions Open Access https://doi.org/10.5194/dwes-2020-37 Preprint. Discussion started: 5 February 2021 c⃝Author(s) 2021. CC BY 4.0 License. in this Pareto front are feasible (and all the network constraints are satisfied). As it is observed the cost changes in the range of 205 [0.424×106 - 4.400×106] $ and rI , nI , MRI and m I criteria changes in the ranges [0.338-0.903], [0.287-0.903], [0.040-0.107] and [0.122-12.856], respectively. In the cost range of [0.424×106-1×106]$, Cost- nI Pareto front shows more and varied solutions, in comparison to other graphs. However, with increase in cost, non-dominated solutions decreases and the current continuity in Pareto front disappears while Cost- rI and Cost-MRI Pareto fronts have better performance. In Cost- m I graph the variety of obtained solutions in the lower and upper bound of Pareto front is lower than other graphs. 210 (a) (b ) (a) (b ) 8 Drinking Water Engineering and Science Discussions Open Access https://doi.org/10.5194/dwes-2020-37 Preprint. Discussion started: 5 February 2021 c⃝Author(s) 2021. CC BY 4.0 License. 215 (c) (d ) Fig. 2 Pareto front of two-objective function optimization of the Two-loop network, (a) Cost- rI , (b) Cost- nI , (c) Cost- MRI , (d) Cost- m I 220 (c) 215 (d ) (d ) Fig. 2 Pareto front of two-objective function optimization of the Two-loop network, (a) Cost- rI , (b) Cost- nI , (c) Cost- MRI , (d) Cost- m I Fig. 2 Pareto front of two-objective function optimization of the Two-loop network, (a) Cost- rI , (b) Cost- nI , (c) Cost- Fig. 2 Pareto front of two-objective function optimization of the Two-loop network, (a) Cost- rI , (b) Cost- nI , (c) Cost- MRI , (d) Cost- m I MRI , (d) Cost- m I 220 Figure 3 shows the surplus pressure of the minimum pressure head requirement in the nodes of Two-loop network for solutions with maximum reliability criteria and minimum cost. As it is observed, the surplus pressure of the nodes in the solutions with minimum cost is lower than the solutions of maximum reliability criteria ( rI , nI ,MRI and m I ). Also, the design based on single-objective function (minimum cost), surplus pressure is closer to the minimum allowed pressure in nodes 3, 6, and 7, showing that these nodes are the critical nodes of the network. 3.1 Example 1: The Two - loop network As a result, if the two-loop network was designed only based on 225 minimum cost, in critical periods such as pipe failures, there would be problems issues at these nodes. Reliability evaluation should be analyzed under all feasible extreme conditions. Failure of multiple pipes as well as the failure of the reservoir connection line during a firefighting event and/or power or pumping station failures should be evaluated 9 9 Drinking Water Engineering and Science Discussions Open Access https://doi.org/10.5194/dwes-2020-37 Preprint. Discussion started: 5 February 2021 c⃝Author(s) 2021. CC BY 4.0 License. https://doi.org/10.5194/dwes-2020-37 Preprint. Discussion started: 5 February 2021 c⃝Author(s) 2021. CC BY 4.0 License. simultaneously. Although an infinite number of failure scenarios are likely, the probability of simultaneous failures in multiple pipes is too low (Tabesh, Tanyimboh et al. 2001). Pipe failures independency can be assumed (Su, Mays et al. 1987) and any 230 simultaneously. Although an infinite number of failure scenarios are likely, the probability of simultaneous failures in multiple pipes is too low (Tabesh, Tanyimboh et al. 2001). Pipe failures independency can be assumed (Su, Mays et al. 1987) and any 230 likely dependency will be negative. For example, if a pipe failure occurs in the network, the pressure will decrease, and consequently the probability of another pipe failure will decrease as well. However, in case the system is a large-scale WDN, the influence of pressure might not be significant. Other pipe failure reasons, such as damages or traffic loadings, may lead to pipe failures that are completely independent events (Shafiqul Islam, Sadiq et al. 2013). pipes is too low (Tabesh, Tanyimboh et al. 2001). Pipe failures independency can be assumed (Su, Mays et al. 1987) and any 230 likely dependency will be negative. For example, if a pipe failure occurs in the network, the pressure will decrease, and consequently the probability of another pipe failure will decrease as well. However, in case the system is a large-scale WDN, the influence of pressure might not be significant. Other pipe failure reasons, such as damages or traffic loadings, may lead to pipe failures that are completely independent events (Shafiqul Islam, Sadiq et al. 2013). 235 Fig. 3.1 Example 1: The Two - loop network 3 Surplus pressure of nodes in two-loop network for solutions of maximum reliability criteria and minimum cost In this paper, to evaluate reliability of the candidate solutions of maximum rI , nI ,MRI and m I criteria, the nodal pressures have been investigated under pipe failure conditions. Table 4 presents the statistical parameters of each pipe in two-loop network under different runs of single-objective optimizations when the objective function is to minimize cost. This table helps the designers to recognize critical and sensitive pipes that have the most probability of failure in the network. For example, maximum 240 and minimum diameters that are allocated to pipe 1 in different runs of GANetXL are 24 and 18 inch, respectively. 235 3 Surplus pressure of nodes in two-loop network for solutions of maximum reliability criteria and minimum cost Fig. 3 Surplus pressure of nodes in two-loop network for solutions of maximum reliability criteria and minimum cost In this paper, to evaluate reliability of the candidate solutions of maximum rI , nI ,MRI and m I criteria, the nodal pressures have been investigated under pipe failure conditions. Table 4 presents the statistical parameters of each pipe in two-loop network under different runs of single-objective optimizations when the objective function is to minimize cost. This table helps the In this paper, to evaluate reliability of the candidate solutions of maximum rI , nI ,MRI and m I criteria, the nodal pressures have been investigated under pipe failure conditions. Table 4 presents the statistical parameters of each pipe in two-loop network under different runs of single-objective optimizations when the objective function is to minimize cost. This table helps the designers to recognize critical and sensitive pipes that have the most probability of failure in the network. For example, maximum 240 have been investigated under pipe failure conditions. Table 4 presents the statistical parameters of each pipe in two loop network under different runs of single-objective optimizations when the objective function is to minimize cost. This table helps the designers to recognize critical and sensitive pipes that have the most probability of failure in the network. For example, maximum 240 and minimum diameters that are allocated to pipe 1 in different runs of GANetXL are 24 and 18 inch, respectively. designers to recognize critical and sensitive pipes that have the most probability of failure in the network. 3.1 Example 1: The Two - loop network For example, maximum 240 and minimum diameters that are allocated to pipe 1 in different runs of GANetXL are 24 and 18 inch, respectively. Table 4. Statistical parameters for diameters obtained for each pipe of two-loop network Pipe numbers Maximum (mm) Minimum (mm) Average (mm) ST. DEV Variance C.V. 1 609.6 457.2 459.35 17.96 322.52 0.04 2 304.8 152.4 248.28 26.35 694.18 0.11 3 457.2 406.4 452.91 14.13 199.65 0.03 4 254 101.6 208.21 41.03 1683.23 0.20 5 609.6 406.4 469.36 37.62 1414.97 0.08 6 508 101.6 251.85 57.47 3302.97 0.23 7 558.8 76.2 217.87 63.47 4028.38 0.29 8 304.8 25.4 68.69 58.51 3423.79 0.85 Table 4. Statistical parameters for diameters obtained for each pipe of two-loop network Table 4. Statistical parameters for diameters obtained for each pipe of two loop network Pipe numbers Maximum (mm) Minimum (mm) Average (mm) ST. DEV Variance C.V. 10 Drinking Water Engineering and Science Discussions Open Access https://doi.org/10.5194/dwes-2020-37 Preprint. Discussion started: 5 February 2021 c⃝Author(s) 2021. CC BY 4.0 License. https://doi.org/10.5194/dwes-2020-37 Preprint. Discussion started: 5 February 2021 c⃝Author(s) 2021. CC BY 4.0 License. According to Table 4, pipe numbers 2, 3 and 5 that have minimum standard deviations and variation coefficients are chosen for According to Table 4, pipe numbers 2, 3 and 5 that have minimum standard deviations and variation coefficients are chosen for According to Table 4, pipe numbers 2, 3 and 5 that have minimum standard deviations and variation coefficients are chosen for failure analysis Pipe 1 belongs to a water transmission line from the reservoir to the network that is important during network 245 failure analysis. Pipe 1 belongs to a water transmission line from the reservoir to the network that is important during network 245 operation. If a failure is considered in this pipe, then the network will be unreliable. That is why this pipe is not taken into account for failure analysis. Figure 4 shows the performance of solutions with maximum reliability criteria under the failure of pipes 2, 3 and 5. failure analysis. Pipe 1 belongs to a water transmission line from the reservoir to the network that is important during network 245 operation. If a failure is considered in this pipe, then the network will be unreliable. That is why this pipe is not taken into account for failure analysis. 3.1 Example 1: The Two - loop network Figure 4 shows the performance of solutions with maximum reliability criteria under the failure of pipes 2, 3 and 5. 250 (a) (b ) 250 (a) 250 (b ) (b ) 11 (c) 255 Fig. 4 Surplus pressure of nodes in two-loop network for solution with maximum reliability criteria under failure of pipes No. (a) 2, (b) 3 and (c) 5 https://doi.org/10.5194/dwes-2020-37 Drinking Water Engineering and Science Discussions Open Access Preprint. Discussion started: 5 February 2021 c⃝Author(s) 2021. CC BY 4.0 License. Drinking Water Engineering and Science Discussions Open Access https://doi.org/10.5194/dwes-2020-37 Preprint. Discussion started: 5 February 2021 c⃝Author(s) 2021. CC BY 4.0 License. (c) 255 Fig. 4 Surplus pressure of nodes in two-loop network for solution with maximum reliability criteria under failure of pipes No. (a) 2 (b) 3 and (c) 5 (c) 255 Fig. 4 Surplus pressure of nodes in two-loop network for solution with maximum reliability criteria under failure of pipes No. (a) 2, (b) 3 and (c) 5 255 (c) Fig. 4 Surplus pressure of nodes in two-loop network for solution with maximum reliability criteria under failure of pipes No. (a) 2, (b) 3 and (c) 5 Figure 4 shows that, nodeNo. 6 encounters with a serious pressure loss with failure in pipe No. 2, 3 and 5 in represented solutions by m I criterion. In represented solutions based on rI , nI and MRI for all the pipes of the network the diameter was 609.6 mm while in the obtained solution with maximum m I , the diameter of pipes No. 4 and 6 was 25.4 mm and other pipes were 260 609.6 mm. Consequently, m I criterion is of lower performance than any other criterion under pipe failure condition. 3.2 Example 2: The Hanoi network Figure 4 shows that, nodeNo. 6 encounters with a serious pressure loss with failure in pipe No. 2, 3 and 5 in represented solutions by m I criterion. In represented solutions based on rI , nI and MRI for all the pipes of the network the diameter was 609.6 mm while in the obtained solution with maximum m I , the diameter of pipes No. 4 and 6 was 25.4 mm and other pipes were 260 609.6 mm. Consequently, m I criterion is of lower performance than any other criterion under pipe failure condition. 3.2 Example 2: The Hanoi network The Hanoi network in Vietnam (Figure. 3.1 Example 1: The Two - loop network 5), ﬁrst presented by Fujiwara and Khang, is a new design as all new pipes are to be selected. The network consists of 32 nodes and 34 pipes organized in three loops. The system is gravity fed by a single reservoir. g ) p y j g g p p selected. The network consists of 32 nodes and 34 pipes organized in three loops. The system is gravity fed by a single reservoir. The network details are given in Table 5. The minimum required pressure head for all nodes is 30 m and the elevation for all 265 nodes is zero. There are six available pipe diameters to be selected for each new pipe and the pipe cost per meter for the six available pipe diameters are listed in Table 6. The network details are given in Table 5. The minimum required pressure head for all nodes is 30 m and the elevation for all 265 nodes is zero. There are six available pipe diameters to be selected for each new pipe and the pipe cost per meter for the six available pipe diameters are listed in Table 6. The network details are given in Table 5. The minimum required pressure head for all nodes is 30 m and the elevation for all 265 nodes is zero. There are six available pipe diameters to be selected for each new pipe and the pipe cost per meter for the six available pipe diameters are listed in Table 6. 12 https://doi.org/10.5194/dwes-2020-37 Drinking Water Engineering and Science Discussions Open Access Preprint. Discussion started: 5 February 2021 c⃝Author(s) 2021. CC BY 4.0 License. Drinking Water Engineering and Science Discussions Open Access https://doi.org/10.5194/dwes-2020-37 Preprint. Discussion started: 5 February 2021 c⃝Author(s) 2021. CC BY 4.0 License. Fig. 5 Layout of Hanoi network 270 Table 5. Network data for the Hanoi network Table 5. Network data for the Hanoi network Node data Pipe data Node Demand (m3/h) Pipe Length (m) 1 -19940 1 100 2 890 2 1350 3 850 3 900 4 130 4 1150 5 725 5 1450 6 1005 6 450 7 1350 7 850 8 550 8 850 9 525 9 800 10 525 10 950 11 500 11 1200 12 560 12 3500 13 940 13 800 14 615 14 500 15 280 15 550 16 310 16 2730 17 865 17 1750 18 1345 18 800 19 60 19 400 20 1275 20 2200 21 930 21 1500 13 Drinking Water Engineering and Science Discussions Open Access https://doi.org/10.5194/dwes-2020-37 Preprint. Discussion started: 5 February 2021 c⃝Author(s) 2021. CC BY 4.0 License. 22 485 22 500 23 1045 23 2650 24 820 24 1230 25 170 25 1300 26 900 26 850 27 370 27 300 28 290 28 750 29 360 29 1500 30 360 30 2000 31 105 31 1600 32 805 32 150 33 860 34 950 Table 6. Pipe sizes and costs for Hanoi network Pipe Diameter (mm) Cost ($/m) 1 304.8 45.726 2 406.4 70.400 3 508.0 98.378 4 609.6 129.333 5 762.0 180.748 6 1016.0 278.280 Table 6. Pipe sizes and costs for Hanoi network In the first step, as a result of single-objective optimization, GA method in GANetXL obtained a minimum cost of 6.097×106$ 275 with 100000 NFEs for this network while in the previous researches the methods of GA(Savic and Walters 1997), Ant Colony Optimization (ACO)(Zecchin, Simpson et al. 2006), and Shuffled Complex Evolution (SCE) Liong and Atiquzzaman (Atiquzzaman and Liong 2004)reported costs of 6.195, 6.134 and 6.22 million$ with 1000000, 25402 and 85571 NFEs, respectively. In the second step, figure 6. (a-d) shows non-dominated solutions of Hanoi network which calculated by NSGA-II considering 280 minimum cost versus maximum reliability criteria and all of the solutions in the Pareto front is feasible. As it is observed in figure 7 minimum values of rI , nI ,MRI and m I are 0.228, 0.256, 0.555 and 0.090 and maximum values are 0.353, 0.353, 0.825 and 19.916, respectively. In the second step, figure 6. (a-d) shows non-dominated solutions of Hanoi network which calculated by NSGA-II considering 280 minimum cost versus maximum reliability criteria and all of the solutions in the Pareto front is feasible. As it is observed in figure 7 minimum values of rI , nI ,MRI and m I are 0.228, 0.256, 0.555 and 0.090 and maximum values are 0.353, 0.353, 0.825 and 19.916, respectively. Cost values change in a range of [6.251×106-10.791×106]$ for Cost- rI and in [6.584×106- 10.969×106]$ Cost- nI space that the increase in Cost- nI to Cost- rI is due j C factor in formula [Eq. (8)] which cause https://doi.org/10.5194/dwes-2020-37 Drinking Water Engineering and Science Discussions Open Access Preprint. Discussion started: 5 February 2021 c⃝Author(s) 2021. CC BY 4.0 License. Cost values change in a range of [6.251×106-10.791×106]$ for Cost- rI and in [6.584×106- 10.969×106]$ Cost- nI space that the increase in Cost- nI to Cost- rI is due j C factor in formula [Eq. (8)] which cause uniformity diameters in the design phase. In this example monotony and variety of represented solutions are observed in all 285 Pareto fronts, the reason can be found in the increase of the network size and possible solutions for network design. 14 14 Drinking Water Engineering and Science Discussions Open Access https://doi.org/10.5194/dwes-2020-37 Preprint. Discussion started: 5 February 2021 c⃝Author(s) 2021. CC BY 4.0 License. (a) (b) (c) (a) (a) (b) (b) 290 (b) (c) (c) 15 15 (d) Fig. 6 Pareto front of two-objective function optimization of the Hanoi network, (a) Cost- rI , (b) Cost- nI , (c) Cost- 295 MRI (d) Cost I https://doi.org/10.5194/dwes-2020-37 Drinking Water Engineering and Science Discussion Open Access Preprint. Discussion started: 5 February 2021 c⃝Author(s) 2021. CC BY 4.0 License. Drinking Water Engineering and Science Discussions Open Access https://doi.org/10.5194/dwes-2020-37 Preprint. Discussion started: 5 February 2021 c⃝Author(s) 2021. CC BY 4.0 License. (d) Fig 6 Pareto front of two-objective function optimization of the Hanoi network (a) Cost- I (b) Cost- I (c) (d) Fig. 6 Pareto front of two-objective function optimization of the Hanoi network, (a) Cost- rI , (b) Cost- nI , (c) Cost- MRI I (d) 295 Fig. 6 Pareto front of two-objective function optimization of the Hanoi network, (a) Cost- rI , (b) Cost- nI , (c) Cost- MRI , (d) Cost- m I Figure 7 shows the surplus pressure in comparison with minimum allowed pressure in the nodes of the Hanoi network for solutions of maximum reliability criteria and minimum cost. In the cost-based optimization, surplus pressure in nodes No. 13, 30 and 31 is less than 1 m which shows that these nodes are the most critical ones of this network. rI , nI and MRI criteria have similar performance for all the nodes, but m I criterion determinates more surplus pressure for most of the nodes than other 300 i i i hi k lik h l k Figure 7 shows the surplus pressure in comparison with minimum allowed pressure in the nodes of the Hanoi network for solutions of maximum reliability criteria and minimum cost. In the cost-based optimization, surplus pressure in nodes No. https://doi.org/10.5194/dwes-2020-37 Drinking Water Engineering and Science Discussions Open Access Preprint. Discussion started: 5 February 2021 c⃝Author(s) 2021. CC BY 4.0 License. 13, 30 and 31 is less than 1 m which shows that these nodes are the most critical ones of this network I I and MRI criteria Figure 7 shows the surplus pressure in comparison with minimum allowed pressure in the nodes of the Hano solutions of maximum reliability criteria and minimum cost. In the cost-based optimization, surplus pressure in n Figure 7 shows the surplus pressure in comparison with minimum allowed pressure in the nodes of the Hanoi network for solutions of maximum reliability criteria and minimum cost. In the cost-based optimization, surplus pressure in nodes No. 13, 30 and 31 is less than 1 m which shows that these nodes are the most critical ones of this network. rI , nI and MRI criteria have similar performance for all the nodes, but m I criterion determinates more surplus pressure for most of the nodes than other 300 criteria in this network unlike the two-loop network. Fig. 7 Nodal surplus pressure of Hanoi network for solutions of maximum reliability criteria and minimum cost Table 7 shows the statistical parameters for each pipe of Hanoi network due to different runs of single-objective optimization by GANetXL. According to this table, Pipes No. 4, 5, 6 and 20 that have standard deviation and variation coefficient equal t 305 zero have been chosen for reliability evaluation when there is a failure in the network. Fig. 7 Nodal surplus pressure of Hanoi network for solutions of maximum reliability criteria and minimum cost Table 7 shows the statistical parameters for each pipe of Hanoi network due to different runs of single-objective optimizations by GANetXL. According to this table, Pipes No. 4, 5, 6 and 20 that have standard deviation and variation coefficient equal to 305 zero have been chosen for reliability evaluation when there is a failure in the network. Nodal surplus pressure of Hanoi network for solutions of maximum reliability criteria and minimum cost Fig. 7 Nodal surplus pressure of Hanoi network for solutions of maximum reliability criteria and minimum cost Table 7 shows the statistical parameters for each pipe of Hanoi network due to different runs of single-objective optimizations by GANetXL. According to this table, Pipes No. 4, 5, 6 and 20 that have standard deviation and variation coefficient equal to 305 zero have been chosen for reliability evaluation when there is a failure in the network. https://doi.org/10.5194/dwes-2020-37 Drinking Water Engineering and Science Discussions Open Access Preprint. Discussion started: 5 February 2021 c⃝Author(s) 2021. CC BY 4.0 License. 16 Drinking Water Engineering and Science Discussions Open Access https://doi.org/10.5194/dwes-2020-37 Preprint. Discussion started: 5 February 2021 c⃝Author(s) 2021. CC BY 4.0 License. Table 7. Statistical parameters for diameters obtained for each pipe of Hanoi network Table 7. Statistical parameters for diameters obtained for each pipe of Hanoi network Pipe numbers Maximum (mm) Minimum (mm) Average (mm) ST. DEV Variance C.V. 1 1016 762 1005.84 49.77 2477.41 0.05 2 1016 762 1010.92 35.56 1264.51 0.04 3 1016 508 1005.84 61.38 3767.73 0.06 4 1016 1016 1016 0 0 0 5 1016 1016 1016 0 0 0 6 1016 1016 1016 0 0 0 7 1016 762 1013.46 25.27 638.71 0.02 8 1016 762 1005.84 49.77 2477.41 0.05 9 1016 508 1008.38 56.28 3167.74 0.06 10 1016 508 889 136.78 18709.64 0.15 11 762 304.8 608.584 47.64 2269.93 0.08 12 762 508 611.124 31.69 1004.13 0.05 13 1016 406.4 495.808 67.82 4599.73 0.14 14 1016 304.8 477.52 107.52 11561.27 0.23 15 762 304.8 387.604 146.47 21453.12 0.38 16 1016 304.8 341.376 110.85 12287.98 0.32 17 508 406.4 447.04 49.77 2477.41 0.11 18 1016 508 662.432 137.61 18937.77 0.21 19 762 406.4 511.048 43.59 1900.38 0.09 20 1016 1016 1016 0 0 0 21 1016 406.4 510.032 53.72 2886.19 0.11 22 1016 304.8 399.288 210.32 44233.20 0.53 23 1016 762 1013.46 25.27 638.71 0.02 24 1016 609.6 824.484 120.18 14444.10 0.15 25 1016 609.6 850.392 126.00 15877.13 0.15 26 1016 406.4 541.528 117.66 13843.59 0.22 27 1016 304.8 414.528 220.58 48656.42 0.53 28 1016 304.8 369.824 135.95 18481.51 0.37 29 1016 304.8 504.952 102.06 10416.50 0.20 30 609.6 304.8 446.024 64.09 4107.35 0.14 31 609.6 304.8 346.456 78.82 6213.15 0.23 32 1016 304.8 799.592 263.14 69244.76 0.33 33 1016 304.8 491.236 163.36 26686.66 0.33 34 762 406.4 514.604 40.42 1633.80 0.08 Pipe numbers Maximum (mm) Minimum (mm) Average (mm) ST. DEV Variance C.V. 17 17 Drinking Water Engineering and Science Discussions Open Access https://doi.org/10.5194/dwes-2020-37 Preprint. Discussion started: 5 February 2021 c⃝Author(s) 2021. CC BY 4.0 License. The results of figure 8.(a) and (b) shows that by failure in pipes No. 4 and 5 the surplus pressure in most of the nodes for solutions of maximum m I criterion is more than solutions with maximum rI , nI and MRI . In effect of pipes No. 4 and 5 failures, 310 nodes reactions to pressure changes are similar because these two pipes are along. https://doi.org/10.5194/dwes-2020-37 Drinking Water Engineering and Science Discussions Open Access Preprint. Discussion started: 5 February 2021 c⃝Author(s) 2021. CC BY 4.0 License. However, due to failure in pipe No. 6, none of the nodes of the network meet lack of pressure and the figure 8.(c) shows that the solutions with maximum nI and MRI criteria has more capability to supply pressure in most of the networks. In figure 8.(d) there is no significant difference in represented solutions with reliability criteria values. The nodes with no values in the graph are those that have negative pressures. 315 (a) (b ) 315 (a) (a) b (b ) 18 18 Drinking Water Engineering and Science Discussions Open Access https://doi.org/10.5194/dwes-2020-37 Preprint. Discussion started: 5 February 2021 c⃝Author(s) 2021. CC BY 4.0 License. https://doi.org/10.5194/dwes-2020-37 Preprint. Discussion started: 5 February 2021 c⃝Author(s) 2021. CC BY 4.0 License. (c) 320 (d ) Figure 8. Surplus pressure of Hanoi network nodes for solutions of maximum reliability criteria when pipes No. (a) 4, (b) 5, (c) 6 and (d) 20 are lost due to failure 3.3 Example 3:The Real- life network 325 Real- life WDN is located in Iran and it has 37 pipes, 24 nodes and a reservoir with a 962 m ﬁxed head. The design purpose of thi t k i i i l t l f it d i i f th i ti diti f th WDN (fi 9) F thi (c) (c) 20 (d ) Figure 8. Surplus pressure of Hanoi network nodes for solutions of maximum reliability criteria when pipes No. (a) 4, (b) 5, (c) 6 and (d) 20 are lost due to failure Table 9. Pipe sizes and costs for Real- life network 3.3 Example 3:The Real- life network 325 9 Layout of Real- life network 335 335 Table 8.Pipes data for real- life WDN Pipe data Pipe 1 2 3 4 5 6 7 8 9 10 11 12 Length (m) 73.287 54.624 705 78.582 254.51 205.44 805.59 723.29 556.56 417.27 367.89 707.14 Pipe 13 14 15 16 17 18 19 20 21 22 23 24 Length (m) 320.95 485.55 226.47 201.47 292 628.8 225.25 323.09 451.71 200.86 897.33 232.56 Pipe 25 26 * 27 * 28 * 29 * 30 * 31 * 32 * 33 * 34 * 35 * 36 * 37 * Length (m) 9999.87 634.90 299.62 409.04 96.62 423.98 427.02 422.15 318.52 274.62 99.06 174.04 192.33 Node data Node Reservoir 1 2 3 4 5 6 7 8 9 10 11 12 Elevation (m) 962 901.5 896.5 895.5 903.5 903.5 902.5 901.5 900 900.5 900.5 903.5 902 Demand (m3/h) -301.176 21.78 7.81 20.56 7.56 22.64 30.13 13.25 32.58 18.86 32.8 14.29 0 Node 13 14 15 16 17 18 19 20 21 22 23 24 Elevation (m) 899 905 898 900.5 900.5 899.5 901.5 902 902 898 900.5 904 Demand (m3/h) 18.79 9.54 7.45 15.8 11.02 16.31 0 0 0 0 0 0 * They are existing pipes in the network and are not considered in the total cost of the network. Table 9. Pipe sizes and costs for Real- life network Table 9. Pipe sizes and costs for Real- life network * They are existing pipes in the network and are not considered in the total cost of the network. 3.3 Example 3:The Real- life network 325 p Real- life WDN is located in Iran and it has 37 pipes, 24 nodes and a reservoir with a 962 m ﬁxed head. The design purpose of this network is municipal water supply of city and improving of the existing condition of the WDN (figure. 9). For this purpose, a series of pipes which have diameters more than 100 mm are used for future conditions (Rasekh, Afshar et al. 2010).For designing this network, polyethylene pipes (PE-80) with Hazen-Williams coefficient of 130 are used. Table 8 presents the nodes and pipes characteristics. Table 9 gives the diameter of these polyethylene pipes as well as cost per unit length. In the design of 330 Real- life WDN is located in Iran and it has 37 pipes, 24 nodes and a reservoir with a 962 m ﬁxed head. The design purpose of this network is municipal water supply of city and improving of the existing condition of the WDN (figure. 9). For this purpose, a series of pipes which have diameters more than 100 mm are used for future conditions (Rasekh, Afshar et al. 2010).For designing this network, polyethylene pipes (PE-80) with Hazen-Williams coefficient of 130 are used. Table 8 presents the nodes and pipes characteristics. Table 9 gives the diameter of these polyethylene pipes as well as cost per unit length. In the design of 330 19 Drinking Water Engineering and Science Discussions Open Access https://doi.org/10.5194/dwes-2020-37 Preprint. Discussion started: 5 February 2021 c⃝Author(s) 2021. CC BY 4.0 License. https://doi.org/10.5194/dwes-2020-37 Preprint. Discussion started: 5 February 2021 c⃝Author(s) 2021. CC BY 4.0 License. the network, nodes pressure and velocity constraints are between 14-60 m and 0.2-2 m/s, respectively (Department of Technical Affairs 2013). There are more constraints in this example than the other ones. the network, nodes pressure and velocity constraints are between 14-60 m and 0.2-2 m/s, respectively (Department of Technical Affairs 2013). There are more constraints in this example than the other ones. Fig. 9 Layout of Real- life network Fig. 9 Layout of Real- life network Fig. 3.3 Example 3:The Real- life network 325 Table 8.Pipes data for real- life WDN Pipe data Pipe 1 2 3 4 5 6 7 8 9 10 11 12 Length (m) 73.287 54.624 705 78.582 254.51 205.44 805.59 723.29 556.56 417.27 367.89 707.14 Pipe 13 14 15 16 17 18 19 20 21 22 23 24 Length (m) 320.95 485.55 226.47 201.47 292 628.8 225.25 323.09 451.71 200.86 897.33 232.56 Pipe 25 26 * 27 * 28 * 29 * 30 * 31 * 32 * 33 * 34 * 35 * 36 * 37 * Length (m) 9999.87 634.90 299.62 409.04 96.62 423.98 427.02 422.15 318.52 274.62 99.06 174.04 192.33 Node data Node Reservoir 1 2 3 4 5 6 7 8 9 10 11 12 Elevation (m) 962 901.5 896.5 895.5 903.5 903.5 902.5 901.5 900 900.5 900.5 903.5 902 Demand (m3/h) -301.176 21.78 7.81 20.56 7.56 22.64 30.13 13.25 32.58 18.86 32.8 14.29 0 Node 13 14 15 16 17 18 19 20 21 22 23 24 Elevation (m) 899 905 898 900.5 900.5 899.5 901.5 902 902 898 900.5 904 Demand (m3/h) 18.79 9.54 7.45 15.8 11.02 16.31 0 0 0 0 0 0 * Th i i i i h k d id d i h l f h k 20 Drinking Water Engineering and Science Discussions Open Access https://doi.org/10.5194/dwes-2020-37 Preprint. Discussion started: 5 February 2021 c⃝Author(s) 2021. CC BY 4.0 License. Pipe Diameter (mm) Cost (Rial/m) Pipe Diameter (mm) Cost (Rial/m) 1 76.6 49560 6 191.8 305200 2 93.8 73360 7 213.2 375200 3 106.6 94360 8 238.8 470400 4 136.4 154000 9 268.6 593600 5 170.6 239680 10 302.8 753200 * pipes cost is 28000 Rial/kg based on PE-80 standards with 10 atm pressure. In the first step, as a result of single-objective optimization using GA in GANetXL, the minimum cost is estimated 7.54×108 340 Rials with 100000 NFEs which shows a cost decrease of 46.14% in comparison to the solution of the consultant company with 14×108 Rials (Rasekh, Afshar et al. 2010). In the second step, the results of figure 10 (a-d) shows that the rI , MRI , m I criteria have better performance than nI criterion for this network in terms of non-dominated solutions. All these three criteria have similar solutions of maximum and minimum cost in the Pareto front. In the first step, as a result of single-objective optimization using GA in GANetXL, the minimum cost is estimated 7.54×108 340 Rials with 100000 NFEs which shows a cost decrease of 46.14% in comparison to the solution of the consultant company with 14×108 Rials (Rasekh, Afshar et al. 2010). In the second step, the results of figure 10 (a-d) shows that the rI , MRI , m I criteria have better performance than nI criterion for this network in terms of non dominated solutions All these three criteria have similar solutions of maximum and minimum https://doi.org/10.5194/dwes-2020-37 Drinking Water Engineering and Science Discussions Open Access Preprint. Discussion started: 5 February 2021 c⃝Author(s) 2021. CC BY 4.0 License. Drinking Water Engineering and Science Discussions Open Access https://doi.org/10.5194/dwes-2020-37 Preprint. Discussion started: 5 February 2021 c⃝Author(s) 2021. CC BY 4.0 License. Fig. 10 Pareto front of two-objective function optimization of the Real- life network, (a) Cost- rI , (b) Cost- nI , Fig. 10 Pareto front of two-objective function optimization of the Real- life network, (a) Cost- rI , (b) Cost- nI , (c) Fig. 10 Pareto front of two-objective function optimization of the Real- life network, (a) Cost- rI , (b) Cost- nI , (c) 355 Cost- MRI , (d) Cost- m I Cost- MRI , (d) Cost- m I Cost- MRI , (d) Cost- m I The results shown in figure 11 demonstrate that in the cost-based optimization, surplus pressure in the nodes number 13 and 23 is less than 1m that explains these nodes are the most critical ones in the network. rI and MRI criteria have similar and more successful performance compared to m I in terms of the surplus pressure for all the nodes in the network. nI has less capability than other criteria to create surplus pressure in the network. 360 The results shown in figure 11 demonstrate that in the cost-based optimization, surplus pressure in the nodes number 13 and 23 is less than 1m that explains these nodes are the most critical ones in the network. rI and MRI criteria have similar and more successful performance compared to m I in terms of the surplus pressure for all the nodes in the network. nI has less capability than other criteria to create surplus pressure in the network. 360 Fi 11 S l f th l lif f l ti f i li bilit it i d i i t Fig. 11 Surplus pressure of the real- life for solutions of maximum reliability criteria and minimum cost Fig. 11 Surplus pressure of the real- life for solutions of maximum reliability criteria and minimum cost Table 10 presents statistical parameters for new pipes of Real- life Network in result of different runs by GANetXL whe optimization approach is cost-based. According to this table pipes No. 3.3 Example 3:The Real- life network 325 All of the solutions in the Pareto front (figure 10) which obtained by NSGA-II is feasible and satisfied 345 In the first step, as a result of single-objective optimization using GA in GANetXL, the minimum cost is estimated 7.54×108 340 Rials with 100000 NFEs which shows a cost decrease of 46.14% in comparison to the solution of the consultant company with 14×108 Rials (Rasekh, Afshar et al. 2010). MRI p, g ( ) r , , m p n for this network in terms of non-dominated solutions. All these three criteria have similar solutions of maximum and minimum cost in the Pareto front. All of the solutions in the Pareto front (figure 10) which obtained by NSGA-II is feasible and satisfied 345 the velocity and pressure constraints. cost in the Pareto front. All of the solutions in the Pareto front (figure 10) which obtained by NSGA-II is feasible and satisfied 345 the velocity and pressure constraints. (a ) (a ) (a) (a ) 21 Drinking Water Engineering and Science Discussions Open Access https://doi.org/10.5194/dwes-2020-37 Preprint. Discussion started: 5 February 2021 c⃝Author(s) 2021. CC BY 4.0 License. 22 (b ) (c ) (d ) (b ) 0 (c ) (b ) 350 350 (c ) (c ) (d ) (d ) 22 https://doi.org/10.5194/dwes-2020-37 Drinking Water Engineering and Science Discussions Open Access Preprint. Discussion started: 5 February 2021 c⃝Author(s) 2021. CC BY 4.0 License. 18 and 21 were chosen to evaluate the performance o reliability criteria under failure condition because these pipes have less standard deviation and coefficient of variation than oth 365 Table 10 presents statistical parameters for new pipes of Real- life Network in result of different runs by GANetXL when optimization approach is cost-based. According to this table pipes No. 18 and 21 were chosen to evaluate the performance of reliability criteria under failure condition because these pipes have less standard deviation and coefficient of variation than other 365 pipes. Moreover, in this network the failure probability should be evaluated in the existing pipes because they have more lifetime in comparison to new pipes. There are different methods accessible to estimate the probability of pipe failure, repair time, and failure return periods. Interested readers should refer to Chapter 18 of Mays (2000). Subsequently, in this study, a random pipe failure has been created using a uniform distribution in the range of [26, 37], that is the pipe numbers for existing pipes (Shafiqul Islam, Sadiq et al. 2013) Finally, the failure of the pipes No. 27 and 34 was analyzed in the network. 370 reliability criteria under failure condition because these pipes have less standard deviation and coefficient of variation than other 365 pipes. Moreover, in this network the failure probability should be evaluated in the existing pipes because they have more lifetime in comparison to new pipes. There are different methods accessible to estimate the probability of pipe failure, repair time, and failure return periods. Interested readers should refer to Chapter 18 of Mays (2000). Subsequently, in this study, a random pipe failure has been created using a uniform distribution in the range of [26, 37], that is the pipe numbers for existing pipes (Shafiqul Islam, Sadiq et al. 2013) Finally, the failure of the pipes No. 27 and 34 was analyzed in the network. 370 Table 10. Statistical parameters for diameters obtained for each new pipe of Real- life network Pipe numbers Maximum (mm) Minimum (mm) Average (mm) ST. DEV Variance C.V. 1 213.2 76.6 125.86 21.14 446.99 0.17 2 170.6 76.6 83.30 16.06 257.93 0.19 3 238.8 76.6 140.96 23.97 574.42 0.17 4 302.8 76.6 93.83 25.70 660.65 0.27 5 213.2 76.6 94.09 20.03 401.31 0.21 Table 10. https://doi.org/10.5194/dwes-2020-37 Drinking Water Engineering and Science Discussions Open Access Preprint. Discussion started: 5 February 2021 c⃝Author(s) 2021. CC BY 4.0 License. Statistical parameters for diameters obtained for each new pipe of Real- life network 23 Drinking Water Engineering and Science Discussions Open Access https://doi.org/10.5194/dwes-2020-37 Preprint. Discussion started: 5 February 2021 c⃝Author(s) 2021. CC BY 4.0 License. 6 302.8 76.6 112.86 37.87 1433.82 0.34 7 136.4 76.6 97.10 15.48 239.59 0.16 8 268.6 76.6 80.66 21.65 468.56 0.27 9 341.2 76.6 100.11 31.71 1005.72 0.32 10 341.2 76.6 97.53 42.00 1763.92 0.43 11 191.8 76.6 106.91 20.70 428.58 0.19 12 191.8 76.6 97.50 17.47 305.09 0.18 13 403.8 76.6 109.44 39.39 1551.55 0.36 14 213.2 76.6 117.36 29.38 863.00 0.25 15 238.8 76.6 113.80 34.39 1182.59 0.30 16 302.8 76.6 94.20 32.44 1052.59 0.34 17 238.8 76.6 122.40 43.05 1853.03 0.35 18 136.4 76.6 91.02 11.28 127.19 0.12 19 191.8 76.6 96.02 15.58 242.86 0.16 20 213.2 76.6 94.89 22.87 522.88 0.24 21 136.4 76.6 90.13 10.38 107.72 0.12 22 302.8 76.6 88.82 32.45 1052.68 0.37 23 268.6 76.6 88.20 31.13 968.85 0.35 24 170.6 76.6 93.74 17.64 311.19 0.19 The results of the investigations in figure 12 shows that only the failure in Pipe No. 18 can influence the The results of the investigations in figure 12 shows that only the failure in Pipe No. 18 can influence the pressure nodes. Consequently, this pipe is one of the most sensitive pipes in this network. However, reliability performance in the failure 375 conditions is similar to no failure conditions in figure 11. Finally, for this network that includes low diameter in existing pipes, nI has not a suitable performance because of making the uniformity in pipes connected to a node leads to the decrease of the diameter of new pipes. Thus, the capability of the surplus pressure decreases due to the increase in head-loss in the pipes. Consequently, this pipe is one of the most sensitive pipes in this network. However, reliability performance in the failure 375 conditions is similar to no failure conditions in figure 11. Finally, for this network that includes low diameter in existing pipes, I h t it bl f b f ki th if it i i t d t d l d t th d f th Consequently, this pipe is one of the most sensitive pipes in this network. However, reliability performance in the failure 375 conditions is similar to no failure conditions in figure 11. https://doi.org/10.5194/dwes-2020-37 Drinking Water Engineering and Science Discussions Open Access Preprint. Discussion started: 5 February 2021 c⃝Author(s) 2021. CC BY 4.0 License. Finally, for this network that includes low diameter in existing pipes, conditions is similar to no failure conditions in figure 11. Finally, for this network that includes low diameter in existing pipes, nI has not a suitable performance because of making the uniformity in pipes connected to a node leads to the decrease of the diameter of new pipes. Thus, the capability of the surplus pressure decreases due to the increase in head-loss in the pipes. nI has not a suitable performance because of making the uniformity in pipes connected to a node leads to the decrease of the diameter of new pipes. Thus, the capability of the surplus pressure decreases due to the increase in head-loss in the pipes. 24 24 24 Drinking Water Engineering and Science Discussions Open Access https://doi.org/10.5194/dwes-2020-37 Preprint. Discussion started: 5 February 2021 c⃝Author(s) 2021. CC BY 4.0 License. (a) (b ) (c) 380 (a) (a) (b ) (c) (c) 25 385 (d ) Fig. 12 Surplus pressure of nodes in Real- life network for solutions of maximum reliability criteria under failure of pipes No. (a) 18, (b) 21, (c) 27 and (d) 34 https://doi.org/10.5194/dwes-2020-37 Drinking Water Engineering and Science Discussions Open Access Preprint. Discussion started: 5 February 2021 c⃝Author(s) 2021. CC BY 4.0 License. Drinking Water Engineering and Science Discussions Open Access https://doi.org/10.5194/dwes-2020-37 Preprint. Discussion started: 5 February 2021 c⃝Author(s) 2021. CC BY 4.0 License. https://doi.org/10.5194/dwes-2020-37 Preprint. Discussion started: 5 February 2021 c⃝Author(s) 2021. CC BY 4.0 License. 385 (d ) Fig. 12 Surplus pressure of nodes in Real- life network for solutions of maximum reliability criteria under failure of pipes No. (a) 18, (b) 21, (c) 27 and (d) 34 385 (d ) 385 Fig. 12 Surplus pressure of nodes in Real- life network for solutions of maximum reliability criteria under failure of pipes No. (a) 18, (b) 21, (c) 27 and (d) 34 390 Fig. 13 Velocity variations in pipes for the solutions of minimum cost and maximum rI , MRI and m I criteria 390 Fig. 13 Velocity variations in pipes for the solutions of minimum cost and maximum rI , MRI and m I criteria Fig. https://doi.org/10.5194/dwes-2020-37 Drinking Water Engineering and Science Discussions Open Access Preprint. Discussion started: 5 February 2021 c⃝Author(s) 2021. CC BY 4.0 License. 13 Velocity variations in pipes for the solutions of minimum cost and maximum rI , MRI and m I criteria Figure 13 shows the velocity variations in the pipes for the solutions with minimum cost and maximum rI , MRI and m I criteria obtained using GA and NSGA-II in GANetXL. As it is observed, when the cost is the basis for the design and optimization of Real- life network, velocity variation are so high in the pipes. This can lead to some issues in the network. But 26 Drinking Water Engineering and Science Discussions Open Access https://doi.org/10.5194/dwes-2020-37 Preprint. Discussion started: 5 February 2021 c⃝Author(s) 2021. CC BY 4.0 License. https://doi.org/10.5194/dwes-2020-37 Preprint. Discussion started: 5 February 2021 c⃝Author(s) 2021. CC BY 4.0 License. in the presented solutions with maximum reliability criteria ( rI , MRI and m I ), velocity variations are not only low but 395 almost uniform. in the presented solutions with maximum reliability criteria ( rI , MRI and m I ), velocity variations are not only low but 395 almost uniform. References Alperovits, E. and U. Shamir (1977). "Design of optimal water distribution systems." Water resources research 13(6): 885-900. Atiquzzaman, M. and S.-Y. Liong (2004). "Using shuffled complex evolution to calibrate water distribution network model." Journal of Civil Engineering (IEB) 32(2): 111-119. Atkinson, S., R. Farmani, F. A. Memon and D. Butler (2014). "Reliability indicators for water distribution system design: 420 Comparison." Journal of Water Resources Planning and Management 140(2): 160-168. B ñ R C Gil J A ll i d J R (2007) A ti l ith f t di t ib ti t k d i S ft ti Atkinson, S., R. Farmani, F. A. Memon and D. Butler (2014). "Reliability indicators for water distribution system design: 420 Comparison." Journal of Water Resources Planning and Management 140(2): 160-168. Baños, R., C. Gil, J. Agulleiro and J. Reca (2007). A memetic algorithm for water distribution network design. Soft computing in industrial applications, Springer: 279-289. Baños, R., J. Reca, J. Martínez, C. Gil and A. L. Márquez (2011). "Resilience indexes for water distribution network design: a Baños, R., J. Reca, J. Martínez, C. Gil and A. L. Márquez (2011). "Resilience indexes for water distribution network design: a performance analysis under demand uncertainty." Water resources management 25(10): 2351-2366. 425 performance analysis under demand uncertainty." Water resources management 25(10): 2351-2366. 425 Cunha, M. d. C. and J. Sousa (1999). "Water distribution network design optimization: simulated annealing approach." Journal of Water Resources Planning and Management 125(4): 215-221. Deb, K., A. Pratap, S. Agarwal and T. Meyarivan (2002). "A fast and elitist multiobjective genetic algorithm: NSGA-II." IEEE transactions on evolutionary computation 6(2): 182-197. g g Deb, K., A. Pratap, S. Agarwal and T. Meyarivan (2002). "A fast and elitist multiobjective genetic algorithm: NSGA-II." IEEE transactions on evolutionary computation 6(2): 182-197. Department of Technical Affairs, B. o. E. a. T. C. f. W. a. W., Ministry of Energy (2013). Design Criteria of Urban and Rural 430 Water Supply and Distribution Systems. No . 117-3, Vice Presidency For Strategic Planning and Supervision. Eusuff, M. M. and K. E. Lansey (2003). "Optimization of water distribution network design using the shuffled frog leaping algorithm." Journal of Water Resources planning and management 129(3): 210-225. Water Supply and Distribution Systems. No . 117 3, Vice Presidency For Strategic Planning and Supervision. Eusuff, M. M. and K. E. Lansey (2003). 4. Conclusions 410 mpeting interests: There is no conflict of interest. ata Availability: All data and models generated or used during this study are proprietary and confidential in nature. Author contribution: Conceptualization and data curation were done by A Moghaddam, formal analysis and methodology carried out by A Moghaddam, R Peirovi and H Rezaee, project administration was done by A Faridhosseini and R Peirovi, Author contribution: Conceptualization and data curation were done by A Moghaddam, formal analysis and methodology carried out by A Moghaddam, R Peirovi and H Rezaee, project administration was done by A Faridhosseini and R Peirovi, visualization and writing have prepared by A Naghi Ziaei, A Moghaddam and R Peirovi. 415 carried out by A Moghaddam, R Peirovi and H Rezaee, project administration was done by A Faridhosseini and R Peirovi, visualization and writing have prepared by A Naghi Ziaei, A Moghaddam and R Peirovi. 415 y g p j y visualization and writing have prepared by A Naghi Ziaei, A Moghaddam and R Peirovi. 415 visualization and writing have prepared by A Naghi Ziaei, A Moghaddam and R Peirovi. 415 4. Conclusions In this paper the performance of a few reliability criteria was evaluated when applying to a two benchmark (T In this paper the performance of a few reliability criteria was evaluated when applying to a two benchmark (Two loop and Hanoi) and one real-life (in Iran) networks. Both the existing pipes and hydraulic constraints were considered in the study in 400 which GANetXL was used as the optimizer. The optimizations were performed taking into account two different objective functions including a cost and reliability. Hanoi) and one real-life (in Iran) networks. Both the existing pipes and hydraulic constraints were considered in the study in 400 which GANetXL was used as the optimizer. The optimizations were performed taking into account two different objective functions including a cost and reliability. The results of cost-oriented optimization showed that the solutions proposed by GANetXL for case study networks give solutions that are either less expensive than or as the same as the ones from literature. . In order to investigate the solutions with maximum The results of cost-oriented optimization showed that the solutions proposed by GANetXL for case study networks give solutions that are either less expensive than or as the same as the ones from literature. . In order to investigate the solutions with maximum values of rI , nI , MRI and m I criteria and finding sensitive and important pipes with the most probability of failure in the 405 network, statistical analysis of single-objective optimization was used. The results showed that rI ,MRI and m I criteria have better performance than nI under failure conditions, especially in real-life networks that include the existing pipes with small diameter and if a WDN was only optimized based on cost, it would be difficult to overcome losses in pipe failure conditions and pressure supply of nodes. values of rI , nI , MRI and m I criteria and finding sensitive and important pipes with the most probability of failure in the 405 network, statistical analysis of single-objective optimization was used. The results showed that rI ,MRI and m I criteria have better performance than nI under failure conditions, especially in real-life networks that include the existing pipes with small diameter and if a WDN was only optimized based on cost, it would be difficult to overcome losses in pipe failure conditions and pressure supply of nodes. References Jayaram, N. and K. Srinivasan (2008). "Performance‐based optimal design and rehabilitation of water distribution networks using life cycle costing." Water Resources Research 44(1). Lansey, K. E. and L. W. Mays (1989). Optimization models for design of water distribution systems. Reliability Analysis of Jacoby, S. L. (1968). "Design of optimal hydraulic networks." Journal of the Hydraulics Division 94(3): 641-662. Jayaram, N. and K. Srinivasan (2008). "Performance‐based optimal design and rehabilitation of water distribution networks using life cycle costing." Water Resources Research 44(1). Lansey K E and L W Mays (1989) Optimization models for design of water distribution systems Reliability Analysis of Jayaram, N. and K. Srinivasan (2008). "Performance‐based optimal design and rehabilitation of water distribution networks using life cycle costing." Water Resources Research 44(1). Lansey, K. E. and L. W. Mays (1989). Optimization models for design of water distribution systems. Reliability Analysis of Water Distribution Systems. Part 1: State-of-the-Art. 450 Water Distribution Systems. Part 1: State-of-the-Art. 450 Lin, M.-D., Y.-H. Liu, G.-F. Liu and C.-W. Chu (2007). "Scatter search heuristic for least-cost design of water distribution networks." Engineering Optimization 39(7): 857-876. Lippai, I., J. P. Heaney and M. Laguna (1999). "Robust water system design with commercial intelligent search optimizers." Journal of Computing in Civil Engineering 13(3): 135-143. y Lin, M.-D., Y.-H. Liu, G.-F. Liu and C.-W. Chu (2007). "Scatter search heuristic for least-cost design of water distribution networks." Engineering Optimization 39(7): 857-876. Lippai, I., J. P. Heaney and M. Laguna (1999). "Robust water system design with commercial intelligent search optimizers." P. Heaney and M. Laguna (1999). "Robust water system design with commercial intelligent search optimizers." mputing in Civil Engineering 13(3): 135-143. Loganathan, G., J. Greene and T. Ahn (1995). "Design heuristic for globally minimum cost water-distribution systems." Journal 455 of Water Resources Planning and Management 121(2): 182-192. Loganathan, G., J. Greene and T. Ahn (1995). "Design heuristic for globally minimum cost water-distribution systems." Journal 455 of Water Resources Planning and Management 121(2): 182-192. Maier, H. R., A. R. Simpson, A. C. Zecchin, W. K. Foong, K. Y. Phang, H. Y. Seah and C. L. Tan (2003). "Ant colony optimization for design of water distribution systems." Journal of water resources planning and management 129(3): 200-209. Mala-Jetmarova, H., A. Barton and A. Bagirov (2014). "Exploration of the trade-offs between water quality and pumping costs in optimal operation of regional multiquality water distribution systems." Journal of Water Resources Planning and 141(6): 04014077. References "Optimization of water distribution network design using the shuffled frog leaping algorithm." Journal of Water Resources planning and management 129(3): 210-225. 27 Drinking Water Engineering and Science Discussions Open Access https://doi.org/10.5194/dwes-2020-37 Preprint. Discussion started: 5 February 2021 c⃝Author(s) 2021. CC BY 4.0 License. Farmani, R., H. J. Henriksen and D. Savic (2009). "An evolutionary Bayesian belief network methodology for optimum Farmani, R., H. J. Henriksen and D. Savic (2009). "An evolutionary Bayesian belief network methodology for optimum management of groundwater contamination." Environmental Modelling & Software 24(3): 303-310. 435 management of groundwater contamination." Environmental Modelling & Software 24(3): 303-310. 435 Farmani, R., D. Savic and G. Walters (2005). "Evolutionary multi-objective optimization in water distribution network design." Engineering Optimization 37(2): 167-183. Fujiwara, O. and D. B. Khang (1990). "A two‐phase decomposition method for optimal design of looped water distribution networks." Water resources research 26(4): 539-549. D. Savic and G. Walters (2005). "Evolutionary multi-objective optimization in water distribution network design." ptimization 37(2): 167-183. Fujiwara, O. and D. B. Khang (1990). "A two‐phase decomposition method for optimal design of looped water distribution networks." Water resources research 26(4): 539-549. 2009). "Harmony search optimisation to the pump-included water distribution network design." Civil Engineering ental Systems 26(3): 211-221. Geem, Z. W. (2009). "Harmony search optimisation to the pump-included water distribution network design." Civil Engineering 440 and Environmental Systems 26(3): 211-221. Geem, Z. W., J. H. Kim and G. Loganathan (2002). "Harmony search optimization: application to pipe network design." International Journal of Modelling and Simulation 22(2): 125-133. Greco, R., A. Di Nardo and G. Santonastaso (2012). "Resilience and entropy as indices of robustness of water distribution 44 y ( ) Geem, Z. W., J. H. Kim and G. Loganathan (2002). "Harmony search optimization: application to pipe network design." International Journal of Modelling and Simulation 22(2): 125-133. Geem, Z. W., J. H. Kim and G. Loganathan (2002). "Harmony search optimization: application to pipe network design." International Journal of Modelling and Simulation 22(2): 125-133. Greco, R., A. Di Nardo and G. Santonastaso (2012). "Resilience and entropy as indices of robustness of water distribution networks " Journal of Hydroinformatics 14(3): 761-771 445 Greco, R., A. Di Nardo and G. Santonastaso (2012). "Resilience and entropy as indices of robustness of water distribution networks." Journal of Hydroinformatics 14(3): 761-771. 445 networks." Journal of Hydroinformatics 14(3): 761-771. 445 Jacoby, S. L. (1968). "Design of optimal hydraulic networks." Journal of the Hydraulics Division 94(3): 641-662. References in optimal operation of regional multiquality water distribution systems." Journal of Water Resources Planning and Management 460 141(6): 04014077. Mala-Jetmarova, H., A. Barton and A. Bagirov (2015). "Impact of water-quality conditions in source reservoirs on the optimal operation of a regional multiquality water-distribution system." Journal of Water Resources Planning and Management 141(10): 04015013. ( ) Mala-Jetmarova, H., A. Barton and A. Bagirov (2015). "Impact of water-quality conditions in source reservoirs on the optimal operation of a regional multiquality water-distribution system." Journal of Water Resources Planning and Management 141(10): 04015013. Miri, S. M. and A. Afshar (2014). "Optimum Layout for Sensors in Water Distribution Networks through Ant Colony Algorithm: 465 A Dual Use Vision." Journal of Water and Wastewater(parallel title ); Ab va Fazilab ( in persian ) 25(3): 67-75. Moghaddam, A., A. Alizadeh, A. Faridhosseini, A. Ziaei and D. F. Heravi (2018). " Optimal design of water distribution networks using simple modifed particle swarm optimization approach." Desalination and Water Treatment In press: 1-12. Mohan, S. and K. J. Babu (2009). "Water distribution network design using heuristics-based algorithm." Journal of Computing and K. J. Babu (2009). "Water distribution network design using heuristics-based algorithm." Journal of Computin ngineering 23(5): 249-257. in Civil Engineering 23(5): 249-257. 470 Murphy, L. and A. Simpson (1992). "Pipe optimization using genetic algorithms." Res. Rep 93: 95. Neelakantan, T., C. Suribabu and S. Lingireddy (2008). "Optimisation procedure for pipe-sizing with break-repair and replacement economics." Water SA 34(2): 217-224. Pandit, A. and J. C. Crittenden (2012). "Index of network resilience (INR) for urban water distribution systems." Nature. p Pandit, A. and J. C. Crittenden (2012). "Index of network resilience (INR) for urban water distribution sys and J. C. Crittenden (2012). "Index of network resilience (INR) for urban water distribution systems." Nature. 6 i i f i bl d ili d i l i f di ib i k b Piratla, K. R. (2016). "Investigation of sustainable and resilient design alternatives for water distribution networks." Urban Water 475 Journal 13(4): 412-425. Piratla, K. R. and S. T. Ariaratnam (2012). "Reliability based optimal design of water distribution networks considering life cycle components." Urban Water Journal 9(5): 305-316. Prasad, T. D. and N.-S. Park (2004). "Multiobjective genetic algorithms for design of water distribution networks." Journal of Piratla, K. R. (2016). "Investigation of sustainable and resilient design alternatives for water distribution networks." Urban Water 475 Journal 13(4): 412-425. R. and S. T. References Ariaratnam (2012). "Reliability based optimal design of water distribution networks considering li ponents." Urban Water Journal 9(5): 305-316. Prasad, T. D. and N.-S. Park (2004). "Multiobjective genetic algorithms for design of water distribution networks." Journal of Water Resources Planning and Management 130(1): 73-82. 480 Water Resources Planning and Management 130(1): 73-82. 480 Quindry, G. E., J. C. Liebman and E. D. Brill (1981). "Optimization of looped water distribution systems." Journal of the Environmental Engineering Division 107(4): 665-679. Raad, D., A. Sinske and J. van Vuuren (2010). "Multiobjective optimization for water distribution system design using a hyperheuristic." Journal of Water Resources Planning and Management 136(5): 592-596. 48 Raad, D., A. Sinske and J. van Vuuren (2010). "Multiobjective optimization for water distribution system design using a hyperheuristic." Journal of Water Resources Planning and Management 136(5): 592-596. Rasekh, A., A. Afshar and M. H. Afshar (2010). "Risk-cost optimization of hydraulic structures: methodology and case study." 485 Water resources management 24(11): 2833-2851. Reca, J., J. Martinez, R. Banos and C. Gil (2008). "Optimal design of gravity-fed looped water distribution networks considering the resilience index." Journal of Water Resources Planning and Management 134(3): 234-238. Rossman, L. A. (2000). "EPANET 2: users manual." Water resources management 24(11): 2833 2851. Reca, J., J. Martinez, R. Banos and C. Gil (2008). "Optimal design of gravity-fed looped water distribution networks considering the resilience index." Journal of Water Resources Planning and Management 134(3): 234-238. Rossman, L. A. (2000). "EPANET 2: users manual." 28 Drinking Water Engineering and Science Discussions Open Access https://doi.org/10.5194/dwes-2020-37 Preprint. Discussion started: 5 February 2021 c⃝Author(s) 2021. CC BY 4.0 License. https://doi.org/10.5194/dwes-2020-37 Preprint. Discussion started: 5 February 2021 c⃝Author(s) 2021. CC BY 4.0 License. Savić, D. A., J. Bicik and M. S. Morley (2011). "A DSS generator for multiobjective optimisation of spreadsheet-based models." 490 Environmental modelling & software 26(5): 551-561. Savic, D. A. and G. A. Walters (1997). "Genetic algorithms for least-cost design of water distribution networks." Journal of water resources planning and management 123(2): 67-77. Shafiqul Islam, M., R. Sadiq, M. J. Rodriguez, H. Najjaran and M. Hoorfar (2013). "Reliability assessment for water supply g ( ) and G. A. Walters (1997). "Genetic algorithms for least-cost design of water distribution networks." Journal of es planning and management 123(2): 67-77. Shafiqul Islam, M., R. Sadiq, M. J. Rodriguez, H. Najjaran and M. Hoorfar (2013). References "Reliability assessment for water supply systems under uncertainties." Journal of Water Resources Planning and Management 140(4): 468-479. 495 systems under uncertainties." Journal of Water Resources Planning and Management 140(4): 468-479. 495 Simpson, A. R., G. C. Dandy and L. J. Murphy (1994). "Genetic algorithms compared to other techniques for pipe optimization." Journal of water resources planning and management 120(4): 423-443. Simpson, A. R. and D. E. Goldberg (1994). "Pipeline optimization via genetic algorithms: From theory to practice." Water pipeline systems: 309-320. Simpson, A. R. and D. E. Goldberg (1994). "Pipeline optimization via genetic algorithms: From theory to practice." Water pipeline systems: 309-320. Su, Y.-C., L. W. Mays, N. Duan and K. E. Lansey (1987). "Reliability-based optimization model for water distribution systems." 500 Journal of Hydraulic Engineering 113(12): 1539-1556. Tabesh, M., T. T. Tanyimboh and R. Burrows (2001). Extended period reliability analysis of water distribution systems based on head driven simulation method. Bridging the Gap: Meeting the World's Water and Environmental Resources Challenges. Tayfur, G. (2017). "Modern optimization methods in water resources planning, engineering and management." Water Resources Management 31(10): 3205-3233. 505 Todini, E. (2000). "Looped water distribution networks design using a resilience index based heuristic approach." Urban water 2(2): 115-122. Vasan, A. and S. P. Simonovic (2010). "Optimization of water distribution network design using differential evolution." Journal of Water Resources Planning and Management 136(2): 279-287. Todini, E. (2000). "Looped water distribution networks design using a resilience index based heuristic approach." Urban water 2(2): 115-122. and S. P. Simonovic (2010). "Optimization of water distribution network design using differential evolution." Journ Resources Planning and Management 136(2): 279 287 and S. P. Simonovic (2010). "Optimization of water distribution network design using differential evolution." Journ Resources Planning and Management 136(2): 279-287. Watanatada, T. (1973). "Least-cost design of water distribution systems." Journal of the Hydraulics Division 99(9): 1497-1513. 510 Zecchin, A. C., A. R. Simpson, H. R. Maier, M. Leonard, A. J. Roberts and M. J. Berrisford (2006). "Application of two ant colony optimisation algorithms to water distribution system optimisation." Mathematical and Computer Modelling 44(5-6): 451- 468. 515 29
https://openalex.org/W3189223905	https://www.frontiersin.org/articles/10.3389/fphys.2021.690341/pdf	English	null	Oxidative Stress, Kinase Activation, and Inflammatory Pathways Involved in Effects on Smooth Muscle Cells During Pulmonary Artery Hypertension Under Hypobaric Hypoxia Exposure	Frontiers in physiology	2,021	cc-by	11,765	Citation: Siques P, Pena E, Brito J and El Alam S (2021) Oxidative Stress, Kinase Activation, and Inﬂammatory Pathways Involved in Effects on Smooth Muscle Cells During Pulmonary Artery Hypertension Under Hypobaric Hypoxia Exposure. Front. Physiol. 12:690341. doi: 10.3389/fphys.2021.690341 REVIEW published: 09 August 2021 doi: 10.3389/fphys.2021.690341 REVIEW published: 09 August 2021 doi: 10.3389/fphys.2021.690341 published: 09 August 2021 doi: 10.3389/fphys.2021.690341 Keywords: hypobaric hypoxia, pulmonary artery smooth muscle cell, oxidative stress, kinases, inﬂammation, high altitude High-altitude exposure results in hypobaric hypoxia, which affects organisms by activating several mechanisms at the physiological, cellular, and molecular levels and triggering the development of several pathologies. One such pathology is high- altitude pulmonary hypertension (HAPH), which is initiated through hypoxic pulmonary vasoconstriction to distribute blood to more adequately ventilated areas of the lungs. Importantly, all layers of the pulmonary artery (adventitia, smooth muscle, and endothelium) contribute to or are involved in the development of HAPH. However, the principal action sites of HAPH are pulmonary artery smooth muscle cells (PASMCs), which interact with several extracellular and intracellular molecules and participate in mechanisms leading to proliferation, apoptosis, and ﬁbrosis. This review summarizes the alterations in molecular pathways related to oxidative stress, inﬂammation, kinase activation, and other processes that occur in PASMCs during pulmonary hypertension under hypobaric hypoxia and proposes updates to pharmacological treatments to mitigate the pathological changes in PASMCs under such conditions. In general, PASMCs exposed to hypobaric hypoxia undergo oxidative stress mediated by Nox4, inﬂammation mediated by increases in interleukin-6 levels and inﬂammatory cell inﬁltration, and activation of the protein kinase ERK1/2, which lead to the proliferation of PASMCs and contribute to the development of hypobaric hypoxia- induced pulmonary hypertension. Edited by: Cristina M. Sena, University of Coimbra, Portugal University of Coimbra, Portugal Reviewed by: Bojun Zhang, Baylor College of Medicine, United States Ji-Wang Chen, University of Illinois at Chicago, United States Jorge G. Farias, University of La Frontera, Chile Correspondence: Eduardo Pena Eduardo.enrique.bio@gmail.com †These authors share ﬁrst authorship Reviewed by: Bojun Zhang, Baylor College of Medicine, United States Ji-Wang Chen, University of Illinois at Chicago, United States Jorge G. Farias, University of La Frontera, Chile Correspondence: Eduardo Pena Eduardo.enrique.bio@gmail.com †These authors share ﬁrst authorship Specialty section: This article was submitted to Redox Physiology, a section of the journal Frontiers in Physiology Received: 02 April 2021 Accepted: 16 July 2021 Published: 09 August 2021 Keywords: hypobaric hypoxia, pulmonary artery smooth muscle cell, oxidative stress, kinases, inﬂammation, high altitude PASMCs AND HYPOBARIC HYPOXIA The structure of smooth muscle tissue is diﬀerent from that of other muscle tissue subtypes, and the spindle shape of mononucleated smooth muscle cells (SMCs) permits close contact among cells in vessel walls (Vodenicharov, 2012). The thickness of the medial layer, which is composed mainly of SMCs, greatly determines pulmonary artery resistance and pressure (Siques et al., 2018). Under hypoxic conditions, the muscularization and contractibility of these vessels are enhanced (Meyrick and Reid, 1979), constituting a key feature of pulmonary vascular remodeling and subsequent hypobaric hypoxia-induced pulmonary hypertension (Pak et al., 2007). This eﬀect has been observed both in chronic hypobaric hypoxia (Penaloza and Arias-Stella, 2007) and, albeit to a lesser extent, in intermittent hypobaric hypoxia (Brito et al., 2015). However, this eﬀect in acute hypobaric hypoxia is controversial (Pak et al., 2007), since some studies showed PASMC proliferation under acute hypoxia, while others did not, as demonstrated in a meta- analysis (Pak et al., 2007). Therefore, more studies are necessary to clarify the eﬀect of acute hypobaric hypoxia exposure on SMCs. As mentioned above, vascular remodeling due to pulmonary artery vasoconstriction is a common characteristic of pulmonary hypertension under hypobaric hypoxia, during which important changes occur in the muscular vessel medial layer that result principally from accumulation of vascular SMCs. Thus, hypobaric hypoxia plays an important role in triggering the excessive proliferation of vascular SMCs and contributing to vascular remodeling (Tan et al., 2017). The processes mentioned are related to oxidative stress, kinase protein, and inﬂammation in SMCs under hypobaric hypoxia and are discussed below. The structure of smooth muscle tissue is diﬀerent from that of other muscle tissue subtypes, and the spindle shape of mononucleated smooth muscle cells (SMCs) permits close contact among cells in vessel walls (Vodenicharov, 2012). The thickness of the medial layer, which is composed mainly of SMCs, greatly determines pulmonary artery resistance and pressure (Siques et al., 2018). Under hypoxic conditions, the muscularization and contractibility of these vessels are enhanced (Meyrick and Reid, 1979), constituting a key feature of pulmonary vascular remodeling and subsequent hypobaric hypoxia-induced pulmonary hypertension (Pak et al., 2007). This eﬀect has been observed both in chronic hypobaric hypoxia (Penaloza and Arias-Stella, 2007) and, albeit to a lesser extent, in intermittent hypobaric hypoxia (Brito et al., 2015). INTRODUCTION The cut-oﬀpoint for all types of pulmonary hypertension is a mean pulmonary arterial pressure (mPAP) ≥25 mm Hg (Galiè et al., 2016); however, the cut-oﬀpoint for HAPH is considered to be an mPAP ≥30 mm Hg according to a consensus reached in 2005 (León-Velarde et al., 2005). This condition is observed in subjects who are chronically exposed to high-altitude and is due to hypoxic pulmonary vasoconstriction and subsequent pulmonary artery remodeling (Sommer et al., 2008; Neupane and Swenson, 2017). INTRODUCTION At high altitude – 3,000 m above sea level – the partial pressure of oxygen (PaO2) is decreased, creating a condition termed hypobaric hypoxia; this condition is characterized by reduced bioavailability of oxygen at the organism, tissue, and cell levels, which clearly alters organism homeostasis (Chavala, 2018; Bilo et al., 2019). Currently, more than 140 million people worldwide At high altitude – 3,000 m above sea level – the partial pressure of oxygen (PaO2) is decreased, creating a condition termed hypobaric hypoxia; this condition is characterized by reduced bioavailability of oxygen at the organism, tissue, and cell levels, which clearly alters organism homeostasis (Chavala, 2018; Bilo et al., 2019). Currently, more than 140 million people worldwide August 2021 \| Volume 12 \| Article 690341 Frontiers in Physiology \| www.frontiersin.org 1 Pulmonary Artery SMCs Under High-Altitude Siques et al. are working and living under the condition of hypobaric hypoxia (Moore et al., 1998; West, 2012), and the consequences associated with this exposure constitute a substantial public health burden (Herrera et al., 2015a; Brito et al., 2020). Importantly, all layers of the pulmonary artery (the adventitia, smooth muscle, and endothelial cell layers) and the surrounding cells (mast cells, ﬁbroblasts, and macrophages) contribute to or are involved in the development of HAPH (Vodenicharov, 2012; Siques et al., 2018). However, some of the principal action sites of HAPH are pulmonary artery smooth muscle cells (PASMCs), which interact with several extracellular and intracellular molecules and participate in diverse mechanisms – based on the evolution and degree of hypoxia – that lead to proliferation, apoptosis, and ﬁbrosis (Udjus et al., 2019). The aims of this review are to discuss the molecular alterations related to oxidative stress, inﬂammation, and protein kinase activation that occur under hypobaric hypoxia-induced pulmonary hypertension, particularly in PASMCs, and to identify current and potential molecular targeted pharmacological treatments. High-altitude exposure results in diﬀerent pathologies, such as acute mountain sickness, chronic mountain sickness (or Monge’s disease), acute cerebral edema (ACE), and high-altitude pulmonary edema (HAPE); however, other important pathology associated with high-altitude exposure is high-altitude pulmonary hypertension (HAPH) (León-Velarde et al., 2005; Robinson et al., 2017). HAPH corresponds to classiﬁcation 3.6 for general pulmonary hypertension (Simonneau et al., 2013). PASMCs AND HYPOBARIC HYPOXIA Abbreviations: HAPH, high-altitude pulmonary hypertension; PASMCs, pulmonary artery smooth muscle cells; PaO2, partial pressure of oxygen; ACE, acute cerebral edema; HAPE, high-altitude pulmonary edema; mPAP, mean pulmonary arterial pressure; SMCs, smooth muscle cells; ET-1, endothelin-1; NO, nitric oxide; HIF-1, hypoxia inducible factor-1; NFATc3, nuclear factor of activated T cells isoform c3; ROS, reactive oxygen species; NADPH, nicotinamide adenine dinucleotide phosphate; TGF-β, transforming growth factor-β; eNOS, endothelial nitric oxide synthase; O2.−, superoxide; .OH, hydroxyl radical; H2O2, hydrogen peroxide; CAT, catalase; GSH-PX/RX, glutathione peroxidase/reductase; HDIs, histone deacetylase inhibitors; FOX, forkhead box; FOXM1, forkhead box M1; FOXO1, forkhead box O1; PLK1, polo-like kinase 1; FOXC1, forkhead box C1; Erg-1, early growth response-1; SOD, superoxide dismutase; MPTP, mitochondrial permeability transition pore; MAPK, mitogen-activated protein kinase; ASK1, apoptosis signal-regulating kinase 1; PPARs, peroxisome proliferator-activated receptors; YAP, yes-associated protein; NF-κB, nuclear factor κB; JAK, Janus kinase; STAT, signal transducer and activator of transcription; TRPM7, transient receptor potential melastatin 7; PKC, protein kinase C; IL-6, interleukin-6; IL6R, membrane-bound interleukin-6 receptor; sIL-6R, soluble interleukin-6 receptor; MKL1, megakaryocytic leukemia 1; TNF-α, tumor necrosis factor alpha; SAHA, suberoylanilide hydroxamic acid; BUR, butyrate; JNK, c-Jun N-terminal kinase. Abbreviations: HAPH, high-altitude pulmonary hypertension; PASMCs, pulmonary artery smooth muscle cells; PaO2, partial pressure of oxygen; ACE, acute cerebral edema; HAPE, high-altitude pulmonary edema; mPAP, mean pulmonary arterial pressure; SMCs, smooth muscle cells; ET-1, endothelin-1; NO, nitric oxide; HIF-1, hypoxia inducible factor-1; NFATc3, nuclear factor of activated T cells isoform c3; ROS, reactive oxygen species; NADPH, nicotinamide adenine dinucleotide phosphate; TGF-β, transforming growth factor-β; eNOS, endothelial nitric oxide synthase; O2.−, superoxide; .OH, hydroxyl radical; H2O2, hydrogen peroxide; CAT, catalase; GSH-PX/RX, glutathione peroxidase/reductase; HDIs, histone deacetylase inhibitors; FOX, forkhead box; FOXM1, forkhead box M1; FOXO1, forkhead box O1; PLK1, polo-like kinase 1; FOXC1, forkhead box C1; Erg-1, early growth response-1; SOD, superoxide dismutase; MPTP, mitochondrial permeability transition pore; MAPK, mitogen-activated protein kinase; ASK1, apoptosis signal-regulating kinase 1; PPARs, peroxisome proliferator-activated receptors; YAP, yes-associated protein; NF-κB, nuclear factor κB; JAK, Janus kinase; STAT, signal transducer and activator of transcription; TRPM7, transient receptor potential melastatin 7; PKC, protein kinase C; IL-6, interleukin-6; IL6R, membrane-bound interleukin-6 receptor; sIL-6R, soluble interleukin-6 receptor; MKL1, megakaryocytic leukemia 1; TNF-α, tumor necrosis factor alpha; SAHA, suberoylanilide hydroxamic acid; BUR, butyrate; JNK, c-Jun N-terminal kinase. PASMCs AND HYPOBARIC HYPOXIA However, this eﬀect in acute hypobaric hypoxia is controversial (Pak et al., 2007), since some studies showed PASMC proliferation under acute hypoxia, while others did not, as demonstrated in a meta- analysis (Pak et al., 2007). Therefore, more studies are necessary to clarify the eﬀect of acute hypobaric hypoxia exposure on SMCs. Hypoxic pulmonary vasoconstriction is the ﬁrst response to hypoxia – and certainly to hypobaric hypoxia – due to the decrease in alveolar oxygen pressure (Moudgil et al., 2005; Siques et al., 2018). This intrinsic vasoconstrictor mechanism is activated to distribute oxygen to more adequately ventilated areas of the lung to improve oxygen delivery (Von Euler and Liljestrand, 1946). However, when alveolar hypoxia is maintained, the hypoxic pulmonary vasoconstriction mechanism is exacerbated and can trigger vascular remodeling (León-Velarde et al., 2010; Tan et al., 2017) and, subsequently, pulmonary hypertension, or HAPH (Penaloza et al., 1964; Xu and Jing, 2009; León-Velarde et al., 2010). As mentioned above, pulmonary hypertension is a disease characterized by pulmonary vascular remodeling leading to a progressive increase in pulmonary vascular resistance; however, when this pathology is maintained, the afterload on the right ventricle is increased, and right ventricular hypertrophy occurs, ultimately resulting in right heart failure and death (Mikhael et al., 2019). As mentioned above, vascular remodeling due to pulmonary artery vasoconstriction is a common characteristic of pulmonary hypertension under hypobaric hypoxia, during which important changes occur in the muscular vessel medial layer that result principally from accumulation of vascular SMCs. Thus, hypobaric hypoxia plays an important role in triggering the excessive proliferation of vascular SMCs and contributing to vascular remodeling (Tan et al., 2017). The processes mentioned are related to oxidative stress, kinase protein, and inﬂammation in SMCs under hypobaric hypoxia and are discussed below. OXIDATIVE STRESS Under physiological conditions, the presence of controlled concentrations of reactive oxygen species (ROS) molecules in the organism is necessary to modulate several pathways involved in maintaining homeostasis and cellular diﬀerentiation (Li et al., 2016). ROS are oxidative and unstable molecules August 2021 \| Volume 12 \| Article 690341 Frontiers in Physiology \| www.frontiersin.org 2 Pulmonary Artery SMCs Under High-Altitude Siques et al. derived from oxygen. The most well-known ROS molecules are superoxide anions (O2.−), hydroxyl radicals (.OH), peroxyl radicals (ROO.), and alkoxy radicals (RO.). Additionally, some non-radicals are either oxidizing or easily converted into oxidative species; these non-radicals include hypochlorous acid (HOCl), ozone (O3), singlet oxygen (1O2), hydrogen peroxide (H2O2), and peroxynitrite (ONOO−) (Li et al., 2016). However, when the level of ROS increases due to environmental or chemical stimuli and is not correspondingly reduced by endogenous antioxidant systems, such as those involving superoxide dismutase (SOD), catalase (CAT), glutathione peroxidase/reductase (GSH-PX/RX), and peroxiredoxin/thioredoxin (Herrera et al., 2015b; Peoples et al., 2019), oxidative stress is triggered. This oxidative stress results in damage to DNA, proteins, and lipids as well as mitochondrial permeability transition pore (MPTP) activation, mitochondrial dysfunction, and cell death (Richardson and Schadt, 2014; Farías et al., 2016, 2017). isoform) is overexpressed, contributing to PASMC proliferation leading to pulmonary hypertension. Moreover, NADPH oxidase expression and subsequent oxidative stress are increased in the lungs of rat exposed to hypobaric hypoxia, eﬀects that are also related to PASMC proliferation through depletion of the transcription factor CREB, triggering pulmonary hypertension (Klemm et al., 2011). Although there have not been many studies showing the speciﬁc expression and activity of the Nox4 isoform related to PASMC proliferation and pulmonary hypertension under hypobaric hypoxia, however, it is important to mention that Nox4 is the principal isoform expressed in PASMCs of patients with pulmonary hypertension (Sturrock et al., 2006). Moreover, a recent study by He et al. (2020) showed that inhibition of oxidative stress induced by hypoxia exposure reduces PASMC proliferation by decreasing hypoxia inducible factor-1α (HIF- 1α) levels and Nox4 expression. In addition, a study in mice exposed to both intermittent and chronic hypobaric hypoxia showed that Nox4 expression in the lungs increased and was higher after chronic than intermittent exposure (Lüneburg et al., 2016). Therefore, based on the above evidence, Nox4 could be related to the PASMC proliferation and pulmonary hypertension induced by hypobaric hypoxia exposure. OXIDATIVE STRESS Despite the apparent paradox – that O2 bioavailability is decreased under hypoxic conditions – there is evidence showing that increased ROS levels trigger oxidative stress in hypobaric hypoxia (Klemm et al., 2011; Lüneburg et al., 2016; Siques et al., 2018). Notably, a study on the lungs of rats exposed to hypobaric hypoxia showed that the H2O2 concentration increased and that SOD and GSH-PX activity decreased, producing oxidative stress (Xu et al., 2016). Interestingly, hypoxia and consequent pulmonary hypertension are related to oxidative stress (Bowers et al., 2004; Guzy and Schumacker, 2006; Demarco et al., 2010). This is supported by the results of a human study showing that ROS levels are two- to threefold higher in patients with pulmonary hypertension than in control subjects (Wong et al., 2013) and an animal study showing that the levels of a plasmatic oxidative biomarker (malondialdehyde) are elevated in rats with pulmonary hypertension induced by intermittent or chronic hypobaric hypoxia exposure (Lüneburg et al., 2016). Moreover, Xu et al. (2016) demonstrated that ROS and H2O2 levels are elevated in PASMCs cultured under hypoxic conditions and that the increases in the levels of these molecules are related to PASMC proliferation, contributing to the development of pulmonary hypertension. Regarding Nox4 pathways, studies have showed that transforming growth factor-β (TGF-β) promotes PASMC proliferation through Nox4 expression (Sturrock et al., 2006; Zhou et al., 2013), and Nox4-mediated production of H2O2 depends on TGF-β levels (Martin-Garrido et al., 2011). Notably, hypoxia also increases TGF-β levels in PASMCs, contributing to proliferation (Zhang et al., 2018). Zhang et al. (2018) also determined that rats exposed to chronic hypobaric hypoxia exhibited increased TGF-β levels and phosphorylation of smad3 (a key downstream molecule in the TGF-β signaling pathway) in pulmonary arteries, contributing to pulmonary hypertension and remodeling. Importantly, that study also showed that primary pulmonary adventitial ﬁbroblasts and macrophages contributed to the TGF-β secretion under this hypoxic condition; these mechanisms could thus play an important role in PASMC proliferation and pulmonary artery remodeling (Zhang et al., 2018). As mentioned before, Nox4 could play an important role in the PASMC proliferation under hypobaric hypoxia-induced pulmonary hypertension. However, it is important to highlight that Nox2 isoform mediating vascular remodeling is also demonstrated in both normobaric (Liu et al., 2005) and hypobaric (Norton et al., 2013) hypoxia-induced pulmonary hypertension. OXIDATIVE STRESS Although – to our knowledge – we did not ﬁnd studies particularly in PASMCs proliferation under this latter condition, which results interesting to consider in future studies. JAK2 An important kinase protein related to pulmonary hypertension is mammalian Janus kinase (JAK). This kinase undergoes autophosphorylation at tyrosine residues, which generates binding sites for phosphorylated signal transducer and activator of transcription (STAT). The JAK family contains four evolutionarily conserved members, JAK1, JAK2, JAK3, and tyrosine kinase 2 (TYK2). It has been demonstrated that mRNA expression of JAK1, JAK2, and JAK3 is increased in PASMCs after hypoxia exposure. In addition, inhibition of JAK2 blocks STAT3 expression in PASMCs under hypoxia (Wang et al., 2005). Interestingly, the levels of tyrosine 705-phosphorylated STAT3 (PY-STAT3) are increased in the lungs of rats with pulmonary hypertension induced by hypobaric hypoxia (Huang et al., 2019). JAK2 is involved in the regulation of various processes relevant to cell survival, proliferation, activation, and diﬀerentiation (O’Shea and Plenge, 2012; Thomas et al., 2015). A recent study demonstrated that the JAK2/STAT3 pathway is involved in the proliferation of PASMCs under hypoxic conditions; additionally, mice with SMC-speciﬁc Jak2 knockout show reduced right ventricular systolic pressure and right ventricular hypertrophy under hypoxic conditions, implying improved function of pulmonary blood vessels (Zhang et al., 2020). Moreover, the same study indicated that the JAK2/STAT3 pathway induces pulmonary artery remodeling by enhancing cyclin A2 expression in PASMCs. An important kinase protein related to pulmonary hypertension is mammalian Janus kinase (JAK). This kinase undergoes autophosphorylation at tyrosine residues, which generates binding sites for phosphorylated signal transducer and activator of transcription (STAT). The JAK family contains four evolutionarily conserved members, JAK1, JAK2, JAK3, and tyrosine kinase 2 (TYK2). It has been demonstrated that mRNA expression of JAK1, JAK2, and JAK3 is increased in PASMCs after hypoxia exposure. In addition, inhibition of JAK2 blocks STAT3 expression in PASMCs under hypoxia (Wang et al., 2005). Interestingly, the levels of tyrosine 705-phosphorylated STAT3 (PY-STAT3) are increased in the lungs of rats with pulmonary hypertension induced by hypobaric hypoxia (Huang et al., 2019). As mentioned above, ET-1 may be associated with oxidative stress, since a study in fetal PASMCs showed that increases in the levels of this protein are related to production of large amounts of ROS and that the use of antioxidants or NADPH oxidase inhibitors induces apoptosis of fetal PASMCs (Wedgwood et al., 2001). The above ﬁndings are supported by a study showing that hypoxia-induced ET-1 promotes SMC proliferation via NADPH oxidase activation, contributing to pulmonary hypertension induced by hypoxia exposure (Wu et al., 2007). ET-1 Endothelin-1 markedly aﬀects SMCs of the vascular wall via its potent vasoconstrictor activity and is related to hypertrophic, mitogenic, and antiapoptotic eﬀects on vascular SMCs (Wu et al., 2007). ET-1 is produced primarily by endothelial cells but can also be secreted by epithelial cells, macrophages, and ﬁbroblasts. Importantly, studies have shown that SMCs can also produce ET-1 (Yanagisawa et al., 1988; Miller et al., 1993; Kosacka and Brzecka, 2021). The activity of ET-1 is increased in hypoxia; this increase is related to the level of HIF-1, since ET-1 contains a HIF-1 binding site at base pair 118 (Yamashita et al., 2001). Additionally, a study has demonstrated that ET-1 may play autocrine and paracrine roles (in SMCs and endothelial cells) in the remodeling of PASMCs during the development of pulmonary hypertension under hypobaric hypoxia (Nakanishi et al., 1999). As previously mentioned, hypobaric hypoxia-induced oxidative stress is also related to protein kinase activation and contributes to the initiation of PASMC proliferation and subsequent pulmonary hypertension. These pathways are discussed in the next section. Peroxisome Proliferator-Activated Receptor to oxidative stress (Barman et al., 2014; Siques et al., 2014; Lüneburg et al., 2016). Therefore, based on the above evidence, NADPH oxidase could play an important role in the oxidative stress that occurs in PASMCs under hypobaric hypoxia. Peroxisome Proliferator-Activated Receptor Other molecules related to oxidative stress in PASMCs are the peroxisome proliferator-activated receptors (PPARs), which regulate several genes involved in cell metabolism, growth, and diﬀerentiation (Lefebvre et al., 2006). PPARγ is expressed in several cells, including pulmonary endothelial cells and PASMCs, where it plays an important role in normal pulmonary vascular function (Green et al., 2011, 2012). A study on human PASMCs exposed to hypoxia determined that PPARγ downregulation is related to increases in Nox4-mediated H2O2 production, ERK1/2 levels, and nuclear factor κB (NF-κB) levels (Lu et al., 2013). However, PPARγ activation attenuates signaling pathways associated with SMC proliferation and remodeling (Nisbet et al., 2010). Moreover, loss of PPARγ expression is associated with the development of pulmonary hypertension induced by hypoxia; this ﬁnding is supported by a study showing that the expression of PPARγ is decreased in the lungs of rodents with pulmonary hypertension induced by chronic hypoxia exposure (Nisbet et al., 2010). Therefore, alterations in PPARγ under hypoxia are related to kinase activation, oxidative stress, and inﬂammation, which could contribute to SMCs proliferation under hypobaric hypoxia. Pathways Related to Oxidative Stress y Reactive oxygen species are also involved in regulation of cell signaling through the mitogen-activated protein kinase (MAPK) pathway, which is associated with a variety of cellular eﬀects, including proliferation, survival, death, and transformation (Son et al., 2013). One study on rats exposed to hypobaric hypoxia showed that the constitutive MAPK termed apoptosis signal- regulating kinase 1 (ASK1), which is known to be critical in the remodeling processes of pulmonary artery hypertension, participates in PASMC proliferation through activation of p38 MAPK and endothelin-1 (ET-1) in response to oxidative stress (Wilson et al., 2020). NADPH Oxidase Regarding the sources of ROS, studies have proposed that mitochondria and nicotinamide adenine dinucleotide phosphate (NADPH) oxidase are the major ROS producers in the cardiovascular system (Daiber et al., 2017; Fukai and Ushio- Fukai, 2020). Additionally, it has been demonstrated the involvement of mitochondria-derived ROS in the development of normobaric hypoxia-induced pulmonary hypertension (Adesina et al., 2015; Snow et al., 2020). Although, previous studies have claimed that the principal source of ROS in the cardiovascular system is the NADPH oxidase complex (Moudgil et al., 2005), and NADPH oxidase-induced oxidative stress may play an important role in pulmonary artery hypertension induced by hypoxia. In addition, studies on PASMCs under normobaric hypoxia in both humans (Ismail et al., 2009) and animals (Mittal et al., 2007) have shown that the NADPH oxidase complex (speciﬁcally the Nox4 Another molecule that could contribute to oxidative stress is nitric oxide (NO), which is an important endogenous vasodilator that act directly in SMCs and is altered under hypobaric hypoxia. One study has revealed that NO bioavailability is decreased in the pulmonary arteries of rats with hypobaric hypoxia-induced pulmonary hypertension (Siques et al., 2014). This phenomenon could be explained by the reaction between NO and NADPH oxidase-produced O2.−, which produces ONOO−, contributing August 2021 \| Volume 12 \| Article 690341 Frontiers in Physiology \| www.frontiersin.org 3 Pulmonary Artery SMCs Under High-Altitude Siques et al. IL-6 The level of IL-6 was demonstrated to be increased in mouse lungs and cultured PASMCs after exposure to normobaric hypoxia, elucidating that IL-6 stimulates human PASMC migration (Savale et al., 2009). In addition, mice with IL-6 deﬁciency exposed to hypoxia showed reductions in pulmonary hypertension, vascular remodeling, and macrophage recruitment in the lung (Savale et al., 2009). Importantly, increases in IL-6 have also been showed to occur in the lungs of mice exposed to chronic hypobaric hypoxia (Maston et al., 2018). JAK2 Moreover, another study determined that circulating ET-1 levels were signiﬁcantly higher in HAPH patients than in healthy high-altitude inhabitants (Kojonazarov et al., 2012), which is supported by the aforementioned study by Nakanishi et al. (1999). JAK2 is involved in the regulation of various processes relevant to cell survival, proliferation, activation, and diﬀerentiation (O’Shea and Plenge, 2012; Thomas et al., 2015). A recent study demonstrated that the JAK2/STAT3 pathway is involved in the proliferation of PASMCs under hypoxic conditions; additionally, mice with SMC-speciﬁc Jak2 knockout show reduced right ventricular systolic pressure and right ventricular hypertrophy under hypoxic conditions, implying improved function of pulmonary blood vessels (Zhang et al., 2020). Moreover, the same study indicated that the JAK2/STAT3 pathway induces pulmonary artery remodeling by enhancing cyclin A2 expression in PASMCs. Endothelin-1 has been found to be related to upregulation of nuclear factor of activated T cells isoform c3 (NFATc3) in the pulmonary arteries of mice under chronic hypobaric hypoxia (De Frutos et al., 2011). NFATc3 is a Ca2+-dependent transcription factor that has been implicated in the PASMC proliferative response and pulmonary hypertension under chronic hypobaric hypoxia in adult mice (De Frutos et al., 2007; Bierer et al., 2011). The role of transcription factors in this process will be discussed later. August 2021 \| Volume 12 \| Article 690341 Frontiers in Physiology \| www.frontiersin.org 4 Pulmonary Artery SMCs Under High-Altitude Siques et al. PTEN The macrophage accumulation and IL-6 production are regulated by PTEN, since a study on rats exposed to hypobaric hypoxia has shown that selective inhibition of PTEN in PASMCs increases their proliferation and impairs pulmonary hypertension; highlighting that depletion of PTEN exacerbated macrophage accumulation and IL-6 production (Horita et al., 2013). PTEN is a phosphatase enzyme that suppresses multiple signaling networks involved in cell proliferation, survival, and inﬂammation (Huang et al., 2019). It also is a speciﬁc regulator and protector against pathological vascular SMC remodeling, which is inactivated in the lungs (Huang et al., 2019) and PASMCs in human pulmonary hypertension (Horita et al., 2013). However, a study on the lungs of rats exposed to hypobaric hypoxia found that PTEN expression was decreased but that the membrane localization of its regulator caveolin-1 (cav-1) was unchanged, which could contribute to hypoxia- induced pulmonary hypertension (Huang et al., 2019). This latter ﬁnding needs further validation in PASMCs proliferation under hypobaric hypoxia. MIF Additionally, both macrophages and endothelial cells can produce a proinﬂammatory cytokine termed macrophage migration inhibitory factor (MIF). Under stress conditions (such as hypoxia), this molecule inhibits the random migration of macrophages and stimulates TNF-α, IL-1β, IL-6, and IL-8 production (Bloom and Bennett, 1966). MIF has been found to be related to proliferation of SMCs (Calandra et al., 1994; Chen et al., 2004; Onodera et al., 2004), and a study on rats under hypobaric hypoxia showed that an increase in MIF induces proliferation of PASMCs through activation of ERK1/2 and JNK proteins, contributing to the pulmonary hypertension (Zhang et al., 2012). PTEN ERK1/2 also mediates the activation of protein kinase C (PKC) and G-protein coupled receptors, and this mechanism induces SMC proliferation under hypoxia (Dempsey et al., 1996; Lannér et al., 2005). Importantly, oxidative agents such as H2O2 and superoxide result in PKC activation independent of the classical PKC cofactor diacylglycerol; indeed, a study showed that PKC activation induced by H2O2 results in the activation of the L-type Ca2+ channel in SMCs, highlighting that inhibition of PKC activity inactivates this channel (Perez-Vizcaino et al., 2010). Therefore, we suggest that hypoxia exposure can alter oxidative status and kinase activation, leading to impairment of Ca2+ homeostasis and contractibility in PASMCs. ERK chronic inﬂammatory responses through the secretion of inﬂammatory cytokines [TNF-α, IL-1β, and interleukin-6 (IL-6)] in the pathogenesis of inﬂammatory lung disease (Lee et al., 2021). Interestingly, macrophages are involved in PASMC proliferation, and studies have demonstrated that the numbers of macrophages in perivascular cells increase under hypobaric hypoxia (Horita et al., 2013; Yu Y. R. et al., 2013). ERK Notably, some molecular channels are also related to protein kinase activation during PASMC proliferation under hypobaric hypoxia, as indicated by a study showing that transient receptor potential melastatin 7 (TRPM7), which acts as a magnesium channel in mammalian cells, is downregulated in PASMCs of patients with pulmonary hypertension and in rats with hypobaric hypoxia-induced pulmonary hypertension. In particular, inhibition of TRPM7 increased the proliferation and apoptosis resistance of PASMCs in vitro through activation of the MEK/ERK pathway (Xing et al., 2019). The role of ERK activation in PASMC proliferation could be supported by the ﬁnding that ERK activity is increased in rats with hypobaric hypoxia exposure-induced pulmonary hypertension by a metabolite derived from the lipoxygenase pathway [12(S)- hydroxyeicosatetraenoic acid (12(S)-HETE)], which is related to PASMC proliferation (Preston et al., 2006). SphK1 Sphingosine kinase-1 (SphK1) is an important lipid kinase that is also related to PASMC proliferation. Its product sphingosine-1-phosphate (S1P) induces PASMC proliferation through activation of early growth response protein-1 (Erg-1), a product of platelet-derived growth factor (PDGF)-induced ERK activation (Sysol et al., 2016). This pathway is also activated during hypoxia-induced PASMC proliferation and pulmonary hypertension, and could be regulated by microRNA (miR-1) (Sysol et al., 2018). Interestingly, a recent study has shown that increases in S1P result in dephosphorylation of yes-associated protein (YAP), a transcriptional regulator, which induces an increase in miR-130a/b contributing to PASMC proliferation and pulmonary hypertension (Shi et al., 2021). The activity of the transcription factor YAP will be discussed later. INFLAMMATION Inﬂammatory processes are present in various types of pulmonary hypertension in both humans and animals, contributing to pulmonary vascular remodeling (Savale et al., 2009; Tuder, 2017). It is important to mention the role of inﬂammatory cells, since in mice exposed to hypoxia, monocytes migrate to the lung perivascular space and diﬀerentiate into interstitial macrophages (Florentin et al., 2018). These cells contribute to the development and progression of acute or Interleukin-6 acts through two mechanisms in organisms to exert its biological eﬀects. For example, activation of the membrane-bound IL-6 receptor (IL6R) is related to the classical signaling pathway that controls homeostatic processes and immunological outcomes, such as host defense against bacterial August 2021 \| Volume 12 \| Article 690341 Frontiers in Physiology \| www.frontiersin.org 5 Pulmonary Artery SMCs Under High-Altitude Siques et al. infections and activation of anti-inﬂammatory and regenerative epithelial pathways (Scheller et al., 2011). On the other hand, IL-6 and its soluble receptor sIL-6R form a complex of the IL-6 trans-signaling pathway that is related to pathological responses such as tissue ﬁbrosis, inﬂammatory arthritis, and cardiovascular disease (Hunter and Jones, 2015). Moreover, the binding of IL- 6 to either the membrane-bound IL6R or sIL-6R triggers the oligomerization of the ubiquitous transmembrane protein gp130, resulting in activation of the JAK/STAT signal transduction pathway (Campbell et al., 2014). Notably, there is evidence showing that gp130 expression is increased in the lungs of rats exposed to hypobaric hypoxia (Huang et al., 2019). collagen type I, by SMCs (Crouch et al., 1989). On the other hand, a study based on this ﬁnding by Yuan et al. (2014) showed that depletion of MKL1 prevents the increased synthesis of endogenous collagen type I mRNA in vascular SMCs under hypoxic conditions (Yuan et al., 2014). Therefore, MKL1 synthesized in SMCs likely participates in pulmonary hypertension induced by hypobaric hypoxia through ﬁbrogenesis and inﬂammation. TRANSCRIPTION FACTORS As mentioned previously, YAP is a transcriptional regulator that, together with WW domain-containing transcription regulator 1 (WWTR1 or TAZ), contributes to PASMC proliferation and subsequent pulmonary hypertension under hypoxic conditions through inhibition of cyclooxygenase 2 (COX2) activity (Dieﬀenbach et al., 2017). Additionally, a recent study has shown that YAP induces the activation of forkhead box M1 (FOXM1), a transcription factor that contributes to PASMC proliferation and pulmonary hypertension (Zhang et al., 2021). A recent study showed the importance of IL6R and its potential prosurvival eﬀects in pulmonary hypertension under hypoxic conditions (Tamura et al., 2018). Tamura et al. (2018) exposed transgenic mice (deﬁcient in IL6R in SMCs) to chronic normobaric hypoxia and found that they exhibited decreased activation of STAT3 and expression of antiapoptotic proteins, such as MCL-1 and BCL2, in the lungs. Moreover, the mice showed decreased collagen deposition, inﬂammatory cell inﬁltration, and pulmonary hypertension (Tamura et al., 2018). Interestingly, inhibition of the pathway involving the IL- 6/sIL-6R complex reduces the pulmonary artery remodeling and pulmonary hypertension that occur under hypobaric hypoxia, highlighting that activation of the IL-6/sIL-6R complex enhances PASMC migration (but not proliferation) and that this migration process requires STAT3 activation and increases in [Ca2+]i (Maston et al., 2018). These results indicate that IL-6, IL6R, and sIL-6R could play an important role in hypoxia-induced PASMC proliferation and pulmonary hypertension. Caspase-1 Another interesting protein is the inﬂammasome eﬀector caspase-1, which mediates inﬂammation through activation of the proinﬂammatory cytokines IL-18 and IL-1β (Udjus et al., 2019). A study showed that caspase-1 deﬁciency reduces pulmonary hypertension in mice under normobaric hypoxia, mainly by reducing PASMC proliferation. In addition, caspase-1 activation likely induces PASMC proliferation via the IL-18/IL-1β and IL-6/STAT3 signaling pathways. Therefore, caspase-1 could be another therapeutic target for hypoxia-induced pulmonary hypertension (Udjus et al., 2019). FOX The forkhead box (FOX) proteins are a superfamily of transcriptional regulators that mediate numerous functions during development and adulthood (Lam et al., 2013). FOXM1 is a transcription factor that plays a crucial role in the maintenance and diﬀerentiation of epithelial cells lining the airways, especially during the embryonic stage, during which it is essential in the formation and proliferation of pulmonary vessels (Raoof and Daoud, 2019). In mice with hypoxia-induced pulmonary artery hypertension, FOXM1 mRNA and protein levels are increased in the lungs; moreover, constitutive SMC-speciﬁc knockdown of FOXM1 inhibits hypoxia-induced pulmonary hypertension and prevents the proliferation of PASMCs, although it does not inhibit PASMC migration (Dai et al., 2018). Importantly, FOXM1 in PASMCs is regulated by HIF-1α under normoxic conditions but stimulated by HIF-2α under hypoxic conditions (Xia et al., 2009). Therefore, FOXM1 can be modulated by HIF-1α and HIF-2α, but its response to oxygen (according to its expression level) and its distribution in tissues diﬀer (Raghavan et al., 2012). Hypoxia upregulates FOXM1 expression in SMCs by increasing the release of growth factors and inﬂammatory cytokines from endothelial cells, contributing to pulmonary artery hypertension (Raoof and Daoud, 2019). MKL1 The transcriptional regulator megakaryocytic leukemia 1 (MKL1) is also an important modulator of SMCs under hypobaric hypoxia (Yuan et al., 2014). MKL1 inhibition attenuates pulmonary hypertension in rats exposed to chronic hypobaric hypoxia (Yuan et al., 2014). It also decreases the expression of inﬂammatory cytokines (TNF-α and IL-6) and chemokines responsible for recruitment of inﬂammatory cells, such as C–C motif chemokine ligand 2 (CCL2) and CCL5, resulting in decreased aggregation of cell types including macrophages, leukocytes, and T lymphocytes in the lungs (Yuan et al., 2014). Another important transcription factor that regulates the cell cycle is forkhead box protein O1 (FOXO1), which has been demonstrated to play a role opposite that of FOXM1 (Sengupta et al., 2013). FOXO1 expression is also decreased in the lungs of humans with another type of pulmonary artery hypertension. Supporting the role of FOXO1 under hypoxia, it has been demonstrated that inhibition of FOXO1 signiﬁcantly increases PASMC proliferation (Savai et al., 2014). There is a relationship among FOXM1, FOXO1, and the cell cycle regulator polo-like kinase 1 (PLK1) (Wilson et al., 2019). The latter has been shown to be an important upregulator of FOXM1 expression, contributing to PASMC proliferation In addition, a study on calves has indicated that hypobaric hypoxia-induced pulmonary hypertension is associated with increased production of extracellular matrix proteins, speciﬁcally August 2021 \| Volume 12 \| Article 690341 Frontiers in Physiology \| www.frontiersin.org 6 Pulmonary Artery SMCs Under High-Altitude Siques et al. in subjects with pulmonary artery hypertension (Wilson et al., 2019). Inhibiting the actions of either PLK1 or FOXM1 in human PASMCs increases the expression of the cyclin-dependent kinase inhibitor 1B (p27kip1) protein, which inhibits the cell cycle progression of PASMCs. In addition, speciﬁc inhibition of FOXO1 increases the expression of FOXM1 (Wilson et al., 2019). exposed to hypobaric hypoxia (Huang et al., 2019). On the other hand, induction of Wnt5a overexpression in recombinant mice signiﬁcantly inhibits β-catenin expression, human PASMC proliferation, and subsequent hypoxia-induced pulmonary hypertension (Yu X. M. et al., 2013; Jin et al., 2015). Therefore, the Wnt/β-catenin pathway could be important to consider in future studies on hypobaric hypoxia-induced PASMC proliferation and pulmonary hypertension. p More studies are necessary to assess the roles of FOX proteins in pulmonary hypertension under hypobaric hypoxia. Wnt/β-Catenin Another pathway that is important to mention is the Wnt/β-catenin pathway. β-Catenin is a transcription factor involved in PASMC proliferation, and its expression can be inhibited by Wnt5a (in a non-canonical Wnt pathway) in hypoxia-induced pulmonary hypertension (Yu X. M. et al., 2013). Wnt5a levels are decreased under hypoxic conditions (Yu X. M. et al., 2013). Therefore, β-catenin activity is predominant under hypoxia, as supported by a recent study showing that the expression of β-catenin is increased in the lungs of rats Erg-1 The redox-sensitive transcription factor Erg-1 regulates cell proliferation, and the levels of this transcription factor are increased in newborn calves with pulmonary artery hypertension and remodeling induced by chronic hypobaric hypoxia. The increases in Erg-1 are related to protein kinase activation and oxidative stress, since oxidant/antioxidant imbalance due to chronic hypobaric hypoxia-induced loss of extracellular SOD activity in the pulmonary artery results in Erg-1 activation and contributes to PASMC proliferation through the ERK1/2 pathway (Hartney et al., 2011). Proposed schematics of PASMC proliferation under hypobaric hypoxia summarizing the molecular pathways and their cellular interactions discussed in this review are depicted in Figures 1, 2, respectively. MKL1 However, a recent study has shown that the levels of forkhead box C1 (FOXC1) are increased in the pulmonary arteries of mice exposed to hypobaric hypoxia and in PASMCs cultured under hypoxic conditions (Yang et al., 2020). In addition, the viability and migration ability of PASMCs transfected with a FOXC1 overexpression plasmid have been found to be increased under hypoxic conditions. Moreover, the circular RNA mmu_circ_0000790 regulates the expression of FOXC1 by binding to miR-374c, establishing an important regulatory axis of vascular remodeling and pulmonary hypertension under hypobaric hypoxia (Yang et al., 2020). Although these new molecules and pathways were originally described in the context of other types of hypertension, they have been demonstrated to also play a role in hypoxic pulmonary hypertension. PHARMACOLOGICAL INTERVENTIONS TO DECREASE PASMCs PROLIFERATION UNDER HYPOXIC CONDITIONS Additionally, we review the current interventions for PASMCs proliferation under hypoxic conditions. FIGURE 1 \| Proposed scheme summarizing the molecular pathways (oxidative stress, protein kinase and inﬂammation) involved in pulmonary artery smooth muscle cell (PASMC) proliferation induced by hypobaric hypoxia exposure. FIGURE 1 \| Proposed scheme summarizing the molecular pathways (oxidative stress, protein kinase and inﬂammation) involved in pulmonary artery smooth muscle cell (PASMC) proliferation induced by hypobaric hypoxia exposure. FIGURE 1 \| Proposed scheme summarizing the molecular pathways (oxidative stress, protein kinase and inﬂammation) involved in pulmonary artery smooth muscle cell (PASMC) proliferation induced by hypobaric hypoxia exposure. August 2021 \| Volume 12 \| Article 690341 August 2021 \| Volume 12 \| Article 690341 7 Frontiers in Physiology \| www.frontiersin.org Siques et al. Pulmonary Artery SMCs Under High-Altitude FIGURE 2 \| Scheme of the principal cellular interactions involved in pulmonary artery smooth muscle cell (PASMC) proliferation under hypobaric hypoxia. FIGURE 2 \| Scheme of the principal cellular interactions involved in pulmonary artery smooth muscle cell (PASMC) proliferation under hypobaric hypoxia. \| Scheme of the principal cellular interactions involved in pulmonary artery smooth muscle cell (PASMC) proliferation under hypoba FIGURE 3 \| Scheme summarizing pharmacological interventions that have been described to decrease pulmonary artery smooth muscle cell (PASMC) proliferation and pulmonary hypertension under hypobaric hypoxia. FIGURE 3 \| Scheme summarizing pharmacological interventions that have been described to decrease pulmonary artery smooth muscle cell (PASMC) proliferation and pulmonary hypertension under hypobaric hypoxia. inhibited the production of ROS in PASMCs (probably by increasing Nrf-2/Trx-1 axis signaling) and suppressed the proliferation of PASMCs. Additionally, the decrease in the proliferation of PASMCs by resveratrol could be explained by the decrease in HIF-1α expression through MAPK/ERK1 and PI3K/AKT inhibition (Xu et al., 2016). In fact, in support of this hypothesis, a recent study by He et al. (2020) showed that resveratrol inhibited hypoxia-induced oxidative stress and the proliferation of rat PASMCs by decreasing HIF-1α levels and Nox4 expression. One interesting study has shown that treatment with histone deacetylase inhibitors (HDIs), such as suberoylanilide hydroxamic acid (SAHA) and butyrate (BUR), increases the expression of endothelial NO synthase (eNOS) in vascular SMCs, producing an elevation in NO that induces upregulation of p21 expression and cleavage of caspase 3 (a protein related to cell cycle arrest and apoptosis) and decreases hypobaric hypoxia- induced vascular SMC remodeling (Tan et al., 2017). REFERENCES hypobaric hypoxia in rats. Biomed. Res. Int. 2015:169841. doi: 10.1155/2015/1 69841 hypobaric hypoxia in rats. Biomed. Res. Int. 2015:169841. doi: 10.1155/2015/1 69841 Adesina, S. E., Kang, B. Y., Bijli, K. M., Ma, J., Cheng, J., Murphy, T. C., et al. (2015). Targeting mitochondrial reactive oxygen species to modulate hypoxia- induced pulmonary hypertension. Free Radic. Biol. Med. 87, 36–47. doi: 10. 1016/j.freeradbiomed.2015.05.042 Calandra, T., Bernhagen, J., Mitchell, R. A., and Bucala, R. (1994). The macrophage is an important and previously unrecognized source of macrophage migration inhibitory factor. J. Exp. Med. 179, 1895–1902. doi: 10.1084/jem.179.6.1895 inhibitory factor. J. Exp. Med. 179, 1895–1902. doi: 10.1084/jem.179.6.1895 Campbell, I. L., Erta, M., Lim, S. L., Frausto, R., May, U., Rose-John, S., et al. (2014). Trans-signaling is a dominant mechanism for the pathogenic actions of interleukin-6 in the brain. J. Neurosci. 34, 2503–2513. doi: 10.1523/jneurosci. 2830-13.2014 Barman, S. A., Chen, F., Su, Y., Dimitropoulou, C., Wang, Y., Catravas, J. D., et al. (2014). NADPH oxidase 4 is expressed in pulmonary artery adventitia and contributes to hypertensive vascular remodeling. Arterioscler. Thromb. Vasc. Biol. 34, 1704–1715. doi: 10.1161/atvbaha.114.303848 Chang, Z., Zhang, P., Zhang, M., Jun, F., Hu, Z., Yang, J., et al. (2019). Aloperine suppresses human pulmonary vascular smooth muscle cell proliferation via inhibiting inﬂammatory response. Chin. J. Physiol. 62, 157–165. doi: 10.4103/ cjp.cjp_27_19 Biol. 34, 1704–1715. doi: 10.1161/atvbaha.114.303848 Bierer, R., Nitta, C. H., Friedman, J., Codianni, S., de Frutos, S., Dominguez- Bautista, J. A., et al. (2011). NFATc3 is required for chronic hypoxia-induced pulmonary hypertension in adult and neonatal mice. Am. J. Physiol. Lung Cell. Mol. Physiol. 301, L872–L880. doi: 10.1152/ajplung.00405.2010 ol. Physiol. 301, L872–L880. doi: 10.1152/ajplung.00405.2010 Chavala, M. L. A. (2018). A journey between high altitude hypoxia and critical patient hypoxia: what can it teach us about compression and the management of critical disease? Med. Intensiva 42, 380–390. doi: 10.1016/j.medin.2017.08.006 Bilo, G., Caravita, S., Torlasco, C., and Parati, G. (2019). Blood pressure at high altitude: physiology and clinical implications. Kardiol. Pol. 77, 596–603. doi: 10.33963/kp.14832 Chen, J., Wang, Y. X., Dong, M. Q., Zhang, B., Luo, Y., Niu, W., et al. (2017). Reoxygenation reverses hypoxic pulmonary arterial remodeling by inducing smooth muscle cell apoptosis via reactive oxygen species-mediated mitochondrial dysfunction. J. Am. Heart Assoc. 6:e005602. doi: 10.1161/jaha. 117.005602 Bloom, B. R., and Bennett, B. (1966). Mechanism of a reaction in vitro associated with delayed-type hypersensitivity. Science 153, 80–82. doi: 10.1126/science. FUNDING This review was funded by FIC GORE-TARAPACA and BIP 3047 7541-0. AUTHOR CONTRIBUTIONS PS, EP, JB, and SEA contributed to conceiving the review, reviewing the literature, writing and preparing the draft, and reviewing and editing the manuscript. All authors have read and agreed with the submitted version of this manuscript. Finally, an interesting study, which requires further validation, showed that reoxygenation results in increased apoptosis in PASMCs, thus reversing hypobaric hypoxia-induced pulmonary artery remodeling, which could be attributed to mitochondrial ROS-mediated mitochondrial dysfunction (Chen et al., 2017). A proposed scheme of the pharmacological approaches is depicted in Figure 3. CONCLUSION hypertension (Gai et al., 2020). Administration of another traditional Tibetan medicine, Tsantan Sumtang, prevented pulmonary hypertension in rats exposed to hypobaric hypoxia (4,500 m), reducing pulmonary vascular cell proliferation probably by suppressing cyclin D1 and cyclin-dependent kinase 4 (CDK4) expression through inhibition of p27kip1 degradation (He et al., 2018). Hypoxia exposure is an important inducer of PASMC proliferation and subsequent remodeling and pulmonary hypertension. Moreover, evidence suggests that hypobaric hypoxia exposure activates several molecular pathways related to oxidative stress (the Nox4 pathway), protein kinases (ERK1/2), and inﬂammation (IL-6 and inﬂammatory cell inﬁltration). These pathways independently or cooperatively lead to further proliferation of PASMCs, contributing to subsequent pulmonary hypertension. In addition, based on the evidence discussed, a summarized array of some diﬀerent pharmacological approaches has been developed to mitigate this proliferative eﬀect. However, further studies are necessary to validate these treatments in people with PASMCs proliferation and subsequent HAPH induce by hypobaric hypoxia. Consistent with the relation between JAK2/STAT3 pathway activation and cell proliferation, administration of tanshinone IIA (Tan), a fat-soluble pharmacologically active ingredient in Danshen, decreases hypoxia-induced PASMC proliferation via apoptosis activation through inhibition of the JAK2/STAT3 pathway both in vivo (in rats) and in vitro (Chen et al., 2014). Aloperine, an alkaloid isolated from Sophora alopecuroides L., possesses a variety of pharmacological activities, such as anti-inﬂammatory and antitumor activities (Chang et al., 2019). Human PASMCs in which proliferation was induced by stimulation with platelet-derived growth factor- BB (PDGF-BB) were treated with aloperine and exhibited decreased proliferation and increased apoptosis. These eﬀects could be explained by the decreased levels of NF-kB, TNF-α, and cyclin E1 and the increased level of p27kip1 (Chang et al., 2019). PHARMACOLOGICAL INTERVENTIONS TO DECREASE PASMCs PROLIFERATION UNDER HYPOXIC CONDITIONS However, importantly, a signiﬁcant decrease in mPAP does not occur (Tan et al., 2017). Other interesting medicines derived from Tibetan plants have been used to reduce hypoxia-induced pulmonary hypertension. Echinacoside, a phenylethanoid glycoside from the Tibetan herbs Lagotis brevituba Maxim and Cistanche tubulosa, shows antiproliferative eﬀects in rat PASMCs under normobaric hypoxia (Gai et al., 2014). In addition, echinacoside was administered to rats under hypobaric hypoxia (4,500 m), and it reduced pulmonary artery remodeling and pulmonary Resveratrol, a plant-derived polyphenolic compound and a phytoestrogen, has been demonstrated to alleviate hypobaric hypoxia-induced pulmonary hypertension through its anti- inﬂammatory, antioxidant, and antiproliferative eﬀects (Xu et al., 2016). This compound decreased neutrophil inﬁltration in the lungs of rats exposed to hypobaric hypoxia and suppressed the expression of inﬂammatory molecules such as phosphorylated NF-kB, IL-6, IL-1β, and TNF-α. Importantly, resveratrol August 2021 \| Volume 12 \| Article 690341 Frontiers in Physiology \| www.frontiersin.org 8 Pulmonary Artery SMCs Under High-Altitude Siques et al. REFERENCES G., Herrera, E. A., Carrasco-Pozo, C., Sotomayor-Zárate, R., Cruz, G., Morales, P., et al. (2016). Pharmacological models and approaches for pathophysiological conditions associated with hypoxia and oxidative stress. Pharmacol. Ther. 158, 1–23. doi: 10.1016/j.pharmthera.2015.11.006 Ismail, S., Sturrock, A., Wu, P., Cahill, B., Norman, K., Huecksteadt, T., et al. (2009). NOX4 mediates hypoxia-induced proliferation of human pulmonary artery smooth muscle cells: the role of autocrine production of transforming growth factor-{beta}1 and insulin-like growth factor binding protein-3. Am. J. Physiol. Lung Cell. Mol. Physiol. 296, L489–L499. doi: 10.1152/ajplung.90488. 2008 Farías, J. G., Molina, V. M., Carrasco, R. A., Zepeda, A. B., Figueroa, E., Letelier, P., et al. (2017). Antioxidant therapeutic strategies for cardiovascular conditions associated with oxidative stress. Nutrients 9:966. doi: 10.3390/nu9090966 Jin, Y., Wang, W., Chai, S., Liu, J., Yang, T., and Wang, J. (2015). Wnt5a attenuates hypoxia-induced pulmonary arteriolar remodeling and right ventricular hypertrophy in mice. Exp. Biol. Med. 240, 1742–1751. doi: 10.1177/ 1535370215584889 Florentin, J., Coppin, E., Vasamsetti, S. B., Zhao, J., Tai, Y. Y., Tang, Y., et al. (2018). Inﬂammatory macrophage expansion in pulmonary hypertension depends upon mobilization of blood-borne monocytes. J. Immunol. 200, 3612–3625. doi: 10.4049/jimmunol.1701287 Fukai, T., and Ushio-Fukai, M. (2020). Cross-talk between NADPH oxidase and mitochondria: role in ROS signaling and angiogenesis. Cells 9:1849. doi: 10. 3390/cells9081849 Klemm, D. J., Majka, S. M., Crossno, J. T. Jr., Psilas, J. C., Reusch, J. E., and Garat, C. V. (2011). Reduction of reactive oxygen species prevents hypoxia-induced CREB depletion in pulmonary artery smooth muscle cells. J. Cardiovasc. Pharmacol. 58, 181–191. doi: 10.1097/FJC.0b013e31821f2773 Gai, X. Y., Tang, F., Ma, J., Zeng, K. W., Wang, S. L., Wang, Y. P., et al. (2014). Antiproliferative eﬀect of echinacoside on rat pulmonary artery smooth muscle cells under hypoxia. J. Pharmacol. Sci. 126, 155–163. doi: 10.1254/jphs.14072fp Kojonazarov, B., Isakova, J., Imanov, B., Sovkhozova, N., Sooronbaev, T., Ishizaki, T., et al. (2012). Bosentan reduces pulmonary artery pressure in high altitude residents. High Alt. Med. Biol. 13, 217–223. doi: 10.1089/ham.201 1.1107 Gai, X., Lin, P., He, Y., Lu, D., Li, Z., Liang, Y., et al. (2020). Echinacoside prevents hypoxic pulmonary hypertension by regulating the pulmonary artery function. J. Pharmacol. Sci. 144, 237–244. doi: 10.1016/j.jphs.2020.09.002 Kosacka, M., and Brzecka, A. (2021). Endothelin-1 and LOX-1 as markers of endothelial dysfunction in obstructive sleep apnea patients. Int. J. Environ. Res. Public Health 18:1319. doi: 10.3390/ijerph18031319 Galiè, N., Humbert, M., Vachiery, J. REFERENCES doi: 10.1111/bph.13403 He, Q., Nan, X., Li, S., Su, S., Ma, K., Li, Z., et al. (2018). Tsantan sumtang alleviates chronic hypoxia-induced pulmonary hypertension by inhibiting proliferation of pulmonary vascular cells. Biomed. Res. Int. 2018:9504158. doi: 10.1155/2018/ 9504158 De Frutos, S., Diaz, J. M., Nitta, C. H., Sherpa, M. L., and Bosc, L. V. (2011). Endothelin-1 contributes to increased NFATc3 activation by chronic hypoxia in pulmonary arteries. Am. J. Physiol. Cell Physiol. 301, C441–C450. doi: 10.1152/ ajpcell.00029.2011 Herrera, E. A., Farías, J. G., Ebensperger, G., Reyes, R. V., Llanos, A. J., and Castillo, R. L. (2015a). Pharmacological approaches in either intermittent or permanent hypoxia: a tale of two exposures. Pharmacol. Res. 101, 94–101. doi: 10.1016/j.phrs.2015.07.011 De Frutos, S., Spangler, R., Alò, D., and Bosc, L. V. (2007). NFATc3 mediates chronic hypoxia-induced pulmonary arterial remodeling with alpha-actin up- regulation. J. Biol. Chem. 282, 15081–15089. doi: 10.1074/jbc.M702679200 Herrera, E. A., Farías, J. G., González-Candia, A., Short, S. E., Carrasco-Pozo, C., and Castillo, R. L. (2015b). 3 Supplementation and intermittent hypobaric hypoxia induce cardioprotection enhancing antioxidant mechanisms in adult rats. Mar. Drugs 13, 838–860. doi: 10.3390/md130 20838 Demarco, V. G., Whaley-Connell, A. T., Sowers, J. R., Habibi, J., and Dellsperger, K. C. (2010). Contribution of oxidative stress to pulmonary arterial hypertension. World J. Cardiol. 2, 316–324. doi: 10.4330/wjc.v2.i10.316 Horita, H., Furgeson, S. B., Ostriker, A., Olszewski, K. A., Sullivan, T., Villegas, L. R., et al. (2013). Selective inactivation of PTEN in smooth muscle cells synergizes with hypoxia to induce severe pulmonary hypertension. J. Am. Heart Assoc. 2:e000188. doi: 10.1161/jaha.113.000188 Dempsey, E. C., Das, M., Frid, M. G., and Stenmark, K. R. (1996). Unique growth properties of neonatal pulmonary vascular cells: importance of time- and site-speciﬁc responses, cell-cell interaction, and synergy. J. Perinatol. 16, S2–S11. Dieﬀenbach, P. B., Haeger, C. M., Coronata, A. M. F., Choi, K. M., Varelas, X., Tschumperlin, D. J., et al. (2017). Arterial stiﬀness induces remodeling phenotypes in pulmonary artery smooth muscle cells via YAP/TAZ-mediated repression of cyclooxygenase-2. Am. J. Physiol. Lung Cell. Mol. Physiol. 313, L628–L647. doi: 10.1152/ajplung.00173.2017 Huang, J., Frid, M., Gewitz, M. H., Fallon, J. T., Brown, D., Krafsur, G., et al. (2019). Hypoxia-induced pulmonary hypertension and chronic lung disease: caveolin-1 dysfunction an important underlying feature. Pulm. Circ. 9:2045894019837876. doi: 10.1177/2045894019837876 Hunter, C. A., and Jones, S. A. (2015). IL-6 as a keystone cytokine in health and disease. Nat. Immunol. 16, 448–457. doi: 10.1038/ni.3153 Farías, J. REFERENCES 153.3731.80 Bowers, R., Cool, C., Murphy, R. C., Tuder, R. M., Hopken, M. W., Flores, S. C., et al. (2004). Oxidative stress in severe pulmonary hypertension. Am. J. Respir. Crit. Care Med. 169, 764–769. doi: 10.1164/rccm.200301-147OC Chen, M., Liu, Y., Yi, D., Wei, L., Li, Y., and Zhang, L. (2014). Tanshinone IIA promotes pulmonary artery smooth muscle cell apoptosis in vitro by inhibiting the JAK2/STAT3 signaling pathway. Cell. Physiol. Biochem. 33, 1130–1138. doi: 10.1159/000358682 Brito, J., Siques, P., and Pena, E. (2020). Long-term chronic intermittent hypoxia: a particular form of chronic high-altitude pulmonary hypertension. Pulm. Circ. 10, 5–12. doi: 10.1177/2045894020934625 Chen, Z., Sakuma, M., Zago, A. C., Zhang, X., Shi, C., Leng, L., et al. (2004). Evidence for a role of macrophage migration inhibitory factor in vascular disease. Arterioscler. Thromb. Vasc. Biol. 24, 709–714. doi: 10.1161/01.ATV. 0000119356.35748.9e Brito, J., Siques, P., Arribas, S. M., López de Pablo, A. L., González, M. C., Naveas, N., et al. (2015). Adventitial alterations are the main features in pulmonary artery remodeling due to long-term chronic intermittent August 2021 \| Volume 12 \| Article 690341 Frontiers in Physiology \| www.frontiersin.org 9 Pulmonary Artery SMCs Under High-Altitude Siques et al. Crouch, E. C., Parks, W. C., Rosenbaum, J. L., Chang, D., Whitehouse, L., Wu, L. J., et al. (1989). Regulation of collagen production by medial smooth muscle cells in hypoxic pulmonary hypertension. Am. Rev. Respir. Dis. 140, 1045–1051. doi: 10.1164/ajrccm/140.4.1045 Hartney, T., Birari, R., Venkataraman, S., Villegas, L., Martinez, M., Black, S. M., et al. (2011). Xanthine oxidase-derived ROS upregulate Egr-1 via ERK1/2 in PA smooth muscle cells; model to test impact of extracellular ROS in chronic hypoxia. PLoS One 6:e27531. doi: 10.1371/journal.pone.0027531 He, L. N., Lan, Y. R., He, G. M., Guo, S. J., Wen, F. Q., and Wang, T. (2020). [Resveratrol inhibits hypoxia-induced oxidative stress and proliferation in pulmonary artery smooth muscle cells through the HIF-1α/NOX4/ROS signaling pathway]. Sheng Li Xue Bao 72, 551–558. Dai, J., Zhou, Q., Tang, H., Chen, T., Li, J., Raychaudhuri, P., et al. (2018). Smooth muscle cell-speciﬁc FoxM1 controls hypoxia-induced pulmonary hypertension. Cell. Signal. 51, 119–129. doi: 10.1016/j.cellsig.2018.08.003 Daiber, A., Di Lisa, F., Oelze, M., Kröller-Schön, S., Steven, S., Schulz, E., et al. (2017). Crosstalk of mitochondria with NADPH oxidase via reactive oxygen and nitrogen species signalling and its role for vascular function. Br. J. Pharmacol. 174, 1670–1689. REFERENCES Interleukin-6 trans-signaling contributes to chronic hypoxia-induced pulmonary hypertension. Pulm. Circ. 8:2045894018780734. doi: 10.1177/2045894018780734 Richardson, A. G., and Schadt, E. E. (2014). The role of macromolecular damage in aging and age-related disease. J. Gerontol. A Biol. Sci. Med. Sci. 69(Suppl. 1), S28–S32. doi: 10.1093/gerona/glu056 Meyrick, B., and Reid, L. (1979). Hypoxia and incorporation of 3H-thymidine by cells of the rat pulmonary arteries and alveolar wall. Am. J. Pathol. 96, 51–70. Robinson, J. C., Abbott, C., Meadows, C. A., Roach, R. C., Honigman, B., and Bull, T. M. (2017). Long-term health outcomes in high-altitude pulmonary hypertension. High Alt. Med. Biol. 18, 61–66. doi: 10.1089/ham.201 6.0098 Mikhael, M., Makar, C., Wissa, A., Le, T., Eghbali, M., and Umar, S. (2019). Oxidative stress and its implications in the right ventricular remodeling secondary to pulmonary hypertension. Front. Physiol. 10:1233. doi: 10.3389/ fphys.2019.01233 Savai, R., Al-Tamari, H. M., Sedding, D., Kojonazarov, B., Muecke, C., Teske, R., et al. (2014). Pro-proliferative and inﬂammatory signaling converge on FoxO1 transcription factor in pulmonary hypertension. Nat. Med. 20, 1289–1300. doi: 10.1038/nm.3695 y Miller, R. C., Pelton, J. T., and Huggins, J. P. (1993). Endothelins–from receptors to medicine. Trends Pharmacol. Sci. 14, 54–60. doi: 10.1016/0165-6147(93)90031- e Mittal, M., Roth, M., König, P., Hofmann, S., Dony, E., Goyal, P., et al. (2007). Hypoxia-dependent regulation of nonphagocytic NADPH oxidase subunit NOX4 in the pulmonary vasculature. Circ. Res. 101, 258–267. doi: 10.1161/ circresaha.107.148015 Savale, L., Tu, L., Rideau, D., Izziki, M., Maitre, B., Adnot, S., et al. (2009). Impact of interleukin-6 on hypoxia-induced pulmonary hypertension and lung inﬂammation in mice. Respir. Res. 10:6. doi: 10.1186/1465-9921-10-6 Scheller, J., Chalaris, A., Schmidt-Arras, D., and Rose-John, S. (2011). The pro- and anti-inﬂammatory properties of the cytokine interleukin-6. Biochim. Biophys. Acta 1813, 878–888. doi: 10.1016/j.bbamcr.2011.01.034 Moore, L. G., Niermeyer, S., and Zamudio, S. (1998). Human adaptation to high altitude: regional and life-cycle perspectives. Am. J. Phys. Anthropol. 107(Suppl. 27), 25–64. Sengupta, A., Kalinichenko, V. V., and Yutzey, K. E. (2013). FoxO1 and FoxM1 transcription factors have antagonistic functions in neonatal cardiomyocyte cell-cycle withdrawal and IGF1 gene regulation. Circ. Res. 112, 267–277. doi: 10.1161/circresaha.112.277442 Moudgil, R., Michelakis, E. D., and Archer, S. L. (2005). Hypoxic pulmonary vasoconstriction. J. Appl. Physiol. 98, 390–403. doi: 10.1152/japplphysiol.00733. 2004 Nakanishi, K., Tajima, F., Nakata, Y., Osada, H., Tachibana, S., Kawai, T., et al. (1999). Expression of endothelin-1 in rats developing hypobaric hypoxia- induced pulmonary hypertension. Lab. Invest. 79, 1347–1357. REFERENCES The heart and pulmonary circulation at high altitudes: healthy highlanders and chronic mountain sickness. Circulation 115, 1132–1146. doi: 10.1161/circulationaha.106.624544 Li, R., Jia, Z., and Trush, M. A. (2016). Deﬁning ROS in biology and medicine. React. Oxyg. Species 1, 9–21. doi: 10.20455/ros.2016.803 Penaloza, D., Arias-Stella, J., Sime, F., Recavarren, S., and Marticorena, E. (1964). The heart and pulmonary circulation in children at high altitudes: physiological, anatomical, and clinical observations. Pediatrics 34, 568–582. Liu, J. Q., Erbynn, E. M., and Folz, R. J. (2005). Chronic hypoxia-enhanced murine pulmonary vasoconstriction: role of superoxide and gp91phox. Chest 128(6 Suppl.), 594S–596S. doi: 10.1378/chest.128.6_suppl.594S Peoples, J. N., Saraf, A., Ghazal, N., Pham, T. T., and Kwong, J. Q. (2019). Mitochondrial dysfunction and oxidative stress in heart disease. Exp. Mol. Med. 51, 1–13. doi: 10.1038/s12276-019-0355-7 Lu, X., Bijli, K. M., Ramirez, A., Murphy, T. C., Kleinhenz, J., and Hart, C. M. (2013). Hypoxia downregulates PPARγ via an ERK1/2-NF-κB-Nox4-dependent mechanism in human pulmonary artery smooth muscle cells. Free Radic. Biol. Med. 63, 151–160. doi: 10.1016/j.freeradbiomed.2013.05.013 Perez-Vizcaino, F., Cogolludo, A., and Moreno, L. (2010). Reactive oxygen species signaling in pulmonary vascular smooth muscle. Respir. Physiol. Neurobiol. 174, 212–220. doi: 10.1016/j.resp.2010.08.009 Lüneburg, N., Siques, P., Brito, J., Arriaza, K., Pena, E., Klose, H., et al. (2016). Long-term chronic intermittent hypobaric hypoxia in rats causes an imbalance in the asymmetric dimethylarginine/nitric oxide pathway and ROS activity: a possible synergistic mechanism for altitude pulmonary hypertension? Pulm. Med. 2016:6578578. doi: 10.1155/2016/6578578 Preston, I. R., Hill, N. S., Warburton, R. R., and Fanburg, B. L. (2006). Role of 12-lipoxygenase in hypoxia-induced rat pulmonary artery smooth muscle cell proliferation. Am. J. Physiol. Lung Cell. Mol. Physiol. 290, L367–L374. doi: 10.1152/ajplung.00114.2005 Raghavan, A., Zhou, G., Zhou, Q., Ibe, J. C., Ramchandran, R., Yang, Q., et al. (2012). Hypoxia-induced pulmonary arterial smooth muscle cell proliferation is controlled by forkhead box M1. Am. J. Respir. Cell Mol. Biol. 46, 431–436. doi: 10.1165/rcmb.2011-0128OC Martin-Garrido, A., Brown, D. I., Lyle, A. N., Dikalova, A., Seidel-Rogol, B., Lassègue, B., et al. (2011). NADPH oxidase 4 mediates TGF-β-induced smooth muscle α-actin via p38MAPK and serum response factor. Free Radic. Biol. Med. 50, 354–362. doi: 10.1016/j.freeradbiomed.2010.11.007 Raoof, I., and Daoud, A. (2019). Hypoxia-inducible factors-α as a regulator for forkhead box protein M1 in pulmonary artery hypertension. Mustansiriya Med. J. 18, 59–62. doi: 10.4103/mj.mj_12_19 Maston, L. D., Jones, D. T., Giermakowska, W., Resta, T. C., Ramiro-Diaz, J., Howard, T. A., et al. (2018). REFERENCES Shi, W., Wang, Q., Wang, J., Yan, X., Feng, W., Zhang, Q., et al. (2021). Activation of yes-associated protein mediates sphingosine-1-phosphate- induced proliferation and migration of pulmonary artery smooth muscle cells and its potential mechanisms. J. Cell. Physiol. 236, 4694–4708. doi: 10.1002/jcp. 30193 Neupane, M., and Swenson, E. R. (2017). High-altitude pulmonary vascular diseases. Adv. Pulm. Hyperten. 15, 149–157. doi: 10.21693/1933-088X-15.3.149 Nisbet, R. E., Bland, J. M., Kleinhenz, D. J., Mitchell, P. O., Walp, E. R., Sutliﬀ, R. L., et al. (2010). Rosiglitazone attenuates chronic hypoxia-induced pulmonary hypertension in a mouse model. Am. J. Respir. Cell Mol. Biol. 42, 482–490. doi: 10.1165/rcmb.2008-0132OC Simonneau, G., Gatzoulis, M. A., Adatia, I., Celermajer, D., Denton, C., Ghofrani, A., et al. (2013). Updated clinical classiﬁcation of pulmonary hypertension. J. Am. Coll. Cardiol. 62(25 Suppl.), D34–D41. doi: 10.1016/j.jacc.2013. 10.029 Siques, P., Brito, J., and Pena, E. (2018). Reactive oxygen species and pulmonary vasculature during hypobaric hypoxia. Front. Physiol. 9:865. doi: 10.3389/fphys. 2018.00865 Norton, C. E., Broughton, B. R., Jernigan, N. L., Walker, B. R., and Resta, T. C. (2013). Enhanced depolarization-induced pulmonary vasoconstriction following chronic hypoxia requires EGFR-dependent activation of NAD(P)H oxidase 2. Antioxid. Redox Signal. 18, 1777–1788. doi: 10.1089/ars.2012.4836 Siques, P., López de Pablo, A. L., Brito, J., Arribas, S. M., Flores, K., Arriaza, K., et al. (2014). Nitric oxide and superoxide anion balance in rats exposed to chronic and long term intermittent hypoxia. Biomed. Res. Int. 2014:610474. doi: 10.1155/2014/610474 Onodera, S., Nishihira, J., Koyama, Y., Majima, T., Aoki, Y., Ichiyama, H., et al. (2004). Macrophage migration inhibitory factor up-regulates the expression of interleukin-8 messenger RNA in synovial ﬁbroblasts of rheumatoid arthritis patients: common transcriptional regulatory mechanism between interleukin-8 and interleukin-1beta. Arthritis Rheum. 50, 1437–1447. doi: 10.1002/art.20190 Snow, J. B., Norton, C. E., Sands, M. A., Weise-Cross, L., Yan, S., Herbert, L. M., et al. (2020). Intermittent hypoxia augments pulmonary vasoconstrictor reactivity through PKCβ/mitochondrial oxidant signaling. Am. J. Respir. Cell Mol. Biol. 62, 732–746. doi: 10.1165/rcmb.2019-0351OC O’Shea, J. J., and Plenge, R. (2012). JAK and STAT signaling molecules in immunoregulation and immune-mediated disease. Immunity 36, 542–550. doi: 10.1016/j.immuni.2012.03.014 Sommer, N., Dietrich, A., Schermuly, R. T., Ghofrani, H. A., Gudermann, T., Schulz, R., et al. (2008). Regulation of hypoxic pulmonary vasoconstriction: basic mechanisms. Eur. Respir. J. 32, 1639–1651. doi: 10.1183/09031936. 00013908 Pak, O., Aldashev, A., Welsh, D., and Peacock, A. (2007). The eﬀects of hypoxia on the cells of the pulmonary vasculature. Eur. Respir. J. REFERENCES L., Gibbs, S., Lang, I., Torbicki, A., et al. (2016). 2015 ESC/ERS guidelines for the diagnosis and treatment of pulmonary hypertension: the joint task force for the diagnosis and treatment of pulmonary hypertension of the European Society of Cardiology (ESC) and the European Respiratory Society (ERS): endorsed by: Association for European Paediatric and Congenital Cardiology (AEPC), International Society for Heart and Lung Transplantation (ISHLT). Eur. Heart J. 37, 67–119. doi: 10.1093/eurheartj/ ehv317 Lam, E. W., Brosens, J. J., Gomes, A. R., and Koo, C. Y. (2013). Forkhead box proteins: tuning forks for transcriptional harmony. Nat. Rev. Cancer 13, 482–495. doi: 10.1038/nrc3539 Lannér, M. C., Raper, M., Pratt, W. M., and Rhoades, R. A. (2005). Heterotrimeric G proteins and the platelet-derived growth factor receptor-beta contribute to hypoxic proliferation of smooth muscle cells. Am. J. Respir. Cell Mol. Biol. 33, 412–419. doi: 10.1165/rcmb.2005-0004OC Green, D. E., Murphy, T. C., Kang, B. Y., Kleinhenz, J. M., Szyndralewiez, C., Page, P., et al. (2012). The Nox4 inhibitor GKT137831 attenuates hypoxia-induced pulmonary vascular cell proliferation. Am. J. Respir. Cell Mol. Biol. 47, 718–726. doi: 10.1165/rcmb.2011-0418OC Lee, J. W., Chun, W., Lee, H. J., Min, J. H., Kim, S. M., Seo, J. Y., et al. (2021). The role of macrophages in the development of acute and chronic inﬂammatory lung diseases. Cells 10:897. doi: 10.3390/cells10040897 g Lefebvre, P., Chinetti, G., Fruchart, J. C., and Staels, B. (2006). Sorting out the roles of PPAR alpha in energy metabolism and vascular homeostasis. J. Clin. Invest. 116, 571–580. doi: 10.1172/jci27989 Green, D. E., Sutliﬀ, R. L., and Hart, C. M. (2011). Is peroxisome proliferator- activated receptor gamma (PPARγ) a therapeutic target for the treatment of pulmonary hypertension? Pulm. Circ. 1, 33–47. doi: 10.4103/2045-8932.78101 León-Velarde, F., Maggiorini, M., Reeves, J. T., Aldashev, A., Asmus, I., Bernardi, L., et al. (2005). Consensus statement on chronic and subacute high altitude diseases. High Alt. Med. Biol. 6, 147–157. doi: 10.1089/ham.2005.6.147 Guzy, R. D., and Schumacker, P. T. (2006). Oxygen sensing by mitochondria at complex III: the paradox of increased reactive oxygen species during hypoxia. Exp. Physiol. 91, 807–819. doi: 10.1113/expphysiol.2006.033506 August 2021 \| Volume 12 \| Article 690341 Frontiers in Physiology \| www.frontiersin.org 10 Pulmonary Artery SMCs Under High-Altitude Siques et al. León-Velarde, F., Villafuerte, F. C., and Richalet, J. P. (2010). Chronic mountain sickness and the heart. Prog. Cardiovasc. Dis. 52, 540–549. doi: 10.1016/j.pcad. 2010.02.012 Penaloza, D., and Arias-Stella, J. (2007). REFERENCES doi: 10.1111/j.1748- 1716.1946.tb00389.x Zhang, N., Dong, M., Luo, Y., Zhao, F., and Li, Y. (2018). Danshensu prevents hypoxic pulmonary hypertension in rats by inhibiting the proliferation of pulmonary artery smooth muscle cells via TGF-β-smad3-associated pathway. Eur. J. Pharmacol. 820, 1–7. doi: 10.1016/j.ejphar.2017.12.010 Wang, G., Qian, G., Zhou, D., and Zhao, J. (2005). JAK-STAT signaling pathway in pulmonary arterial smooth muscle cells is activated by hypoxia. Cell Biol. Int. 29, 598–603. doi: 10.1016/j.cellbi.2005.03.014 Zhang, Q., Li, W., Zhu, Y., Wang, Q., Zhai, C., Shi, W., et al. (2021). Activation of AMPK inhibits Galectin-3-induced pulmonary artery smooth muscle cells proliferation by upregulating hippo signaling eﬀector YAP. Mol. Cell. Biochem. 476, 3037–3049. doi: 10.1007/s11010-021-0 4131-3 Wedgwood, S., Dettman, R. W., and Black, S. M. (2001). ET-1 stimulates pulmonary arterial smooth muscle cell proliferation via induction of reactive oxygen species. Am. J. Physiol. Lung Cell. Mol. Physiol. 281, L1058–L1067. doi: 10.1152/ajplung.2001.281.5.L1058 Zhou, Y., Yan, H., Guo, M., Zhu, J., Xiao, Q., and Zhang, L. (2013). Reactive oxygen species in vascular formation and development. Oxid. Med. Cell. Longev. 2013:374963. doi: 10.1155/2013/374963 West, J. B. (2012). High-altitude medicine. Am. J. Respir. Crit. Care Med. 186, 1229–1237. doi: 10.1164/rccm.201207-1323CI Wilson, J. L., Wang, L., Zhang, Z., Hill, N. S., and Polgar, P. (2019). Participation of PLK1 and FOXM1 in the hyperplastic proliferation of pulmonary artery smooth muscle cells in pulmonary arterial hypertension. PLoS One 14:e0221728. doi: 10.1371/journal.pone.0221728 Conﬂict of Interest: The authors declare that the research was conducted in the absence of any commercial or ﬁnancial relationships that could be construed as a potential conﬂict of interest. Wilson, K. S., Buist, H., Suveizdyte, K., Liles, J. T., Budas, G. R., Hughes, C., et al. (2020). Apoptosis signal-regulating kinase 1 inhibition in in vivo and in vitro models of pulmonary hypertension. Pulm. Circ. 10:2045894020922810. doi: 10.1177/2045894020922810 Publisher’s Note: All claims expressed in this article are solely those of the authors and do not necessarily represent those of their aﬃliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher. Wong, C. M., Bansal, G., Pavlickova, L., Marcocci, L., and Suzuki, Y. J. (2013). Reactive oxygen species and antioxidants in pulmonary hypertension. Antioxid. Redox Signal. 18, 1789–1796. doi: 10.1089/ars.2012.4568 Wu, W., Platoshyn, O., Firth, A. L., and Yuan, J. X. (2007). REFERENCES 30, 364–372. doi: 10.1183/09031936.00128706 August 2021 \| Volume 12 \| Article 690341 Frontiers in Physiology \| www.frontiersin.org 11 Pulmonary Artery SMCs Under High-Altitude Siques et al. Son, Y., Kim, S., Chung, H. T., and Pae, H. O. (2013). Reactive oxygen species in the activation of MAP kinases. Methods Enzymol. 528, 27–48. doi: 10.1016/b978-0- 12-405881-1.00002-1 Xing, J., Wang, M., Hong, J., Gao, Y., Liu, Y., Gu, H., et al. (2019). TRPM7 channel inhibition exacerbates pulmonary arterial hypertension through MEK/ERK pathway. Aging 11, 4050–4065. doi: 10.18632/aging.102036 Sturrock, A., Cahill, B., Norman, K., Huecksteadt, T. P., Hill, K., Sanders, K., et al. (2006). Transforming growth factor-beta1 induces Nox4 NAD(P)H oxidase and reactive oxygen species-dependent proliferation in human pulmonary artery smooth muscle cells. Am. J. Physiol. Lung Cell. Mol. Physiol. 290, L661–L673. doi: 10.1152/ajplung.00269.2005 Xu, D., Li, Y., Zhang, B., Wang, Y., Liu, Y., Luo, Y., et al. (2016). Resveratrol alleviate hypoxic pulmonary hypertension via anti-inﬂammation and anti- oxidant pathways in rats. Int. J. Med. Sci. 13, 942–954. doi: 10.7150/ijms. 16810 Xu, X. Q., and Jing, Z. C. (2009). High-altitude pulmonary hypertension. Eur. Respir. Rev. 18, 13–17. doi: 10.1183/09059180.00011104 Sysol, J. R., Chen, J., Singla, S., Zhao, S., Comhair, S., Natarajan, V., et al. (2018). Micro-RNA-1 is decreased by hypoxia and contributes to the development of pulmonary vascular remodeling via regulation of sphingosine kinase 1. Am. J. Physiol. Lung Cell. Mol. Physiol. 314, L461–L472. doi: 10.1152/ajplung.00057. 2017 Yamashita, K., Discher, D. J., Hu, J., Bishopric, N. H., and Webster, K. A. (2001). Molecular regulation of the endothelin-1 gene by hypoxia. Contributions of hypoxia-inducible factor-1, activator protein-1, GATA-2, AND p300/CBP. J. Biol. Chem. 276, 12645–12653. doi: 10.1074/jbc.M011344200 J. Biol. Chem. 276, 12645–12653. doi: 10.1074/jbc.M011344200 Sysol, J. R., Natarajan, V., and Machado, R. F. (2016). PDGF induces SphK1 expression via Egr-1 to promote pulmonary artery smooth muscle cell proliferation. Am. J. Physiol. Cell. Physiol. 310, C983–C992. doi: 10.1152/ajpcell. 00059.2016 Yanagisawa, M., Kurihara, H., Kimura, S., Tomobe, Y., Kobayashi, M., Mitsui, Y., et al. (1988). A novel potent vasoconstrictor peptide produced by vascular endothelial cells. Nature 332, 411–415. doi: 10.1038/332411a0 Yang, L., Liang, H., Meng, X., Shen, L., Guan, Z., Hei, B., et al. (2020). mmu_circ_0000790 is involved in pulmonary vascular remodeling in mice with HPH via microRNA-374c-mediated FOXC1. Mol. Ther. Nucleic Acids 20, 292–307. doi: 10.1016/j.omtn.2019.12.027 Tamura, Y., Phan, C., Tu, L., Le Hiress, M., Thuillet, R., Jutant, E. M., et al. (2018). REFERENCES Ectopic upregulation of membrane-bound IL6R drives vascular remodeling in pulmonary arterial hypertension. J. Clin. Invest. 128, 1956–1970. doi: 10.1172/ jci96462 Yu, X. M., Wang, L., Li, J. F., Liu, J., Li, J., Wang, W., et al. (2013). Wnt5a inhibits hypoxia-induced pulmonary arterial smooth muscle cell proliferation by downregulation of β-catenin. Am. J. Physiol. Lung Cell. Mol. Physiol. 304, L103–L111. doi: 10.1152/ajplung.00070.2012 Tan, X., Feng, L., Huang, X., Yang, Y., Yang, C., and Gao, Y. (2017). Histone deacetylase inhibitors promote eNOS expression in vascular smooth muscle cells and suppress hypoxia-induced cell growth. J. Cell. Mol. Med. 21, 2022– 2035. doi: 10.1111/jcmm.13122 Yu, Y. R., Mao, L., Piantadosi, C. A., and Gunn, M. D. (2013). CCR2 deﬁciency, dysregulation of Notch signaling, and spontaneous pulmonary arterial hypertension. Am. J. Respir. Cell Mol. Biol. 48, 647–654. doi: 10.1165/ rcmb.2012-0182OC Thomas, S. J., Snowden, J. A., Zeidler, M. P., and Danson, S. J. (2015). The role of JAK/STAT signalling in the pathogenesis, prognosis and treatment of solid tumours. Br. J. Cancer 113, 365–371. doi: 10.1038/bjc.2015.233 Yuan, Z., Chen, J., Chen, D., Xu, G., Xia, M., Xu, Y., et al. (2014). Megakaryocytic leukemia 1 (MKL1) regulates hypoxia induced pulmonary hypertension in rats. PLoS One 9:e83895. doi: 10.1371/journal.pone.0083895 Tuder, R. M. (2017). Pulmonary vascular remodeling in pulmonary hypertension. Cell Tissue Res. 367, 643–649. doi: 10.1007/s00441-016-2539-y Udjus, C., Cero, F. T., Halvorsen, B., Behmen, D., Carlson, C. R., Bendiksen, B. A., et al. (2019). Caspase-1 induces smooth muscle cell growth in hypoxia- induced pulmonary hypertension. Am. J. Physiol. Lung Cell. Mol. Physiol. 316, L999–L1012. doi: 10.1152/ajplung.00322.2018 Zhang, B., Shen, M., Xu, M., Liu, L. L., Luo, Y., Xu, D. Q., et al. (2012). Role of macrophage migration inhibitory factor in the proliferation of smooth muscle cell in pulmonary hypertension. Mediators Inﬂamm. 2012:840737. doi: 10.1155/ 2012/840737 Vodenicharov, A. (2012). “Structure and function of smooth muscle with special reference to mast cells,” in Current Basic and Pathological Approaches to the Function of Muscle Cells and Tissues – From Molecules to Humans, ed. H. Sugi (London: InTech Open), 346–362. Zhang, L., Wang, Y., Wu, G., Rao, L., Wei, Y., Yue, H., et al. (2020). Blockade of JAK2 protects mice against hypoxia-induced pulmonary arterial hypertension by repressing pulmonary arterial smooth muscle cell proliferation. Cell Prolif. 53:e12742. doi: 10.1111/cpr.12742 Von Euler, U. S., and Liljestrand, G. (1946). Observations on the pulmonary arterial blood pressure in the cat. Acta Physiol. Scand. 12, 301–320. REFERENCES Hypoxia divergently regulates production of reactive oxygen species in human pulmonary and coronary artery smooth muscle cells. Am. J. Physiol. Lung Cell. Mol. Physiol. 293, L952–L959. doi: 10.1152/ajplung.00203.2007 Copyright © 2021 Siques, Pena, Brito and El Alam. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms. Xia, L. M., Huang, W. J., Wang, B., Liu, M., Zhang, Q., Yan, W., et al. (2009). Transcriptional up-regulation of FoxM1 in response to hypoxia is mediated by HIF-1. J. Cell. Biochem. 106, 247–256. doi: 10.1002/jcb.21996 August 2021 \| Volume 12 \| Article 690341 Frontiers in Physiology \| www.frontiersin.org 12
https://openalex.org/W4310768760	https://zenodo.org/records/7404922/files/Duygusal%20Emek%20Kavramsal%20Bir%20Derleme.pdf	Turkish	null	Duygusal Emek Kavramsal Bir Derleme	Zenodo (CERN European Organization for Nuclear Research)	2,021	cc-by	7,470	ÖRGÜTSEL DAVRANIŞ ÇALIŞMALARI DERGİSİ I ÖRGÜTSEL DAVRANIŞ ÇALIŞMALARI DERGİSİ I Cilt / Volume: 1 Sayı / Issue: 1 Yıl / Year: 2021 Kurucu ve İmtiyaz Sahibi / Founder & Owner Dr. İsmail ALİCİ Editörler / Editors Prof. Dr. Kubilay ÖZYER Dr. İsmail ALİCİ Yabancı Dil Editörü / Foreign Language Editor Öğr. Gör. Nour HUSSEIN Yazışma Adresi / Mail Address Prof. Dr. Kubilay ÖZYER Tokat Gaziosmanpaşa Üniversitesi Taşlıçiftlik Yerleşkesi İktisadi ve İdari Bilimler Fakültesi İşletme Bölümü 60150 TOKAT Tel: +90 356 252 16 16 – 2363 Fax: +90 356 252 16 73 E-Posta / E-Mail: obstudies@yahoo.com I JOURNAL OF ORGANIZATIONAL BEHAVIOR STUDIES JOURNAL OF ORGANIZATIONAL BEHAVIOR STUDIE JOURNAL OF ORGANIZATIONAL BEHAVIOR STUDIES (OBSTUDIES) ÖRGÜTSEL DAVRANIŞ ÇALIŞMALARI DERGİSİ JOURNAL OF ORGANIZATIONAL BEHAVIOR STUDIES (OBSTUDIES) ÖRGÜTSEL DAVRANIŞ ÇALIŞMALARI DERGİSİ ÖRGÜTSEL DAVRANIŞ ÇALIŞMALARI DERGİSİ Örgütsel Davranış Çalışmaları Dergisi (OBSTUDIES) yılda iki kez yayımlanan hakemli, bilimsel ve uluslararası bir dergidir. Örgütsel davranış alanına ilişkin teorik ve görgül makalelere, yer verilen dergimizin temel amacı, bu alanlarda akademik bilginin üretimi ve paylaşımına katkı sağlamaktır. Dergimizde “Türkçe” ve “İngilizce” olmak üzere iki dilde makale yayımlanmaktadır. Dergiye yayımlanmak üzere gönderilen yazılar, belirtilen yazım kurallarına uygun olarak hazırlanmalıdır. Dergiye yayımlanmak üzere gönderilen yazılar, daha önce yayımlanmamış ve yayımlanmak üzere gönderilmemiş olmalıdır. Dergide yayımlanan yazılarda belirtilen görüşler, yazarlara ait olup Örgütsel Davranış Çalışmaları Dergisi’nin görüşlerini yansıtmaz. Örgütsel Davranış Çalışmaları Dergisinde yayımlanmış yazıların tüm yayın hakları saklı olup, dergimizin adı belirtilmeden hiçbir alıntı yapılamaz. Journal of Organizational Behavior Studies (OBSTUDIES) is a peer-reviewed, scientific and international journal published biannually. The main purpose of our journal, which includes theoretical and empirical articles on the field of organizational behavior, is to contribute to the production and sharing of academic knowledge in these fields. Our journal publishes articles in two languages, "Turkish" and "English". Manuscripts sent to the journal for publication should be prepared in accordance with the specified spelling rules. Manuscripts submitted for publication to the journal must not have been previously published or sent for publication. The views expressed in the articles published in the journal belong to the authors and do not reflect the views of the Journal of Organizational Behavior Studies. All publication rights of the articles published in the Journal of Organizational Behavior Studies are reserved, and no citation can be made without specifying the name of our journal. ÖRGÜTSEL DAVRANIŞ ÇALIŞMALARI DERGİSİ ÖRGÜTSEL DAVRANIŞ ÇALIŞMALARI DERGİSİ II II JOURNAL OF ORGANIZATIONAL BEHAVIOR STUDIES ÖRGÜTSEL DAVRANIŞ ÇALIŞMALARI DERGİSİ JOURNAL OF ORGANIZATIONAL BEHAVIOR STUDIES Bilim Kurulu Members of the Science Board Prof. Dr. Usman GHANI IM Sciences, Pakistan Prof. Dr. Hasan GÜL 19 Mayıs Üniversitesi, Türkiye Prof. Dr. Asep HERMAWAN Trisakti Üniversitesi, Endonezya Prof. Dr. Mesut IDRIZ Sharjah Üniversitesi, BAE Prof. Dr. Farzand Ali JAN Brains Institute Peshawar, Pakistan Prof. Dr. Shah JEHAN Iqra National Üniversitesi, Pakistan Prof. Dr. Kubilay ÖZYER Tokat Gaziosmanpaşa Üniversitesi, Türkiye Prof. Dr. Hasan TAĞRAF Sivas Cumhuriyet Üniversitesi, Türkiye Prof. Dr. Sema YİĞİT Ordu Üniversitesi, Türkiye Prof. Dr. Leyla YUSRAN Trisakti Üniversitesi, Endonezya Doç. Dr. Elmira ADİYETOVA Atırau Üniversitesi, Kazakistan Doç. Dr. Müslüme AKYÜZ Malatya Turgut Özal Üniversitesi, Türkiye Doç. Dr. Umut KOÇ Eskişehir Osmangazi Üniversitesi, Türkiye Doç. Dr. Ufuk ORHAN Mersin Üniversitesi, Türkiye Dr. Öğr. Üyesi Mehmet BİÇER Kilis 7 Aralık Üniversitesi, Türkiye Dr. Öğr. Üyesi M. Said DÖVEN Tokat Gaziosmanpaşa Üniversitesi, Türkiye Dr. Öğr. Üyesi Berdybekova Aiman LESBEKOVNA G. Kazakistan Devlet Pedagoji Üniversitesi, Kazakistan Dr. Öğr. Üyesi Nazmiye Ülkü PEKKAN Tarsus Üniversitesi, Türkiye Dr. Öğr. Üyesi Seyil NAJİMUDİNOVA, Manas Üniversitesi, Kırgizistan Dr. Öğr. Üyesi Ayşe Ersoy YILDIRIM Malatya Turgut Özal Üniversitesi, Türkiye Dr. Öğr. Üyesi Arif YILDIZ Adıyaman Üniversitesi, Türkiye Bilim Kurulu Members of the Science Board III ÖRGÜTSEL DAVRANIŞ ÇALIŞMALARI DERGİSİ IV Prof. Dr. Köksal BÜYÜK Anadolu Üniversitesi Doç. Dr. Müslüme AKYÜZ Malatya Turgut Özal Üniversitesi Doç. Dr. Öznur AZİZOĞLU Hacettepe Üniversitesi Doç. Dr. Ayşe Gökçen KAPUSUZ Selçuk Üniversitesi Doç. Dr. Ufuk ORHAN Mersin Üniversitesi Dr. Öğr. Üyesi Selim ÇAKAR Tokat Gaziosmanpaşa Üniversitesi Dr. Öğr. Üyesi Yusuf DEMİR Sivas Cumhuriyet Üniversitesi Dr. Öğr. Üyesi Alptekin DEVELİ Tokat Gaziosmanpaşa Üniversitesi Dr. Öğr. Üyesi Onur DİRLİK Eskişehir Osmangazi Üniversitesi Dr. Öğr. Üyesi Musa Said DÖVEN Tokat Gaziosmanpaşa Üniversitesi Dr. Öğr. Üyesi Ersin IRK Tokat Gaziosmanpaşa Üniversitesi Dr. Öğr. Nazmiye Ülkü PEKKAN Tarsus Üniversitesi Doç. Dr. Müslüme AKYÜZ Malatya Turgut Özal Üniversitesi Prof. Dr. Köksal BÜYÜK Anadolu Üniversitesi Doç. Dr. Öznur AZİZOĞLU Hacettepe Üniversitesi Doç. Dr. Ayşe Gökçen KAPUSUZ Selçuk Üniversitesi Dr. Öğr. Üyesi Selim ÇAKAR Tokat Gaziosmanpaşa Üniversitesi Doç. Dr. Ufuk ORHAN Mersin Üniversitesi Dr. Öğr. Üyesi Alptekin DEVELİ Tokat Gaziosmanpaşa Üniversitesi Dr. Öğr. Üyesi Yusuf DEMİR Sivas Cumhuriyet Üniversitesi Dr. Öğr. Nazmiye Ülkü PEKKAN Tarsus Üniversitesi IV JOURNAL OF ORGANIZATIONAL BEHAVIOR STUDIES İçindekiler Table of Contents Sayfa No. Page Number ARAŞTIRMA MAKALELERİ / RESEARCH ARTICLES How Collaborative Climate and Its Components Affects The Psychological Empowerment? İşbirlikçi İklim ve Bileşenleri Psikolojik Güçlendirmeyi Nasıl Etkiler? Ayşe GÖKÇEN KAPUSUZ, Nazmiye Ülkü PEKKAN, Alptekin DEVELİ, Mustafa Fedai ÇAVUŞ 1-13 Do Multiple Role Playing From Components of Glass Ceiling Influence Job Performance? Cam Tavan Sendromunun Bileşenlerinden Çoklu Rol Üstlenme İş Performansını Etkiler Mi? Müslüme AKYÜZ 14-23 The Role of Emotional Intelligence In Job Performance of Airline Industry Employees of Pakistan Pakistan Havayolu Sektörü Çalışanlarının İş Performansında Duygusal Zekanın Rolü Yasir ALAM, Maria KHALID, Faiza GOHAR 24-36 The Role of Intellectual Capital In Innovation with The Mediation Effect of Organizational Capital: An Empirical Study of Pakistani Banks İnovasyonda Entellektüel Sermayenin Rolü ve Örgütsel Sermayenin Aracılık Rolü: Pakistan Bankalarında Ampirik Bir Çalışma Zohaib ALI, Yasir ALAM 37-50 DERLEME MAKALELER / REVIEW ARTICLES Duygusal Emek: Kavramsal Bir Derleme Emotional Labor: A Conceptual Review Ahmet Serhat ANAÇ, Kubilay ÖZYER 51-64 İçindekiler Table of Contents Sayfa No. Page Number ARAŞTIRMA MAKALELERİ / RESEARCH ARTICLES How Collaborative Climate and Its Components Affects The Psychological Empowerment? İşbirlikçi İklim ve Bileşenleri Psikolojik Güçlendirmeyi Nasıl Etkiler? Ayşe GÖKÇEN KAPUSUZ, Nazmiye Ülkü PEKKAN, Alptekin DEVELİ, Mustafa Fedai ÇAVUŞ 1-13 Do Multiple Role Playing From Components of Glass Ceiling Influence Job Performance? Cam Tavan Sendromunun Bileşenlerinden Çoklu Rol Üstlenme İş Performansını Etkiler Mi? Müslüme AKYÜZ 14-23 The Role of Emotional Intelligence In Job Performance of Airline Industry Employees of Pakistan Pakistan Havayolu Sektörü Çalışanlarının İş Performansında Duygusal Zekanın Rolü Yasir ALAM, Maria KHALID, Faiza GOHAR 24-36 The Role of Intellectual Capital In Innovation with The Mediation Effect of Organizational Capital: An Empirical Study of Pakistani Banks İnovasyonda Entellektüel Sermayenin Rolü ve Örgütsel Sermayenin Aracılık Rolü: Pakistan Bankalarında Ampirik Bir Çalışma Zohaib ALI, Yasir ALAM 37-50 DERLEME MAKALELER / REVIEW ARTICLES Duygusal Emek: Kavramsal Bir Derleme Emotional Labor: A Conceptual Review Ahmet Serhat ANAÇ, Kubilay ÖZYER 51-64 Journal of Organizational Behavior Studies Örgütsel Davranış Çalışmaları Dergisi Accepted / Kabul Tarihi: 29.12.2021 Review Article/ Derleme Makalesi Accepted / Kabul Tarihi: 29.12.2021 Review Article/ Derleme Makalesi Received / Geliş Tarihi: 13.12.2021 Published / Yayın Tarihi: 31.12.2021 1 Bu çalışma, Ahmet Serhat ANAÇ’ın, Prof. Dr. Kubilay ÖZYER danışmanlığında hazırladığı ve Tokat Gaziosmanpaşa Üniversitesi Lisansüstü Eğitim Enstitüsü'nce 2021 yılında kabul edilmiş olan "İşe Bağlılığın İş Performansı ve İşten Ayrılma Niyetine Etkisinde Duygusal Emeğin Aracılık Rolü" başlıklı doktora tezinden türetilmiştir. DUYGUSAL EMEK: KAVRAMSAL BİR DERLEME1 EMOTIONAL LABOR: A CONCEPTUAL REVIEW Öğr. Gör. Dr. Ahmet Serhat ANAÇ Tokat Gaziosmanpaşa Üniversitesi ORCID iD: 0000-0002-7786-3029, e-Posta: serhatanac@hotmail.com Prof. Dr. Kubilay ÖZYER Tokat Gaziosmanpaşa Üniversitesi ORCID iD: 0000-0002-9654-7750, e-Posta: kozyer@yahoo.com ÖRGÜTSEL DAVRANIŞ ÇALIŞMALARI DERGİSİ V Anahtar Kelimeler: Duygusal emek, yüzeysel rol yapma, derinden rol yapma, samimi davranış. 1. Giriş Son yıllarda, hizmet sektöründe faaliyet gösteren işletmelerin sayısı hem ülkemizde hem de dünyada giderek artış göstermektedir. Dolayısıyla istihdam edilen işgücünün önemli bir bölümü bu sektörde yer alan işletmelerde çalışmaktadır. Bu oranın yüksekliği, hizmet sektörünü ülke ekonomileri için daha önemli kılmakta; iyi ve kaliteli hizmet sunma yollarının belirlenebilmesiyle alakalı olarak örgütlerde ve araştırmacılarda merak uyandırabilmektedir. İşletmelerin rekabet üstünlüğü elde etmeleri verdikleri hizmetin kalitesiyle doğru orantılıdır. İnsan kaynağının yoğun olarak kullanıldığı sektörde çalışanların mesleki bilgileri, deneyimleri, tutum ve davranışları verilen hizmetin kalitesi ve müşteri memnuniyeti bakımından oldukça büyük önem arz etmektedir. Kaliteli hizmet vermek, kalifiye çalışanlarla mümkün olmaktadır. Özellikle müşterilerle devamlı yüz yüze çalışmak durumunda olan çalışanlar sürekli olarak duygu durumlarını düzenlemekte, gerçekte hissettikleri duyguların yerine örgütün sergilenmesini istediği duyguları göstermek zorunda kalmaktadırlar. Bu durum da çalışanların yıpranmalarına, bağlılıklarının azalmasına ve işten ayrılma davranışı sergilemelerine sebep olmaktadır. Cilt / Vol.: 1, Sayı / Is.: 1, Yıl / Year: 2021, Sayfa / Pages: 51-64 Cilt / Vol.: 1, Sayı / Is.: 1, Yıl / Year: 2021, Sayfa / Pages: 51-64 2. Kavram ve Tanım İnsanlar sosyal varlıklardır ve yaşamlarının başından sonuna kadar yakın çevreleriyle ve diğer insanlarla etkileşim içerisinde bulunmak zorundadırlar. “İnsan” unsurunun bulunduğu her yerde ise insana özgü bir özellik olan “duygu” olgusu ön plana çıkmaktadır. Duygu, özellikle 19. yüzyılın başından bu yana, üzerinde çalışılan, araştırılan, ve faklı araştırmacılar tarafından üzerinde farklı görüşler sunulan bir kavram olmuştur. 20. yüzyılın sonuna doğru ise duyguların gösterim biçimleri, çalışma yaşamı ve örgütler için önemi ve duyguların insanların davranışlarındaki yeri literatürde sıklıkla tartışılmaya başlanmıştır (Dursun vd., 2014:13). Duygu kavramı araştırmacılar tarafından farklı şekillerde tanımlanmıştır. Duygu, bir durum, bir kişi ya da herhangi bir şeye karşı sergilenen yoğun hislerdir (Akçay ve Çoruk, 2012:81 - 82). Başka bir tanıma göre ise fiziksel ve bilişsel unsurlara sahip olan ve davranışlara etki eden güçlü hisler olarak belirtilmektedir (Can vd., 2015:96). Tablo 1’de farklı araştırmacılar tarafından yapılmış duygusal emek tanımları bulunmaktadır. Tablo 1 Duygusal Emek Tanımları Yazar Tanım Hochschild (1983) İnsanlarda (çalışanlarda) kolaylıkla gözlenebilen duyguların, jest, mimik ve bedensel olarak sergilenmesini yönetmektir. England ve Farkas (1986) Müşterilerle empati kurarak onları anlamaya çalışmak için sarf edilen çabadır. Ashfort ve Humphrey (1993) Örgüt tarafından çalışanlardan sergilemeleri istenen duyguların müşterilere gösterilmesi davranışıdır. Morris ve Feldman (1996) Hizmet faaliyetinin ifası esnasındaki etkileşimde, örgütün çalışanlardan sergilemelerini istedikleri duyguları gösterirken harcadıkları emek, çaba ve bu emeği ve çabayı göstermeden önceki planlama sürecidir. Steinberg ve Figart (1999) Hizmete konu olan etkileşim esnasında çalışanlar tarafından sergilenen ve çoğu zaman gerçekte yaşanmayan duygusal bir tutumdur. Grandey (2000) Örgütün veya bir meslek grubunun çalışanlarının duygusal gösterim kurallarını yerine getirmek ve kendi duygularını, örgütün sergilenmesini istediği duygularla örtüştürmek amacıyla çeşitli yollarla (rol yapma, bastırma, abartma, vb.) duyguları dönüştürmesidir. Tablo 1 Duygusal Emek Tanımları Yazar Tanım Hochschild (1983) İnsanlarda (çalışanlarda) kolaylıkla gözlenebilen duyguların, jest, mimik ve bedensel olarak sergilenmesini yönetmektir. England ve Farkas (1986) Müşterilerle empati kurarak onları anlamaya çalışmak için sarf edilen çabadır. Ashfort ve Humphrey (1993) Örgüt tarafından çalışanlardan sergilemeleri istenen duyguların müşterilere gösterilmesi davranışıdır. Morris ve Feldman (1996) Hizmet faaliyetinin ifası esnasındaki etkileşimde, örgütün çalışanlardan sergilemelerini istedikleri duyguları gösterirken harcadıkları emek, çaba ve bu emeği ve çabayı göstermeden önceki planlama sürecidir. Steinberg ve Figart (1999) Hizmete konu olan etkileşim esnasında çalışanlar tarafından sergilenen ve çoğu zaman gerçekte yaşanmayan duygusal bir tutumdur. Grandey (2000) Örgütün veya bir meslek grubunun çalışanlarının duygusal gösterim kurallarını yerine getirmek ve kendi duygularını, örgütün sergilenmesini istediği duygularla örtüştürmek amacıyla çeşitli yollarla (rol yapma, bastırma, abartma, vb.) duyguları dönüştürmesidir. ABSTRACT Bu çalışmada duygusal emek kavramının tanımı ve alt boyutları, kavramla ilgili yaklaşımlar, diğer değişkenlerle olan ilişkiler yer almaktadır. Duygusal emek kavramının literatüre kazandıran Amerikalı sosyolog Arlie R. Hochschild’in bu kavrama yaklaşımı ile sonrasındaki araştırmacıların yaklaşımları kronolojik olarak sunulmuştur. Daha sonra bu yaklaşımların karşılaştırmaları yapılarak farklı yönleri ortaya konmuştur. Çalışmada ayrıca duygusal emeğin örgütler için önemine, çalışanlara ve örgütlere etkilerine ve sonuçlarına yer verilmiştir. Çalışmanın sonunda ise çalışanlar, örgütler ve araştırmacılar için öneriler sunulmuştur. The study includes the definition of the concept of emotional labor and its sub-dimensions, approaches related to the concept, relations with other variables. The approach of the American sociologist Arlie R. Hochschild, who introduced the concept of emotional labor to the literature, and the approaches of latter researchers are presented chronologically. Then, these approaches are compared and their different aspects are revealed. The study also includes the importance of emotional labor for organizations, its effects and results on employees and organizations. At the end of the study, suggestions are presented for employees, organizations, and researchers. Vol. / Cilt: 1, Is. / Sayı: 1, Yıl / Year: 2021, Pages /Sayfa: 51-64 Keywords: Emotional labor, surface acting, deep acting, expression of naturally felt emotions. Anahtar Kelimeler: Duygusal emek, yüzeysel rol yapma, derinden rol yapma, samimi davranış. 1 Bu çalışma, Ahmet Serhat ANAÇ’ın, Prof. Dr. Kubilay ÖZYER danışmanlığında hazırladığı ve Tokat Gaziosmanpaşa Üniversitesi Lisansüstü Eğitim Enstitüsü'nce 2021 yılında kabul edilmiş olan "İşe Bağlılığın İş Performansı ve İşten Ayrılma Niyetine Etkisinde Duygusal Emeğin Aracılık Rolü" başlıklı doktora tezinden türetilmiştir. Tablo 1 52 Anaç ve Özyer Kruml ve Geddes (2000) Çalışanların gerçekte hissettikleri duygularını bastırıp farklı duygular sergileme ve davranışta bulunma çabalarıdır. Diefendorff ve Gosserand (2003) Örgütlerin amaç ve hedeflerine ulaşabilmeleri için müşterilerin duygularını etkilemeye yönelik sergilenen düzenlenmiş duygusal davranışlardır. Noon ve Blyton (2007) İçinde düzenlenmiş duygu veya duygular barındıran sözlü veya sözsüz araçlarla müşterilerde farklı duyguların uyandırılması için sergilenen davranışlardır. Kaynak: Yazarlar tarafından oluşturulmuştur. Kruml ve Geddes (2000) Çalışanların gerçekte hissettikleri duygularını bastırıp farklı duygular sergileme ve davranışta bulunma çabalarıdır. Diefendorff ve Gosserand (2003) Örgütlerin amaç ve hedeflerine ulaşabilmeleri için müşterilerin duygularını etkilemeye yönelik sergilenen düzenlenmiş duygusal davranışlardır. Noon ve Blyton (2007) İçinde düzenlenmiş duygu veya duygular barındıran sözlü veya sözsüz araçlarla müşterilerde farklı duyguların uyandırılması için sergilenen davranışlardır. Kaynak: Yazarlar tarafından oluşturulmuştur. Küresel dünya, örgütleri sürdürülebilir rekabet ortamına dâhil olmaya, yüksek motivasyonla, yüksek verimlilikle ve yüksek kaliteyle hizmet vermeye zorlamaktadır. Özellikle “an”a odaklanan yeni ekonomik ve yönetimsel koşullar ve sürekliliğin anlık tüketici tatmini üzerinden sağlanabilen yapısı, örgütlerin özellikle hizmet sektöründe çalışanlar üzerinde yeni kontrol ve yönetim biçimlerini ortaya çıkarmaktadır (Kart, 2011:215 çalışanlar üzerinde yeni kontrol ve yönetim biçimlerini ortaya çıkarmaktadır (Kart, 2011:215 - 217). 21. yüzyılda teknolojinin gelişmesi bazı iş ve iş kollarını yok ederken, özellikle hizmet yoğun ekonomilerde daha öncekilerden farklı çalışma alanları ortaya çıkarmıştır. Tüketicilerle çalışanların nihai olarak yüz yüze geldiği hizmet sektöründe duygular, modlar ve bunların gösterimleri oldukça önemli hale gelmiştir (Ashforth ve Humphrey, 1993:88; Grandey, 2000:95; Morris ve Feldman, 1996:986; Scott ve Barnes, 2011:116). Yoğun rekabet koşullarının gerektirdiği hizmet kalitesi, verimlilik ve üretkenlik kriterlerini karşılamak zorunda olan çalışanların, örgütün amaç ve hedefleri ile örtüşen biçimde iş süreçlerine duygularını da katmaları örgütler tarafından daha fazla talep edilen bir durum olarak karşımıza çıkmaktadır (Vural ve Beril, 2003:15 – 18). Dolayısıyla, kaliteli hizmet ve müşteri memnuniyeti çalışanların gayretleriyle ve örgütün kendilerinden gerçekleştirmelerini istedikleri davranışlarla ve bu davranışları ne derecede karşıladıklarıyla ilgili olmaktadır. Öyle ki, hizmet sektöründe çalışanlar ancak duygularını en iyi şekilde sergileyip davranışlarına yansıtarak kaliteli hizmet sunabilmekte ve müşteri memnuniyetini arttırabilmektedirler (Basım ve Beğenirbaş, 2012:77). Duyguların, insanların ve yaşamın önemli bir unsuru olmasının yanında, örgütler ve çalışma hayatı için de oldukça önemli bir girdi olduğunu öngören ilk ve alan yazınındaki öncü çalışma Amerikalı bir sosyolog olan Arlie R. Hochschild tarafından 1983 yılında “The Managed Heart: Commercialization of Human Feeling” adlı kitapta incelenmiştir (Keleş ve Tuna, 2016:379; Sharma ve Sharma, 2014:30; Kesen ve Akyüz, 2016:235). Tablo 1 Hochschild (1983) yazmış olduğu kitapta kolayca gözlenebilen duyguların mimik, jest ve bedensel ifadelerle sergilenmesini yönetmek şeklinde tanımladığı “duygusal emek” kavramını literatüre kazandırmıştır. Hochschild, duygusal emek ile ilgili iki farklı yaklaşıma dikkat çekmektedir. Bunların ilki duyguların, dolayısıyla duygusal emeğin, belirli bir ücret karşılığında satın alınabilecek ve böylelikle değiştirilebilecek bir değer olduğu, diğeri ise çalışanlardan farklı koşullar altında farklı duygu sergilemelerini gerektiren duygu gösterim kuralları olduğu şeklinde ifade edilmektedir (Sreedevi vd., 2014:18). 2.1. Duygu Gösterim Kuralları Çalışanların hizmet sunarlarken gerçekte hangi duygulara sahip oldukları ve/veya hangi duygulara sahip olmadıkları büyük önem arz etmektedir. Bundan daha önemlisinin ise, gözlemlenebilmesi sebebiyle bu duyguların sonucunda ne tür davranışların ortaya çıktığının takip edilerek gereken önlemlerin alınmasıdır (Buckner, 2012:29). Duygu gösterim kuralları, sosyal kurallar ve normlar ile örgütsel ve mesleki kuralların bir fonksiyonu olarak ifade edilmektedir. Sosyal normlar farklı topluluk ve kültürlerde farklı 53 Cilt / Vol.: 1, Sayı / Is.: 1, Yıl / Year: 2021, Sayfa / Pages: 51-64 şekillerde ortaya çıkabilmektedir. Kimi kültürlerde çalışanların hizmet sunumu esnasında yakın ve samimi davranışlar sergilemeleri beklenirken, diğer bir kültürde bu tür davranış biçimleri ciddiyetsizlik ve hatta saygısızlık olarak algılanabilmektedir (Ashfort ve Humpherey, 1993:91 - 92). Bu durum kişilere, mekâna veya zamana göre de farklılık gösterebilmektedir. Örneğin kendi evinde ayağını masanın kenarına vuran bir kimse çektiği ağrı sonucunda lanet okuyabiliyorken, aynı durumu kilisede yaşadığı zaman benzer tepkiyi verememektedir (Woo, 2013:44 - 45). Bu kurallar farklı meslek gruplarına göre de değişiklik gösterebilmektedir. Örneğin müşterilerle sürekli iletişim halinde bulunan bir satış temsilcisinden veya bir hostesten müşterilere karşı güler yüzlü ve anlayışlı olmaları beklenirken, bir tahsil memurundan mesafeli ve olumsuz davranış sergilemesi beklenebilmektedir. Yani duygu gösterim kurallarının net bir biçimde tanımlanması ve standartlaştırılması mümkün olmayabilmektedir (Türkay ve Kayıkçı, 2017:57). Duygusal emeğin öncüllerinden olan duygu gösterim kuralları, örgütün çalışanlardan hangi duyguları sergilemeleri, hangi duyguları sergilememeleri gerektiğini düzenleyen kurallar şeklinde tanımlanmaktadır (Morris ve Feldman, 1996:988). Duygu gösterim kuralları örgütlerin müşterilerinin yüzlerinde görmeyi arzuladıkları ifadelerin yansıması olarak ifade edilebilmektedir. Hatta bu kurallar, örgütler tarafından hazırlanan personel kitapçıklarında veya eğitim araçlarında açıkça belirtilmektedir. Örgütler, yüksek kaliteli hizmet ve müşteri memnuniyeti hedeflerini gerçekleştirmek adına, belirledikleri gösterim kurallarının, iş gereksinimlerinin ve çalışanların duygu gösterimlerinin örtüşmesini istemektedirler (Grandey, 2000:103). Örgüt çerçevesinden bakıldığında, çalışanların sergileyecekleri olumlu gösterim kurallarının müşteri memnuniyetini arttırabileceği, olumsuz gösterim kurallarının ise müşteri memnuniyetsizliği ile sonuçlanabileceği düşünülmektedir (Becker, 2010:18). Duygu gösterim kuralları, olumsuz duyguların bastırılarak, olumlu duyguların ön plana çıkartılmasını öngörmektedir. Gerçekte hissedilmeyen duyguların taklidi biçiminde de nitelendirilen yüzeysel davranışla ya da çalışanlar tarafından duyguların gerçekten gösterilme arzusu veya hissedilme çabası biçiminde ifade edilen derinden davranışla sergilenebilmektedir (Ceylan, 2017:124). Bunların yanında Ashforth ve Humphrey’e (1993) göre çalışanların içinden gelen, kendiliğinden oluşan biçimde sergilediği davranış türü olan samimi davranış da eklenebilmektedir. 3. Duygusal Emek Yaklaşımları Örgütlerin, hizmet kalitesinin ve müşteri memnuniyetinin arttırılması adına çalışanlarının duygularını yönetme istekleri ve bu yöndeki çabaları, duygusal emek kavramını ilgi çekici hale getirmiştir (Köse ve Oral, 2011:165; Basım ve Beğenirbaş, 2012:78). Duygusal emekle ilgili literatür incelendiğinde bu kavramla ilgili farklı yaklaşımların olduğu ve çalışmaların özellikle dört ana yaklaşım etrafında toplandığı görülmektedir:  Hochschild’in (1983) duygusal emek yaklaşımı  Ashforth ve Humphrey’in (1993) duygusal emek yaklaşımı  Morris ve Feldman’ın (1996) duygusal emek yaklaşımı  Grandey’in (2000) duygusal emek yaklaşımı Bu yaklaşımlar aşağıda sırasıyla ayrıntılı olarak ele alınmıştır. 54 54 Anaç ve Özyer 3.1. Hochschild’in Duygusal Emek Yaklaşımı Duygusal emek kavramını literatüre kazandıran Amerikalı sosyolog Hochschild, “The Managed Heart: Commercialization of Human Feeling (Yönetilen Kalp: İnsan Duygularının Ticarileşmesi)” isimli kitabında duygusal emeği “işin gereklerine uyum sağlayabilmek amacıyla açık bir şekilde mimik ve bedensel ifadeler oluşturarak hislerin yönetilmesi süreci” olarak tanımlamaktadır (Fettahlıoğlu vd., 2016:422; Basım vd., 2013:1490). Hochschild (1983), duygusal emeği “ne hissediyorum?” ve “ne hissetmeliyim”in arasında bir yerde konumlandırmaktadır. Burada, çalışanların müşterilerde belirli duygusal durumları yaratmak adına kendi duygularını değiştirmeleri, dizginlemeleri, dolayısıyla yönetmeleri durumuna dikkat çekilmektedir (Hochschild, 1983:4 – 7). Hochschild duygusal emeği, yüzeysel rol yapma ve derinden rol yapma olmak üzere iki alt boyutta ele almıştır. Bu boyutlara duygu düzenleme stratejileri de denilmektedir. Bunlardan ilki olan yüzeysel rol yapma, çalışanların örgütsel duygu gösterim kuralları çerçevesinde, gerçekte hissetmedikleri duygu veya duyguları sergilemeleri, gerçekte hissettikleri duyguları sergilememeleri ya da bu duyguları bastırmaları, sergiledikleri duyguların niceliksel olarak arttırılması ya da azaltılması şeklinde tanımlanabilmektedir (Hochschild, 1983:33; Chau vd., 2009:1151). Kısacası, duyguların düzenlenerek davranışa dökülmesi olarak da ifade edilmektedir (Brotheridge ve Lee, 2003:366). Mesela çalışan, ağlamak istediği anda olumsuz bir duygu sergileyerek üzgün bir dış görünüş yansıtmamak adına bilinçli bir şekilde gülümserse, eğleniyormuş ve mutluymuş gibi davranırsa, bu çalışan yüzeysel rol yapmış sayılmaktadır (Ashforth ve Humphrey, 1993:92 - 93). Hochschild’e (1983) göre duygusal emeğin diğer alt boyutu olan derinden rol yapma ise çalışanın sergilemesi istenen duyguları gerçekte hissetmeye ve yaşamaya çalışması olarak tanımlanmaktadır. Başka bir ifadeyle çalışanın sadece görünüşünü değil, hislerini de değiştirmeye ve deneyimlemeye çalışması olarak ifade edilmektedir (Hideg ve Van Kleef, 2017:2). Hochschild duygusal emeğin, yüzeysel rol yapma ve derinden rol yapma davranışlarıyla sergileneceğinden bahsetmiş ve duygusal emeğin olumsuz etkileri ve sonuçlarına odaklanmıştır. Sonraki yıllarda farklı araştırmacılar duygusal emekle alakalı değişik yaklaşımlar ortaya koyarak kavramı genişletmiş ve geliştirmişlerdir (Duman, 2017:32 – 33). 3.3. Morris ve Feldman Duygusal Emek Yaklaşımı (1996) Morris ve Feldman (1996), duygusal emeği “kişilerarası etkileşim sürecinde örgütün sergilenmesini arzu ettiği duyguları göstermede sarf edilen çaba, planlama ve kontrol” şeklinde tanımlamışlardır. Buna göre çalışanlar, örgütün sergilemesini istediği duyguyu gerçekte hissediyor olsalar dahi bu hislerin hangi davranışlara dönüşüp karşı tarafa yansıtılacağına çalışanların kendi kararları değil, duygu gösterim kuralları belirlemekte ve bu şekilde aslında duygusal emek sarf edilmektedir (Morris ve Feldman, 1996:987; Beğenirbaş ve Meydan, 2012:164 – 165; Polatcı ve Özyer, 2015:133;). Morris ve Feldman, duygusal emek literatürüne farklı yaklaşımlar getirmiş ve önemli katkılarda bulunmuşlardır. Ashforth ve Humphrey’in yaklaşımını büyük oranda benimseyen Morris ve Feldman duygusal emekte çaba kavramına odaklanmışlardır. Ancak Ashforth ve Humphrey’in aksine çalışanların gerçekte hissettiği duygular ile örgütün duygu sergileme kuralları arasında uyuşmazlığın olmadığı zamanlarda dahi, gerçekte hissedilen duygunun örgütün sergilenmesini istediği biçimde ifade edilmesini sağlamak amacıyla belirli miktarda çabaya gereksinim duyulduğunu belirtmişlerdir. Çalışanların, ister yüzeysel rol yaptıkları, ister derinden rol yaptıkları, isterlerse de samimi duygularını sergiledikleri her türlü duygusal emek davranışında çaba göstermeleri gerektiğini ifade etmişlerdir (Güngör, 2009:170). Morris ve Feldman’ın (1996) bu yaklaşımı önceki iki yaklaşımdan farklı olarak, duygusal emeğe etki eden bireysel ve örgütsel faktörler üzerinde yoğunlaşmaktadır. Duygusal emeği yüzeysel rol yapma, derinden rol yapma ve samimi davranış türleri biçiminde incelemek yerine duygusal emeğin sürecine ilişkin boyutları ele almışlardır. Duygusal emek düzeyinin durumsallık gösterebileceğini ve bu durumsal faktörlere göre de farklılık göstereceğini ileri sürmüşlerdir. Bu farklılık yaratan öğeleri de “Duygusal Emek Boyutları” olarak ele almışlardır. Buna göre duygusal emeğin dört boyutu olduğunu ifade etmişlerdir (Morris ve Feldman, 1996:987). Bu boyutlar aşağıdaki gibidir. 3.2. Ashforth ve Humphrey Duygusal Emek Yaklaşımı Ashforth ve Humphrey Hochschild’in yaklaşımına yakın fakat farklı bir yaklaşım ortaya koymuşlardır. Ashforth ve Humphrey duygusal emeği örgüt tarafından gösterilmesi istenen duyguların, çalışanlar tarafından sergilenmesi olarak tanımlamaktadırlar (Ashforth ve Humphrey, 1993:88). Bu yaklaşımın Hochschild’in yaklaşımından farkları bulunmaktadır. Hochschild, davranışın öncülü olan ve davranışın altında yatan duygulara odaklanırken, Ashforth ve Humphrey doğrudan davranışın kendisine odaklanmaktadırlar. Yani davranışı etkileyen duygu yerine, duygunun gösterildiği ve karşı tarafa yansıtıldığı davranış ön planda olmaktadır. Çünkü dışarıdan gözlemlenebilen ve müşteriler için önemli olan kısmı davranışlardır. Buna göre, çalışanların duygu gösterimlerinin ve dolayısıyla davranışlarının hizmet kalitesi üzerinde doğrudan ve güçlü bir etkiye sahip olduğu söylenebilmektedir (Pugh, 2001:1024). Yüzeysel ve derinden rol yapma davranışları, duyguların devamlı olarak yönetilmesini gerektirmektedir. Çalışanların duygularını ve davranış kurallarını sürekli tekrar etmeleri, bu davranışları çalışanların alışkanlık haline getirmelerine sebebiyet 55 Cilt / Vol.: 1, Sayı / Is.: 1, Yıl / Year: 2021, Sayfa / Pages: 51-64 vereceğinden dolayı duygusal emeğin daha az emek gerektiren veya hiç emek gerektirmeyen davranışları da kapsadığı düşünülmektedir (Ashforth ve Humphrey, 1993:94). Bundan dolayı bu davranışlara ek olarak Ashforth ve Humphrey, yüzeysel ve derinden rol yapmanın yanına üçüncü bir duygusal emek alt boyutu olarak “samimi duyguların gösterilmesi”ni eklemişlerdir. Bu yaklaşımdaki yeni alt boyut duygusal emeğin olumsuz etkilerinin yanında olumlu etkilerinin de olabildiğini öne sürmektedir (Saltık ve Asunakutlu, 2016:112). Ashforth ve Humphrey duygusal emek teorisine, Hochschild’den farklı olarak, iki önemli katkıda bulunmuşlardır. Bunlardan ilki, duygusal emeği gözlemlenebilen davranışlara bağlı kalarak açıklamaya çalışmaları, diğeri ise çalışanlardan sergilemeleri istenen samimi davranışların gözlemlenebilmesi için çalışanların örgütün belirlediği duygu gösterim kuralları ile ne derecede uyum gösterdiğinin anlaşılması gerektiğini savunmaları şeklinde ifade edilmektedir (Aslan ve Çaldağ, 2010:10 – 12). 3.3.3. Sergilenmesi Gereken Duyguların Çeşitliliği Gösterilmesi gereken duyguların farklılığı olarak ifade edilmektedir. Gösterilmesi gereken duyguların sayısı ne kadar fazlaysa ihtiyaç duyulan duygusal emek de o kadar fazla olmaktadır. Çeşitliliği artan duygu gösterim kurallarına uyum sağlamak isteyen çalışanlar daha fazla çaba harcamaktadırlar (Wharton ve Erickson, 1993:463 – 464). 3.3.2. Davranış Kurallarına Gösterilen Dikkat Çalışanların işin gerektirdiği gösterim kurallarını sergilerken gösterdiği dikkat seviyesi olarak tanımlanmaktadır. Dikkat gerektiren işler, daha fazla fiziksel ve psikolojik çabaya ihtiyaç duyduğu için sarf edilen duygusal emek de fazla olmaktadır (Sutton ve Rafaeli, 1988:477). Anaç ve Özyer Anaç ve Özyer 3.3.4. Duygusal Uyumsuzluk Çalışanların gerçekte hissettiği duygularla, örgüt tarafından sergilenmesi istenen duyguların birbirinden farklı olması sonucu yaşanılan durum olarak ifade edilmektedir (Middleton, 1989:199). Duygusal uyumsuzluk, Morris ve Feldman (1996) tarafından duygusal emeğin bir boyutu olarak kabul edilirken, bazı araştırmacılar tarafından duygusal emeğin bir sonucu olarak kabul edilmektedir (Aslan ve Çaldağ, 2011:15). 3.3.1. Duygusal Davranış Gösterim Sıklığı Çalışanların, örgütün kendilerinden sergilemelerini istediği duyguları ne sıklıkta sergilemek durumunda oldukları ile ilişkilendirilmiştir. Çalışanlar müşterilerle ne kadar sık etkileşimde bulunurlarsa, sergilemek zorunda oldukları örgütsel gösterim kurallarının sayısı da o kadar artmaktadır. 56 Tablo 2 Karşılaştırmalı Duygusal Emek Yaklaşımları Hochschild (1983) Ashforth ve Humphrey (1993) Morris ve Feldman (1996) Grandey (2000) Model Kişi duygularını yönetir. Model Kişi davranışlarını yönetir. Model Bireysel ve örgütsel faktörler belirleyici Model İlk üç yaklaşımın karması Tanım İnsanlarda (çalışanlarda) kolaylıkla gözlenebilen duyguların, jest, mimik ve bedensel olarak sergilenmesini yönetmektir. Tanım Örgüt tarafından çalışanlardan sergilemeleri istenen duyguların müşterilere gösterilmesi davranışıdır. Tanım Hizmet faaliyetinin ifası esnasındaki etkileşimde, örgütün çalışanlardan sergilemelerini istedikleri duyguları gösterirken harcadıkları emek, çaba ve bu emeği ve çabayı göstermeden önceki planlama sürecidir. Tanım Çalışanların duygularını örgütün sergilenmesini istediği duygularla örtüştürmeleri amacıyla çeşitli yollarla (rol yapma, bastırma, abartma, vb.) duyguları dönüştürmesidir. Alt boyutlar Yüzeysel Rol Yapma Derinden Rol Yapma Alt boyutlar Yüzeysel Rol Yapma Derinden Rol Yapma Samimi Davranış Alt boyutlar Duygu Gösterim Sıklığı Gösterim Kuralları için Harcanan Dikkat Duygusal Çelişki Alt boyutlar Yüzeysel Rol Yapma Derinden Rol Yapma Genel Özellik Çalışanlar için olumsuz Örgütler için olumlu Genel Özellik Çalışanlar davranışlarını yönetirler. Çalışanlar üzerinde olumlu ve olumsuz sonuçları olabilir. Genel Özellik Bireysel ve örgütsel faktörler duygusal emeği etkiler. Her duygu seviyesi için belirli bir çaba gerekir. Genel Özellik Kendisinden önceki yaklaşımların hepsini barındırır. Duygusal emeğin duygu düzenleme süreci olduğunu savunur. Kaynak: Yazarlar tarafından oluşturulmuştur. Karşılaştırmalı Duygusal Emek Yaklaşımları Karşılaştırmalı Duygusal Emek Yaklaşımları Karşılaştırmalı Duygusal Emek Yaklaşımları 3.4. Grandey Duygusal Emek Yaklaşımı (2000) Duygusal emek hakkındaki bir başka önemli yaklaşım da Grandey (2000) tarafından geliştirilmiştir. Grandey, duygusal emek kavramının önceki yaklaşımların herhangi biriyle tam olarak açıklanamayacağını belirtmiş ve duygusal emeğin ancak bu yaklaşımların tümünün senteziyle kavramsallaştırılabileceğini ve anlaşılabileceğini ifade etmiştir (Grandey, 2000:98). Grandey daha önceki araştırmacıların yaklaşımlarında yer alan önemli kavramları kendi modeline ekleyerek yorumlamış ve duygusal emeği çalışanların duygu ve davranışlarını örgütün amaçlarına hizmet edecek biçimde düzenleme sürecinde harcadıkları çaba olarak tanımlamıştır (Saltık ve Asunakutlu, 2016:112). Grandey’e göre duygusal emek bir süreç olarak düşünülmekte ve bu süreçteki tüm adımlar uygun bir şekilde analiz edilirse duygusal emeğin çalışanlar üzerinde yarattığı sonuçlar olumlu hale getirilebilmektedir. Bundan dolayı Grandey, psikoloji yazınında bir süreç olarak belirtilen “duygu düzenlemesi” kavramını kullanarak duygusal emeği “duygu ayarlama süreci” biçiminde ifade etmektedir (Gosserand, 2003:14). Duygu düzenlemesi, bireylerin herhangi bir durum karşısında hangi duyguları, nerede ve nasıl yansıtmaları gerektiğiyle alakalı yaptıkları duygu ayarlamaları olarak tanımlanmaktadır. Çalışanlar, örgütün sergilenmesini istediği gösterim kurallarına uymak için duygularını düzenlemekte ve olaylara karşı optimum tepkiler gösterebilmektedirler (Chau, 2007:28). Duygu düzenlemesi, duygusal emeğin tanımında yer alan duygusal davranış sergileme kurallarına uyulması bağlamında işlevsel bir süreç olarak düşünülmektedir. Duygu düzenleme süreci girdiler ve çıktılar ilişkisinde incelenmektedir. Çalışanlara çevreden gelen uyarıcılar girdiler, bahsi geçen uyarıcılara tepki olarak sergilenen duygu ve davranışlar da çıktılar olarak adlandırılmaktadır. Girdiler duyguların oluşumundan önceki süreci, çıktılar ise duyguların davranışlara dönüşmesinden sonraki süreci kapsamaktadır. Bundan dolayı duygu düzenleme süreci duyguların oluşumundan önce yapılan “öncül odaklı düzenleme” ve duyguları yaşadıktan sonra yapılan “tepki odaklı 57 Cilt / Vol.: 1, Sayı / Is.: 1, Yıl / Year: 2021, Sayfa / Pages: 51-64 düzenleme” şeklinde sınıflandırılmaktadır. Öncül odaklı düzenleme, bireylerin durumlar ve olaylara etkisi ve bunlar için aldıkları önlemlere odaklanırken, tepki odaklı düzenleme önceden önlem alınamayan durum ve olaylar sonucunda sergilenen duyguların etkisine odaklanmaktadır (Gross, 1998:274 – 282). düzenleme” şeklinde sınıflandırılmaktadır. Öncül odaklı düzenleme, bireylerin durumlar ve olaylara etkisi ve bunlar için aldıkları önlemlere odaklanırken, tepki odaklı düzenleme önceden önlem alınamayan durum ve olaylar sonucunda sergilenen duyguların etkisine odaklanmaktadır (Gross, 1998:274 – 282). Duygusal emek yaklaşımlarının (türleri, tanımları, temel boyutları ve genel özellikleri) karşılaştırmaları Tablo 2’de gösterilmiştir. 4. Duygusal Emek Boyutları Duygusal emek, hizmet sektöründe kalitenin arttırılmasını ve kalitenin sonraki süreçlerde düzeyinin belirlenmesini sağlayan en önemli faktörlerdendir. Duygusal emeğin farklı alt boyutları bulunmaktadır. Bunlardan en çok kullanılanları: • Yüzeysel Rol Yapma • Derinden Rol Yapma • Samimi Davranış şeklinde ifade edilebilir (Polatcı ve Özyer, 2015:133 – 134). İlerleyen bölümde bu boyutlar detaylıca açıklanmıştır. 58 58 Anaç ve Özyer 4.1. Yüzeysel Rol Yapma Gerçekte hissedilmeyen duyguların hissediliyormuş gibi gösterilmesi olarak tanımlanmaktadır (Eroğlu, 2014:149). Daha açık bir ifadeyle çalışanların örgütsel duygu gösterim kuralları çerçevesinde, gerçekte hissetmedikleri duygu veya duyguları sergilemeleri, gerçekte hissettikleri duyguları sergilememeleri ya da bu duyguları bastırmaları, sergiledikleri duyguların niceliksel olarak arttırılması ya da azaltılması şeklinde tanımlanabilmektedir (Hochschild, 1983:37; Ashforth ve Humphrey, 1993:92 – 93; Grandey, 2000:97; Hu ve Shi, 2015:1132). Bu davranışı başarmanın yolu ise, jest, mimik, ses tonu gibi sözlü – sözlü olmayan ifadelerin dikkatlice kullanımından geçmektedir. Bu durum, maske takmak olarak da nitelendirilmektedir (Ashforth ve Humphrey, 1993:92). Yani çalışanlar, duyguları gerçekte hissetmemekte, sadece “öyleymiş gibi” davranmaktadırlar (Glomb, 2004:701). Yüzeysel rol yapmada, çalışanlar gerçek duygularını değil, davranışlarını değiştirmektedirler. Çalışanlar bunu, örgütün sergilenmesini arzu ettiği kurallara uyup işlerinde kalıcı olmak adına yapmaktadırlar (Yang ve Chang, 2008:881). Yüzeysel rol yapma, genellikle olumsuz duyguların bastırılıp davranışa dönüştürülmemesi veya olumlu duygular içindeymiş gibi davranılmasıyla ilgili olmaktadır. Çalışanlardan genelde müşterilere karşı olumlu duygular sergilemeleri beklenmektedir. Gerçekte kötü bir ruh hali içerisindeki bir çalışanın müşteriyle etkileşim içerisindeyken gülümsemesi, olumlu duyguları taklit etmesi ve dışarıya olumlu bir izlenim sunarak, gerçekte hissedilen olumsuz duyguları gizlemesi bu tür davranışa örnek olarak verilebilmektedir (Diefendorff vd., 2005:340). 4.3. Samimi Davranış Çalışanların hizmet sunumu esnasında gerçekten hissederek sergiledikleri davranışlar olarak tanımlanan samimi davranış, duygusal uyum olarak da adlandırılmaktadır (Chu ve Murrmann, 2006:1182; Beğenirbaş ve Çalışkan, 2014:111). Hochschild’in (1983) yaklaşımında belirttiği yüzeysel rol yapma ve derinden rol yapmaya ek olarak Ahsforth ve Humphrey (1993), samimi davranış kavramını geliştirmişlerdir. Samimi davranış, yüzeysel rol yapma ve derinden rol yapmadan farklı olarak rol yapmayı gerektirmemekte, bireyin doğal duygularıyla örtüşmektedir (Bağcı, 2014:74). Yüzeysel rol yapma ve derinden rol yapma duyguların açık bir biçimde veya stratejik olarak düzenlenmesini gerektirirken, samimi davranış duyguların bilinçli veya stratejik olarak düzenlenmediği, doğal bir biçimde ve gerçekten hissedilen duygulardan oluşmaktadır (Anderson, 2014:21). Samimi davranışta, çalışanlar örgütün sergilenmesini arzu ettiği davranışlarla gerçekte hissettikleri duygular uyuştuğundan dolayı yüzeysel rol yapma ve derinden rol yapmada olduğu gibi rol yapma ihtiyacı duymamaktadırlar. Böylece daha az duygusal emek harcamakta ve duygusal emeğin negatif sonuçlarından daha az etkilenmektedirler (Dahling ve Perez, 2010:575). Örnek olarak, ciddi bir hastalığa sahip hastası için üzülüp kaygılanan bir doktor veya iyi görünmeyen, kötü bir gün geçirdiğini düşündüğü bir müşterisi için ilgi gösteren bir barmen gerçekte hissettikleri duyguları sergileyebilmektedir (Chu ve Murrmann, 2006:1182). 4.2. Derinden Rol Yapma Derinden rol yapma, bireyin sergilemek zorunda olduğu duyguyu bizzat hissetmeye çalışması olarak tanımlanmaktadır (Ashforth ve Humphrey, 1993:93). Daha farklı bir ifadeyle, gerçekte hissedilen duyguların sergilenmeden önce örgütün arzu ettiği biçimde bilişsel olarak değiştirilmesi şeklinde tanımlanmaktadır. Derinden rol yapmada çalışanlar, örgütün sergilenmesini istediği duyguyla kendi duygularını eşleştirmek amacıyla değiştirmektedirler (Grandey, 2000:100). Derinden rol yapma aynı zamanda duygusal çalışma stratejisi ve derin seviye duygusal düzenleme olarak da adlandırılmaktadır. Derinlemesine davranış sergileyen çalışanlar, müşteriyle etkileşime girmeden önce gerekli duyguları özümsemekte ve tecrübe etmek için çaba sarf etmektedirler (Schmidt ve Diestel, 2014: 1453). Derinden rol yapma ile yüzeysel rol yapma arasında farklar bulunmaktadır. Yüzeysel rol yapmada sadece davranışlar düzenlenirken, derinden rol yapmada yalnızca davranışlar değil duygular da davranış kurallarına uyumlu hale getirilmektedir. Bu durumun empatik duyarlılıkla da ilgisi vurgulanmaktadır (Ashforth ve Humphrey, 1993:88). Çalışanların müşterilere karşı empati kurmaları, onu zorlu ve agresif hale getiren şartları anlamaya çalışarak pozitif duygularla yaklaşmaları derin davranışa örnek olarak gösterilebilmektedir (Aslan ve Çaldağ, 2011:27). Derinden rol yapmada, yüzeysel rol yapmadaki gibi duygularla davranışlar arasında uyumsuzluk bulunmamaktadır. Yüzeysel davranıştan ayrıştığı bir diğer yön ise bilişsel değişimdir. Bundan dolayı da herhangi bir durum karşısındaki stres, kaygı gibi duygusal etkileri azaltmaktadır (Ensey, 2012:11). Pozitif duyguların hissedilmesi veya hissedilmesi 59 Cilt / Vol.: 1, Sayı / Is.: 1, Yıl / Year: 2021, Sayfa / Pages: 51-64 için gayret gösterilmesi, insanın ruh ve beden sağlığı için faydalı olmaktadır (Bhomwick ve Mulla, 2016:48). Scott ve Barnes (2011:118) derin davranışın bir bütün olarak ele alındığında çalışanlar üzerindeki negatif etkileri azalttığını, pozitif etkileri arttırdığını belirtmektedirler. Derinden rol yapmanın olumsuz sonuçları da bulunmaktadır. Derinlemesine davranış sergileyen çalışanın, gerçek duygularıyla bağlantısı kopmaktadır. Çalışanların, örgüt tarafından sergilenmesi istenen duyguları kendi duygularıymış gibi yaşamaları ve bunları içselleştirmeleri, ruh sağlığının olumsuz etkilenmesine ve tükenmişliğe neden olabilmektedir (Fineman, 2008:30). 5. Duygusal Emeğin Öncül ve Ardıllarına Yönelik Literatür İncelemesi Duygusal emek, son yıllarda araştırmacıların ilgisini çeken bir kavram olmaktadır. Duygusal emeğin birçok örgütsel davranış değişkeniyle ilişkisi incelenmiş ve sonuçları örgütler için yol gösterici olmuştur. Bunlar arasında önemli olduğu düşünülerek aşağıdakiler verilmiştir. Örgütün tarafından istenen duyguların müşterilere gösterilmesi davranışı olarak tanımlanan duygusal emek, çalışanlar için yıpratıcı olabilmektedir. Duygusal emek davranışı sürekli duygu düzenlemesi gerektirmektedir. Bunun sonucunda çalışanların motivasyonları, performansları ve verimlilikleri azalabilmekte, işten ayrılma niyetleri artabilmektedir (Beğenirbaş ve Çalışkan, 2014:115). Türkay ve diğerleri (2011) yapmış oldukları çalışmada işe bağlılıkla duygusal emek arasında pozitif yönlü anlamlı bir ilişki tespit etmişlerdir. Gülova ve diğerleri de (2013) 60 Anaç ve Özyer yaptıkları araştırmada işe bağlılık ile duygusal emek alt boyutları olan yüzeysel rol yapma ve derinden rol yapma arasında pozitif yönlü ve anlamlı ilişkiler olduğu sonucuna ulaşmışlardır. yaptıkları araştırmada işe bağlılık ile duygusal emek alt boyutları olan yüzeysel rol yapma ve derinden rol yapma arasında pozitif yönlü ve anlamlı ilişkiler olduğu sonucuna ulaşmışlardır. Beğenirbaş ve Çalışkan (2014) yapmış oldukları çalışmada duygusal emeğin alt boyutlarından yüzeysel rol yapma ile iş performansı arasında negatif yönlü derinden rol yapma ile iş performansı arasında pozitif yönlü ilişki tespit etmişlerdir. Shuck ve diğerleri (2013) 214 çocuk gelişim uzmanı üzerinde yaptıkları araştırmada duygusal emekle iş performansı arasında pozitif yönlü bir ilişki tespit etmişlerdir. Çelik ve Yıldız (2016) Gaziantep’te bulunan kamu hastaneleri ve özel hastanelerde çalışan 3000 örneklemin bulunduğu çalışmalarında duygusal emeğin alt boyutlarından olan yüzeysel rol yapma ile işten ayrılma niyeti arasında pozitif yönlü, derinden rol yapma ile işten ayrılma niyeti arasında ise negatif yönlü bir ilişki belirlemişlerdir. Bartram ve diğerleri (2012) Avustralya’da görev yapan hemşireler üzerine yaptıkları çalışmada duygusal emek ile işten ayrılma niyeti arasında pozitif yönlü bir ilişkinin olduğunu ifade etmişlerdir. Duygusal emek ve tükenmişlik arasındaki ilişkileri inceleyen çalışmalar incelendiğinde genellikle duygusal emek ile tükenmişlik arasında pozitif yönlü bir ilişki bulunmuştur (Köse vd., 2011:181). Kesen ve Akyüz (2016:244) yaptıkları araştırmada duygusal emeğin alt boyutlarından derinden rol yapma ve samimi davranışın prososyal motivasyonu olumlu yönde etkilediği, yüzeysel rol yapma ile prososyal motivasyon arasında anlamlı bir ilişki bulunmadığı sonucu ulaşmışlardır. Tur rehberlerinin duygusal emek sergilemeleri ile iş tatminleri arasındaki ilişkiyi inceleyen çalışmalarında İrigüler ve Güler (2016:121) yüzeysel rol yapma ile iş tatmini arasında negatif yönlü, derinlemesine davranış ile pozitif yönlü bir ilişki tespit etmişlerdir. 6. Sonuç ve Öneriler Örgütler insanlardan, insanlar da duygulardan yoksun düşünülemez. Bundan dolayı örgütleri oluşturan insanların duygu, niyet ve davranışları oldukça büyük önem arz etmektedir. Çalışanların duygu ve davranışlarının örgütler tarafından şekillendirilme ve yönetilme çabaları, bunların olumlu ve olumsuz sonuçları, duygusal emek kavramını örgütsel davranış yazını açısından oldukça önemli hale getirmiştir. Duygusal emeğin özellikle yüz yüze etkileşimin yoğun olduğu hizmet sektöründe daha fazla sergilendiği görülebilmektedir. Hizmet sektörünün her geçen gün büyüdüğü dünyada ve ülkemizde, duygusal emeğe etki eden faktörlerin, duygusal emeğin boyutlarının, sebeplerinin ve sonuçlarının araştırılması örgütlere önemli katkılar sağlayabilecektir. Müşterilerin beklentileri doğrultusunda, çalışanların örgüt tarafından sergilenmesi istenen duyguları yöneterek müşterilere göstermeleri davranışı olarak tanımlanan duygusal emek işletmeler için büyük öneme sahiptir ve örgütler ve çalışanlar açısından olumlu ve olumsuz birtakım sonuçları bulunmaktadır. Örgütler kaliteli hizmetin tanımlamasını yapmakta, çalışanlardan da bu tanımlamaya paralel davranışlar sergilemesini istemektedirler. Örgütlerin çalışanlardan sergilemelerini istedikleri duygularla çalışanların 61 Cilt / Vol.: 1, Sayı / Is.: 1, Yıl / Year: 2021, Sayfa / Pages: 51-64 gerçekte hissettikleri duyguların örtüşmediği durumlarda ise duygusal uyumsuzluk ortaya çıkmakta ve dolayısıyla çalışanlar duygusal emek sergilemektedirler. Çalışanların sürekli olarak gerçekte hissettikleri duyguları değil de örgüt tarafından istenen gösterim kurallarını uygulamaları, kendi düşünce, duygu ve değerleriyle çelişebilmektedir. Sürekli olarak duygu düzenlemesi yapmak zorunda olan çalışanlar işlerini anlamlandıramamakta ve çalışanların işlerine olan bağlılıkları azalabilmektedir. Duygusal emek kavramıyla ilgili şimdiye kadar yapılmış çalışmalar genellikle sağlık, turizm ve eğitim sektörü çalışanlarına yönelik olarak yapılmıştır. Yeni yapılacak çalışmaların farklı sektör çalışanları üzerinde yapılmasının sonuçların genellenebilir olması açısından önemli olacağı düşünülmektedir. Bunların yanında kamu kurumlarında ve/veya gardiyanlık, tahsilat memurluğu gibi genellikle negatif duygu sergilenmesini gerektiren meslek grubu çalışanları üzerinde yapılacak çalışmaların sonuçlarının önceki çalışmaların sonuçlarıyla karşılaştırılmasının örgütler ve sonraki çalışmalar için yol gösterici olacağı öngörülmektedir. Ayrıca duygularını düzenlemek için emek harcayan çalışanları yöneten yöneticilerin duygusal emek düzeylerini ölçen çalışmaların yapılmasının, çalışanlar, yöneticiler ve örgütler için faydalı sonuçlar ortaya koyacağı düşünülmektedir. Duygusal emeği etkileyen faktörler ile duygusal emeğin yarattığı olumlu – olumsuz tüm sonuçlar yakından takip edilmelidir. Duygusal emek sergilenmesini gerektiren meslek ve meslek grupları incelenmeye devam edilmelidir. Kaynakça Akçay, C. ve Çoruk, A., 2012, Çalışma Yaşamında Duygular ve Yönetimi: Kavramsal Bir İnceleme. Eğitimde Politika Analizi Dergisi, 1 (1), 3-25. Altın Gülova, A., Palamutçuoğlu, B. ve Palamutçuoğlu, A., 2013 Duygusal Emek İle İşe Bağlılık Arasındaki İlişkide Amir Desteğinin Rolü: Üniversitede Öğrenci İşleri Personeline Yönelik Bir Çalışma, Dokuz Eylül Üniversitesi İktisadi ve İdari Bilimler Fakültesi Dergisi, 2 (28), 41 – 74 Arıkan Saltık, I. 6. Sonuç ve Öneriler ve Asunakutlu, T., 2016, Duygusal Emek ve Kültür: Konaklama Sektöründe Karşılaştırmalı Bir Araştırma. Mehmet Akif Ersoy Üniversitesi Sosyal Bilimler Enstitüsü Dergisi, 8 (14), 109 – 132. Ashforth, B. E. ve Humphrey, R. H. 1993, Emotional labor in service roles - The influence of identity. Academy of Management Review, 18, 88–115. Aslan, Ş. ve Çaldağ, M. A., (2011). “Duygusal Emek Davranışları, Yönetimde Birey ve Örgüt Odaklı Davranışlar”, kitabının içinde bölüm Editörler: Özcan Yeniçeri ve Yavuz Demirel, Bursa: Ekin Yayınevi. Basım, N. ve Beğenırbaş, M., 2012, Çalışma Yaşamında Duygusal Emek: Bir Ölçek Uyarlama Çalışması. Yönetim ve Ekonomi, 19 (1), 77 – 90. Basım, N., Beğenırbaş, M. ve CAN YALÇIN, R., 2013, Effects of Teacher Personalities on Emotional Exhaustion: Mediating Role of Emotional Labor. Educational Sciences: Theory & Practice, 13 (3), 1488- 1496. Becker, W., 2010, Shared Display Rules and Emotional Labor In Work Teams. US: The University of Arizona. Beğenırbaş, M. ve Çalışkan, A., 2014, Duygusal Emeğin İş Performansı ve İşten Ayrılma Niyetine Etkisinde Kişilerarası Çarpıklığın Aracılık Rolü. Business and Economics Research Journal, 5 (2), 109-127. Beğenirbaş, M. ve Meydan, C. H., 2012, Duygusal Emeğin Örgütsel Vatandaşlik Davranışıyla İlişkisi: Öğretmenler Üzerinde Bir Araştırma. Gazi Üniversitesi İktisadi ve İdari Bilimler Fakültesi Dergisi, 14 (3), 159 – 181. Bhomwıck, S. ve Mulla, Z., 2016, Emotional labour of policing: Does authenticity play a role?. International Journal of Police Science & Management, 18 (1), 47–60. Brotherıdge, M. ve LEE, T., 2003, Development and validation of the Emotional Labour Scale. Journal of Occupational and Organizational Psychology, 76, 365–379. Buckner, J. E., 2012, Emotional Labor And Authentic Leadership, Louisiana Tech University, Ruston 62 Anaç ve Özyer Can, H., Aşan Azızoğlu, Ö. ve Mıskı Aydın, E., 2015, Örgütsel Davranış, Ankara: Siyasal Kitabevi. Çelık, M. ve Yıldız, B., 2016, Duygusal Emek Düzeyinin İşten Ayrılma Niyeti Üzerindeki Etkisinde Stresin Aracılık Rolü, Uluslararası Sosyal Araştırmalar Dergisi, 9 (45), 734 – 745. Ceylan, A.,2 017, Öğretmenlerin Duygusal Emeklerinin Sosyo - Demografik Değişkenler Yönünden Özellikleri: Batman İli Araştırması. Batman Üniversitesi Yaşam Bilimleri Dergisi , 7 (2), 122 – 132. Chau, S., 2007, Examining The Emotional Labor Process: A Moderated Model of Emotional Labor And Its Effects On Job Performance And Turnover. US: University of Akron. Chau, S.L., Dahlıng, J.J., Levy, P.E. ve Dıefendorff, J.M., 2009, A Predictive Study of Emotional Labor and Turnover, Journal of Organizational Behaviour, 30 (8), 1151- 1163. Chu, K. H. L. ve Murrmann, S. K. 6. Sonuç ve Öneriler 2006, Development and Validation Of The Hospitality Emotional Labor Scale, Tourism Management, 27 (6), 1181-1191. Dahlıng, J. ve Perez, L., 2010, Older worker, different actor? Linking age and emotional labor strategies. Personality and Individual Differences , 48, 574–578. Dıefendorf, J. ve Gosserand, R., 2003, Understanding the emotional labor process: a control theory perspective. Journal of Organizational Behavior, 24, 945-959. Dıefendorff, J. M., Gosserand, R. H. ve CROYLE, M. H., 2005, The Dimensionality and Antecedents of Emotional Labor Strategies. Journal of Vocational Behaviour, 66 (2), 339 – 357. Duman, N., 2017, Duygusal Emek: Bir Literatür Değerlendirmesi. International Journal Entrepreneurship and Management Inquiries, 1 (1), 29 – 39. Dursun, S., Bayram, N. ve AYTAÇ, S., 2014, Duygusal Emek Ölçeği Türkçe Formunun Geçerlilik ve Güvenilirlik Çalışması. “İş, Güç” Endüstri İlişkileri ve İnsan Kaynakları Dergisi, 16 (3), 10 – 18. England, P. ve Farkas, G., 1986, Households, Employment, and Gender: A Social, Economic, and Demographic View, New York: Aldine Pub. Co. Ensey, C., 2012, Avoiding The Slater Slide: Examining The Relationship Between Emotional Labor And Counterproductive Work Behaviors. US: Alliant International University. Eroğlu Gün, Ş., 2014, Örgütlerde Duygusal Emek ve Tükenmişlik İlişkisi Üzerine Bir Araştırma, Pamukkale Üniversitesi Sosyal Bilimler Enstitüsü Dergisi, 19, 147 – 160. Fettahlıoğlu, Ö. O., BIYIKBEYİ, T., GÜLER, B., DEMİR, S., 2016, Duygusal Emeğin Üretkenlik Karşıtı Davranışa Etkisi ve Sosyal Bağllılık İlişkisi: Çağrı Merkezinde Bir Uygulama. The Journal of Academic Social Science Studies, 51, 421 – 434. Fıneman, S., 2008, The Emotional Organization: Passion and Power. Uk: Wiley – Blackwell. Glomb, T., Arvey, R. ve STEEL, P., 2002, Office Sneers, Snipes, & Stab Wounds: Antecedents, Consequences, and Implications of Workplace Violence and Aggression. Frontiers of Industrial and Organizational Psychology: Emotions and work, 227 – 259. Grandey, A. A. 2000, Emotional regulation in the workplace: A new way to conceptualize emotional labor. Journal of Occupational Health Psychology, 5, 95–110. Gross, J., 1998, The Emerging Field of Emotion Regulation: An Integrative Review. Review of General Psychology, 2 (3), 271-299. Grosserand, H., 2003, An Examination of Individual and Organizational Factors Related To Emotional Labor. US: Louisiana State University. Güngör, M., 2009, Duygusal Emek Kavramı: Süreci ve Sonuçları. Kamu – İş, 11 (1), 167 – 184. Hıdeg, I. ve Van Kleef, G. A., 2017, When Expressions of Fake Emotions Elicit Negative Reactions: The Role of Observers' Dialectical Thinking, Journal of Organizational Behavior, 38 (8), 1 – 17. Hochschıld, A. 6. Sonuç ve Öneriler R., 1983, The Managed Heart: The Commercialization of Human Feeling, Los Angeles: University of California Press Hu, X. ve Shı, J., 2015, Employees’ surface acting in interactions with leaders and peers. Journal of Organizational Behavior , 36, 1132–1152. İrigüler, F. ve Güler, M. E., 2016, Emotional Labor of Tourist Guides: How Does It Affect Their Job Satisfaction and Burnout Levels?. Journal of Yasar University, 11 (42), 113 – 123. 63 63 Cilt / Vol.: 1, Sayı / Is.: 1, Yıl / Year: 2021, Sayfa / Pages: 51-64 Kart, E., 2011, Bir Duygu Yönetimi Süreci Olarak Duygusal Emeğin Çalışanlar Üzerindeki Etkisi. Çalışma ve Toplum Ekonomi ve Hukuk Dergisi, 3 (30), 215 – 230. Keleş, Y. ve Tuna, M., 2016, Örgütsel Adaletin Duygusal Emek Üzerindeki Etkisi: Antalya’daki Beş Yıldızlı Otel İşletmelerinde Bir Araştırma. İşletme Araştırmaları Dergisi, 8 (2), 376 – 406. Kesen, M. ve Akyüz, B., 2016, Duygusal Emek Ve Prososyal Motivasyonun İşe Gömülmüşlüğe Etkisi: Sağlık Çalışanları Üzerine Bir Uygulama. Ç.Ü. Sosyal Bilimler Enstitüsü Dergisi, 25 (2), 233-250. Köse, S., Oral, L. ve TÜRESIN, H., 2011, Duygusal Emek Davranışlarının İşgörenlerin Tükenmişlik Düzeyleri ile İlişkisi Üzerine Sağlık Sektöründe Bir Araştırma. İşletme Fakültesi Dergisi, 12 (2), 165-185. Köse, S., Oral, L., 2011, Hekimlerin Duygusal Emek Kullanımı İle İş Doyumu Ve Tükenmişlik Düzeyleri Arasındaki İlişkiler Üzerine Bir Araştırma. Süleyman Demirel Üniversitesi İktisadi ve İdari Bilimler Fakültesi Dergisi , 16 (2), 463 - 492. Kruml, S. ve Geddes, D., 2000, Exploring The Dimensions of Emotional Labor: The Heart of Hochschild’s Work. Management Communication Quarterly, 14 (1), 8 – 49. Morrıs, J. A. ve Feldman, D. C. 1996. The dimensions, antecedents, and consequences of emotional labor. Academy of Management Review, 21: 986 - 1010. Noon, M. ve Blyton, P., 2007, The Realities of Work: Experiencing Work and Employment in Contemporary Society, 3rd Edition, New York: Palgrave. Polatcı, S. ve Özyer, K., 2015, Duygusal Emek Stratejilerinin Duygusal Zekânın Tükenmişliğe Etkisindeki Aracılık Rolü. AİBÜ Sosyal Bilimler Enstitüsü Dergisi, 15 (3), 131-156. Pugh, D., 2001, Service With a Smile: Emotional Contagion in the Service Encounter. Academy of Management Journal , 44 (5), 1018 – 1027. Schmıdt, K. H. ve Dıestel, S., 2014, Are Emotional Labour Strategies by Nurses Associated with Psychological Costs? A Cross-Sectional Survey, International Journal of Nursing Studies, 51 (11), 1450-1461. Scott, B. A. ve Barnes, C. M. 2011. A Multilevel Field Investigation of Emotional Labor, Affect, Work Withdrawal, and Gender. 6. Sonuç ve Öneriler Academy of Management Journal, 54(1), 116 - 136. Sharma, S. ve Sharma, S., 2014, Emotional Labour And Counterproductive Workplace Behaviour: Evidence From Banking Sector in India. Journal of Organisation & Human Behaviour, 3 (4), 29 – 38. Shuck, A. L., Shuck, B. ve Reıo JR., T. G., 2013, Emotional Labor and Performance in the Field of Child Life: Initial Model Exploration and Implications for Practice, Children’s Health Care, 42, 168–190. Sreedevı, V., Annamalaı, S. ve And Ramalıngam, S., 2014, Emotional labour models: differing perspectives. International Journal of Enterprise Network Management , 6 (1), 14-29. Steınberg, R. ve Fıgart, D., 1999, Emotional Demands at Work: A Job Content Analysis. The Annals of the American Academy of Political and Social Science, 561, 177 – 191. Sutton, R. ve Rafaelı, A., 1988, Untangling the Relationship between Displayed Emotions and Organizational Sales: The Case of Convenience Stores. The Academy of Management Journal , 31 (3), 461-487. Türkay, O., Ünal, A. ve Taşar, O., 2011, Motivasyonel ve Yapısal Etkenler Altında Duygusal Emeğinn İşe Bağlılığa Etkisi, ZKÜ Sosyal Bilimler Dergisi, 7 (14), 201 – 222. Türkay, O., Yalçın Kayıkçı, M., 2017, İşletmelerde Duygu Gösterim Kurallari ve Örgütsel Vatandaşliğin Duygusal Emek Üzerine Etkileri, Ekonomik ve Sosyal Araştırmalar Dergisi, 13(2), 53 – 73. Vural, Z. ve Berıl, A. 2003. Kurum Kültürü, İstanbul: İletişim Yayınları. Wharton, A. ve Erıckson, R., 1993, Managing Emotions On The Job And At Home: Understanding The Consequences of Multiple Emotional Roles. Academy of Management Review , 18 (3), 457-486. Woo, V., 2013, The Perks and Pitfalls of Emotional Labor: A Conservation of Resources Approach, Us: Hofstra University. Yang, F. H. ve Chang, C. C., 2008, Emotional Labour, Job Satisfaction and Organizational Commitment Amongst Clinical Nurses: A Questionnaire Survey. International Journal of Nursing Studies, 45 (6), 879 – 887. 64 64
https://openalex.org/W2120901362	https://bioone.org/journals/mountain-research-and-development/volume-31/issue-3/MRD-JOURNAL-D-11-00058.1/Mountains-of-Southern-China-as-Plant-Museums-and-Plant-Cradles/10.1659/MRD-JOURNAL-D-11-00058.1.pdf	English	null	Mountains of Southern China as “Plant Museums” and “Plant Cradles”: Evolutionary and Conservation Insights	Mountain research and development	2,011	cc-by	9,184	Downloaded From: https://bioone.org/journals/Mountain-Research-and-Development on 23 Oct 2024 Terms of Use: https://bioone.org/terms-of-use Introduction The occurrence of areas or centers of endemism is commonly attributed to the existence of suitable refugia in which plant lineages survived while others evolved during the late Neogene and Quaternary global cooling. In China, several studies performed since the 1980s have identified the mountains of central and southern China as the main centers of endemism in the country. A recent work studied the patterns of endemism separately for palaeoendemics and neoendemics and found that these tend to be located in different mountain ranges. Whereas, young endemics are preferably located in the mountain ranges of the eastern fringe of the Tibetan Plateau (‘‘plant cradles’’), old endemics tend to occur in the mountains of central, south central, and southeastern China (‘‘plant museums’’), although there are some exceptions. This pattern seems to be related to the different geological history of the mountain ranges. The eastern fringe of the Tibetan Plateau clearly constitutes the ‘‘evolutionary front’’ of China, probably due to the uninterrupted uplift of the plateau from the late Neogene. In contrast, the relative tectonic stability in central and southern China during most of the Tertiary may have maximized the persistence of relict plant lineages. These results have significant implications in the conservation of the endemic flora, which are briefly discussed. Mountains are often linked to high levels of plant diversity (Ko¨ rner and Ohsawa 2005), and this is widely recognized to be a direct consequence of the geological history of mountainous regions (eg orogenic processes, past climatic changes) as well as of their climatic and topographic heterogeneity (Huber et al 2005; Me´dail and Diadema 2009). More contemporary events also explain why mountains usually harbor more plant diversity than lowlands, and these are mainly related to socioeconomic issues: mountains have often forcibly become the ‘‘last refuge’’ for wildlife because of dense human settlements and intensive agricultural practices in the lowlands. However, mountain biodiversity is at present at high risk because of intensification and changes of land use (departing from traditional management), global climate change (shrinkage of glaciers), or the introduction of non- native species (Huber et al 2005; Spehn et al 2010). Thus, mountains should be regarded as prime conservation targets. Introduction Mountains anywhere stand out for their levels of endemism, and this has been mostly attributed to the role played by them as refugia during the Neogene and Quaternary global cooling; their topographical diversity should have provided enough eco- environmental stability, buffering them against extreme climate fluctuations. Interestingly, these features not only allowed plants to persist but also propelled evolution and speciation (Fjeldsa˚ and Lovett 1997; Tzedakis et al 2002; Me´dail and Diadema 2009). Therefore, distributional patterns of endemism are a commonly used tool for inferring the location of glacial refugia (Lo´pez-Pujol et al 2011 and China is not an exception to this close connection between mountains and plant diversity. Its mountainous areas, especially those located in the southern part of the country, harbor most of China’s plants, both in terms of total species richness and endemism. According to Tang et al (2006), about 92% of the country’s plant genera and 90.5% of endemic plant genera are located in just 10 areas, all of them in southern Keywords: China; centers of endemism; conservation; neoendemics; palaeoendemics; species cradles; species museums. Mountains of Southern China as ‘‘Plant Museums’’ and ‘‘Plant Cradles’’: Evolutionary and Conservation Insights Jordi Lo´pez-Pujol 1,2, Fu-Min Zhang 1, Hai-Qin Sun 1, Tsun-Shen Ying 1, and Song Ge 1 * Corresponding authors: jlopez@ibb.csic.es; gesong@ibcas.ac.cn 1 State Key Laboratory of Systematic and Evolutionary Botany, Institute of Botany, Chinese Academy of Sciences, Beijing, China 2 BioC-GReB, Botanic Institute of Barcelona (CSIC-ICUB), Barcelona, Spain Open access article: please credit the authors and the full source. Jordi Lo´pez-Pujol 1,2, Fu-Min Zhang 1, Hai-Qin Sun 1, Tsun-Shen Ying 1, and Song Ge 1 * Corresponding authors: jlopez@ibb.csic.es; gesong@ibcas.ac.cn p g j p g g 1 State Key Laboratory of Systematic and Evolutionary Botany, Institute of Botany, Chinese Academy of Sciences, B 2 BioC-GReB, Botanic Institute of Barcelona (CSIC-ICUB), Barcelona, Spain Open access article: please credit the authors and the full source. mountain regions. China has one of the highest percentages of mountainous lands in the world, with over 50% (Kohler et al 2010). These lands comprise the highest mountain range on Earth (the Himalaya), perhaps the most rugged one (the Hengduan Mountains, with altitudinal gradients of up to 5000 m), and also a wide array of mountainous types that range from cold and dry (eg Kunlun Mountains or the Tian Shan) to warm and humid ranges (eg Nanling or the Taiwanese Central Mountain Range). In addition, Chinese mountains also show a marked heterogeneity with respect to mountain formation: some are very old and highly eroded, and their origin can be traced back to the early Tertiary or before (Hsu¨ 1983; Wang 1985), whereas others were not built until very recent times (eg Hengduan Mountains, probably uplifted during the last 3 million years; An et al 2001). Mountains of Southern China as “Plant Museums” and “Plant Cradles”: Evolutionary and Conservation Insights Authors: López-Pujol, Jordi, Zhang, Fu-Min, Sun, Hai-Qin, Ying, Tsun- Shen, and Ge, Song Source: Mountain Research and Development, 31(3) : 261-269 Published By: International Mountain Society URL: https://doi.org/10.1659/MRD-JOURNAL-D-11-00058.1 Mountains of Southern China as “Plant Museums” and “Plant Cradles”: Evolutionary and Conservation Insights Authors: López-Pujol, Jordi, Zhang, Fu-Min, Sun, Hai-Qin, Ying, Tsun- Shen, and Ge, Song Source: Mountain Research and Development, 31(3) : 261-269 Published By: International Mountain Society URL: https://doi.org/10.1659/MRD-JOURNAL-D-11-00058.1 BioOne Complete (complete.BioOne.org) is a full-text database of 200 subscribed and open-access titles in the biological, ecological, and environmental sciences published by nonprofit societies, associations, museums, institutions, and presses. Your use of this PDF, the BioOne Complete website, and all posted and associated content indicates your acceptance of BioOne’s Terms of Use, available at www.bioone.org/terms-of-use. Usage of BioOne Complete content is strictly limited to personal, educational, and non - commercial use. Commercial inquiries or rights and permissions requests should be directed to the individual publisher as copyright holder. BioOne sees sustainable scholarly publishing as an inherently collaborative enterprise connecting authors, nonprofit publishers, academic institutions, research libraries, and research funders in the common goal of maximizing access to critical research. MountainNotes Mountain Research and Development (MRD) An international, peer-reviewed open access journal published by the International Mountain Society (IMS) www.mrd-journal.org Mountain Research and Development (MRD) An international, peer-reviewed open access journal published by the International Mountain Society (IMS) www.mrd-journal.org Mountain Research and Development (MRD) An international, peer-reviewed open access journal published by the International Mountain Society (IMS) www.mrd-journal.org Mountain Research and Development Vol 31 No 3 Aug 2011: 261–269 http://dx.doi.org/10.1659/MRD-JOURNAL-D-11-00058.1 2011 by the authors 261 Downloaded From: https://bioone.org/journals/Mountain-Research-and-Development on 23 Oct 2024 Terms of Use: https://bioone.org/terms-of-use Accepted: July 2011 011: 261–269 http://dx.doi.org/10.1659/MRD-JOURNAL-D-11-00058.1 2011 by the authors 261 nd-Development on 23 Oct 2024 Mountain Research and Development Vol 31 No 3 Aug 2011: 261–269 http://dx.doi.org/10.1659/MRD-JOURNAL-D-11-00058.1 2011 by the authors 261 Downloaded From: https://bioone.org/journals/Mountain-Research-and-Development on 23 Oct 2024 Terms of Use: https://bioone.org/terms-of-use t Vol 31 No 3 Aug 2011: 261–269 http://dx.doi.org/10.1659/MRD-JOURNAL-D-11-00058.1 2011 by the authors 261 Mountain-Research-and-Development on 23 Oct 2024 MountainNotes argyrophylla, Eucommia ulmoides, Ginkgo biloba, Glyptostrobus pensilis, Metasequoia glyptostroboides (Figure 2), and glyptostroboides (Figure 2), and Taiwania cryptomerioides are just the most conspicuous examples (indeed, they are regarded as ‘‘living fossils’’) of the Tertiary thermophilous flora that once spanned most of the Northern Hemisphere (the ‘‘boreotropical flora’’; Kubitzki and Krutzsch 1996). The work of Lo´pez- Pujol et al (2011) confirms the strong relictual character of the Chinese flora; 39.1% of their endemics database was classified as palaeoendemics and 60.9% as neoendemics; in the Mediterranean Basin, for instance, the percentage of relict endemics is much lower (,23%; Verlaque et al 1997). Other empirical data are also supportive of the antiquity of Chinese flora and show that China has a disproportionately large number of taxa that belong to the oldest plant taxonomic groups, that is, pteridophytes, gymnosperms, magnolids, and ranunculids (Qian and Ricklefs 1999; Qian 2001), that China has a significantly higher number of ancient endemic genera and families than the United States (Qian 2001), and, finally, that most of the endemic genera of spermatophytes of China (,240) are of ancient origin (Ying et al 1993; Wu et al 2007). Chinese mountains: places for both plant persistence and evolution references therein). The strong endemic character of Chinese mountain flora has been stressed since the 1980s, and a big picture of the main areas of endemism (ie the hypothesized refugia) has been drawn in several studies (eg Ying and Zhang 1984; Ying et al 1993; Wang and Zhang 1994; Ying 2001; Tang et al 2006). Recently, Lo´pez-Pujol et al (2011) went one step further and, in addition to delimiting the areas of endemism at species level (and not at genus level as in the above- mentioned works), they divided their endemic species database into relict endemics (palaeoendemics) and young endemics (neoendemics) to examine whether these areas are mainly places of plant survival (‘‘plant museums’’) or areas from where differentiation and speciation occur (‘‘plant cradles’’). references therein). Accepted: July 2011 The strong endemic character of Chinese mountain flora has been stressed since the 1980s, and a big picture of the main areas of endemism (ie the hypothesized refugia) has been drawn in several studies (eg Ying and Zhang 1984; Ying et al 1993; Wang and Zhang 1994; Ying 2001; Tang et al 2006). Recently, Lo´pez-Pujol et al (2011) went one step further and, in addition to delimiting the areas of endemism at species level (and not at genus level as in the above- mentioned works), they divided their endemic species database into relict endemics (palaeoendemics) and young endemics (neoendemics) to examine whether these areas are mainly places of plant survival (‘‘plant museums’’) or areas from where differentiation and speciation occur (‘‘plant cradles’’). Mountain Research and Development 262 Downloaded From: https://bioone.org/journals/Mountain-Research-and-Development on 23 Oct 2024 Terms of Use: https://bioone.org/terms-of-use Chinese mountains: places for both plant persistence and evolution Third, the existence of such southern refugia is consistent with the Quaternary vegetation reconstructions for China, which suggests the occurrence of extensive temperate and subtropical forests in these areas during the glacial periods (eg An et al 1990; Winkler and Wang 1993; Wang and Sun 1994; Harrison et al 2001; Ni et al 2010), quite a contrast to that which occurred in North America but especially in Europe (Adams and Faure 1997; Prentice et al 2000). southeast Yunnan, southwest Guangxi, and southwest Guizhou are the most significant refugia for palaeoendemics in China (Figure 1). Chinese mountains: places for both plant persistence and evolution Centers of endemism as glacial refugia Up to 20 areas of endemism were delimited by Lo´pez-Pujol et al (2011), all located in central and southern China, which coincide with the main mountain ranges of these regions (Figure 1), and broadly consistent with the above-mentioned earlier studies. Several lines of evidence indicate that southern Chinese mountains acted as ‘‘true’’ refugia. Second, almost all the centers of endemism (19 of 20; Figure 1) identified by Lo´pez-Pujol et al (2011) contained both palaeoendemic and neoendemic taxa, as would be expected for glacial refuges; in fact, assemblages of relict and recently speciated taxa have been reported in other mountainous regions around the world, with a widely recognized role as refugia (eg the tropical Andes, the mountains of tropical Africa, South African Cape region, or the Mediterranean mountains; see references within Lo´pez-Pujol et al 2011). However, the 2 classes of endemics are not evenly distributed within each center of endemism in southern China; a clear preponderance of neoendemics exists in the eastern section (ie the First, it should be noted that China harbors more than 30,000 vascular species, much more than those found in Europe (about 11,500) and the United States (,18,000), territories with almost the same land area and a comparable floristic richness and composition up until the Middle Miocene (,14–15 million years ago [Ma]) (Axelrod et al 1996). Many lineages that were widely distributed along the Northern Hemisphere are today endemic to meridional Chinese mountains, thanks to a much lesser extinction rate experienced there. Cathaya This article aims at summarizing the state of the art in research on patterns of plant endemism in Chinese mountains and their evolutionary aspects, that is, the role of mountains as ‘‘museums’’ and/or ‘‘cradles.’’ Such information is of great value for setting up a conservation agenda for Chinese mountains. http://dx.doi.org/10.1659/MRD-JOURNAL-D-11-00058.1 MountainNotes FIGURE 2 Type tree of Metasequoia glyptostroboides at Moudao Town (Hubei Province, Three Gorges Region). (Photo by Qin Leng) FIGURE 2 Type tree of Metasequoia glyptostroboides at Moudao Town (Hubei Province, Three Gorges Region). (Photo by Qin Leng) FIGURE 2 Type tree of Metasequoia glyptostroboides at Moudao Town (Hubei Province, Three Gorges Region). (Photo by Qin Leng) the excellent review by Qiu et al (2011), a general pattern of multiple refugia appears to emerge from the DNA-based data, and, interestingly, shows a broad congruence with the centers of endemism identified by Lo´pez-Pujol et al (2011). Chinese mountains: places for both plant persistence and evolution A similar congruence is also evident in the Mediterranean Basin, where the areas of endemism (Me´dail and Que´zel 1997) clearly match the phylogeographically defined refugia (Me´dail and Diadema 2009). The article by Qiu et al (2011) also depicts a very significant outcome: Chinese plant taxa tend to show a combination of high genetic diversity and strong divergence among populations, fairly compatible with a scenario of numerous and extensive Quaternary refugia in southern China. the excellent review by Qiu et al (2011), a general pattern of multiple refugia appears to emerge from the DNA-based data, and, interestingly, shows a broad congruence with the centers of endemism identified by Lo´pez-Pujol et al (2011). A similar congruence is also evident in the Mediterranean Basin, where the areas of endemism (Me´dail and Que´zel 1997) clearly match the To conclude, it can be stated that the numerous mountain ranges in the southern section of China, coupled with the limited ice coverage during the Quaternary glaciations (large ice caps were only present in the Tibetan Plateau; Shi 2002) and the lack of barriers to southward migration enabled the country to serve as a refuge for many Tertiary plant lineages as well as a region for plant evolution and speciation. Hengduan Mountains), although palaeoendemics are dominant in the areas to the east (Figure 1), which confirms the hypothesis raised by several Chinese botanists (Wu and Wu 1996; Li and Li 1997) and seems to be related to the different geological and tectonic history of the involved areas: the Tibetan Plateau is the most significant ‘‘evolutionary front’’ of China, probably due to its uninterrupted uplift from the late Neogene (Li and Fang 1999; An et al 2001), which enhanced southeast Yunnan, southwest Guangxi, and southwest Guizhou are the most significant refugia for palaeoendemics in China (Figure 1). Third, the existence of such southern refugia is consistent with the Quaternary vegetation reconstructions for China, which suggests the occurrence of extensive temperate and subtropical forests in these areas during the glacial periods (eg An et al 1990; Winkler and Wang 1993; Wang and Sun 1994; Harrison et al 2001; Ni et al 2010), quite a contrast to that which occurred in North America but especially in Europe (Adams and Faure 1997; Prentice et al 2000). southeast Yunnan, southwest Guangxi, and southwest Guizhou are the most significant refugia for palaeoendemics in China (Figure 1). Mountain Research and Development 263 Downloaded From: https://bioone.org/journals/Mountain-Research-and-Development on 23 Oct 2024 Terms of Use: https://bioone.org/terms-of-use The Hengduan Mountains: a paradigmatic example of ‘‘species cradle’’ At the same time, however, these mountains have become subject to increasingly intense human pressures; intensification of agriculture, hydroelectric development, and other industrial activities (eg mining), and the rise of tourism activities have all gone well beyond the limits of sustainable development. consequence of) sustainable development in mountain regions. Local mountain communities have always relied on their natural resources; both their agricultural practices and their overall social organization have generally been respectful of mountain biodiversity and mountain environments (Xu et al 2005; Ramakrishnan 2007). Thus, the design of conservation strategies for mountain biodiversity needs to take into account sustainable consequence of) sustainable development in mountain regions. Local mountain communities have always relied on their natural resources; both their agricultural practices and their overall social organization have generally been respectful of mountain biodiversity and mountain environments (Xu et al 2005; Ramakrishnan 2007). Thus, the design of conservation strategies for mountain biodiversity needs to take into account sustainable development. The mountains of southern China, as discussed in the former paragraphs, support very rich and heterogeneous plant diversity from the point of view of evolution. At the same time, however, these mountains have become subject to increasingly intense human pressures; intensification of agriculture, hydroelectric development, and other industrial activities (eg mining), and the rise of tourism activities have all gone well beyond the limits of sustainable development. development. The mountains of southern China, as discussed in the former paragraphs, support very rich and heterogeneous plant diversity from the point of view of evolution. At the same time, however, these mountains have become subject to increasingly intense human pressures; intensification of agriculture, hydroelectric development, and other industrial activities (eg mining), and the rise of tourism activities have all gone well beyond the limits of sustainable development. 2001), have greatly contributed to the appearance of many new species (mainly by allopatric and ecological speciation). Radiations are common here, even in primitive genera such as Rhododendron (276 of ,1000 of the world’s species are present in the Hengduan Mountains, and 132 are strictly endemic; Zhang et al 2009). However, these mountains also allowed (although in a lower scale; Figure 1) the preservation of relict elements: Taiwania cryptomerioides and Cunninghamia lanceolata, 2 of the oldest lineages of spermatophytes (.100 Ma; LePage 2009; Ran et al 2010), found in the Hengduan Mountains a suitable refugia. Three Gorges Region: a paradigmatic example of ‘‘species museum’’ The Three Gorges Region (which is also known as the central China Mountains, namely Daba and Wu Mountains) is also one of the richest areas in plant diversity in China, with ,6400 plant species (Lo´pez-Pujol and Ren 2009). This region, in contrast to the Hengduan Mountains, contains more relict than recently evolved taxa (Lo´pez-Pujol et al 2011). One of the most interesting areas within this hotspot is the so-called Conservation of Chinese mountainous flora: challenges and recommendations Conservation considerations under an evolutionary framework The need to protect the ecological and evolutionary processes that create and maintain biodiversity (eg preserving those areas characterized by topographic, climatic, and environmental heterogeneity and those providing migratory corridors and landscape connectivity) and not exclusively the current patterns of this biodiversity (eg simply preserving those species included in a Red List) has become a more important consideration when designing conservation strategies (Cowling and Pressey 2001; Mace et al 2003; Davis et al 2008). Because the processes that trigger species diversification and those that promote their maintenance may not always completely coincide, distribution patterns for relict and newly generated taxa can differ to some extent (Fjeldsa˚ 1994; Silva and Bates 2002), as also revealed in China in the study of Lo´pez-Pujol et al (2011). In this context, a debate is The Hengduan Mountains: a paradigmatic example of ‘‘species cradle’’ Despite their inaccessibility, these mountains have been severely logged and overgrazed, and numerous dams have been planned or are already in place (CI 2007; Morell 2008). metasequoia area, a region of about 800 km2 in the juncture of Hubei and Chongqing (ie the Fangdou Mountains), where there are still natural populations of Metasequoia glyptostroboides (Figure 2). In this very small land extension, at least 550 species of vascular plants occur, including many ‘‘living fossils’’ in addition to M. glyptostroboides, for example, Cunninghamia lanceolata, Eucommia ulmoides, Keteleeria davidiana, Pseudolarix amabilis, Taiwania cryptomerioides, Tapiscia sinensis, and Tetracentron sinense (Hu 1980), all of them with fossil records dating back to the early Tertiary or even before (Manchester et al 2009). Not within the metasequoia area but nearby are occurrences of other very ancient plants, such as Cathaya argyrophylla, and some truly wild populations of Ginkgo biloba. Unfortunately, human activities such as hydroelectric development and farming represent growing challenges for the conservation of the Three Gorges Region relict flora (Lo´pez-Pujol and Ren 2009). metasequoia area, a region of about 800 km2 in the juncture of Hubei and Chongqing (ie the Fangdou Mountains), where there are still natural populations of Metasequoia glyptostroboides (Figure 2). In this very small land extension, at least 550 species of vascular plants occur, including many ‘‘living fossils’’ in addition to M. glyptostroboides, for example, Cunninghamia lanceolata, Eucommia ulmoides, Keteleeria davidiana, Pseudolarix amabilis, Taiwania cryptomerioides, Tapiscia sinensis, and Tetracentron sinense (Hu 1980), all of them with fossil records dating back to the early Tertiary or even before (Manchester et al 2009). Not within the metasequoia area but nearby are occurrences of other very ancient plants, such as Cathaya argyrophylla, and some truly wild populations of Ginkgo biloba. Unfortunately, human activities such as hydroelectric development and farming represent growing challenges for the conservation of the Three Gorges Region relict flora (Lo´pez-Pujol and Ren 2009). The Hengduan Mountains: a paradigmatic example of ‘‘species cradle’’ The Hengduan Mountains sensu lato (that is, the Hengduan, Daxue, and Min Mountains) (Figure 3) are undoubtedly the richest area for plant diversity in China but also one of the most significant (if not the most significant) of the North Temperate Zone (it has gained international recognition as a world ‘‘biodiversity hotspot’’). Spanning up to 5 different provinces (Yunnan, Tibet, Sichuan, Qinghai, Gansu), it harbors some 12,000 species, of which ,3500 are endemic (Myers et al 2000). Its varied and highly dissected topography (combining precipitous mountains, river valleys, and deep gorges) formed during the Pliocene and Pleistocene (An et al to be related to the different geological and tectonic history of the involved areas: the Tibetan Plateau is the most significant ‘‘evolutionary front’’ of China, probably due to its uninterrupted uplift from the late Neogene (Li and Fang 1999; An et al 2001), which enhanced differentiation. In contrast, the relative tectonic stability in central and southern China during most of the Tertiary (with a few exceptions, such as Hainan, Taiwan, or western Guangdong; Hsu¨ 1983; Wang 1985; Liang and Li 2005) may have maximized the persistence of relict plant lineages. The Three Gorges Region, the northeastern corner of Guangxi, and the area formed by Fourth (and last), the glacial refugia identified through areas of endemism seem to coincide with those inferred from the still few but increasingly more frequently carried out molecular phylogeographic studies on Chinese plant species. In http://dx.doi.org/10.1659/MRD-JOURNAL-D-11-00058.1 Mountain Research and Development 263 Downloaded From: https://bioone.org/journals/Mountain-Research-and-Development on 23 Oct 2024 Terms of Use: https://bioone.org/terms-of-use MountainNotes FIGURE 3 A view of Hengduan Mountains. The highest peak in the picture is Mount Gongga (7556 masl). Some deep gorges (of about only 1000 masl) can be found just a few km away. (Photo by Jordi Lo´pez-Pujol) consequence of) sustainable development in mountain regions. Local mountain communities have always relied on their natural resources; both their agricultural practices and their overall social organization have generally been respectful of mountain biodiversity and mountain environments (Xu et al 2005; Ramakrishnan 2007). Thus, the design of conservation strategies for mountain biodiversity needs to take into account sustainable development. The mountains of southern China, as discussed in the former paragraphs, support very rich and heterogeneous plant diversity from the point of view of evolution. Mountain Research and Development 264 Downloaded From: https://bioone.org/journals/Mountain-Research-and-Development on 23 Oct 2024 Terms of Use: https://bioone.org/terms-of-use Conservation of Chinese mountainous flora: challenges and recommendations In addition, the Tibetan Plateau is one of the regions in China with less population density and also one of the most inaccessible, which limits the a priori potential impact of human activities. However, the astonishing rise of domestic tourism in China (which has tripled in just 15 years; NBSC 2009) is bringing visitors even to the most remote places, including the Hengduan Mountains; for example, the new airport of Jiuzhaigou-Huanglong, opened in 2003, has contributed to the increase in the number of visitors in Jiuzhaigou Nature Reserve, from only 32,000 in 1984 to more than 2 million at present (Hendrickson 2009), with pervasive effects on the reserve’s biodiversity (eg Zhu et al. 2006; Morell 2008). Also, intensive logging during the second half of the 20th century destroyed large areas of Hengduan’s primary forests (Morell 2008). Other areas rich in young endemics (Guangdong) are among the most populated regions of China, which makes their preservation a very difficult task. The extirpation of relict taxa would imply the loss of unique, irreplaceable evolutionary history (or ‘‘phylogenetic diversity’’; Faith 1992) (Figure 4). For instance, their potential economic uses could be definitively lost because these taxa generally do not have close living relatives. In radiating areas, in contrast, floras tend to be phylogenetically clustered, that is plant taxa usually have many close relatives (Cowling and Pressey 2001). Thus, the extinction of a given taxon will not involve a significant loss of evolutionary history because of the presence of its congeners (Mace et al 2003) (Figure 4). However, evolutionary fronts are regarded as very important for assuring the maximum levels of both present and future biodiversity (Erwin 1991), because biodiversity is rapidly replaced after any extinction event (Crozier 1997). growing about whether priority should be given to areas of active speciation (‘‘species cradles’’) or, on the contrary, to those acting mainly as a repository of ancestral taxa (‘‘species museums’’) (Erwin 1991; Fjeldsa˚ 1994; Crozier 1997; Cowling and Pressey 2001). One may be especially interested in preserving the ‘‘evolutionary fronts’’ or the ‘‘species pumps,’’ that is, those places that contribute most to the growth of the Tree of Life, but one may also focus on areas with assemblages of isolated lineages that carry unique gene combinations and biological characters (Erwin 1991; Mace et al 2003). When focusing on China, we should have in mind 2 important evolutionary considerations. Mountain Research and Development 265 Downloaded From: https://bioone.org/journals/Mountain-Research-and-Development on 23 Oct 2024 Terms of Use: https://bioone.org/terms-of-use Conservation of Chinese mountainous flora: challenges and recommendations Protecting biodiversity and natural ecosystems is a necessary prerequisite to guarantee (and is also a http://dx.doi.org/10.1659/MRD-JOURNAL-D-11-00058.1 Mountain Research and Development 264 Downloaded From: https://bioone.org/journals/Mountain-Research-and-Development on 23 Oct 2024 Terms of Use: https://bioone.org/terms-of-use MountainNotes FIGURE 4 Hypothetical phylogram of 8 plant species (A–H). Species with the gray branches are probably the result of a recent radiation and then share most of their evolutionary history; thus, extinction of one of them would not be very serious. In contrast, species with black branches (G and H) are relict lineages which share very little phylogenetic history. The loss of any of them will be very serious based on criteria of irreplaceability. compared with palaeoendemics: young taxa are mostly concentrated in the eastern fringe of the Tibetan Plateau (most of the 1u latitude 3 1u longitude cells richest in neoendemics are located there; Lo´pez-Pujol et al 2011). These features may confer significant advantages concerning conservation purposes. For example, protecting relatively few areas might ensure the conservation of a considerable portion of the neoendemic taxa. In addition, the Tibetan Plateau is one of the regions in China with less population density and also one of the most inaccessible, which limits the a priori potential impact of human activities. However, the astonishing rise of domestic tourism in China (which has tripled in just 15 years; NBSC 2009) is bringing visitors even to the most remote places, including the Hengduan Mountains; for example, the new airport of Jiuzhaigou-Huanglong, opened in 2003, has contributed to the increase in the number of visitors in Jiuzhaigou Nature Reserve, from only 32,000 in 1984 to more than 2 million at present (Hendrickson 2009), with pervasive effects on the reserve’s biodiversity (eg Zhu et al. 2006; Morell 2008). Also, intensive logging during the second half of the 20th century destroyed large areas of Hengduan’s primary forests (Morell 2008). Other areas rich in young endemics (Guangdong) are among the most populated regions of China, which makes their preservation a very difficult task. compared with palaeoendemics: young taxa are mostly concentrated in the eastern fringe of the Tibetan Plateau (most of the 1u latitude 3 1u longitude cells richest in neoendemics are located there; Lo´pez-Pujol et al 2011). These features may confer significant advantages concerning conservation purposes. For example, protecting relatively few areas might ensure the conservation of a considerable portion of the neoendemic taxa. Conservation of Chinese mountainous flora: challenges and recommendations Anthropogenic habitat destruction caused, for example, the local extinction of the relict Euryodendron excelsum (a species that belongs to a monotypic genus endemic to China) from Guangxi and the shrinkage of the Guangdong populations (Shen et al 2009). in the level of investigation; only the recent work of Huang et al (2011) uses the species level, although the scale of the operational geographic units (provincial) invalidates its use for identifying areas of endemism. However, because botanical exploration has grown significantly in recent years (mainly because of the rise of funding and scientific personnel at Chinese universities and research institutes), and molecular studies are expanding at an even faster pace, it can be anticipated that we will soon be able to obtain a reasonably precise location of glacial refugia and to decipher the evolutionary patterns of the Chinese flora. The localization of refugia would also benefit from obtaining more palaeoecological data, which are at present still scarce especially in latitudes between 25u and 30uN (Ni et al 2010) where most of the hypothetical refugia were located (Figure 1). in western China (Figure 5). Hence, it is mandatory to enlarge the PAs within or around the centers of endemism as well as to ensure their connectivity through biological corridors. It is also desirable that their funding be increased to avoid departures from their intended purpose (Yu 2010). The current ex situ strategies also fall short in their protection of the endemic, in western China (Figure 5). Hence, it is mandatory to enlarge the PAs within or around the centers of endemism as well as to ensure their connectivity through biological corridors. It is also desirable that their funding be increased to avoid departures from their intended purpose (Yu 2010). The current ex situ strategies also fall short in their protection of the endemic, evolutionarily significant flora: very few botanical gardens (BG) in China are located in the southwest (Huang 2010), where some of the major areas of endemism are found (Figure 1), although the network is adequate in southeastern China. The threatened flora of China (to which most of the endemic species belong) is still poorly represented in the BGs (for example, fewer than 40% of the taxa included in the China Species Red List are cultivated; Huang 2010). Mountain Research and Development 266 Downloaded From: https://bioone.org/journals/Mountain-Research-and-Development on 23 Oct 2024 Terms of Use: https://bioone.org/terms-of-use Conservation of Chinese mountainous flora: challenges and recommendations On one hand, and as discussed above, China maintains a very rich sample of relict lineages, generally consisting of unique, very deep evolutionary lineages, in some cases that represent entire families or even higher taxonomic levels (eg Ginkgoales or Trochodendrales). On the other hand, China harbors probably the largest ‘‘evolutionary front’’ of the world’s North Temperate Zone, the Hengduan Mountains, which are still active because the Tibetan Plateau is uplifting (Zhang et al 2000) and which have experienced extensive explosive radiations in the last million years (Liu and Tian 2007). Areas rich in palaeondemics, in contrast, seem to be more dispersed throughout the country according to the distribution maps specific for palaeoendemics and neoendemics provided in the article of Lo´pez- Pujol et al (2011), and this might be because old endemics showed significantly wider distribution ranges (more than double) as reported by the same researchers. Whatever the reason, this feature, which was also reported in South America and sub-Saharan Africa Leaving aside this evolutionary debate (which today still remains as an open question in conservation biology; Mace et al 2003; Isaac et al 2007), the protection of both palaeoendemic and neoendemic plant taxa in China represents a great challenge in terms of practical conservation. Regarding neoendemics, their centers of abundance are much more defined http://dx.doi.org/10.1659/MRD-JOURNAL-D-11-00058.1 Mountain Research and Development 265 Downloaded From: https://bioone.org/journals/Mountain-Research-and-Development on 23 Oct 2024 Terms of Use: https://bioone.org/terms-of-use MountainNotes (Fjeldsa˚ 1994; Fjeldsa˚ and Lovett 1997), confers some ‘‘degree of protection’’ from human perturbations (but, ironically, not from the evolutionary forces; Erwin 1991). Despite the fact that some palaeoendemic centers are located in relatively remote sites (eg Shennongjia), most of the mountains where they can be found are generally close to the densely populated and cultivated areas of southern China. Anthropogenic habitat destruction caused, for example, the local extinction of the relict Euryodendron excelsum (a species that belongs to a monotypic genus endemic to China) from Guangxi and the shrinkage of the Guangdong populations (Shen et al 2009). (Fjeldsa˚ 1994; Fjeldsa˚ and Lovett 1997), confers some ‘‘degree of protection’’ from human perturbations (but, ironically, not from the evolutionary forces; Erwin 1991). Despite the fact that some palaeoendemic centers are located in relatively remote sites (eg Shennongjia), most of the mountains where they can be found are generally close to the densely populated and cultivated areas of southern China. Recommendations for conservation Recommendations for conservation Because the refugia of the past are the areas best suited to being the refugia of the future, and this is especially significant in the context of global climate change (Bhagwat and Willis 2008; Me´dail and Diadema 2009), it is urgent that a conservation agenda is launched, which should ensure: (1) the conservation of the taxonomic entities that inhabit the mountain centers of endemism; (2) the preservation of their landscape features, keeping them as intact as possible to conserve those processes that maintain and generate biodiversity; and (3) development of mountain regions in an ecologically sustainable manner. Improving both in situ and ex situ conservation: Irrespective of the type of endemics, and even though commendable progress has been made in recent decades, the current in situ and ex situ measures implemented by the Chinese governmental agencies in charge of biodiversity conservation are still insufficient. Despite being well above the average for developing countries, the current network of protected areas (PA), which covers 18% of the country’s total land area (MacKinnon and Xie 2008), has 3 major deficiencies: (1) lack of budget; (2) PAs are too small, which is especially true in eastern China (eg the 20% of the total number of PAs account for just 0.13% of the total area, which means an average of less than 4 km2 per PA; MacKinnon and Xie 2008); and (3) plant hotspots coverage is inadequate (Figure 5). Although more than 80% of the network protects mountainous areas (Tang et al 2006) and thus one can be satisfied with this situation, a large part of these protected areas is found in the northern part of the Tibetan Plateau, Involving the local people and raising environmental awareness: For various reasons, in China, there is a strong correlation between mountains and indigenous people (the shaoshu minzu or China’s minority nationalities), especially in the southwest (Xu et al 2005). These cultures have been stewards of the land for centuries or even millennia and are used to managing it in a sustainable way. Indeed, there is a traceable link between biodiversity conservation and ethnic minorities in China. Conservation of Chinese mountainous flora: challenges and recommendations Another weakness is the size of the collections, which usually are too small to be representative of the species’ genetic diversity; for example, the only ex situ collection of Picea neoveitchii (a narrow endemic species of Qinling Mountains) consists of just 2 individuals cultivated in Xi’an Botanical Garden (Zhang 2007). Recommendations for conservation The conservation of Holy Places, Holy Mountains, and Holy Trees by several ethnic minorities is well known (Yang et al 2004; Xu et al 2005): for example, in the Tibetan ‘‘sacred mountains,’’ such the Khawa Karpo in northwest Yunnan, the biodiversity is better conserved than in the Involving the local people and raising environmental awareness: For various reasons, in China, there is a strong correlation between mountains and indigenous people (the shaoshu minzu or China’s minority nationalities), especially in the southwest (Xu et al 2005). These cultures have been stewards of the land for centuries or even millennia and are used to managing it in a sustainable way. Indeed, there is a traceable link between biodiversity conservation and ethnic minorities in China. The conservation of Holy Places, Holy Mountains, and Holy Trees by several ethnic minorities is well known (Yang et al 2004; Xu et al 2005): for example, in the Tibetan ‘‘sacred mountains,’’ such the Khawa Karpo in northwest Yunnan, the biodiversity is better conserved than in the Achieving a deeper scientific knowledge: Despite the recent efforts to study mountain plant diversity in China, there are still significant gaps, especially in the remote western ranges (Li et al 2011); botanical exploration in this region was scarce until the 1980s, and comprehensive inventories had to wait another 2 decades (eg Kelley 2001). Thus, the available studies on endemism or taxonomic richness (eg Ying et al 1993; Tang et al 2006) were performed at genus level to avoid the performed at genus level to avoid the biases caused by regional differences http://dx.doi.org/10.1659/MRD-JOURNAL-D-11-00058.1 MountainNotes FIGURE 5 A map of nature reserves in China at the end of 2008. Elaborated by Lu Zhang, Research Center for Eco-Environmental Sciences, Beijing, as requested by the authors. FIGURE 5 A map of nature reserves in China at the end of 2008. Elaborated by Lu Zhang, Research Center for Eco-Environmental Sciences, Beijing, as requested by the authors. (high-rank cadres, local officials, the public in general) is also crucial, but this should be particularly geared toward the indigenous and local communities, which are in direct contact with the plant diversity. REFERENCES Adams JM, Faure H, editors. 1997. Review and Atlas of Palaeovegetation: Preliminary Land Ecosystem Maps of the World Since the Last Glacial Maximum. Oak Ridge, TN: Oak Ridge National Laboratory. www.esd.ornl.gov/projects/ qen/; accessed on 8 June 2011. An Z, Kutzbach JE, Prell WL, Porter SC. 2001. An Z, Kutzbach JE, Prell WL, Porter SC. 2001. Evolution of Asian monsoons and phased uplift of the Himalaya-Tibetan plateau since Late Miocene times. Nature 411:62–66. Evolution of Asian monsoons and phased uplift of the Himalaya-Tibetan plateau since Late Miocene times. Nature 411:62–66. An Z, Wu X, Lu Y, Zhang D, Sun X, Dong G. 1990. A preliminary study on the paleoenvironment change of China during the last 20,000 years. In: Liu T, editor. Loess, Quaternary Geology and Global Change, Part II [in Chinese]. Beijing, China: Science Press, pp 1–26. An Z, Wu X, Lu Y, Zhang D, Sun X, Dong G. 1990. A preliminary study on the paleoenvironment change of China during the last 20,000 years. In: Recommendations for conservation One example of how to proceed is the case of Acer yangbiense, the rarest maple of the world (with only 4 individuals remaining in the southwest corner of the Hengduan Mountains): the local communities of the mountain where the trees are located were not only promptly informed but also invited to take part in their management (BGCI 2008). nonsacred mountains (Salick et al 2007). The establishment of modern conservation measures (mainly setting PAs) often creates conflict with the local population (eg by denying access to territory and environmental services, or even their resettlement); moreover, corruption among local officials may result in the failure of compensation schemes, whereas new opportunities, such as the benefits from non-timber forests products and cultural (ethnic) tourism, often enrich outsiders (Xu et al 2005; Xu and Melick 2007). Therefore, it not only is essential to engage local people in conservation activities as well as in the planning and management of the PAs but also to grant them an active role in the local economies. The strengthening of environmental awareness at all levels subsidized by a postdoctoral grant (EX2005-0922) from the Spanish Ministry of Education and Science (MEC) for the period 02/2006-01/2008, and by the Chinese Academy of Sciences (KZCX2-YW-414). Mountain Research and Development 267 Downloaded From: https://bioone.org/journals/Mountain-Research-and-Development on 23 Oct 2024 Terms of Use: https://bioone.org/terms-of-use ACKNOWLEDGMENTS g g , y Liu T, editor. Loess, Quaternary Geology and Global Change, Part II [in Chinese]. Beijing, China: Science Press, pp 1–26. The authors thank Qin Leng for providing the picture of Metasequoia glyptostroboides in Figure 2 and Lu Zhang for his help in elaborating Figure 5. The authors also thank Gemma Rawsthorne for revising the English and editor Anne Zimmermann for her useful comments. This study has been partially Axelrod DI, Al-Shehbaz I, Raven PH. 1996. History of the modern flora of China. In: Zhang A, Wu S, editors. Floristic Characteristics and Diversity of East Asian Plants. Beijing, China: China Higher Education Press, pp 43–55. http://dx.doi.org/10.1659/MRD-JOURNAL-D-11-00058.1 Mountain Research and Development 267 Downloaded From: https://bioone.org/journals/Mountain-Research-and-Development on 23 Oct 2024 Terms of Use: https://bioone.org/terms-of-use Mountain Research and Development 267 Downloaded From: https://bioone.org/journals/Mountain-Research-and-Development on 23 Oct 2024 Terms of Use: https://bioone.org/terms-of-use MountainNotes Kohler T, Giger M, Hurni H, Ott C, Wiesmann U, Wymann von Dach S, Maselli D. 2010. Mountains and climate change: A global concern. Mountain Research and Development 30:53–55. global scheme of plant functional types. Palaeogeography, Palaeoclimatology, Palaeoecology 289:44–61. global scheme of plant functional types. Palaeogeography, Palaeoclimatology, Palaeoecology 289:44–61. BGCI [Botanic Gardens Conservation International]. 2008. Workshop on the conservation of Acer yangbiense in Dali. Botanic Gardens Conservation International. www.bgci.org/ china_en/news/0545/; accessed on 7 June 2011. g p yp Palaeogeography, Palaeoclimatology, Palaeoecology 289:44–61. Prentice IC, Jolly D, BIOME 6000 participants. 2000. Mid-Holocene and glacial maximum vegetation geography of the northern continents and Africa. Journal of Biogeography 27:507–519. Qian H. 2001. A comparison of generic endemism of vascular plants between East Asia and North America. International Journal of Plant Sciences 162:191–199. Ko¨rner C, Ohsawa M, editors. 2005. Mountain systems. In: Hassan R, Scholes R, Ash N, editors. Ecosystems and Human Well-being, Current State and Trends, Volume 1. Findings of the Condition and Trends Working Group of the Millennium Ecosystem Assessment. Washington, DC: Island Press, pp 681–716. Bhagwat SA, Willis KJ. 2008. Species persistence in northerly glacial refugia of Europe: A matter of chance or biogeographical traits? Journal of Biogeography 35:464–482. CI [Conservation International]. 2007. Biodiversity Hotspots. Mountains of Southwest China. C ti I t ti l Ecosystem Assessment. Washington, DC: Island Press, pp 681–716. CI [Conservation International]. 2007. Biodiversity Hotspots. Mountains of Southwest China. Conservation International. www. biodiversityhotspots.org/xp/hotspots/china/ Pages/default.aspx; accessed on 3 June 2011. Qian H, Ricklefs RE. 1999. ACKNOWLEDGMENTS A Case Study on Biodiversity Protection and Local Sustainable Development in Jiuzhaigou, China [MSc Dissertation]. Lund, Sweden: Lund University Center for Sustainable Studies, Lund University. www.lumes.lu.se/database/alumni/ 07.09/thesis/Hendrickson_Cary_Thesis_2009. pdf; accessed on 2 June 2011. Hsu¨ J. 1983. Late Cretaceous and Cenozoic vegetation in China, emphasizing their connections with North America. Annals of the Missouri Botanical Garden 70:490–508. Hendrickson CY. 2009. To Develop or Conserve? A Case Study on Biodiversity Protection and Local Sustainable Development in Jiuzhaigou, China [MSc Dissertation]. Lund, Sweden: Lund University Center for Sustainable Studies, Lund University. www.lumes.lu.se/database/alumni/ 07.09/thesis/Hendrickson_Cary_Thesis_2009. pdf; accessed on 2 June 2011. Mace GM, Gittleman JL, Purvis A. 2003. Preserving the Tree of Life. Science 300:1707– 1709. Spehn EM, Rudmann-Maurer K, Ko¨rner C, Maselli D, editors. 2010. Mountain Biodiversity and Global Change. Basel, Switzerland: GMBA-DIVERSITAS. MacKinnon J, Xie Y. 2008. Regional Action Plan for the Protected Areas of East Asia. 2006–2010. Gland, Switzerland: IUCN. Tang Z, Wang Z, Zheng C, Fang J. 2006. Biodiversity in China’s mountains. Frontiers in Ecology and the Environment 4:347–352. Hsu¨ J. 1983. Late Cretaceous and Cenozoic vegetation in China, emphasizing their connections with North America. Annals of the Missouri Botanical Garden 70:490–508. Manchester SR, Chen Z-D, Lu A-M, Uemura K. 2009. Eastern Asian endemic seed plant genera and their paleogeographic history throughout the Northern Hemisphere. Journal of Systematics and Evolution 47:1–42. Tzedakis PC, Lawson IT, Frogley MR, Hewitt GM, Preece RC. 2002. Buffered tree population changes in a Quaternary refugium: Evolutionary implications. Science 297:2044–2047. Hu SY. 1980. The Metasequoia flora and its phytogeographic significance. Journal of Arnold Arboretum 61:41–94. changes in a Quaternary refugium: Evolutionary implications. Science 297:2044–2047. Verlaque R, Me´dail F, Que´zel P, Babinot JF. 1997. Ende´misme ve´ge´tal et pale´oge´ographie dans le bassin Me´diterrane´en [in French with English abstract]. Geobios 21:159–166. Verlaque R, Me´dail F, Que´zel P, Babinot JF. 1997. Ende´misme ve´ge´tal et pale´oge´ographie dans le bassin Me´diterrane´en [in French with English abstract]. Geobios 21:159–166. Wang H, editor. 1985. Atlas of the Palaeogeography of China. Beijing, China: Cartographic Publishing House. W H S Zh Y L 1994 Th bi di i d Huang H. 2010. Ex situ plant conservation: A key role of Chinese botanic gardens in implementing China’s Strategy for Plant Conservation. BGjournal 7:14–19. Me´dail F, Diadema K. 2009. Glacial refugia influence plant diversity patterns in the Mediterranean Basin. Journal of Biogeography 36: 1333–1345. Wang H, editor. 1985. Atlas of the Palaeogeography of China. Beijing, China: Cartographic Publishing House. ACKNOWLEDGMENTS A comparison of the taxonomic richness of vascular plants in China and the United States. The American Naturalist 154: 160–181. Kubitzki K, Krutzsch W. 1996. Origins of East and South East Asian plant diversity. In: Zhang A, Wu S, editors. Floristic Characteristics and Diversity of East Asian Plants. Beijing, China: China Higher Education Press, pp 56–70. Conservation International. www. biodiversityhotspots.org/xp/hotspots/china/ Pages/default.aspx; accessed on 3 June 2011. Cowling RM, Pressey RL. 2001. Rapid plant diversification: Planning for an evolutionary future. Proceedings of the National Academy of Sciences of the United States of America 98:5452–5457 Cowling RM, Pressey RL. 2001. Rapid plant diversification: Planning for an evolutionary future. Proceedings of the National Academy of Sciences of the United States of America 98:5452–5457. Crozier RH. 1997. Preserving the information content of species: Genetic diversity, phylogeny, and conservation worth. Annual Review of Ecology and Systematics 28:243–268. Qiu Y-X, Fu C-X, Comes HP. 2011. Plant molecular phylogeography in China and adjacent regions: Tracing the genetic imprints of Quaternary climate and environmental change in the world’s most diverse temperate flora. Molecular Phylogenetics and Evolution 59:225–244. Qiu Y-X, Fu C-X, Comes HP. 2011. Plant molecular phylogeography in China and adjacent regions: Tracing the genetic imprints of Quaternary climate and environmental change in the world’s most diverse temperate flora. Molecular Phylogenetics and Evolution 59:225–244. Ramakrishnan PS. 2007. Sustainable mountain development: The Himalayan tragedy. Current Science 92:308–316. LePage BA. 2009. Earliest occurrence of Taiwania (Cupressaceae) from the early Cretaceous of Alaska: Evolution, biogeography, and paleoecology. Proceedings of the Academy of Natural Sciences of Philadelphia 158:129–158. Crozier RH. 1997. Preserving the information content of species: Genetic diversity, phylogeny, and conservation worth. Annual Review of Ecology and Systematics 28:243–268. Li J, Fang X. 1999. Uplift of the Tibetan Plateau and environmental changes. Chinese Science Bulletin 44:2117–2124. Ramakrishnan PS. 2007. Sustainable mountain development: The Himalayan tragedy. Current Science 92:308–316. y Davis EB, Koo MS, Conroy C, Patton JL, Moritz C. 2008. The California Hotspots Project: Identifying regions of rapid diversification of mammals. Molecular Ecology 17:120–138. Erwin TL. 1991. An evolutionary basis for conservation strategies. Science 253:750–752. Faith DP. 1992. Conservation evaluation and phylogenetic diversity. Biological Conservation 61: 1–10. Davis EB, Koo MS, Conroy C, Patton JL, Moritz C. 2008. The California Hotspots Project: Identifying regions of rapid diversification of mammals. Molecular Ecology 17:120–138. Li R, Dao Z, Li H. 2011. ACKNOWLEDGMENTS Seed plant species diversity and conservation in the northern Gaoligong Mountains in western Yunnan, China. Mountain Research and Development 31:160–165. Ran J-H, Gao H, Wang X-Q. 2010. Fast evolution of the retroprocessed mitochondrial rps3 gene in Conifer II and further evidence for the phylogeny of gymnosperms. Molecular Phylogenetics and Evolution 54:136–149. g Erwin TL. 1991. An evolutionary basis for conservation strategies. Science 253:750–752. Li X-W, Li J. 1997. The Tanaka-Kaiyong line: An important floristic line for the study of the flora of East Asia. Annals of the Missouri Botanical Garden 84:888–892. Faith DP. 1992. Conservation evaluation and phylogenetic diversity. Biological Conservation 61: 1–10. Salick J, Amend A, Anderson D, Hoffmeister K, Gunn B, Fang Z. 2007. Tibetan sacred sites conserve old growth trees and cover in the eastern Himalayas. Biodiversity and Conservation 16:693– 706. Fjeldsa˚ J. 1994. Geographical patterns for relict and young species of birds in Africa and South America and implications for conservation priorities. Biodiversity and Conservation 3:207–226. Liang X-Q, Li X-H. 2005. Late Permian to Middle Triassic sedimentary records in Shiwandashan Basin: Implication for the Indosinian Yunkai Orogenic Belt, South China. Sedimentary Geology 177:297–320. Shen S-K, Wang Y-H, Wang B-Y, Ma H-Y, Shen G-Z, Han Z-W. 2009. Distribution, stand characteristics and habitat of a critically endangered plant Euryodendron excelsum H. T. Chang (Theaceae): implications for conservation. Plant Species Biology 24:133–138. Fjeldsa˚ J, Lovett JC. 1997. Geographical patterns of old and young species in African forest biota: The significance of specific montane areas as evolutionary centres. Biodiversity and Conservation 6:325–346. Liu J-Q, Tian B. 2007. Origin, evolution, and systematics of Himalaya endemic genera. Newsletter of Himalayan Botany 40:20–27. Shi Y. 2002. Characteristics of late Quaternary monsoonal glaciation on the Tibetan Plateau and in East Asia. Quaternary International 97–98:79–91. Harrison SP, Yu G, Takahara H, Prentice IC. 2001. Lo´pez-Pujol J, Ren M-X. 2009. Biodiversity and the Three Gorges Reservoir: A troubled marriage. Journal of Natural History 43:2765–2786. Harrison SP, Yu G, Takahara H, Prentice IC. 2001. Diversity of temperate plants in east Asia. Nature 413:129–130. Silva JMC, Bates JM. 2002. Biogeographic patterns and conservation in the South American Cerrado: A tropical savanna hotspot. BioScience 52:225–233. Lo´pez-Pujol J, Zhang F-M, Sun H-Q, Ying T-S, Ge S. 2011. Centres of plant endemism in China: Places for survival or for speciation? Journal of Biogeography 38:1267–1280. Hendrickson CY. 2009. To Develop or Conserve? Mountain Research and Development 268 Downloaded From: https://bioone.org/journals/Mountain-Research-and-Development on 23 Oct 2024 Terms of Use: https://bioone.org/terms-of-use Mountain Research and Development 269 Downloaded From: https://bioone.org/journals/Mountain-Research-and-Development on 23 Oct 2024 Terms of Use: https://bioone.org/terms-of-use ACKNOWLEDGMENTS Xu J, Ma ET, Tashi D, Fu Y, Lu Z, Melick D. 2005. Integrating sacred knowledge for conservation: Cultures and landscapes in Southwest China. Ecology and Society 10(2):7. www.ecologyandsociety.org/ vol10/iss2/art7/; accessed on 2 June 2011. Xu J, Melick D. 2007. Towards Community-Driven Conservation in Southwest China: Reconciling State and Local Perceptions. ICRAF Working Paper Number 52. Beijing, China: ICRAF China. www. worldagroforestry.org/downloads/publications/ PDFS/WP15421.PDF; accessed on 3 June 2011. Yang Y, Tian K, Hao J, Pei S, Yang Y. 2004. Biodiversity and biodiversity conservation in Yunnan, China. Biodiversity and Conservation 13: 813–826. Ying T-S. 2001. Species diversity and distribution pattern of seed plants in China [in Chinese with English abstract]. Biodiversity Science 9:393– 398. Yang Y, Tian K, Hao J, Pei S, Yang Y. 2004. Biodiversity and biodiversity conservation in Yunnan, China. Biodiversity and Conservation 13: 813–826. Yunnan, China. Biodiversity and Conservation 13: 813–826. Zhang D, Liu F, Bing J. 2000. Eco-environmental effects of the Qinghai-Tibet Plateau uplift during the Quaternary in China. Environmental Geology 39: 1352–1358. Ying T-S. 2001. Species diversity and distribution pattern of seed plants in China [in Chinese with English abstract]. Biodiversity Science 9:393– 398. Quaternary in China. Environmental Geology 39: 1352–1358. with English abstract]. Biodiversity Science 9:393– 398. Zhang D-C, Zhang Y-H, Boufford DE, Sun H. 2009. Elevational patterns of species richness and endemism for some important taxa in the Hengduan Mountains, southwestern China. Biodiversity and Conservation 18:699–716. Zhang D-C, Zhang Y-H, Boufford DE, Sun H. 2009. Elevational patterns of species richness and endemism for some important taxa in the Hengduan Mountains, southwestern China. Biodiversity and Conservation 18:699–716. Zhu Z, Bao W, Pang X, Yan X, Li Y. 2006. Tourism effect on species composition and diversity of understory plants in Abies fargesii var. faxoniana forest in Jiuzhaigou, Sichuan [in Chinese with English abstract]. Biodiversity Science 14:284– 291. Ying T-S, Zhang Z-S. 1984. Endemism in the flora of China: Studies on the endemic genera [in Chinese with English abstract]. Acta Phytotaxonomica Sinica 22:259–268. Ying T-S, Zhang Y-L, Boufford DE. 1993. The endemic genera of seed plants of China. Beijing, China: Science Press. Yu X. 2010. Biodiversity conservation in China: Barriers and future actions. International Journal of Environmental Studies 67:117–126. Ying T-S, Zhang Z-S. 1984. Endemism in the flora of China: Studies on the endemic genera [in Chinese with English abstract]. Acta Phytotaxonomica Sinica 22:259–268. Ying T-S, Zhang Y-L, Boufford DE. 1993. ACKNOWLEDGMENTS Me´dail F, Que´zel P. 1997. Hot-spots analysis for conservation of plant biodiversity in the Mediterranean Basin. Annals of the Missouri Botanical Garden 84:112–127. Huang J-H, Chen J-H, Ying J-S, Ma K-P. 2011. Features and distribution patterns of Chinese endemic seed plant species. Journal of Systematics and Evolution 49:81–94. Wang H-S, Zhang Y-L. 1994. The bio-diversity and characters of spermatophytic genera endemic to China [in Chinese with English abstract]. Acta Botanica Yunnanica 16:209–220. Morell V. 2008. Letting 1000 forests bloom. Science 320:1442–1443. Huber UM, Bugmann HKM, Reasoner MA. 2005. Global Change and Mountain Regions. An Overview of Current Knowledge. Dordrecht, The Netherlands: Springer. Myers N, Mittermeier RA, Mittermeier CG, Da Wang P, Sun X. 1994. Last glacial maximum in China: Comparison between land and sea. Catena 23:341–353. Myers N, Mittermeier RA, Mittermeier CG, Da Fonseca GAB, Kent J. 2000. Biodiversity hotspots for conservation priorities. Nature 403:853–858. Isaac NJB, Turvey ST, Collen B, Waterman C, Baillie JEM. 2007. Mammals on the EDGE: Conservation priorities based on threat and phylogeny. Plos ONE 2(3):e296. doi:10.1371/ journal.pone.0000296. NBSC [National Bureau of Statistics of China]. 2009 Chi S i i l Y b k 2009 B iji NBSC [National Bureau of Statistics of China]. 2009. China Statistical Yearbook 2009. Beijing, Winkler MG, Wang PK. 1993. The late-Quaternary vegetation and climate of China. In: Wright HE Jr, Kutzbach JE, Webb T III, Ruddiman WF, Street- Perrot FA, Bartlein PJ, editors. Global Climates Since the Last Glacial Maximum. Minneapolis, MN: University of Minnesota Press, pp 221–261. NBSC [National Bureau of Statistics of China]. 2009. China Statistical Yearbook 2009. Beijing, China: China Statistics Press. China: China Statistics Press. Ni J, Yu G, Harrison SP, Prentice IC. 2010. Palaeovegetation in China during the late Quaternary: Biome reconstructions based on a Kelley S. 2001. Plant hunting on the rooftop of the world. Arnoldia 61:2–13. http://dx.doi.org/10.1659/MRD-JOURNAL-D-11-00058.1 Mountain Research and Development 268 Downloaded From: https://bioone.org/journals/Mountain-Research-and-Development on 23 Oct 2024 Terms of Use: https://bioone.org/terms-of-use MountainNotes Zhang D. 2007. Plant Germplasm Resources and Conservation. Shanghai, China: Shanghai Scientific and Technical Publishers. Wu Z, Sun H, Zhou Z, Peng H, Li D. 2007. Origin and differentiation of endemism in the flora of China. Frontiers of Biology in China 2:125–143. Wu Z, Wu S. 1996. A proposal for a new floristic kingdom (realm). In: Zhang A, Wu S, editors. Floristic Characteristics and Diversity of East Asian Plants. Beijing, China: China Higher Education Press, pp 3–42. http://dx.doi.org/10.1659/MRD-JOURNAL-D-11-00058.1 ACKNOWLEDGMENTS The Ying T-S, Zhang Y-L, Boufford DE. 1993. The endemic genera of seed plants of China. Beijing, China: Science Press. Zhu Z, Bao W, Pang X, Yan X, Li Y. 2006. Tourism effect on species composition and diversity of understory plants in Abies fargesii var. faxoniana forest in Jiuzhaigou, Sichuan [in Chinese with English abstract]. Biodiversity Science 14:284– 291. Yu X. 2010. Biodiversity conservation in China: Yu X. 2010. Biodiversity conservation in China: Barriers and future actions. International Journal of Environmental Studies 67:117–126. Barriers and future actions. International Journal of Environmental Studies 67:117–126. http://dx.doi.org/10.1659/MRD-JOURNAL-D-11-00058.1
https://openalex.org/W1599763040	https://research.rug.nl/files/42647936/art_3A10.1007_2Fs11136_015_1018_3.pdf	English	null	Do individuals with and without depression value depression differently? And if so, why?	Quality of life research	2,015	cc-by	9,253	University of Groningen University of Groningen University of Groningen Do individuals with and without depression value depression differently? Papageorgiou, Katerina; Vermeulen, Karin M.; Schroevers, Maya J.; Stiggelbout, Anne M.; Buskens, Erik; Krabbe, Paul F. M.; van den Heuvel, Edwin; Ranchor, Adelita V. Published in: Quality of Life Research DOI: 10.1007/s11136-015-1018-3 IMPORTANT NOTE: You are advised to consult the publisher's version (publisher's PDF) if you wish to cite from it. Please check the document version below. Document Version Publisher's PDF, also known as Version of record Document Version Publisher's PDF, also known as Version of record Publication date: 2015 Link to publication in University of Groningen/UMCG research database Citation for published version (APA): Papageorgiou, K., Vermeulen, K. M., Schroevers, M. J., Stiggelbout, A. M., Buskens, E., Krabbe, P. F. M., van den Heuvel, E., & Ranchor, A. V. (2015). Do individuals with and without depression value depression differently? And if so, why? Quality of Life Research, 24(11), 2565-2575. https://doi.org/10.1007/s11136- 015-1018-3 Copyright Other than for strictly personal use, it is not permitted to download or to forward/distribute the text or part of it without the consent of the author(s) and/or copyright holder(s), unless the work is under an open content license (like Creative Commons). Copyright strictly personal use, it is not permitted to download or to forward/distribute the text or part of it without the consent of th or copyright holder(s), unless the work is under an open content license (like Creative Commons). The publication may also be distributed here under the terms of Article 25fa of the Dutch Copyright Act, indicated by the “Taverne” license. More information can be found on the University of Groningen website: https://www.rug.nl/library/open-access/self-archiving-pure/taverne- amendment. Take-down policy If you believe that this document breaches copyright please contact us providing details, and we will remove access to the work immediately and investigate your claim. Take-down policy If you believe that this document breaches copyright please contact us providing details, and we will remove access to the work immediately and investigate your claim. Take-down policy If you believe that this document breaches copyright please contact us providing details, and we will remove access to and investigate your claim. Downloaded from the University of Groningen/UMCG research database (Pure): http://www.rug.nl/research/portal. For technical reasons the number of authors shown on this cover page is limited to 10 maximum. Downloaded from the University of Groningen/UMCG research database (Pure): http://www.rug.nl/research/portal. For technical reasons the number of authors shown on this cover page is limited to 10 maximum. Download date: 24-10-2024 Qual Life Res (2015) 24:2565–2575 DOI 10.1007/s11136-015-1018-3 Abstract Purpose Health state valuations, used to evaluate the effectiveness of healthcare interventions, can be obtained either by the patients or by the general population. The general population seems to value somatic conditions more negatively than patients, but little is known about valua- tions of psychological conditions. This study examined whether individuals with and without depression differ in their valuations of depression and whether perceptions re- garding depression (empathy, perceived susceptibility, stigma, illness perceptions) and individual characteristics (mastery, self-compassion, dysfunctional attitudes) bias valuations of either individuals with or without depression. Methods In an online study, a general population sample used a time-trade-off task to value 30 vignettes describing depression states (four per participant) and completed questionnaires on perceptions regarding depression and individual characteristics. Participants were assigned to depression groups (with or without depression), based on the PHQ-9. A generalized linear mixed model was used to Results The sample (N = 1268) was representative of the Dutch population on age, gender, education and residence. We found that for mild depression states, individuals with depression (N = 200) valued depression more negatively than individuals without depression (N = 1068) (p = .007). Variables related to perceptions of depression and individual characteristics were not found to affect valuations of either individuals with or individuals without depression. Conclusion Since the general population values depres- sion less negatively, using their perspective might result in less effectiveness for interventions for mild depression. Perceptions of depression or to individual characteristics did not seem to differentially affect valuations made by either individuals with or without depression. Keywords Depression Discrepancies Valuations Health-related quality of life Time-trade-off & Adelita V. Ranchor a.v.ranchor@umcg.nl Katerina Papageorgiou a.papageorgiou@umcg.nl 1 Health Psychology Section, University Medical Center Groningen, University of Groningen, Groningen, The Netherlands 2 Department of Epidemiology, University Medical Center Groningen, University of Groningen, Groningen, The Netherlands 3 Department of Medical Decision Making, Leiden University Medical Center, Leiden, The Netherlands Do individuals with and without depression value depression differently? And if so, why? Katerina Papageorgiou1 • Karin M. Vermeulen2 • Maya J. Schroevers1 • Anne M. Stiggelbout3 • Erik Buskens2 • Paul F. M. Krabbe2 • Edwin van den Heuvel2 • Adelita V. Ranchor1 Accepted: 8 May 2015 / Published online: 3 June 2015 Accepted: 8 May 2015 / Published online: 3 June 2015 The Author(s) 2015. This article is published with open access at Springerlink.com assess discrepancies in valuations and identify their determinants. 3 Department of Medical Decision Making, Leiden University Medical Center, Leiden, The Netherlands Introduction & Adelita V. Ranchor a.v.ranchor@umcg.nl Katerina Papageorgiou a.papageorgiou@umcg.nl 1 Health Psychology Section, University Medical Center Groningen, University of Groningen, Groningen, The Netherlands 2 Department of Epidemiology, University Medical Center Groningen, University of Groningen, Groningen, The Netherlands 3 Department of Medical Decision Making, Leiden University Medical Center, Leiden, The Netherlands Depression is a prevalent condition [1] that imposes a high burden on health-related quality of life [2–4]. Various types of interventions have been found effective in alleviating depression [5–7]. Economic evaluations of healthcare in- terventions are currently necessary for policy decisions. Cost-utility analysis is a type of economic evaluation in which health state valuations (also called ‘‘utility values,’’ ‘‘utilities’’ or ‘‘preferences’’) are the recommended metric measures to assess the effectiveness of the intervention [8]. Currently, cost-utility analyses are primarily used in the evaluation of interventions for somatic conditions but are 1 Health Psychology Section, University Medical Center Groningen, University of Groningen, Groningen, The Netherlands 2 Department of Epidemiology, University Medical Center Groningen, University of Groningen, Groningen, The Netherlands 2 Department of Epidemiology, University Medical Center Groningen, University of Groningen, Groningen, The Netherlands 3 Department of Medical Decision Making, Leiden University Medical Center, Leiden, The Netherlands 12 3 Qual Life Res (2015) 24:2565–2575 2566 less frequently used in evaluating interventions for mental conditions such as depression [9]. values to use in evaluations more critical. To inform this choice of perspective, it is also important to understand the underlying mechanisms of the discrepancies, which was the second aim of this study. Relevant research, although limited, exists regarding valuations for somatic conditions, in which a number of factors have been suggested as in- ﬂuencing over- and/or under-valuations by the patients or by the general population [20, 21]. Patients’ adaptation to their conditions [22] as well as the general population’s attention to the negative aspects of the conditions [23] have both been found to explain why patients value somatic conditions less negatively than the population. For de- pression, some suggestions have thus far been made con- cerning factors that account for observed discrepancies in valuations between individuals with and without depres- sion. The effect of these suggested factors has never been empirically investigated. Introduction Among such factors, dysfunc- tional attitudes (for example, black and white thinking) have been suggested to relate to downward bias in valua- tions of individuals with depression, whereas stigma to- ward depression (for example: ‘‘They could snap out of depression if they really wanted’’) has been proposed to account for upward biases in valuations of individuals without depression [18]. A question that has generated ongoing discussion, with possibly serious policy implications, is whose values should be used in evaluating health conditions: those ex- pressed by the patients experiencing the health states or those expressed by the general population imagining them [10]. This issue has been extensively debated [11] in light of evidence showing that patients and the general popula- tion differ in their valuations of health conditions. Theoretically, the general population should be less biased in their valuations due to lack of self-interest; on the other hand, patients should be better informed on their condi- tions. Studies on valuations of somatic conditions show that somatically ill patients tend to value their states less negatively than the general population [12, 13], although this ﬁnding is not consistent [14, 15]. Consequently, if a condition is valued more negatively by the general population, more favorable conclusions can be expected regarding the efﬁcacy of an intervention for a somatic condition when using the valuation of the population rather than the patient perspective. Much less is known about valuations of depression [16] and the role of individuals’ own state of depression in making such valuations. Two relevant studies have inves- tigated depression valuations, both showing that indi- viduals who experience depression value hypothetical depression states more negatively that those who do not [17, 18]. This less negative valuation of depression by the general population entails that evaluations of depression interventions, in contrast to interventions for somatic conditions, would result in conclusions of less efﬁcacy when using the population’s perspective. However, current evidence is insufﬁcient to draw ﬁrm conclusions, given the limited number of studies and their methodological short- comings. In the previous studies, discrepancies were ob- served between individuals without depression and individuals experiencing severe levels of depression. Fur- thermore, a limited number of hypothetical depression states were included (mild, moderate and severe depres- sion), which may not have reﬂected the variety of different states of depression, characterized by different levels of dysfunction across health-related quality of life domains. Vignettes of depression health states be related with the discrepancies in how they value de- pression. Our purpose was to determine whether the valuations of individuals with depression or the general population were affected by subjectivity in perceptions of depression, by individual characteristics related to depres- sion or by both. We considered this question important for determining whose valuations can be considered more valid and legitimate. be related with the discrepancies in how they value de- pression. Our purpose was to determine whether the valuations of individuals with depression or the general population were affected by subjectivity in perceptions of depression, by individual characteristics related to depres- sion or by both. We considered this question important for determining whose valuations can be considered more valid and legitimate. Thirty vignettes with descriptions of different depression states were developed. Each participant was randomly presented four of a total of 30 vignettes, and the order of presentation was randomized. Depression states were based on the McSad depression-speciﬁc classiﬁcation system [25], previously translated and validated in Dutch [26]. The McSad consists of six depression-speciﬁc health-related quality of life domains: Emotion, Self-appraisal, Cogni- tion, Physiology, Behavior and Role function. Each domain could be categorized using one of the four levels of dys- function, namely ‘‘no,’’ ‘‘mild,’’ ‘‘moderate’’ or ‘‘severe’’ dysfunction. Different combinations of the four levels across the six domains make it possible to generate 4096 (46) different states. For example, the state with the proﬁle ‘‘343433’’ describes moderate dysfunction in the Emotion, Cognition, Behavior and Role functioning domains and severe dysfunctioning in the Self-appraisal and Physiology domains. To include depression states with different combinations of the four levels of dysfunction across the six domains, we used the orthogonal procedure (SPSS), which generated 25 states based on the McSad classiﬁca- tion system. Another ﬁve states, more consistent with re- spect to level of dysfunction across the six domains, were added. Table 1 provides an example of the description of a depression state (‘‘222222’’). Through the current study, we aimed to investigate whether individuals with and without depression differ in their valuations of depression. To overcome the limitations of previous studies on this subject, we conducted a large- scale valuation study using a wide range of hypothetical depression states based on a standard depression-speciﬁc classiﬁcation system, and we elicited valuations using the time-trade-off (TTO) method. Vignettes of depression health states Furthermore, we examined whether potential differences can be related to group dif- ferences in perceptions of depression, by individual char- acteristics or by both. Introduction Moreover, discrepancies were more prominent when a rating scale was used rather than the standard gamble method, although the latter method has been considered more appropriate [19]. Thus, the ﬁrst aim of this study was to investigate whether discrepancies in valuations of de- pression occur between individuals with and without depression. Following and expanding on this line of thought, we hypothesized that potential discrepancies in valuations of depression can be explained by differences between indi- viduals with and without depression with regard to how they perceive depression and with regard to their individual characteristics. We expected that depression, like other mental conditions with an unclear nature and cause, can be more prone to subjectivity in how it is perceived by those with no experience of it. More speciﬁcally, we expected that individuals without depression might differ from those with depression with regard to stigma (for example, whether depression is a ‘‘real’’ disease), empathy (the degree to which the experience of being in depression can be under- stood), perceived susceptibility (whether depression can occur to oneself) and illness perceptions (whether depres- sion is treatable, what its consequences are). We, therefore, hypothesized that this subjectivity in the perceptions of depression by individuals who do not experience it could be related to the discrepancies in valuations of depression be- tween individuals with and without depression. In addition, we expected that individuals who experience depression would differ from those who do not experience depression with regard to individual characteristics known to relate to depression, speciﬁcally levels of mastery, self-compassion and dysfunctional attitudes. We expected lower levels of mastery and self-compassion and higher level of dysfunc- tional attitudes among individuals with depression com- pared with individuals without depression, and we hypothesized that these differences in characteristics might The existence of discrepancies in valuations made by individuals with and without depression, if found, may carry policy implications, making the question of whose 123 123 Qual Life Res (2015) 24:2565–2575 2567 Participants and procedures This project consisted of a cross-sectional nationwide study conducted among members of the Dutch population. Par- ticipants included in this study were at least 18 years old, were able to understand Dutch and provided informed consent. Recruitment was conducted via an existing panel of an international company specialized in sampling for marketing and academic research [24]. To ensure repre- sentativeness of the sample, we used a stratiﬁed sampling strategy with strata based on gender, age, educational background and place of residence (across the 12 provinces of the Netherlands). In the ﬁrst part of the survey, we collected data on participants’ demographic characteristics and health statuses. In the second part, after being trained in the valuations tasks, each participant valued a total of four vignettes describing different states of depression. In the third part, participants ﬁlled out a questionnaire related to their experiences with chronic conditions and depres- sion, their perceptions of depression and their individual characteristics. The study protocol was reviewed by the Medical Ethical Committee of the University Medical Center Groningen, and a waiver was provided (M12.119685). We pilot-tested the instruments with 200 participants. Data regarding the difﬁculty of the tasks, completion time and the construct validity of valuations supported the feasibility of the study protocol. These data were only used for pilot purposes. Primary data collection took place online using software developed for this study by the same company that recruited the participants. The 30 depression states were rated in terms of their severity (as mild, moderate or severe) by three mental healthcare professionals (two clinical psychologists and one psychiatrist), all specialized and experienced in working with individuals with depression. They ﬁrst in- dependently rated the 30 states, and inter-rater agreement was strong (Kendall’s W = .773) [27]. Initial discrepancies among the ratings were resolved through discussion, achieving a ﬁnal consensus of four states rated as ‘‘mild,’’ 17 as ‘‘moderate’’ and nine as ‘‘severe.’’ Valuation elicitation procedure To elicit valuations of depression, we used the TTO method, which has been shown to be a reliable and valid method [19] in previous evaluations of mental health conditions [28, 29]. In the TTO, participants are asked to imagine and choose between two options. The ﬁrst option involves living in the health state described for another 10 years and then dying. The second option involves full recovery from this state, but living for less than 10 years. The number of years in the second option varies, until the point of indifference between Choice A and Choice B is reached. In the current study, participants were trained in the use of the TTO method using a vignette describing asthma. A 123 12 3 Qual Life Res (2015) 24:2565–2575 2568 Table 1 Example of two vignettes describing McSad depression state ‘‘222222’’ and ‘‘444444’’ Emotion You feel more down than usual and you don’t enjoy things as usual Self-appraisal You don’t feel very good about yourself these days and you often see the down side of everything Cognition You have some trouble concentrating and remembering these days, and it seems harder to make decisions Physiology Your sleep is a little troublesome these days. You don’t have quite the normal get up and go and you gave less of an appetite Behavior Things are more of a chore these days and at times you feels sluggish or agitated Role function You are able to function okay at work, home, school or with friends, but often don’t enjoy what you are doing Emotion You feel terribly down or sad all the time. You don’t enjoy anything and you feel desperate Self-appraisal You feel worthless and see absolutely no hope for yourself, or you don’t know why people even care about you, or you feel very guilty about the past and see no future for yourself Cognition You feel like your mind is shut down, overloaded or racing. You can’t read or watch TV and you can’t make even little decisions Physiology Your sleep is terrible these days and you don’t feel rested. You have absolutely no energy and you feel constantly tired. You have no interest in food and you have lost a lot of weight over the last month Behavior You can’t do anything. Explanatory factors of potential discrepancies between individuals with and without depression p g g Upon completion, participants were asked to use a ﬁve- point scale (with 1 being ‘‘not difﬁcult at all’’ and 5 being ‘‘very difﬁcult’’) to indicate difﬁculty with understanding the task. Empathy was assessed using the Emotional subscale of Batson and colleagues’ Empathy scale [35].The six items of the scale are words describing emotions (e.g., ‘‘sympa- thy’’), and participants are asked to what degree they ex- perienced each emotion (from 1: ‘‘not at all’’ to 6: ‘‘very much’’) when reading a vignette describing a woman who experiences depression. Total scores range from 6 to 36, with higher scores representing higher levels of empathy. The Perceived Susceptibility scale [36] was used to assess to what extent participants perceived themselves suscepti- ble to depression. The scale includes four items (e.g., ‘‘What do you believe is the chance that you will develop depression in your lifetime?’’) with varying answer cate- gories. Total scores range from 4 to 22, with higher scores indicating higher perceived susceptibility to depression. Cognitive representations of depression were assessed Valuation elicitation procedure You are completely shut down, or extremely agitated, or, you don’t see any perspective, and wonder whether you wouldn’t be better off dead Role function You had to stop work, or you do nothing at home, or have completely withdrawn from everything You had to stop work, or you do nothing at home, or have completely withdrawn from everything provided on a four-point scale, ranging from 0 (‘‘not at all’’) to 3 (‘‘almost every day’’). Item scores are summed up to calculate a total score, which ranges from 0 to 27, with higher scores representing higher levels of depression. The PHQ-9 has established psychometric properties among the general population [31]. The cutoff of 10 has been rec- ommended to indicate the presence of major depression [31, 33]. Results on the PHQ-9 in the current study demonstrated good internal consistency (Cronbach’s alpha: .92), comparable to what has been previously reported [31, 34]. ping-pong iteration method was used [30]. An interactive horizontal scale representing the life years (0–10) was used as a visual aid. In the primary valuation tasks, the ping- pong method was skipped, and participants could use the interactive scale directly and report the point of indiffer- ence between Choice A and Choice B. We chose this op- tion because the pilot testing indicated that using the ping- pong method for multiple states was tiring for participants. The value attached to the health state was calculated based on the TTO as 1 - (xmax/10), where xmax is the maximum years the participant was willing to trade to live free of depression. The value could therefore range from 0 to 1, with lower values representing more negative valuations. Analysis plan To answer our ﬁrst research question concerning discrep- ancies in valuations of depression between individuals with and without depression, we ﬁrst deﬁned the two depression groups, described them and compared them to each other and to the general population, using Chi-square and t tests. We then aimed to test the effect of the depression group (deﬁned as the predictor variable) on valuations (deﬁned as the outcome variable). Given that the valuations represent a scale from 0 to 1 using discrete increments of .05, the frequency distribution of the valuations was approximated with a binomial distribution in which the proportion rep- resents a 0–1 scale. A binomial distribution would ap- proximate a normal distribution when no ﬂoor or ceiling effects are present. The binomial proportions were modeled using a logit link function. To account for the correlation among the four valuations per subject, a random intercept for subjects was included in the binomial proportion. The effect of the depression group on valuations, including an interaction term with the severity of the depression state, was tested using the Wald test, controlled for demographic variables that were related to both the depression group and the valuations, as well as for the depression states within the severity groups. The mean valuations were estimated for the ‘‘with depression’’ and the ‘‘without depression’’ groups separately for mild, moderate and severe states of depression. To reduce the possibility of biases in valuations due to inappropriate completion of the TTO task, we re- peated the analyses, excluding those participants who were not willing to trade any years or who traded the same number of years for all four of the vignettes presented to them. When the effect of the depression group was signiﬁcant, for all severity states or just one severity group, we re- peated the same analysis for the relevant severity groups, by examining the effect of variables related to perceptions of depression and individual characteristics on the valua- tions. Both the main effect of these variables and their interaction with the depression group were investigated. We used the basic model previously described, with valuations as the outcome variable and demographic vari- ables and depression group as predictor variables. To this model, we added each of the six variables (empathy, per- ceived susceptibility, illness perceptions, mastery, self- compassion and dysfunctional attitudes) one by one as predictors, to examine for main effects. Analysis plan To examine in- teractions, for each of these six variables, we added an interaction term with depression group to our model. Our hypothesis was that certain variables would affect the valuations of either the ‘‘with depression’’ or the ‘‘without depression’’ groups, resulting in discrepancies in valuations between the two groups. For example, dysfunctional Deﬁning depression groups As validated Dutch versions were not available for the scales assessing Stigma for Depression, Perceived susceptibility and Em- pathy, we used established translation guidelines [44, 45] to forward–backward translate these scales into Dutch and then back into English Deﬁning depression groups Deﬁning depression groups To assign participants to one of the two depression groups (‘‘with depression’’ or ‘‘without depression’’), we used the cutoff of 10 on the Patient Health Questionnaire—9-item scale (PHQ-9) [31]. The PHQ-9 is an established scale, previously validated in the Dutch language [32], consisting of nine items corresponding to DSM depression symptoms, such as reduced interest or difﬁculty sleeping. The ques- tions relate to the frequency of depressive symptoms ex- perienced during the last 2 weeks, and answers are 123 123 Qual Life Res (2015) 24:2565–2575 2569 using the subscales of Consequences, Treatment Control and Timeline Cyclical from the revised Illness Perception Questionnaire (IPQr) [37], Dutch version [38]. The number of items differs per subscale (example item: ‘‘Depression does not have much effect on one’s life’’). Total scores for the Consequences, Treatment Control and Timeline Cycli- cal subscales range from 6 to 30, 5 to 25 and 4 to 20, respectively. The Stigma for Depression scale [39], used to assess personal stigma for depression, includes nine items (e.g., ‘‘People with depression could snap out of it if they wanted’’), with a tenth item added by our group following content validity assessments. The scale’s total score ranges from 0 to 40, with higher scores indicating higher stigma. Mastery was measured by means of the Dutch version [40] of the Pearlin Mastery scale [41]. The scale includes seven items (e.g., ‘‘I have little control over things that happen to me’’), with ﬁve answer categories and total scores ranging from 7 to 35. Higher scores indicate higher mastery. The Brief Self-Compassion scale, available in Dutch [42], was used to assess self-compassion. It includes 12 items, to be answered on a seven-point scale. The total score is com- puted as the mean of item scores and, therefore, ranges from 1 to 7, with higher scores indicating higher self- compassion. Dysfunctional attitudes were assessed using the Perfectionism subscale of the Dutch version of the Revised Dysfunctional Attitudes scale [43]. The scale consists of eleven items (e.g., ‘‘If I fail at my work, then I am a failure as a person’’), and answer categories range from 1 to 7. Total scores ranges from 11 to 77, with higher scores indicating less dysfunctional attitudes. Validated Dutch versions were used for all of the scales. Sample From the 2278 individuals who were invited to participate, 1268 (55.6 %) completed the survey. A ﬂowchart of dropouts for the different stages of the survey is presented in Fig. 1. Table 2 provides a description of the ﬁnal sam- ple. Chi-square tests conﬁrmed the representativeness of the sample within the Dutch population with regard to age, gender, place of residence and educational background. Using the PHQ-9 cutoff score of 10, 200 participants (15.8 %) were assigned to the ‘‘with depression’’ group, a percentage relatively higher from the 10 % previously re- ported [34]. In the whole sample, the majority of par- ticipants reported either no (44.2 %) or minor (24.5 %) difﬁculties with understanding the valuation task. Additional information Information was collected on participants’ gender, age, nationality, place of residence, education, family situation, occupation and religious background. Generic health status was also assessed by means of the EQ-5D ? C variant health status classiﬁcation system and visual analogue scale [46, 47]. Additional questions concerned previous diagnoses of a list of 23 chronic conditions, other than depression. Furthermore, we collected self-reported infor- mation concerning previous diagnoses of depression by a psychiatrist or a psychologist, previous depressive episodes or experience with depression via close friends or family. This information was collected to better describe our sample and to control for their effect when searching for discrepancies in valuations based on depression status. 12 3 2570 Qual Life Res (2015) 24:2565–2575 Discrepancies in valuations between individuals with and without depression attitudes might affect valuations of individuals with de- pression, resulting in lower valuations. Therefore, we ex- amined the depression group (‘‘with depression’’ or ‘‘without depression’’) as a moderator in the relationship between the aforementioned variables and valuations of depression. For this purpose, we examined the interaction effect of each of the aforementioned variables, with the depression group as a predictor of depression valuations. Among the participant variables, age, family situation and experience with chronic conditions were found to be cor- related with depression group (‘‘with depression’’ or ‘‘without depression’’) and with valuations. We, therefore, controlled for these three variables in our analysis. Addi- tionally, we controlled for the different depression states within the same severity group. Results of the generalized linear mixed models (GLMM) analysis demonstrated a signiﬁcant interaction effect between the severity of the depression states and the depression group on valuations of depression (F = 3.22; p = .04). This interaction effect indicated that the effect of the presence of depression on valuations was different among depression states of dif- ferent severity. A signiﬁcant and clinically meaningful difference in estimated mean values between ‘‘with de- pression’’ and ‘‘without depression’’ groups was observed for mild, but not for moderate or severe depression states (see Table 3). Individuals with depression valued mild depression lower, and thus more negatively, compared with individuals without depression. When we repeated the same analysis excluding participants who were not willing to trade any years (N = 104, 8.2 %) or who traded the same number of years for all four of the states presented to them (N = 154, 12.1 %), only minor changes in the results were observed. The interaction effect between the severity of depression states and the depression group on valuations of depression remained signiﬁcant, and similar differences were found in estimated mean values between the two depression groups. The analysis was conducted using the GLIMMIX pro- cedure of the SAS/STAT software [48], a generalized linear mixed model that extends the class of generalized linear models (GLMs) by incorporating normally dis- tributed random effects. This procedure was considered appropriate, given the structure of our data, speciﬁcally the correlations among the responses. A difference of .05 in valuations was considered meaningful [49, 50], and sta- tistical signiﬁcance was assessed at the .05 level. Explanatory factors for discrepancies between individuals with and without depression Depression groups differed in their average scores on all of the variables related to perceptions of depression and in- dividual characteristics, except for stigma and perceptions concerning treatment control (see Table 3). As hy- pothesized, individuals without depression reported lower levels of empathy and perceived susceptibility, and they perceived depression to be related to fewer consequences and a more cyclical timeline. Individuals with depression showed lower levels of mastery and self-compassion and higher levels of dysfunctional attitudes. Fig. 1 Study ﬂow diagram 1 3 After controlling for age, family situation, experience with chronic conditions and the depression states within the severity groups, we examined the variables related to perceptions of depression and individual characteristics as predictors of depression valuation, separately for each level of depression (mild, moderate and severe). Self-compas- sion (p = .045) and dysfunctional attitudes (p = .013) 12 Qual Life Res (2015) 24:2565–2575 2571 Table 2 Description of the two depression groups With depression Without depression Comparea Sample (%) Population (%) Compareb N = 200 % N = 1068 % General Gender Female 119 59.5 529 49.5 v2(1) = 6.70 p = .011 51.1 51 p = .76 Age Mean (SD) 42.04 (16.17) 47.53 (17.50) t = 4.35 p \ .000 18–35 71 35.5 287 26.9 v2(3) = 26.79 p \ .000 28.2 27.0 p = .98 36–50 75 37.5 296 27.7 29.3 28.0 51–65 38 19.0 264 24.7 23.8 25.0 65? a Compare ‘‘with depression’’ and ‘‘without depression’’ groups b Compare sample (whole) to the general Dutch population Compare sample (whole) to the general Dutch population Explanatory factors for discrepancies between individuals with and without depression In this study, we compared indi- viduals with and without depression overall, rather than subgroups, and we found that discrepancies do exist for valuations of mild depression. This ﬁnding is more valu- able than those from previous studies, as the research question seeks to compare and to identify whether dis- crepancies occur between those who experience depression and those who do not, to understand the implications of choosing one of the two perspectives. Additionally, we examined the effect of the severity of the valued state. were found to signiﬁcantly affect valuations of mild de- pression, with higher levels of self-compassion and less dysfunctional attitudes related to higher, and thus less negative, valuations. Perceptions concerning consequences of depression (p = .020) and treatment control (p = .001) were found to affect valuations of moderate depression states, with higher valuations related to stronger beliefs that depression has major consequences and that treatment can control depression. Empathy (p = .022) and perceived susceptibility (p = .002) were found to affect valuations of severe depression states, with higher valuations related to less stigma for depression and lower levels of perceived susceptibility to depression. When examined, the interac- tion term for each of these predictor variables with the depression group was never found to be signiﬁcant, meaning that for none of these variables was their asso- ciation with valuations different for the two groups. This result held for all of the different levels of depression severity. Therefore, the depression groups was not found to moderate the effect of the predictor variables on valuations, or in other words, the effect of the predictor variables on depression valuations was not different between indi- viduals with and without depression. This implies that discrepancies in valuations did not seem to relate to dif- ferences between individuals with and without depression with regard to their perceptions of depression or to their individual characteristics. y This study was the ﬁrst to investigate whether observed discrepancies between individuals with and without de- pression in valuations of depression can be explained by factors related to either the perceptions of depression by those who do not experience it (e.g., stigma or low em- pathy) or to individual characteristics known to be related to depression (e.g., low mastery or higher dysfunctional attitudes). Explanatory factors for discrepancies between individuals with and without depression 16 8.0 221 20.7 18.7 20.0 Education Low 76 38.0 340 31.8 v2(2) = 5.67 p = .059 32.8 34.3 p = .95 Middle 86 43.0 446 41.8 42.0 42.1 High 38 19.0 282 26.4 25.2 23.5 Family situation With partner 109 54.5 704 65.9 v2(1) = 9.54 p = .002 Single/other 91 45.5 364 34.1 Children Yes 95 54.7 650 60.9 v2(1) = 12.41 p = .001 Occupation Paid job 50 25.0 413 23.7 v2(2) = 13.62 p = .001 Education 26 13.0 118 13.7 Other 124 57.0 537 62.6 Religion Yes 63 31.5 443 41.5 v2(1) = 7.00 p = .009 General health Chronic conditions Yes 141 70.5 546 51.1 v2(1) = 25.48 p \ .000 EQ-6D VAS Mean (SD) 63.17 (18.53) 80.99 (15.22) t = 12.813 p = .000 Depression Under treatment 72 36.0 59 8.1 v2(1) = 101.5 p \ .000 Diagnosed-ever 108 54.0 179 16.8 v2(1) = 133.4 p = .000 Episode-ever 152 76.0 370 34.6 v2(1) = 118.9 p \ .000 Experience-others 113 56.5 323 30.1 v2(1) = 51.47 p \ .000 a Compare ‘‘with depression’’ and ‘‘without depression’’ groups b Compare sample (whole) to the general Dutch population Table 2 Description of the two depression groups 123 123 2572 Qual Life Res (2015) 24:2565–2575 Table 3 Estimated mean valuationsa for the ‘‘with depression’’ and the ‘‘without depression’’ group, separately for groups of depression states of different severity Severity of the depression state With depression Without depression p value M (SD) M (SD) Mild .693 .776 .007 Moderate .653 .709 .080 Severe .604 .659 .111 a Based on the GLMM; controlled for age, family situation, experience with chronic conditions and the states within severity groups a Based on the GLMM; controlled for age, family situation, experience with chronic conditions and the states within severity groups depression and their individual characteristics, these dif- ferences do not account for the observed discrepancies in their valuations. Our results are in line with and add speciﬁcity to results from previous studies showing more negative valuations of hypothetical depression states by individuals who experience depression. Previous studies report that discrepancies are dependent on the severity of the hypothetical depression state, with discrepancies mainly observed for hypothetical states of mild severity. In addition, discrepancies were previously reported only be- tween individuals without depression and a speciﬁc sub- group of individuals with depression, rather than between the two groups [17, 18]. Explanatory factors for discrepancies between individuals with and without depression As expected, individuals with and without de- pression differed with respect to how they perceive de- pression, with individuals without depression holding lower levels of empathy, higher levels of perceived sus- ceptibility, lower perceived consequences of depression and the perception of a less cyclical timeline. With respect to their individual characteristics, individuals without de- pression showed higher levels of mastery and self-com- passion and lower levels of dysfunctional attitudes. However, the effect of these factors on how the individuals value depression was not different between individuals with and without depression for any of the three severity levels of the hypothetical depression states. This ﬁnding implies that, although discrepancies occur between Table 3 Estimated mean valuationsa for the ‘‘with depression’’ and the ‘‘without depression’’ group, separately for groups of depression states of different severity Discussion Results of this study answer our ﬁrst research question concerning discrepancies between individuals with and without depression in valuations of depression states. Speciﬁcally, results show that individuals with and without depression differ in their valuations of mild hypothetical depression states, with individuals with depression valuing mild states of depression more negatively than those without depression. For moderate and severe depression states, no discrepancies between individuals with and without depression were observed. Concerning the second research question, results show that, although individuals with and without depression differ in their perceptions of 12 123 123 2573 Qual Life Res (2015) 24:2565–2575 However, given that the TTO is undoubtedly a cognitively demanding task, the presence of an interviewer might en- sure more accuracy in comprehension. Nevertheless, the pilot testing, the logical relationships of valuations with severity of depression and participants’ reports concerning the difﬁculty of the task are reassuring. Finally, we found no signiﬁcant interaction effects of variables related to perceptions of depression and individual characteristics with depression group on valuations. This led us to con- clude that the tested interaction variables cannot explain the observed discrepancies in depression valuations be- tween individuals with and without depression. However, it cannot be ruled out that the interaction effects were not signiﬁcant due to the limited power of our study. individuals with and without depression in their percep- tions of depression and their individual characteristics and discrepancies also occur in their valuations of depression, the earlier differences cannot explain the latter, as hypothesized. One of the strengths of the current study is the inclusion a large sample of participants, representative of the Dutch population with respect to age, gender, place of residence and educational background. Furthermore, the sample in- cluded a group of participants with depression that varied with respect to the years since diagnosis and the state of treatment. In contrast to previous studies, we have included a large number of depression states, ensuring a large variance with respect to the level of dysfunction in the various domains, thus supporting the generalizability of our ﬁndings. Our ﬁndings are of major importance for at least two reasons. Given that mild depression is valued differently according to the population being asked, we can expect less favorable outcomes if valuations are obtained from the general population, rather than from individuals with de- pression. Discussion This effect can be more profound in the case of interventions speciﬁcally targeting mild depression [55]. In contrast, in the case of interventions targeting individuals with moderate and severe depression, the choice of population does not seem to have an impact. Nevertheless, given that the population’s perspective correlates with more negative valuations in the case of somatic conditions, choosing the population’s perspective may entail not pri- oritizing mental health interventions in the agenda of health policy makers. Secondly, although we have hypothesized that various variables related to perceptions of depression and individual characteristics would explain discrepancies in perception, we found no evidence for their role. One possible conclusion is that discrepancies are not caused by upward biases in the general population, related to how they perceive depression, neither by downward biases of the individuals experiencing depressions, related to their individual characteristics. Rather, the discrepancies simply reﬂect genuine differences in how patients and the general population value depression, which is affected by experi- ences with depression itself. Future studies could therefore focus on different types of factors, such as focusing illusion [23], with individuals with depression focusing on the aspects in their health-related quality of life that are the most negatively affected by depression, while individuals without depression also consider aspects not affected by the condition. g When interpreting the results of the current study, cer- tain limitations should be considered. First, although we used the recommended cutoff score of an established scale to assess depression [51], inclusion of participants diag- nosed with depression by a healthcare professional might have been more accurate. In addition, our results might have been affected by the fact that participants recognized to experience depression might not have been aware that their symptoms can constitute a diagnosis of depression. Furthermore, following the standard TTO formulation, we asked participants to imagine living in the described states of depression for another 10 years. Yet if, as in reality, depression were described as a condition with ﬂuctuations between depression and remission, different valuations might have been obtained. The range of observed mean values between the mildest and the most severe depression states was rather small (a range of .15 for the ‘‘without depression’’ group and of .13 for the ‘‘with depression’’ group). In previous studies, the differences between the mildest and the most severe states were also relatively small. In the study by Gerhards et al. References 1. Kessler, R., Berglund, P., Demler, O., Jin, R., Koretz, D., Merikangas, K., et al. (2003). The epidemiology of major de- pressive disorder—Results from the National Comorbidity Sur- vey Replication (NCS-R). JAMA: Journal of the American Medical Association, 289(23), 3095–3105. 18. Pyne, J. M., Fortney, J. C., Tripathi, S., Feeny, D., Ubel, P., & Brazier, J. (2009). How bad is depression? Preference score es- timates from depressed patients and the general population. Health Services Research, 44(4), 1406–1423. 19. Green, C., Brazier, J., & Deverill, M. (2000). Valuing health- related quality of life—A review of health state valuation tech- niques. Pharmacoeconomics, 17(2), 151–165. 2. Ustun, T., Ayuso-Mateos, J., Chatterji, S., Mathers, C., & Mur- ray, C. (2004). Global burden of depressive disorders in the year 2000. British Journal of Psychiatry, 184, 386–392. 20. Ubel, P. A., Loewenstein, G., & Jepson, C. (2003). Whose quality of life? A commentary exploring discrepancies between health state evaluations of patients and the general public. Quality of Life Research, 12(6), 599–607. 3. Kruijshaar, M., Hoeymans, N., Bijl, R., Spijker, J., & Essink-Bot, M. (2003). Levels of disability in major depression—Findings from the Netherlands Mental Health Survey and Incidence Study (NEMESIS). Journal of Affective Disorders, 77(1), 53–64. 21. Stiggelbout, A. M., & de Vogel-Voogt, E. (2008). Health state utilities: A framework for studying the gap between the imagined and the real. Value in Health, 11(1), 76–87. 4. Papakostas, G., Petersen, T., Mahal, Y., Mischoulon, D., Nierenberg, A., & Fava, M. (2004). Quality of life assessments in major depressive disorder: A review of the literature. General Hospital Psychiatry, 26(1), 13–17. 22. McTaggart-Cowan, H., Tsuchiya, A., O’Cathain, A., & Brazier, J. (2011). Understanding the effect of disease adaptation informa- tion on general population values for hypothetical health states. Social Science and Medicine, 72(11), 1904–1912. 5. Cuijpers, P., Sijbrandij, M., Koole, S. L., Andersson, G., Beek- man, A. T., & Reynolds, C. F. (2013). The efﬁcacy of psy- chotherapy and pharmacotherapy in treating depressive and anxiety disorders: a meta-analysis of direct comparisons. World Psychiatry, 12(2), 137–148. 23. Peeters, Y., Vlieland, T. P. M. V., & Stiggelbout, A. M. (2012). Focusing illusion, adaptation and EQ-5D health state descrip- tions: the difference between patients and public. Health Ex- pectations, 15(4), 367–378. 6. Barth, J., Munder, T., Gerger, H., Nueesch, E., Trelle, S., Znoj, H., et al. (2013). Comparative efﬁcacy of seven psy- chotherapeutic interventions for patients with depression: A network meta-analysis. Discussion [17], the largest difference was 12.52 in a 100-point rating scale and was observed in the group with no depression, while the smallest difference was 4.98, observed in the group with severe depression. In the study by Pyne et al. [18], max- imum difference in standard gamble-based valuations be- tween the mild and severe depression state was observed in the general population group (.24), whereas minimum differences were found in the group with severe depression (.19). This ﬁnding might be explained by the condition- speciﬁc nature of the description, but might also have made it more difﬁcult to recognize discrepancies in valuations between the two groups. Interview-based administration of the TTO task should also be discussed. Computer-based administration of the TTO is currently the standard, and based on previous studies [52–54], we expected that online administration would be valid and thus cost-effective. Future investigations could possibly examine discrep- ancies in valuations of depression between the general population and inpatients of mental health institutions, as patients’ awareness of their diagnosis as well as their perceptions of depression might differ. Future studies should also examine to what degree the discrepancies ob- served in the current study can actually affect conclusions 12 3 2574 Qual Life Res (2015) 24:2565–2575 of Health Care Programmes. Oxford: Oxford Medical Publications. regarding the effectiveness of depression interventions. More research could also elaborate on our ﬁnding that valuations of depression are affected by the individual’s experience with other chronic conditions. Because de- pression often co-occurs with somatic conditions [56, 57] and special psychological interventions are developed and offered to patients with somatic condition [58], further exploration of the effect of individuals’ somatic conditions on valuations of depression is warranted. Subsequent studies could use qualitative and mixed-methods ap- proaches to identify potential determinants of group dis- crepancies in valuations not considered in this study. Furthermore, examination of whether our ﬁndings con- cerning discrepancies and their determinants also apply to other mental health conditions would be of relevance. 9. Pirraglia, P. A., Rosen, A. B., Hermann, R. C., Olchanski, N. V., & Neumann, P. (2004). Cost-utility analysis studies of depression management: A systematic review. American Journal of Psy- chiatry, 161(12), 2155–2162. 10. Gold, M., Siegel, J., Siegel, J., et al. (1996). Cost-effectiveness in health and medicine. New York: Oxford University Press. 11. De Wit, G. A., Busschbach, J. J. V., & De Charro, F. T. (2000). Discussion Sensitivity and perspective in the valuation of health status: Whose values count? Health Economics, 9(2), 109–126. 12. Peeters, Y., & Stiggelbout, A. M. (2010). Health state valuations of patients and the general public analytically compared: A meta- analytical comparison of patient and population health state uti- lities. Value in Health, 13(2), 306–309. 13. Ubel, P. A., Loewenstein, G., Schwarz, N., & Smith, D. (2005). Misimagining the unimaginable: The disability paradox and health care decision making. Health Psychology, 24(4), S57–S62. 14. Krabbe, P. F. M., Tromp, N., Ruers, T. J. M., & van Riel, P. L. C. M. (2011). Are patients’ judgments of health status really dif- ferent from the general population? Health and Quality of Life Outcomes, 9, 31. Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http:// creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. 15. Dolders, M. G. T., Zeegers, M. P. A., Groot, W., & Ament, A. (2006). A meta-analysis demonstrates no signiﬁcant differences between patient and population preferences. Journal of Clinical Epidemiology, 59(7), 653–664. 16. Mohiuddin, S., & Payne, K. (2014). Utility values for adults with unipolar depression: Systematic review and meta-analysis. Med- ical Decision Making, 34(5), 666–685. 17. Gerhards, S. A. H., Evers, S. M. A. A., Sabel, P. W. M., & Huibers, M. J. H. (2011). Discrepancy in rating health-related quality of life of depression between patient and general population. Quality of Life Research, 20(2), 273–279. References Plos Medicine, 10(5), e1001454. 24. Survey Sampling International. Survey sampling international. http://www.surveysampling.nl/. 25. Bennett, K. J., Torrance, G. W., Boyle, M. H., & Guscott, R. (2000). Cost-utility analysis in depression: The McSad utility measure for depression health states. Psychiatric Services, 51(9), 1171–1176. 7. Jakobsen, J. C., Hansen, J. L., Simonsen, S., Simonsen, E., & Gluud, C. (2012). Effects of cognitive therapy versus interper- sonal psychotherapy in patients with major depressive disorder: A systematic review of randomized clinical trials with meta-ana- lyses and trial sequential analyses. Psychological Medicine, 42(7), 1343–1357. 26. Papageorgiou, K., Vermeulen, K. M., Schroevers, M. J., Buskens, E., & Ranchor, A. V. (2013). Testing the McSad depression speciﬁc classiﬁcation system in patients with somatic conditions: Validity and performance. Health and Quality of Life Outcomes, 11, 125. 8. Drummond, M. F., Sculpher, M. J., Torrance, G. W., O’Brien, B. J., & Stoddart, G. L. (2005). Methods for the economic evaluation 12 3 Qual Life Res (2015) 24:2565–2575 2575 27. Schmidt, R. (1997). Managing Delphi surveys using nonpara- metric statistical techniques. Decision Sciences, 28(3), 763–774. 44. Guillemin, F., Bombardier, C., & Beaton, D. (1993). Cross-cul- tural adaptation of health-related quality-of-life measures—Lit- erature-review and proposed guidelines. Journal of Clinical Epidemiology, 46(12), 1417–1432. 28. Koenig, H., Guenther, O. H., Angermeyer, M. C., & Roick, C. (2009). Utility assessment in patients with mental disorders va- lidity and discriminative ability of the time trade-off method. Pharmacoeconomics, 27(5), 405–419. 45. Sousa, V. D., & Rojjanasrirat, W. (2011). Translation, adaptation and validation of instruments or scales for use in cross-cultural health care research: A clear and user-friendly guideline. Journal of Evaluation in Clinical Practice, 17(2), 268–274. 29. Papageorgiou, K., Vermeulen, K. M., Leijten, F. R., Buskens, E., Ranchor, A. V., & Schroevers, M. J. (2014). Valuation of de- pression co-occurring with a somatic condition: feasibility of the time trade-off task. Health Expectations: An International Jour- nal of Public Participation in Health Care and Health Policy, doi:10.1111/hex.12303. 46. Krabbe, P., Stouthard, M., Essink-Bot, M., & Bonsel, G. (1999). The effect of adding a cognitive dimension to the EuroQol multiattribute health-status classiﬁcation system. Journal of Clinical Epidemiology, 52(4), 293–301. 30. Lenert, L. A., Cher, D. J., Goldstein, M. K., Bergen, M. R., & Garber, A. (1998). The effect of search procedures on utility elicitations. Medical Decision Making, 18(1), 76–83. 47. Wolfs, C. A. G., Dirksen, C. D., Kessels, A., Willems, D. C. M., Verhey, F. R. References J., & Severens, J. L. (2007). Performance of the EQ-5D and the EQ-5D ? C in elderly patients with cognitive impairments. Health and Quality of Life Outcomes, 5, 33. 31. Kroenke, K., & Spitzer, R. L. (2002). The PHQ-9: A new de- pression diagnostic and severity measure. Psychiatric Annals, 32(9), 509–515. p y f f 48. SAS Institute Inc. (2000–2004). SAS 9.1.3 help and documenta- tion. Cary, NC: SAS Institute Inc. ( ) 32. Zuithoff, N. P. A., Vergouwe, Y., King, M., Nazareth, I., van Wezep, M. J., Moons, K. G. M., et al. (2010). The Patient Health Questionnaire-9 for detection of major depressive disorder in primary care: consequences of current thresholds in a crosssec- tional study. BMC Family Practice, 11, 98. y 49. Obrien, B., & Drummond, M. (1994). Statistical versus quanti- tative signiﬁcance in the socioeconomic evaluation of medicines. Pharmacoeconomics, 5(5), 389–398. 50. Dolan, P., & Roberts, J. (2002). To what extent can we explain time trade-off values from other information about respondents? Social Science and Medicine, 54(6), 919–929. 33. Wittkampf, K. A., Naeije, L., Schene, A. H., Huyser, J., & van Weert, H. C. (2007). Diagnostic accuracy of the mood module of the Patient Health Questionnaire: A systematic review. General Hospital Psychiatry, 29(5), 388–395. 51. Kocalevent, R. D., Hinz, A. F. A. U., & Brahler, E. (2013). Standardization of the depression screener Patient Health Ques- tionnaire (PHQ-9) in the general population. General Hospital Psychiatry,. doi:10.1016/j.genhosppsych.2013.04.006. 1873-7714 (Electronic); 0163-8343 (Linking). 34. Martin, A., Rief, W., Klaiberg, A., & Braehler, E. (2006). Va- lidity of the Brief Patient Health Questionnaire Mood Scale (PHQ-9) in the general population. General Hospital Psychiatry, 28(1), 71–77. 52. Buitinga, L., Braakman-Jansen, L. M. A., Taal, E., & van de Laar, M. A. F. J. (2011). A computer time trade-off: A feasible and reliable alternative for the interview time trade-off in rheumatoid arthritis. Clinical and Experimental Rheumatology, 29(5), 783–789. 35. Batson, C., Fultz, J., & Schoenrade, P. (1987). Distress and em- pathy—2 Qualitatively distinct vicarious emotions with different motivational consequences. Journal of Personality, 55(1), 19–39. 53. van Nooten, F. E., Koolman, X., & Brouwer, W. B. F. (2009). The inﬂuence of subjective life expectancy on health state valuations using a 10 year TTO. Health Economics, 18(5), 549–558. 36. Gerend, M., Aiken, L., & West, S. (2004). Personality factors in older women’s perceived susceptibility to diseases of aging. Journal of Personality, 72(2), 243–270. 37. References Moss-Morris, R., Weinman, J., Petrie, K., Horne, R., Cameron, L., & Buick, D. (2002). The revised Illness Perception Ques- tionnaire (IPQ-R). Psychology and Health, 17(1), 1–16. 54. Lenert, L. A., Sturley, A., & Watson, M. E. (2002). iMPACT3: Internet-based development and administration of utility elicita- tion protocols. Medical Decision Making, 22(6), 464–474. 38. Illness Perception Questionnaire Website. http://www.uib.no/ipq/ 55. Jonkers, C. C. M., Lamers, F., Evers, S. M. A. A., Bosma, H., Metsemakers, J. F., & Van Eijk, J. T. M. (2009). Economic evaluation of a minimal psychological intervention in chronically ill elderly patients with minor or mild to moderate depression: A randomized trial (the DELTA-study). International Journal of Technology Assessment in Health Care, 25(4), 497–504. 39. Grifﬁths, K., Christensen, H., Jorm, A., Evans, K., & Groves, C. (2004). Effect of web-based depression literacy and cognitive- behavioural therapy interventions on stigmatising attitudes to depression—Randomised controlled trial RID B-5555-2009. British Journal of Psychiatry, 185, 342–349. 40. Kempen, G., Ranchor, A., Ormel, J., Van Sonderen, E., Van Jaarsveld, C., & Sanderman, R. (2005). Perceived control and long-term changes in disability in late middle-aged and older persons: An eight-year follow-up study. Psychology and Health, 20(2), 193–206. 56. Moussavi, S., Chatterji, S., Verdes, E., Tandon, A., Patel, V., & Ustun, B. (2007). Depression, chronic diseases, and decrements in health: results from the World Health Surveys. Lancet, 370(9590), 851–858. 57. Noel, P. H., Williams, J. W., Unutzer, J., Worchel, J., Lee, S., Cornell, J., et al. (2004). Depression and comorbid illness in elderly primary care patients: Impact on multiple domains of health status and well-being. Annals of Family Medicine, 2(6), 555–562. 41. Pearlin, L., & Schooler, C. (1978). Structure of coping. Journal of Health and Social Behavior, 19(1), 2–21. 42. Raes, F., Pommier, E., Neff, K. D., & Van Gucht, D. (2011). Construction and factorial validation of a short form of the Self- Compassion Scale. Clinical Psychology and Psychotherapy, 18(3), 250–255. 58. Beltman, M. W., Voshaar, R. C. O., & Speckens, A. E. (2010). Cognitive-behavioural therapy for depression in people with a somatic disease: meta-analysis of randomised controlled trials. British Journal of Psychiatry, 197(1), 11–19. 43. de Graaf, L. E., Roelofs, J., & Huibers, M. J. H. (2009). Mea- suring dysfunctional attitudes in the general population: The dysfunctional attitude scale (form A) revised. Cognitive Therapy and Research, 33(4), 345–355. 12 123
https://openalex.org/W2930976687	https://scholarcommons.sc.edu/cgi/viewcontent.cgi?article=2112&context=eche_facpub	English	null	Intelligent Machine Learning: Tailor-Making Macromolecules	Polymers	2,019	cc-by	9,677	Intelligent Machine Learning: Tailor-Making Macromolecules Intelligent Machine Learning: Tailor-Making Macromolecules Yousef Mohammadi See next page for additional authors Follow this and additional works at: https://scholarcommons.sc.edu/eche_facpub Part of the Chemical Engineering Commons University of South Carolina University of South Carolina Scholar Commons Scholar Commons University of South Carolina University of South Carolina Scholar Commons Scholar Commons Faculty Publications Chemical Engineering, Department of Author(s) Author(s) Yousef Mohammadi, Mohammad Reza Saeb, Alexander Penlidis, Esmaiel Jabbari, Florian J. Stadler, Philippe Zinck, and Krzysztof Matyjaszewski Publication Info Publication Info Published in Polymers, Volume 11, Issue 4, 2019, pages 1-13. This Article is brought to you by the Chemical Engineering, Department of at Scholar Commons. It has been accepted for inclusion in Faculty Publications by an authorized administrator of Scholar Commons. For more information, please contact digres@mailbox.sc.edu. Author(s) Author(s) Yousef Mohammadi, Mohammad Reza Saeb, Alexander Penlidis, Esmaiel Jabbari, Florian J. Stadler, Philippe Zinck, and Krzysztof Matyjaszewski Intelligent Machine Learning: Tailor-Making Macromolecules Yousef Mohammadi 1,, Mohammad Reza Saeb 2, Alexander Penlidis 3, , Esmaiel Jabbari 4 , Florian J. Stadler 5,* , Philippe Zinck 6 and Krzysztof Matyjaszewski 7,* Author(s) Author(s) This article is available at Scholar Commons: https://scholarcommons.sc.edu/eche_facpub/1107 This article is available at Scholar Commons: https://scholarcommons.sc.edu/eche_facpub/1107 Yousef Mohammadi 1,, Mohammad Reza Saeb 2, Alexander Penlidis 3, , Esmaiel Jabbari 4 , Florian J. Stadler 5,* , Philippe Zinck 6 and Krzysztof Matyjaszewski 7,* 1 Petrochemical Research and Technology Company (NPC-rt), National Petrochemical Company (NPC), Tehran P.O. Box 14358-84711, Iran 1 Petrochemical Research and Technology Company (NPC-rt), National Petrochemical Company (NPC), Tehran P.O. Box 14358-84711, Iran Petrochemical Research and Technology Company (NPC-rt), National Petrochemical Company (NPC), Tehran P.O. Box 14358-84711, Iran 2 Department of Resin and Additives, Institute for Color Science and Technology, Tehran P.O. Box 16765-654 Iran; mrsaeb2008@gmail.com 2 Department of Resin and Additives, Institute for Color Science and Technology, Tehran P.O. Box 16765- Iran; mrsaeb2008@gmail.com 2 Department of Resin and Additives, Institute for Color Science and Technology, Tehran P.O. Box 16765-654, Iran; mrsaeb2008@gmail.com 3 Department of Chemical Engineering, Institute for Polymer Research (IPR), University of Waterloo, Waterloo, ON N2L 3G1, Canada 4 Biomimetic Materials and Tissue Engineering Laboratory, Department of Chemical Engineering, University of South Carolina Columbia, Columbia, SC 29208, USA; JABBARI@cec.sc.edu 5 College of Materials Science and Engineering, Shenzhen Key Laboratory of Polymer Science and Technology Guangdong Research Center for Interfacial Engineering of Functional Materials, Nanshan District Key Lab for Biopolymers and Safety Evaluation, Shenzhen University, Shenzhen 518055, China 6 Unity of Catalysis and Solid State Chemistry, University of Lille, CNRS, Bât C7, Cité Scientiﬁque, 59652 Villeneuve d’Ascq Cédex, France; philippe.zinck@univ-lille1.fr 7 Department of Chemistry, Carnegie Mellon University, Pittsburgh, PA 15213, USA * Correspondence: mohammadi@npc-rt.ir (Y.M.); penlidis@uwaterloo.ca (A.P.); fjstadler@szu.edu.cn (F.J.S.); matyjaszewski@cmu.edu (K.M.); Tel.: +98-214-478-7510 (Y.M.); +1-519-888-4567 (A.P.); +86-755-8671-3986 (F.J.S.); +1-412-268-3209 (K.M.) Keywords: microstructure; Kinetic Monte Carlo; living copolymerization; oleﬁn block copolymers; artiﬁcial intelligence; ethylene; machine learning; genetic algorithms 1. Introduction The ubiquity of polymers in daily life, especially in packaging and consumer products, gives them a special importance among different types of materials [1–3]. However, the rapid growth of both humanity and the wealth level has led to an ever-increasing demand for resources. Since the mid-1980s, more polymers have been produced in volume than steel (the second-highest-volume material). Among polymeric materials, approximately 70% are thermoplastics, 50% of which are polyoleﬁns; the latter part is dominated by far by polyethylene (PE), polypropylene (PP), and their copolymers. Decades of research have developed important and widely used materials from high-pressure polymerization (low-density polyethylene (LDPE) [4]), (catalyzed) low-pressure polymerization (high-density polyethylene (HDPE), linear low-density polyethylene (LLDPE), and PP [5–12]), and single-site-catalyzed polymerization (metallocene HDPE/LLDPE/PP [13,14]). The property range of polyoleﬁns has been signiﬁcantly expanded by the more recent and seminal introduction of chain-shuttling polymerization by Dow Chemical Company, where a chain-shuttling agent (CSA) can transport a chain from one catalyst to another, thus allowing for the production of block copolymers inexpensively on an industrial scale [15]. In fact, although the ﬁrst catalyst is capable of engaging a large number of comonomer units, the second one incorporates small amounts of comonomer. Hence, two distinct types of polyethylene chains, i.e., soft and hard copolymers, are simultaneously formed due to the difference in comonomer consumption tendency of the catalysts. Cross-shuttling of active centers among different living and dormant chains in the reaction medium via CSA molecules results in the synthesis of multi-block polyethylene chains. This invention permits designing oleﬁn block copolymers (OBCs) with properties that allow for using not only different grades as commodity polymers but also as specialty engineering applications [16]. The multi-block structure of OBCs yields a dual character of thermoplastic processability as well as elastomeric solid state properties. More importantly, the properties can be tuned by varying the concentration and type of the catalysts and monomers, as well as the concentration of the chain-shuttling/transfer agent, ultimately leading to different comonomer contents of the hard and soft segments as well as their average lengths [15,17]. These multifaceted properties can be tuned over a wide range, but this ﬁne-tuning is not that trivial, as the ﬁrst attempt for a usable OBC required 1600 syntheses via high throughput setups with fast molar mass distribution and thermal analyses [15]. Received: 13 February 2019; Accepted: 10 March 2019; Published: 1 April 2019 Abstract: Nowadays, polymer reaction engineers seek robust and effective tools to synthesize complex macromolecules with well-deﬁned and desirable microstructural and architectural characteristics. Over the past few decades, several promising approaches, such as controlled living (co)polymerization systems and chain-shuttling reactions have been proposed and widely applied to synthesize rather complex macromolecules with controlled monomer sequences. Despite the unique potential of the newly developed techniques, tailor-making the microstructure of macromolecules by suggesting the most appropriate polymerization recipe still remains a very challenging task. In the current work, two versatile and powerful tools capable of effectively addressing the aforementioned questions have been proposed and successfully put into practice. The two tools are established through the amalgamation of the Kinetic Monte Carlo simulation approach and machine learning techniques. The former, an intelligent modeling tool, is able to model and visualize the intricate inter-relationships of polymerization recipes/conditions (as input variables) and microstructural features of the produced macromolecules (as responses). The latter is capable of precisely predicting optimal copolymerization conditions to simultaneously satisfy all predeﬁned microstructural features. The effectiveness of the proposed intelligent modeling and optimization techniques for solving this extremely important ‘inverse’ engineering problem was successfully examined by investigating the possibility of tailor-making the microstructure of Oleﬁn Block Copolymers via chain-shuttling coordination polymerization. Keywords: microstructure; Kinetic Monte Carlo; living copolymerization; oleﬁn block copolymers; artiﬁcial intelligence; ethylene; machine learning; genetic algorithms Polymers 2019, 11, 579; doi:10.3390/polym11040579 www.mdpi.com/journal/polymers Polymers 2019, 11, 579 2 of 13 1. Introduction The reason behind this is that a whole set of independently controllable factors interact with each other non-trivially and non-intuitively and, consequently, simple predictions of molecular architecture from the synthesis variables are not readily possible. Hence, in comparison to classical polyoleﬁn synthesis, the complexities of OBC polymerization require a highly advanced understanding of polymerization. This was ﬁrst attempted by Zhang et al. [18], who proposed a kinetic model for chain-shuttling polymerization that was able to correlate average molecular weight, comonomer content, and number of blocks with such synthesis factors as monomer ratio, catalyst composition, and CSA level. While this model was successful in obtaining a rough idea of the structure of the OBC, more powerful tools are required for gaining an in-depth understanding of the microstructure. This has been achieved by recent Kinetic Monte Carlo (KMC) simulations of chain-shuttling polymerizations [19–24]. Due to the multifarious types of reactions occurring simultaneously, including cross-propagation and cross-shuttling reactions, setting up these simulations is not an easy task. The KMC simulation was designed to track every monomer, CSA, and catalyst molecule as well as every single chain in the virtual reactor in order to obtain the microstructural features of the OBC. These features are represented by average number of linkage points per chain (LP), ethylene sequence length (ESL) of hard and soft blocks, average degree of polymerization (DPn) of hard and soft blocks, longest ethylene sequence length (LES) of hard and soft blocks, comonomer content (C8%) of hard and soft blocks, and hard block percent (HB%) [19–24]. While the results on the microstructure of OBCs provided by the KMC simulator are believed to be very accurate, the developed KMC algorithm cannot suggest polymerization variables that can yield a certain desired microstructure. In other words, the KMC simulator receives X (x1, x2, x3, . . . , xn) 3 of 13 Polymers 2019, 11, 579 as an input of polymerization variables and yields microstructural arrays of Y (y1, y2, y3, . . . , ym), i.e., the direction is X→Y (modeling). For practical applications, however, and in order to address a much more complex and interesting problem, engineers need copolymers with certain microstructures, which cannot be simply deduced from this kind of algorithm or by other means (e.g., deterministic models) due to the complexity of the microstructure. 1. Introduction In other words, there is no such thing as a single-valued functional relationship, i.e., a one-to-one relationship between polymerization factors and microstructural features. Hence, the inverse pathway, i.e., selecting synthesis conditions from desired microstructural properties in the direction Y→X (optimization), constitutes an ill-posed problem. p p Classical mathematical modeling methodologies are frequently very complex, time-consuming, and ill-conditioned when confronting very complex systems. The advent of Artiﬁcial Intelligence (AI) Modeling and Optimization techniques has opened new possibilities to scientists/engineers working on nonlinear processes. The use of Artiﬁcial Intelligence approaches with evolutionary learning in different areas of Science and Engineering, and notably in Materials Science [25–31], is offering a viable pathway to mimicking the evolution of complex system performance. This makes AI-based methods good candidates for identifying the mechanisms and inter-relations of complex polymerization kinetics. Since polymerization reactions proceed by a series of probability-controlled steps, and at any given time, new molecules are generated, converted, activated, deactivated, and transferred from one class/situation to another, polymerization systems can be better understood and, consequently, their behavior better tracked, by using evolutionary approaches. The aim of the current study is the establishment and development of unique and versatile modeling and optimization tools capable of handling precise predictions and intricate manipulations of microstructural features of complex macromolecules via amalgamation of KMC simulation approaches and Computational Intelligence techniques. Making this possible requires a deeper understanding of the interplay between different phenomena in the reacting system, culminating in the inter-relationships between polymerization recipes/conditions and ﬁnal microstructural properties. Two powerful and effective tools, including an Intelligent Modeling Tool (IMT) and an Intelligent Optimization Tool (IOT), are successfully developed and introduced. To construct and implement these tools, the KMC simulation approach has been hybridized with appropriate Artiﬁcial-Intelligence-based modeling and optimization techniques. The modeling and optimization of Chain-Shuttling Coordination (co)polymerization (invented by Dow Chemical Company) for the production of OBCs have been selected as a sufﬁciently complex case study to ‘challenge’ the use and validity of the proposed tools. 2. Model Development Figure 1 id i f h h f i f IOT d IMT i i d provides a view of how the functions of IOT and IMT are intertwined. Figure 1. Illustrative description of the Intelligent Modeling Tool (IMT) and the Intelligent Optimization Tool (IOT) developed for tailoring the microstructure of olefin block copolymers (OBCs). KMC, Kinetic Monte Carlo. Figure 1. Illustrative description of the Intelligent Modeling Tool (IMT) and the Intelligent Optimization Tool (IOT) developed for tailoring the microstructure of oleﬁn block copolymers (OBCs). KMC, Kinetic Monte Carlo. Figure 1. Illustrative description of the Intelligent Modeling Tool (IMT) and the Intelligent Optimization Tool (IOT) developed for tailoring the microstructure of olefin block copolymers (OBCs). KMC, Kinetic Monte Carlo. Figure 1. Illustrative description of the Intelligent Modeling Tool (IMT) and the Intelligent Optimization Tool (IOT) developed for tailoring the microstructure of oleﬁn block copolymers (OBCs). KMC, Kinetic Monte Carlo. Application of the IMT and IOT makes it possible to produce new grades of OBCs, with preset characteristics, without numerous trials but only via computer simulations ‘reverse-engineering’ the correct recipe to produce the desired OBC. This kind of approach-finding way to connect 10 microstructural features with highly complex inter-relations of seven synthesis factors, or to determine the possible synthesis variables to obtain a desired set of microstructural features, was tested for the OBC synthesis in references [19–24]. The basic scheme developed here can be used in numerous additional ways, spanning from biology to chemistry to physics, provided that some understanding of a possible mechanism exists that can link the input and output variables. These (simulation) calculations can be completed within an acceptable run time (for instance, ca. 45 years in case of KMC versus a few seconds in case of the IMT [22]). The IMT (X→Y) combines the advantages of ANN and KMC methodologies in a synergistic I h ll h I l l l b f bl Application of the IMT and IOT makes it possible to produce new grades of OBCs, with preset characteristics, without numerous trials but only via computer simulations ‘reverse-engineering’ the correct recipe to produce the desired OBC. This kind of approach-ﬁnding way to connect 10 microstructural features with highly complex inter-relations of seven synthesis factors, or to determine the possible synthesis variables to obtain a desired set of microstructural features, was tested for the OBC synthesis in references [19–24]. 2. Model Development The KMC simulator developed previously [19–24] acted as a virtual reactor to synthesize chains and look at microstructural aspects of OBCs. The microstructural variables of OBCs can be classiﬁed into two categories: (i) those that directly correspond to topological and architectural characteristics of chains, such as LP, DPnSOFT, DPnHARD, ESLSOFT, and ESLHARD; and (ii) those that indirectly determine the ultimate properties of OBCs, such as LESSOFT, LESHARD, C8%SOFT, C8%HARD, and HB%. There is some evidence that thermal, rheological, mechanical, as well as phase separation properties of OBCs in the melt-state are governed by the second category; for instance, crystallization is controlled to a large extent by HB%, LESSOFT, and LESHARD of OBCs. Hence, 10 kinetic parameters (two homo-propagation, two cross-propagation, two shuttling to virgin CSA, two shuttling to polymeryl CSA, and two transfer to hydrogen rate constants) as well as the amounts of seven reactants (a solvent, two catalysts, two monomers, hydrogen, and CSA) lead to a set of 10 (output) responses (the abovementioned microstructural features) calculated from six distribution functions (molar mass distribution of both blocks, block number distribution of both blocks, comonomer distribution of both blocks), which are partially dependent on each other. This alone provides sufﬁcient evidence for why simple single-valued one-to-one relationships cannot be found for a complex system such as OBC synthesis. 4 of 13 Polymers 2019, 11, 579 The hybridization of KMC with machine learning tools enables the prediction and tailoring of the microstructure and ultimate properties of OBCs. In this regard, two different intelligent tools were developed: (i) an Intelligent Modeling Tool (IMT); and (ii) an Intelligent Optimization Tool (IOT). The IMT is a hybrid of the KMC simulator with an intelligent modeler, such as an Artiﬁcial Neural Network (ANN) or a Fuzzy Logic System, while the IOT is an intelligent optimizer, combining KMC with, for example, a Genetic Algorithm (GA) or Swarm Intelligence. Figure 1 provides a view of how the functions of IOT and IMT are intertwined. Polymers 2018, 10, x FOR PEER REVIEW 4 of 13 were developed: (i) an Intelligent Modeling Tool (IMT); and (ii) an Intelligent Optimization Tool (IOT). The IMT is a hybrid of the KMC simulator with an intelligent modeler, such as an Artificial Neural Network (ANN) or a Fuzzy Logic System, while the IOT is an intelligent optimizer, combining KMC with, for example, a Genetic Algorithm (GA) or Swarm Intelligence. 2. Model Development The basic scheme developed here can be used in numerous additional ways, spanning from biology to chemistry to physics, provided that some understanding of a possible mechanism exists that can link the input and output variables. These (simulation) calculations can be completed within an acceptable run time (for instance, ca. 45 years in case of KMC versus a few seconds in case of the IMT [22]). manner. In a nutshell, the IMT relies on calculating a number of viable scenarios. For instance, polymerization recipes and conditions (X) are systematically defined to cover the whole variable space considered. Then, the KMC simulator calculates the microstructural features (Y) of all defined scenarios separately, and these are subsequently used to train several ANNs. Obviously, the KMC simulator and intelligent modeler interact in an offline manner. This results in various black boxes, which can intelligently predict any microstructural pattern with high accuracy within the covered factor space. Considering the time-intensiveness of the KMC-algorithm, the IMT offers a fast solution to obtain the required simulation results, which ultimately allows for determining a significantly wider range of polymerization recipes with their corresponding microstructural patterns, while in parallel determining the relationships between operational conditions and microstructural properties. In the case of the IOT (Y→X), the KMC simulator is continuously recalled by the GA optimizer via an online route in obvious contrast to the IMT step (X→Y) for which the KMC ANN ‘modeler’ The IMT (X→Y) combines the advantages of ANN and KMC methodologies in a synergistic manner. In a nutshell, the IMT relies on calculating a number of viable scenarios. For instance, polymerization recipes and conditions (X) are systematically deﬁned to cover the whole variable space considered. Then, the KMC simulator calculates the microstructural features (Y) of all deﬁned scenarios separately, and these are subsequently used to train several ANNs. Obviously, the KMC simulator and intelligent modeler interact in an ofﬂine manner. This results in various black boxes, which can intelligently predict any microstructural pattern with high accuracy within the covered factor space. Considering the time-intensiveness of the KMC-algorithm, the IMT offers a fast solution to obtain the required simulation results, which ultimately allows for determining a signiﬁcantly wider range of polymerization recipes with their corresponding microstructural patterns, while in parallel determining the relationships between operational conditions and microstructural properties. 2. Model Development via an online route, in obvious contrast to the IMT step (X→Y), for which the KMC-ANN ‘modeler’ works offline. The GA optimizer randomly generates a number of polymerization recipes and sends the recipes for error evaluation to the KMC simulator. When the KMC simulator “runs” these recipes/conditions, it resembles a virtual reactor that individually receives polymerization recipes as g p p p p In the case of the IOT (Y→X), the KMC simulator is continuously recalled by the GA optimizer via an online route, in obvious contrast to the IMT step (X→Y), for which the KMC-ANN ‘modeler’ works ofﬂine. The GA optimizer randomly generates a number of polymerization recipes and 5 of 13 Polymers 2019, 11, 579 sends the recipes for error evaluation to the KMC simulator. When the KMC simulator “runs” these recipes/conditions, it resembles a virtual reactor that individually receives polymerization recipes as inputs and visualizes microstructural features as outputs. Then, it returns the resulting microstructural characteristics, i.e., LP, DPnSOFT, DPnHARD, ESLSOFT, ESLHARD, LESSOFT, LESHARD, C8%SOFT, C8%HARD, and HB%, to the GA optimizer online through a closed-loop circuit. The GA optimizer subsequently invokes genetic operators, including selection, sorting, mating, crossover, and mutation, based on the feedback it received to generate the next generation (polymerization recipes); ﬁnally, the next generations are considered accordingly by sending them to the KMC simulator and receiving the corresponding microstructural characteristics; these steps are iterative. The process continues until a generation is evolved by the IOT, which contains the recipe required for production of the target OBC having a predeﬁned (desirable) microstructure, i.e., the global optimum. Thus, the IOT allows for ‘dialing-in’ variables and obtaining the required synthesis conditions, which is the inverse way to the above-described IMT approach. The output of the IOT is either a unique recipe or several polymerization recipes (Pareto fronts) corresponding to the preset target microstructural patterns. As a result, well-deﬁned microstructures can be tailored depending on customer demand. p g Two separate computer programs were written in the PASCAL programming language (Lazarus 1.2.4 IDE) and compiled into 64-bit executable codes using FPC 2.6.2. The ﬁrst program, i.e., the intelligent modeler, was established based on Artiﬁcial Neural Networks as a powerful black-box modeling technique and appropriately set to recalling our in-house KMC simulator code in an ofﬂine mode. 2. Model Development On the other hand, the second code, i.e., the intelligent optimizer, was written based on the Non-dominated Sorting Genetic Algorithm (NSGA-II); NSGA-II is an extremely (we would like to think, the most) powerful heuristic multi-objective search strategy to communicate with the KMC simulator via an online route [32–35]. A subroutine based on the “Mother-of-all Pseudo-Random Number Generators” algorithm was employed to produce the required random numbers during the modeling and optimization steps [36]. The random number generation subroutine satisﬁed the tests of uniformity and serial correlation with high resolution. The cycle length of the random number generator was 3 × 1047. All modeling and optimization steps were performed with a desktop computer with Intel Core i7-3770K (3.50 GHz), 32 GB of memory (2133 MHz), under the Windows 7 Ultimate 64-bit operating system. 3. Results and Discussion Table 1 Typical OBCs randomly selected in the ‘master plot’ of the IMT and synthesized by the Table 1 summarizes the randomly selected points in Figure 2 representing ﬁve OBCs and their topology- and property-related characteristics yielded by the IMT and KMC operators. It can be recognized that the IMT has appropriately learned the polymerization behavior and successfully predicted both types of molecular characteristics of randomly selected OBCs, as ‘approved’ by the KMC simulator. A closer look at the statistics in Table 1, bearing in mind the very low errors in predictions of the IMT, assures that such an intelligent machine can reliably be applied in ‘anticipating’ two categories of molecular features, leading to sophisticated architecture-property correlations. The IMT consists of 10 individual prediction mechanisms, each responsible for one molecular characteristic of the OBC, which are integrated into an intelligent toolbox towards the prediction of molecular characteristics of OBCs with ‘surgical’ precision. Table 1. Typical OBCs randomly selected in the master plot of the IMT and synthesized by the KMC simulator under specified operating conditions together with their topology- and property-related molecular characteristics. Topology-microstructural Characteristics Property-microstructural Characteristics DPnSOFT DPnHARD ESLSOFT ESLHARD LP C8%SOFT C8%HARD LESSOFT LESHARD HB% Point A: { MR = 0.75 and CC = 0.539 and log(CSA) = 0.313 } KMC 172.88 30.65 25.10 28.92 36.51 3.55 0.70 96.46 55.19 13.30 ANN 169.31 122.07 24.67 31.76 37.26 3.12 0.51 87.49 61.51 14.17 Error % 0.01 2.45 0.92 0.34 1.01 0.88 1.07 2.61 0.14 1.53 Point B: { MR = 0.75 and CC = 0.792 and log(CSA) = 1.886 } KMC 122.41 92.11 23.63 10.52 64.87 2.17 9.23 72.12 16.45 5.97 ANN 160.86 102.50 24.58 13.87 63.72 2.62 8.87 78.00 18.84 6.63 Error % 0.10 0.28 2.04 0.41 1.55 0.92 2.02 1.71 0.05 1.16 Point C: { MR = 0.75 and CC = 0.859 and log(CSA) = -2.594 } KMC 3301.32 13068.35 46.70 649.53 0.01 2.94 0.17 306.40 3183.62 2.52 The points A–E are several typical examples of OBCs “synthesized”, “characterized”, and identiﬁed by the IMT. In the master plot (Figure 2 and Table 1), different classes of OBCs are distinguished, which are yielded from a feed having a monomer molar ratio (MR) of 0.75 and different catalyst compositions (CCs) and CSA levels. These grades are different from each other considering their dissimilar blocky natures. 3. Results and Discussion CSA, chain-shuttling agent; DP, average degree of polymerization; ESL, ethylene sequence length; LES, longest ethylene sequence length. Figure 2. A “ﬁngerprint” of chain-shuttling copolymerization decoded by the IMT (which visualizes different classes of OBCs all having a monomer molar ratio of 0.75, but a different blocky nature). Values (50 to 5) appearing at the bottom of the plot close to the black curves represent copolymers having a speciﬁed hard block percent (HB%). CSA, chain-shuttling agent; DP, average degree of polymerization; ESL, ethylene sequence length; LES, longest ethylene sequence length. Figure 2. A “fingerprint” of chain-shuttling copolymerization decoded by the IMT (which visualizes different classes of OBCs all having a monomer molar ratio of 0.75, but a different blocky nature). Values (50 to 5) appearing at the bottom of the plot close to the black curves represent copolymers having a specified hard block percent (HB%). CSA, chain-shuttling agent; DP, average degree of polymerization; ESL, ethylene sequence length; LES, longest ethylene sequence length. Figure 2. A “ﬁngerprint” of chain-shuttling copolymerization decoded by the IMT (which visualizes different classes of OBCs all having a monomer molar ratio of 0.75, but a different blocky nature). Values (50 to 5) appearing at the bottom of the plot close to the black curves represent copolymers having a speciﬁed hard block percent (HB%). CSA, chain-shuttling agent; DP, average degree of polymerization; ESL, ethylene sequence length; LES, longest ethylene sequence length. Table 1 summarizes the randomly selected points in Figure 2 representing five OBCs and their topology- and property-related characteristics yielded by the IMT and KMC operators. It can be recognized that the IMT has appropriately learned the polymerization behavior and successfully predicted both types of molecular characteristics of randomly selected OBCs, as ‘approved’ by the KMC simulator. A closer look at the statistics in Table 1, bearing in mind the very low errors in predictions of the IMT, assures that such an intelligent machine can reliably be applied in ‘anticipating’ two categories of molecular features, leading to sophisticated architecture-property correlations. The IMT consists of 10 individual prediction mechanisms, each responsible for one molecular characteristic of the OBC, which are integrated into an intelligent toolbox towards the prediction of molecular characteristics of OBCs with ‘surgical’ precision. 3. Results and Discussion The power of the IMT (X→Y) in modeling the microstructure of OBCs has already been recently conﬁrmed by visualizing new grades of OBCs [22]. The “ﬁngerprint” of chain-shuttling copolymerization is envisioned using the developed IMT in terms of architecture-related (LP, DPnSOFT, DPnHARD, ESLSOFT, and ESLHARD) and property-related (LESSOFT, LESHARD, C8%SOFT, C8%HARD, and HB%) characteristics in a phase diagram like that shown in Figure 2. Figure 2 is like an operational ‘map’ for or ‘window’ into the process. Three-dimensional plots of hard and soft segments for each microstructural feature, for instance DPnSOFT and DPnHARD, are plotted ﬁrst and subsequently intersected and projected onto a two-dimensional surface. Then, the intersections are determined, at which the studied speciﬁc microstructural characteristic takes on the same values for soft and hard blocks, and separately marked by different colors, e.g., the thick blue line denotes the intersection of soft and hard blocks’ lengths (DPnSOFT = DPnHARD). Obviously, the contribution of hard and soft characteristics is inversely changing in the ‘shadowed’ and ‘unshadowed’ areas (see the legends of Figure 2). For instance, the area distinguished by blue color denotes DPnSOFT > DPnHARD, while the uncolored area denotes DPnSOFT<DPnHARD. By putting all these microstructural information pieces together, a master diagram is obtained. Such a master diagram provides a fast survey of the architecture-property topological trends. For a more detailed look at such relationships, i.e., a deeper understanding of changes in the blocky nature of architecture-related and property-related microstructural features of OBCs, and for ﬁne-tuning such characteristics, one can “zoom in” to the three-dimensional plots or corresponding contour plots. For instance, ﬁve points, A, B, C, D, and E, 6 of 13 Polymers 2019, 11, 579 in the master plot are randomly considered and their polymerization recipes are read from the master plot. Eventually, the KMC simulator gives microstructural features under the conditions proposed by the IMT, corroborating the authenticity of predictions based on the Artiﬁcial Intelligence approach. Polymers 2018, 10, x FOR PEER REVIEW 6 of 13 Figure 2. A “fingerprint” of chain-shuttling copolymerization decoded by the IMT (which visualizes different classes of OBCs all having a monomer molar ratio of 0.75, but a different blocky nature). Values (50 to 5) appearing at the bottom of the plot close to the black curves represent copolymers having a specified hard block percent (HB%). 3. Results and Discussion The x-axis of Figure 2 refers to the ratio of the two different catalysts, where a low catalyst composition means having more of the catalyst-producing hard blocks. The log(CSA) level on the y-axis refers to the amount of chain-shuttling agent in the reaction and, thus, to the probability of shuttling events. As can be observed, 10 key characteristics (see Table 1) determining the blocky nature of OBCs are compared in Figure 2. All these together provide an OBC with an “identiﬁcation card”, which is unique to this special recipe. Exploiting these unique capabilities of the IMT, it is possible to develop new grades of OBC suited to different applications. 7 of 13 Polymers 2019, 11, 579 Table 1. Typical OBCs randomly selected in the ‘master plot’ of the IMT and synthesized by the KMC simulator under speciﬁed operating conditions together with their topology- and property-related molecular characteristics. 3. Results and Discussion Topology-Microstructural Characteristics Property-Microstructural Characteristics DPnSOFT DPnHARD ESLSOFT ESLHARD LP C8%SOFT C8%HARDLESSOFT LESHARD HB% Point A: {MR = 0.75 and CC = 0.539 and log(CSA) = 0.313} KMC 172.88 30.65 25.10 28.92 36.51 3.55 0.70 96.46 55.19 13.30 ANN 169.31 122.07 24.67 31.76 37.26 3.12 0.51 87.49 61.51 14.17 Error % 0.01 2.45 0.92 0.34 1.01 0.88 1.07 2.61 0.14 1.53 Point B: {MR = 0.75 and CC = 0.792 and log(CSA) = 1.886} KMC 122.41 92.11 23.63 10.52 64.87 2.17 9.23 72.12 16.45 5.97 ANN 160.86 102.50 24.58 13.87 63.72 2.62 8.87 78.00 18.84 6.63 Error % 0.10 0.28 2.04 0.41 1.55 0.92 2.02 1.71 0.05 1.16 Point C: {MR = 0.75 and CC = 0.859 and log(CSA) = −2.594} KMC 3301.32 13068.35 46.70 649.53 0.01 2.94 0.17 306.40 3183.62 2.52 ANN 3555.20 13176.48 47.54 666.29 0.64 3.10 0.25 313.26 3249.55 4.53 Error % 0.67 2.90 1.80 2.03 0.85 0.32 0.45 1.99 1.47 3.52 Point D: {MR = 0.75 and CC = 0.202 and log(CSA) = −2.184} KMC 1873.32 18132.94 20.32 434.77 0.14 5.03 0.26 111.81 1842.66 44.80 ANN 1967.33 18094.88 20.84 456.82 0.33 5.26 0.19 125.13 1927.41 45.15 Error % 0.25 1.02 1.12 2.68 0.26 0.47 0.39 3.87 1.89 0.61 Point E: {MR = 0.75 and CC = 0.229 and log(CSA) = 2.397} KMC 94.37 148.96 23.26 12.54 71.58 6.24 13.02 34.72 9.61 35.05 ANN 67.66 117.61 22.84 16.01 72.82 6.00 12.53 30.50 17.83 35.52 Error % 0.07 0.84 0.90 0.42 1.67 0.49 2.75 1.23 0.18 0.82 LV 5.13 7.12 2.34 2.19 0.00 1.89 0.12 3.29 5.84 1.35 HV 38085.60 3736.87 48.97 825.43 74.28 50.72 17.90 347.30 4480.15 58.38 LV and HV are the lowest and highest values of a given response among simulated scenarios, respectively. MR and CC stand for Monomer Molar Ration and Catalyst Composition, respectively. ANN, artiﬁcial neural network. Property-Microstructural Characteristics Several more results stemming from the master plot (Figure 2) need to be discussed. For instance, while it is counterintuitive that soft blocks (with higher comonomer content) can have a longer ethylene sequence than hard blocks, it becomes immediately obvious that such a possibility exists only for high log(CSA) levels, which in turn leads to very short blocks (few monomer units). 3. Results and Discussion This in combination with the higher concentration of soft-block-producing catalyst leads to much longer soft blocks than hard blocks, which means that the average ethylene sequence in the hard block is short, as it cannot be longer than the block itself. Catalyst compositions close to zero lead to much longer hard than soft blocks and vice versa for CC close to 1, as the CSA causes shuttling of the chain to a random catalyst center, which in the case of CC being very different from 1 means that the block lengths are very different from each other. Thus, the possibilities of producing OBCs with previously unexplored structures as well as exact prediction of both their architecture- and property-related molecular features are good examples of the exceptional capabilities of the IMT. Similar to the case of the IMT, but through a completely different route, the reliability of the IOT has also been tested. For this purpose, two tailored OBCs, OBC1 and OBC2 shown in Table 2, are deﬁned and “animated” using the IOT. OBC1 is a multi-block copolymer in which the weight percent of hard blocks has to take the value of 30; and, at the same time, the average ESL of hard blocks is 10 times bigger than the corresponding parameter deﬁned for the soft blocks. In contrast, OBC2 is a highly uniform OBC having a hard block percent of 45. It is worth mentioning that, in a highly uniform OBC, 1-octene units are distributed evenly in both soft and hard blocks. The objectives assigned to each case, as well as the Pareto optimal fronts representing the best solutions with optimal operational conditions, are provided in Table 2. As can be seen, virtual synthesis of OBC1 and OBC2 required concurrent control of three and seven microstructural characteristics of the copolymer chains, respectively, as can be realized by the deﬁned objectives in each case. The IOT iteratively generates and analyzes billions of solutions to ﬁnd the chain microstructure closest to the target OBC1 and OBC2. 8 of 13 Polymers 2019, 11, 579 Polymers 2019, 11, 579 To validate the authenticity of the IOT results, the optimum polymerization recipes obtained in each case were entered into the KMC simulator to virtually synthesize OBC1 and OBC2 and determine their actual microstructural characteristics. As can be seen in Table 2, errors in predicting target microstructural features are below 3.0% for almost all quantities. 3. Results and Discussion Objective 3: 45 % − HB MIN Optimum Input Variables Optimum Input Variables MR CC Log(CSA Level) 0.21389 0.44755 -0.79549 MR CC Log(CSA Level) 0.44058 0.29194 -1.41288 Optimum Responses Optimum Responses IOT KMC Error (%) IOT KMC Error (%) C8%SOFT 46.943 5 44.0155 5.99667 C8%SOFT 16.122 1 19.2357 6.37665 C8%HARD 3.6627 4 3.73534 0.40827 C8%HARD 1.2874 3 1.14977 0.77419 HB% 30.000 0 30.3236 0.56760 HB% 44.999 4 44.7886 0.36974 ESLSOFT 2.7857 7 2.53169 0.54488 ESLSOFT 5.3312 7 5.58988 0.5546 ESLHARD 27.873 5 29.5826 0.20761 ESLHARD 85.922 4 89.3081 0.41127 DPnSOFT 327.71 9 310.251 0.04587 DPnSOFT 678.95 9 692.488 0.03553 DPnHARD 362.89 7 475.781 3.02657 DPnHARD 2853.9 4 2765.08 2.38255 NC8SOFT 123.61 7 136.605 3.46589 NC8SOFT 126.31 4 133.236 1.84699 0.0 1.0E-3 2.0E-3 3.0E-3 4.0E-3 5.0E-3 6.0E-3 0 1E-5 2E-5 3E-5 4E-5 5E-5 6E-5 Objective 2 Objective 1 Optimum Input Variables Optimum Input Variables MR 0.213890 MR 0.44058 CC 0.44755 CC 0.29194 Log(CSA Level) −0.79549 Log(CSA Level) −1.41288 Optimum Responses Optimum Responses IOT KMC Error (%) IOT KMC Error (%) C8%SOFT 46.9435 44.0155 5.99667 C8%SOFT 16.1221 19.2357 6.37665 C8%HARD 3.66274 3.73534 0.40827 C8%HARD 1.28743 1.14977 0.77419 HB% 30.0000 30.3236 0.56760 HB% 44.9994 44.7886 0.36974 ESLSOFT 2.78577 2.53169 0.54488 ESLSOFT 5.33127 5.58988 0.5546 ESLHARD 27.8735 29.5826 0.20761 ESLHARD 85.9224 89.3081 0.41127 DPnSOFT 327.719 310.251 0.04587 DPnSOFT 678.959 692.488 0.03553 DPnHARD 362.897 475.781 3.02657 DPnHARD 2853.94 2765.08 2.38255 NC8SOFT 123.617 136.605 3.46589 NC8SOFT 126.314 133.236 1.84699 NC8HARD 20.5723 17.7899 0.17786 NC8HARD 32.2173 31.8509 0.02342 The microstructural characteristics of OBC1 and OBC2 are shown in Figures 3 and 4, respectively As can be observed, both the evolution and end-of-batch characteristics of the intelligently synthesized OBCs are precisely and thoroughly predicted and depicted. (A) OBC1 (B) OBC2 Objective 1: MIN\|HB% −30\| Objective 1: MIN DP So f t n 1+N So f t C8 ! −ESLSo f t Objective 2: MIN ESLSo f t ESLHard −0.1 Objective 2: MIN DP Hard n 1+N Hard C8 ! −ESLHard NC8: the average number of comonomer units per block. Objective 3: MIN\|HB% −45\| C N  + 8 1 Objective 2: 1.0 −       Hard Soft ESL ESL MIN Objective 2: Hard Hard C Hard n ESL N DP MIN −         + 8 1 NC8: the average number of comonomer units per block. 3. Results and Discussion Objective 3: 45 % − HB MIN Optimum Input Variables Optimum Input Variables MR CC Log(CSA Level) 0.21389 0.44755 -0.79549 MR CC Log(CSA Level) 0.44058 0.29194 -1.41288 Optimum Responses Optimum Responses IOT KMC Error (%) IOT KMC Error (%) C8%SOFT 46.943 5 44.0155 5.99667 C8%SOFT 16.122 1 19.2357 6.37665 C8%HARD 3.6627 4 3.73534 0.40827 C8%HARD 1.2874 3 1.14977 0.77419 HB% 30.000 0 30.3236 0.56760 HB% 44.999 4 44.7886 0.36974 ESLSOFT 2.7857 7 2.53169 0.54488 ESLSOFT 5.3312 7 5.58988 0.5546 ESLHARD 27.873 5 29.5826 0.20761 ESLHARD 85.922 4 89.3081 0.41127 DPnSOFT 327.71 9 310.251 0.04587 DPnSOFT 678.95 9 692.488 0.03553 DPnHARD 362.89 7 475.781 3.02657 DPnHARD 2853.9 4 2765.08 2.38255 NC8SOFT 123.61 7 136.605 3.46589 NC8SOFT 126.31 4 133.236 1.84699 20 572 32 217 0.0 1.0E-3 2.0E-3 3.0E-3 4.0E-3 5.0E-3 6.0E-3 0 1E-5 2E-5 3E-5 4E-5 5E-5 6E-5 Objective 2 Objective 1 j j Soft C N  + 8 1 Objective 2: 1.0 −       Hard Soft ESL ESL MIN Objective 2: Hard Hard C Hard n ESL N DP MIN −         + 8 1 NC8: the average number of comonomer units per block. 3. Results and Discussion (A) OBC1 (B) OBC2 Objective 1: 30 % − HB MIN Objective 1: Soft S f Soft n ESL DP MIN −     (A) OBC1 (B) OBC2 Objective 1: MIN\|HB% −30\| Objective 1: MIN DP So f t n 1+N So f t C8 ! −ESLSo f t Objective 2: MIN ESLSo f t ESLHard −0.1 Objective 2: MIN DP Hard n 1+N Hard C8 ! −ESLHard NC8: the average number of comonomer units per block. Objective 3: MIN\|HB% −45\| Soft C N  + 8 1 Objective 2: 1.0 −       Hard Soft ESL ESL MIN Objective 2: Hard Hard C Hard n ESL N DP MIN −         + 8 1 NC8: the average number of comonomer units per block. Objective 3: 45 % − HB MIN Optimum Input Variables Optimum Input Variables MR CC Log(CSA Level) 0.21389 0.44755 -0.79549 MR CC Log(CSA Level) 0.44058 0.29194 -1.41288 Optimum Responses Optimum Responses IOT KMC Error (%) IOT KMC Error (%) C8%SOFT 46.943 5 44.0155 5.99667 C8%SOFT 16.122 1 19.2357 6.37665 C8%HARD 3.6627 4 3.73534 0.40827 C8%HARD 1.2874 3 1.14977 0.77419 HB% 30.000 0 30.3236 0.56760 HB% 44.999 4 44.7886 0.36974 ESLSOFT 2.7857 7 2.53169 0.54488 ESLSOFT 5.3312 7 5.58988 0.5546 ESLHARD 27.873 5 29.5826 0.20761 ESLHARD 85.922 4 89.3081 0.41127 DPnSOFT 327.71 9 310.251 0.04587 DPnSOFT 678.95 9 692.488 0.03553 DPnHARD 362.89 7 475.781 3.02657 DPnHARD 2853.9 4 2765.08 2.38255 NC8SOFT 123.61 7 136.605 3.46589 NC8SOFT 126.31 4 133.236 1.84699 0.0 1.0E-3 2.0E-3 3.0E-3 4.0E-3 5.0E-3 6.0E-3 0 1E-5 2E-5 3E-5 4E-5 5E-5 6E-5 Objective 2 Objective 1 Objective 1: 30 % HB MIN Objective 1: Soft C ESL N MIN  + 8 1 Objective 2: 1.0 −       Hard Soft ESL ESL MIN Objective 2: Hard Hard C Hard n ESL N DP MIN −         + 8 1 NC8: the average number of comonomer units per block. 3. Results and Discussion The percentage of error in prediction of microstructural features (highlighted in Turquoise) is deﬁned as the ratio of the difference between the outcomes of the IOT (highlighted in yellow) and KMC (highlighted in green) under the conditions proposed by the IOT (numerator) to the difference between minimum and maximum values of that characteristic for stochastically synthesized macromolecules (denominator) multiplied by 100. Polymers 2018, 10, x FOR PEER REVIEW 8 of 13 under the conditions proposed by the IOT (numerator) to the difference between minimum and maximum values of that characteristic for stochastically synthesized macromolecules (denominator) multiplied by 100. Table 2. The chain microstructure of the OBC1 and OBC2 macromolecules tailored using the IOT. The constraints/objectives used in the multi-objective optimization, Pareto fronts showing the best l ti d ti i t i bl d b th IOT f th th i f OBC1 d OBC2 Polymers 2018, 10, x FOR PEER REVIEW 8 of 13 under the conditions proposed by the IOT (numerator) to the difference between minimum and maximum values of that characteristic for stochastically synthesized macromolecules (denominator) multiplied by 100. Table 2. The chain microstructure of the OBC1 and OBC2 macromolecules tailored using the IOT. The constraints/objectives used in the multi-objective optimization, Pareto fronts showing the best Table 2. The chain microstructure of the OBC1 and OBC2 macromolecules tailored using the IOT. The constraints/objectives used in the multi-objective optimization, Pareto fronts showing the best solution, and optimum input variables proposed by the IOT for the synthesis of OBC1 and OBC2 are provided. Microstructural features controlled by the IOT highlighted in yellow; KMC optimizer outputs obtained at identical optimum polymerization recipes highlighted in green; errors in the IOT predictions highlighted in Turquoise. provided. Microstructural features controlled by the IOT highlighted in yellow; KMC optimizer outputs obtained at identical optimum polymerization recipes highlighted in green; errors in the IOT predictions highlighted in Turquoise. (A) OBC1 (B) OBC2 Objective 1: 30 % − HB MIN Objective 1: Soft Soft Soft n ESL DP MIN −     solution, and optimum input variables proposed by the IOT for the synthesis of OBC1 and OBC2 are provided. Microstructural features controlled by the IOT highlighted in yellow; KMC optimizer outputs obtained at identical optimum polymerization recipes highlighted in green; errors in the IOT predictions highlighted in Turquoise. 3. Results and Discussion Objective 3: 45 % − HB MIN Optimum Input Variables Optimum Input Variables MR CC Log(CSA Level) 0.21389 0.44755 -0.79549 MR CC Log(CSA Level) 0.44058 0.29194 -1.41288 Optimum Responses Optimum Responses IOT KMC Error (%) IOT KMC Error (%) C8%SOFT 46.943 5 44.0155 5.99667 C8%SOFT 16.122 1 19.2357 6.37665 C8%HARD 3.6627 4 3.73534 0.40827 C8%HARD 1.2874 3 1.14977 0.77419 HB% 30.000 0 30.3236 0.56760 HB% 44.999 4 44.7886 0.36974 ESLSOFT 2.7857 7 2.53169 0.54488 ESLSOFT 5.3312 7 5.58988 0.5546 ESLHARD 27.873 5 29.5826 0.20761 ESLHARD 85.922 4 89.3081 0.41127 DPnSOFT 327.71 9 310.251 0.04587 DPnSOFT 678.95 9 692.488 0.03553 DPnHARD 362.89 7 475.781 3.02657 DPnHARD 2853.9 4 2765.08 2.38255 NC8SOFT 123.61 7 136.605 3.46589 NC8SOFT 126.31 4 133.236 1.84699 0.0 1.0E-3 2.0E-3 3.0E-3 4.0E-3 5.0E-3 6.0E-3 0 1E-5 2E-5 3E-5 4E-5 5E-5 6E-5 Objective 2 Objective 1 Optimum Input Variables Optimum Input Variables MR 0.213890 MR 0.44058 CC 0.44755 CC 0.29194 Log(CSA Level) −0.79549 Log(CSA Level) −1.41288 Optimum Responses Optimum Responses IOT KMC Error (%) IOT KMC Error (%) C8%SOFT 46.9435 44.0155 5.99667 C8%SOFT 16.1221 19.2357 6.37665 C8%HARD 3.66274 3.73534 0.40827 C8%HARD 1.28743 1.14977 0.77419 HB% 30.0000 30.3236 0.56760 HB% 44.9994 44.7886 0.36974 ESLSOFT 2.78577 2.53169 0.54488 ESLSOFT 5.33127 5.58988 0.5546 ESLHARD 27.8735 29.5826 0.20761 ESLHARD 85.9224 89.3081 0.41127 DPnSOFT 327.719 310.251 0.04587 DPnSOFT 678.959 692.488 0.03553 DPnHARD 362.897 475.781 3.02657 DPnHARD 2853.94 2765.08 2.38255 NC8SOFT 123.617 136.605 3.46589 NC8SOFT 126.314 133.236 1.84699 NC8HARD 20.5723 17.7899 0.17786 NC8HARD 32.2173 31.8509 0.02342 The microstructural characteristics of OBC1 and OBC2 are shown in Figures 3 and 4, respective As can be observed, both the evolution and end-of-batch characteristics of the intelligently synthesize OBCs are precisely and thoroughly predicted and depicted. DPnSOFT 327.71 9 310.251 0.04587 DPnSOFT 678.95 9 692.488 0.03553 DPnHARD 362.89 7 475.781 3.02657 DPnHARD 2853.9 4 2765.08 2.38255 NC8SOFT 123.61 7 136.605 3.46589 NC8SOFT 126.31 4 133.236 1.84699 DPnSOFT 327.71 9 310.251 0.04587 DPnSOFT 678.95 9 692.488 0.03553 DPnHARD 362.89 7 475.781 3.02657 DPnHARD 2853.9 4 2765.08 2.38255 NC8SOFT 123.61 136 605 3 46589 NC8SOFT 126.31 133 236 1 84699 The microstructural characteristics of OBC1 and OBC2 are shown in Figures 3 and 4, respectively. As can be observed, both the evolution and end-of-batch characteristics of the intelligently synthesized OBCs are precisely and thoroughly predicted and depicted. (A) OBC1 0.0 1.0E-3 2.0E-3 3.0E-3 4.0E-3 5.0E-3 6.0E-3 0 1E-5 2E-5 3E-5 4E-5 5E-5 6E-5 Objective 2 Objective 1 0.0 1.0E-3 2.0E-3 3.0E-3 4.0E-3 5.0E-3 6.0E-3 0 1E-5 2E-5 3E-5 4E-5 5E-5 6E-5 Objective 2 Objective 1 j \| \| Objective 1 .0E-3 3.0E-3 Objective Objective 1 0E 3 3.0E 3 Objective Optimum Input Variables Optimum Input Variables Optimum Input Variables Optimum Input Variables DPnSOFT 327.71 9 310.251 0.04587 DPnSOFT 678.95 9 692.488 0.03553 DPnHARD 362.89 7 475.781 3.02657 DPnHARD 2853.9 4 2765.08 2.38255 NC8SOFT 123.61 7 136.605 3.46589 NC8SOFT 126.31 4 133.236 1.84699 DPnSOFT 327.71 9 310.251 0.04587 DPnSOFT 678.95 9 692.488 0.03553 DPnHARD 362.89 7 475.781 3.02657 DPnHARD 2853.9 4 2765.08 2.38255 NC8SOFT 123.61 136 605 3 46589 NC8SOFT 126.31 133 236 1 84699 The microstructural characteristics of OBC1 and OBC2 are shown in Figures 3 and 4, respectively. As can be observed, both the evolution and end-of-batch characteristics of the intelligently synthesized OBCs are precisely and thoroughly predicted and depicted. 9 of 13 and 4, of the Polymers 2019, 11, 579 The microstru respectively As c respectively. As can be observed, both the evolution and end of batch characteristics of the intelligently synthesized OBCs are precisely and thoroughly predicted and depicted. Figure 3. The identification card of intelligently synthesized OBC1 demonstrating all instantaneous and cumulative microstructural features of virtually produced chains. Figure 3. The identiﬁcation card of intelligently synthesized OBC1 demonstrating all instantaneous and cumulative microstructural features of virtually produced chains. intelligently synthesized OBCs are precisely an thoroughly predicted and depicted. Figure 3. The identification card of intelligently synthesized OBC1 demonstrating all instantaneous and cumulative microstructural features of virtually produced chains. Figure 3. The identiﬁcation card of intelligently synthesized OBC1 demonstrating all instantaneous and cumulative microstructural features of virtually produced chains. 10 of 13 1 f 13 10 of 13 10 f 13 Polymers 2019, 11, 579 P l 2018 10 FOR Polymers 2019, 11, 579 10 of 13 Polymers 2018, 10, x FOR PEER REVIEW 10 of 13 Figure 4. The identification card of intelligently synthesized OBC2 demonstrating all instantaneous and cumulative microstructural features of virtually produced chains. Figure 4. The identiﬁcation card of intelligently synthesized OBC2 demonstrating all instantaneous and cumulative microstructural features of virtually produced chains. Based on these ﬁndings, it is now possible to precisely model and optimize the synthesis of Figure 4. (A) OBC1 The identification card of intelligently synthesized OBC2 demonstrating all instantaneous d l ti i t t l f t f i t ll d d h i Figure 4. The identiﬁcation card of intelligently synthesized OBC2 demonstrating all instantaneous and cumulative microstructural features of virtually produced chains. y p Based on these findings, it is now possible to precisely model and optimize the synthesis of olefinic block copolymers by using the developed intelligent tools. The IOT and IMT make it possible to determine the features required to synthesize a polymer with the desired molecular structure or to visualize the factor space for an overview of the possible combinations of molecular Based on these ﬁndings, it is now possible to precisely model and optimize the synthesis of oleﬁnic block copolymers by using the developed intelligent tools. The IOT and IMT make it possible to determine the features required to synthesize a polymer with the desired molecular structure or to visualize the factor space for an overview of the possible combinations of molecular features as a function of polymerization variables. features as a function of polymerization variables. The approach introduced herein is obviously bound to the specific catalyst/chain-shuttling system for producing OBCs. However, it is clear that the method works sufficiently well for OBC th i hi h i f th t l th f ti ki ti i t f i The approach introduced herein is obviously bound to the speciﬁc catalyst/chain-shuttling system for producing OBCs. However, it is clear that the method works sufﬁciently well for OBC synthesis, which is one of the most complex syntheses from a reaction kinetics point of view. Considering this, 11 of 13 Polymers 2019, 11, 579 the IMT and IOT can readily be adapted to many other polymerization schemes by modifying the reaction kinetics model in the KMC simulator. Furthermore, any scheme that links any given (input) X to any given (response) Y could be modeled by a modiﬁed version of the KMC simulator. 4. Conclusions and P.Z.; Data Curation, Y.M.; Writing—Original Draft Preparation, Y.M. and M.R.S.; Writing—Review and Editing, Y.M., M.R.S., A.P., E.J., F.J.S., and P.Z.; Visualization, Y.M.; Supervision, A.P., F.J.S., and K.M.; Project Administration, Y.M., A.P., and K.M. Author Contributions: Conceptualization, Y.M., A.P., E.J., and F.J.S.; Methodology, Y.M., A.P., and K.M.; Software, Y.M.; Validation, Y.M., P.Z., and K.M.; Formal Analysis, Y.M., A.P., E.J., and P.Z.; Investigation, Y.M., M.R.S., E.J., F.J.S., and P.Z.; Resources, Y.M. and P.Z.; Data Curation, Y.M.; Writing—Original Draft Preparation, Y.M. and M.R.S.; Writing—Review and Editing, Y.M., M.R.S., A.P., E.J., F.J.S., and P.Z.; Visualization, Y.M.; Supervision, A.P., F.J.S., and K.M.; Project Administration, Y.M., A.P., and K.M. Funding: This research received no external funding. Funding: This research received no external funding. Conﬂicts of Interest: The authors declare no conﬂict of interest. Author Contributions: Conceptualization, Y.M., A.P., E.J., and F.J.S.; Methodology, Y.M., A.P., and K.M.; Software, Y.M.; Validation, Y.M., P.Z., and K.M.; Formal Analysis, Y.M., A.P., E.J., and P.Z.; Investigation, Y.M., M.R.S., E.J., F.J.S., and P.Z.; Resources, Y.M. and P.Z.; Data Curation, Y.M.; Writing—Original Draft Preparation, Y.M. and M.R.S.; Writing—Review and Editing, Y.M., M.R.S., A.P., E.J., F.J.S., and P.Z.; Visualization, Y.M.; Supervision, A.P., F.J.S., and K.M.; Project Administration, Y.M., A.P., and K.M. 4. Conclusions Two powerful tools were introduced and successfully implemented to model and optimize the microstructural aspects of complex macromolecules. The newly developed computational techniques were established based on hybridization of molecular simulation approaches and Machine Learning techniques. The strategy made it possible to construct intelligent modeling and optimization tools capable of learning and decision-making. Undoubtedly, when applying these tools, polymer reaction engineers not only can effectively discover the complex inter-relationships between polymerization conditions and ﬁnal architectural characteristics, but will also have the opportunity to adjust rather precisely the polymerization inputs in an attempt to synthesize predeﬁned microstructures in detail. Chain-shuttling coordination copolymerization, an intricate polymerization system, has been chosen as the ﬁrst test case to challenge the proposed intelligent modeling and optimization tools. The results obtained clearly showed that the IMT was capable of meticulously patterning the molecular landscape of OBCs in terms of operating conditions, including monomer molar ratio, catalyst composition, and CSA level. IMT was effectively put into practice to ‘crack’ the inter-relationship between operating conditions and micro-molecular characteristics and/or ﬁnal properties of interest. By superimposing all microstructural information pieces together, a master diagram is obtained that provides a fast survey of the recipe-architecture-property topological trends. In contrast, the IOT was able to accurately predict the input/operating factors in response to predeﬁned micro-molecular/architectural characteristics of the target OBC chains. To precisely evaluate the accuracy and performance of the proposed IOT, two target OBCs were designed ﬁrst. Then, the IOT was implemented and applied to concurrently optimize three and seven molecular characteristics of the predeﬁned OBC chains, respectively. The results obtained demonstrated that the IOT was able to successfully handle multi-objective optimization problems and simultaneously control various micro-molecular/architectural characteristics, resulting in negligible errors calculated for the objective functions. Although the unique capabilities of the proposed techniques were successfully tested and examined with a complex living shuttling coordination copolymerization case study, they can be employed by both academic and industrial experts to model and optimize all types of macromolecular reactions and other reactive systems. Both the IMT and IOT, as intelligent computational tools, have the potential to guide polymer chemists and engineers towards the realization of advanced ‘living and thinking’ materials. Author Contributions: Conceptualization, Y.M., A.P., E.J., and F.J.S.; Methodology, Y.M., A.P., and K.M.; Software, Y.M.; Validation, Y.M., P.Z., and K.M.; Formal Analysis, Y.M., A.P., E.J., and P.Z.; Investigation, Y.M., M.R.S., E.J., F.J.S., and P.Z.; Resources, Y.M. 1. Sun, H.; Kabb, C.P.; Sims, M.B.; Sumerlin, B.S. Architecture-transformable polymers: Reshaping the future of stimuli-responsive polymers. Prog. Polym. Sci. 2019, 89, 61–75. [CrossRef] 2. Lutz, J.-F.; Lehn, J.M.; Meijer, E.W.; Matyjaszewski, K. From precision polymers to complex materials and systems. Nat. Rev. Mater. 2016, 1, 16024. [CrossRef] 3. Lutz, J.-F.; Ouchi, M.; Liu, D.R.; Sawamoto, M. Sequence-Controlled Polymers. Science 2013, 341, 1238149. [CrossRef] References 1. Sun, H.; Kabb, C.P.; Sims, M.B.; Sumerlin, B.S. Architecture-transformable polymers: Reshaping the future of stimuli-responsive polymers. Prog. Polym. Sci. 2019, 89, 61–75. [CrossRef] p p y g y , , [ ] 2. Lutz, J.-F.; Lehn, J.M.; Meijer, E.W.; Matyjaszewski, K. From precision polymers to complex materials and systems. Nat. Rev. Mater. 2016, 1, 16024. [CrossRef] 3. Lutz, J.-F.; Ouchi, M.; Liu, D.R.; Sawamoto, M. Sequence-Controlled Polymers. Science 2013, 341, 1238149. [CrossRef] 12 of 13 Polymers 2019, 11, 579 4. Fawcett, E.W.; Gibson, R.O.; Perrin, M.W.; Paton, J.G.; Williams, E.G. Improvements in or Relating to the Polymerization of Ethylene. Great Britain Patent 471,590A, 6 September 1937. 5. Natta, G. From the Stereospeciﬁc Polymerization to the Asymmetric Autocatalytic Synthesis of Macromolecules. In Nobel Lecture; Elsevier Publishing Company: Amsterdam, the Netherlands, 1963; pp. 27–60. pp 6. Ziegler, K. Consequences and Development of an Invention. In Nobel Lecture; Elsevier Publishing Company: Amsterdam, The Netherlands, 1963; pp. 6–24. 7. Ziegler, K. Alumino-organic Syntheses in the Field of Oleﬁnic Hydrocarbons. Angew. Chem. 1952, 64, 323–329. [CrossRef] 8. Ziegler, K.; Holzkamp, E.; Breil, H.; Martin, H. The Mühlheim Low-pressure Polyethylene Process. Angew. Chem. Int. Ed. 1955, 67, 541–547. [CrossRef] 9. Natta, G. Stereospeciﬁc Catalysis of Isotactic Polymers. Angew. Chem. 1956, 68, 393–403. [CrossRef] 10. Natta, G.; Crespi, G. Properties of Ethylene-Propylene Copolymers and of Terpolymers Containing Unsaturation. J. Polym. Sci. A Gen. Pap. 1962, 61, 83–95. [CrossRef] 11. Natta, G.; Pino, P.; Mazzanti, G.; Giannini, U. A Crystalline Organometallic Complex Containing Titanium and Aluminum. J. Am. Chem. Soc. 1957, 79, 2975–2976. [CrossRef] 2. Ziegler, K.; Breil, H.; Holzkamp, E.; Martin, H. A Process for Preparing High-Molecular Polyethyle Patent DE973626C, 17 November 1953. 13. Lai, S.-Y.; Wilson, J.R.; Knight, G.W.; Stevens, J.C.; Chum, P.-W.S. Elastic Substantially Linear Oleﬁn Polymers. U.S. Patent 5,272,236, 21 December 1993. 14. Brant, P.; Canich, J.A.M.; Dias, A.J.; Bamberger, R.L.; Licciardi, G.F.; Henrichs, P.M. Long Chain Branched Polymers and a Process to Make Long Chain Branched Polymers. International Patent Application. WO 94/07930, 14 April 1994. 15. Arriola, D.J.; Carnahan, E.M.; Hustad, P.D.; Kuhlman, R.L.; Wenzel, T.T. Catalytic Production of Oleﬁn Block Copolymers via Chain Shuttling Polymerization. Science 2006, 312, 714–719. [CrossRef] 16. INFUSE™Oleﬁn Block Copolymers-Product Selection Guide. Available online: http://www.dow.com/ scripts/litorder.asp?ﬁlepath=elastomers/pdfs/noreg/788-08201.pdf (accessed on 4 March 2019). 7. Kuhlman, R.L.; Klosin, J. Tuning Block Compositions of Polyethylene Multi-Block Copolymers by Cata Selection. Macromolecules 2010, 43, 7903–7904. [CrossRef] 18. References Zhang, M.; Karjala, T.W.; Jain, P. Modeling of α-Oleﬁn Copolymerization with Chain-Shuttling Chemistry Using Dual Catalysts in Stirred-Tank Reactors: Molecular Weight Distributions and Copolymer Composition. Ind. Eng. Chem. Res. 2010, 49, 8135–8146. [CrossRef] 19. Mohammadi, Y.; Ahmadi, M.; Saeb, M.R.; Khorasani, M.M.; Yang, P.; Stadler, F.J. A Detailed Model on Kinetics and Microstructure Evolution during Copolymerization of Ethylene and 1-octene: From Coordinative Chain Transfer to Chain Shuttling Polymerization. Macromolecules 2014, 47, 4778–4789. [CrossRef] 20. Saeb, M.R.; Khorasani, M.M.; Ahmadi, M.; Mohammadi, Y.; Stadler, F.J. A Uniﬁed Picture of Hard-soft Segmental Development along Oleﬁn Chain Shuttling Copolymerization. Polymer 2015, 76, 245–253. [CrossRef] 21. Ahmadi, M.; Saeb, M.R.; Mohammadi, Y.; Khorasani, M.M.; Stadler, F.J. A Perspective on Modeling and Characterization of Transformations in the Blocky Nature of Oleﬁn Block Copolymers. Ind. Eng. Chem. Res. 2015, 54, 8867–8873. [CrossRef] 22. Saeb, M.R.; Mohammadi, Y.; Kermaniyan, T.S.; Zinck, P.; Stadler, F.J. Unspoken Aspects of Chain Shuttling Reactions: Patterning the Molecular Landscape of Oleﬁn Multi-block Copolymers. Polymer 2017, 116, 55–75. [CrossRef] 23. Mohammadi, Y.; Saeb, M.R.; Penlidis, A.; Jabbari, E.; Zinck, P.; Stadler, F.J.; Matyjaszewski, K. Intelligent Monte Carlo: A New Paradigm for Inverse Polymerization Engineering. Macromol. Theory Simul. 2018, 27, 1700088. [CrossRef] 24. D’hooge, D.R. In Silico Tracking of Individual Species Accelerating Progress in Macromolecular Engineering and Design. Macromol. Rapid Commun. 2018, 39, 1800057. [CrossRef] [PubMed] 5. Nosengo, N. Can Artiﬁcial Intelligence Create the Next Wonder Material? Nature 2016, 533, 22–25. [Cross 26. Raccuglia, P.; Elbert, K.C.; Adler, P.D.F.; Falk, C.; Wenny, M.B.; Mollo, A.; Zeller, M.; Friedler, S.A.; Schrier, J.; Norquist, A.J. Machine-learning-assisted Materials Discovery Using Failed Experiments. Nature 2016, 533, 73–76. [CrossRef] [PubMed] 13 of 13 13 of 13 Polymers 2019, 11, 579 27. Adams, N.; Schubert, U.S. From Data to Knowledge: Chemical Data Management, Data Mining, and Modeling in Polymer Science. J. Comb. Chem. 2004, 6, 12–23. [CrossRef] [PubMed] 28. AbuOmar, O.; Nouranian, S.; King, R.; Bouvard, J.L.; Toghiani, H.; Lacy, T.E.; Pittman, C.U. Data mining and knowledge discovery in materials science and engineering. Adv. Eng. Inform. 2013, 27, 615–624. [CrossRef] 29. Zhao, X.; Rosen, D.W. A data mining approach in real-time measurement for polymer additive manufacturing process with exposure controlled projection lithography. J. Manuf. Syst. 2017, 43, 271–286. [CrossRef] 30. Mohammadi, Y.; Penlidis, A. Polymerization Data Mining: A Perspective. Adv. Theory Simul. 2018, 1800144. [CrossRef] 31. Mohammadi, Y.; Penlidis, A. References “Optimulation” in Chemical Reaction Engineering: Oxidative Coupling of Methane as a Case Study. Ind. Eng. Chem. Res. 2018, 57, 8664–8678. [CrossRef] 32. Azari, R.; Garshasbi, S.; Amini, P.; Rashed-Ali, H.; Mohammadi, Y. Multi-objective Optimization of Building Envelope Design for Life Cycle Environmental Performance. Energy Build. 2016, 126, 524–534. [CrossRef] 33. Garshasbi, S.; Kurnitski, J.; Mohammadi, Y. A Hybrid Genetic Algorithm and Monte Carlo Simulation Approach to Predict Hourly Energy Consumption and Generation by a Cluster of Net Zero Energy Buildings. Appl. Energy 2016, 179, 626–637. [CrossRef] 34. Hosseinnezhad, M.; Saeb, M.R.; Garshasbi, S.; Mohammadi, Y. Realization of Manufacturing Dye-sensitized Solar Cells with Possible Maximum Power Conversion Efﬁciency and Durability. Sol. Energy 2017, 149, 314–322. [CrossRef] 35. Baghaei, B.; Saeb, M.R.; Jafari, S.H.; Khonakdar, H.A.; Rezaee, B.; Goodarzi, V.; Mohammadi, Y. Modeling and Closed-loop Control of Particle Size and Initial Burst of PLGA Biodegradable Nanoparticles for Targeted Drug Delivery. J. Appl. Polym. Sci. 2017, 134, 45145. [CrossRef] 36. Marsaglia, G. Random Number Generators. J. Mod. Appl. Stat. Methods 2003, 2, 2–13. [CrossRef] © 2019 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).
https://openalex.org/W4384079729	https://link.springer.com/content/pdf/10.1007/978-3-031-36911-7_7.pdf	English	null	Gene Order Phylogeny via Ancestral Genome Reconstruction Under Dollo	Lecture notes in computer science	2,023	cc-by	4,622	Keywords: large phylogeny problem · inferring gene-order HTUs · Dollo’s law · gene proximities · maximum weight matching · plant orders Keywords: large phylogeny problem · inferring gene-order HTUs · Dollo’s law · gene proximities · maximum weight matching · plant orders ⃝ ( ) K. Jahn and T. Vinaˇr (Eds.): RECOMB-CG 2023, LNBI 13883, pp. 100–111, 2023. https://doi.org/10.1007/978-3-031-36911-7_7 Department of Mathematics and Statistics, University of Ottawa, 150 Louis Pasteur Pvt., Ottawa, ON K1N 6N5, Canada sankoff@uottawa.ca Abstract. We present a proof of principle for a new kind of step- wise algorithm for unrooted binary gene-order phylogenies. This method incorporates a simple look-ahead inspired by Dollo’s law, while simul- taneously reconstructing each ancestor (sometimes referred to as hypo- thetical taxonomic units “HTU”). We ﬁrst present a generic version of the algorithm illustrating a necessary consequence of Dollo characters. In a concrete application we use generalized oriented gene adjacencies and maximum weight matching (MWM) to reconstruct fragments of mono- ploid ancestral genomes as HTUs. This is applied to three ﬂowering plant orders while estimating phylogenies for these orders in the process. We discuss how to improve on the extensive computing times that would be necessary for this method to handle larger trees. Gene Order Phylogeny via Ancestral Genome Reconstruction Under Dollo Qiaoji Xu and David Sankoﬀ(B) Department of Mathematics and Statistics, University of Ottawa, 150 Louis Pasteur Pvt., Ottawa, ON K1N 6N5, Canada sankoff@uottawa.ca Department of Mathematics and Statistics, University of Ottawa, 150 Louis Pasteur Pvt., Ottawa, ON K1N 6N5, Canada sankoff@uottawa.ca 1 Introduction In formal phylogenetics, Dollo models postulate that a character can only appear once in the course of evolution, although it may disappear from several descen- dent lineages. This idea has a number of combinatorial consequences, and sug- gests an algorithmic inference of phylogeny that diﬀers signiﬁcantly from stan- dard approaches. In this paper, we present a proof of principle for such an algo- rithm in the context of unrooted binary trees. Our approach is basically hierarchical, agglomerative, combining pairs of input taxa or already constructed subtrees but, crucially, does make use, via Dollo, of a limited amount of information from outside these pairs. The input to the generic version of our algorithm consists of n taxa, some- times referred to as observed taxonomic units - “OTUs”, corresponding to the terminal vertices of the phylogeny to be constructed, each of which represented by a set of distinct characters, where the sets may overlap with each other to varying degrees. The strategy is to construct n −2 output sets, each contain- ing some of the same characters, corresponding to the non-terminal vertices, c ⃝The Author(s) 2023 Gene Order Phylogeny via Ancestral Genome Reconstruction Under Dollo 101 or HTUs, of a binary tree constructed at the same time, such that the Dollo condition, or at least some of its important consequences, is maintained. In contrast to the generic version of the algorithm, for any speciﬁc formulation of a Dollo model, the algorithm must be modiﬁed so that the output sets conform to the requirements of the problem at hand. These modiﬁcations generally mean that Dollo is no longer a suﬃcient condition, but remains a necessary condition for a character to be included in an output set. In the main part of this paper, the elements of an input set are all the proxim- ities between oriented genes in one genome. The elements of each output set are chosen from a very large number of proximities, namely those satisfying the Dollo condition, but narrowed down to include only those consistent with an optimal linear ordering, i.e. fragments of chromosome. We have previously invoked these concepts in studying the “small phylogeny problem”, where the tree topology is given (e.g., [1–5]), but here we concentrate on the “large phylogeny problem”, where we actually construct this phylogeny. We apply our method to three plant orders to compare the results with known phylogenies, with almost total agreement. 1 Introduction Finally we discuss possible improvements in eﬃciency and extensions beyond unrooted binary branching phylogenies. 2 Dollo’s Law in the Context of Unrooted Binary Trees The idea that a character is gained only one time and can never be regained if it is lost is realized in an unrooted tree by the property that the set of vertices containing the character are connected. This is a necessary and suﬃcient condi- tion, valid both for terminal vertices (or degree 1) and internal (ancestral) ones (degree 3 in an unrooted binary tree). Fig. 1. Necessary condition for characters to appear at internal vertex of binary branch- ing tree. Light shaded character (small square) appears in all three trees (triangles) subtended by the internal vertex (circle). Dark shaded character appears in only two of the trees. Unshaded character appears in only one subtree so does not aﬀect inter- nal vertex. The shaded characters are “phylogenetically validated” with respect to the internal vertex. The unshaded one is not validated. Fig. 1. Necessary condition for characters to appear at internal vertex of binary branch- ing tree. Light shaded character (small square) appears in all three trees (triangles) subtended by the internal vertex (circle). Dark shaded character appears in only two of the trees. Unshaded character appears in only one subtree so does not aﬀect inter- nal vertex. The shaded characters are “phylogenetically validated” with respect to the internal vertex. The unshaded one is not validated. Fig. 1. Necessary condition for characters to appear at internal vertex of binary branch- ing tree. Light shaded character (small square) appears in all three trees (triangles) subtended by the internal vertex (circle). Dark shaded character appears in only two of the trees. Unshaded character appears in only one subtree so does not aﬀect inter- nal vertex. The shaded characters are “phylogenetically validated” with respect to the internal vertex. The unshaded one is not validated. 102 Q. Xu and D. Sankoﬀ Formally, the connectedness condition can be satisﬁed by a set of non- terminal nodes of a tree. For phylogenetic inference, however, we require that a character be present in the input set of at least two terminal vertices, i.e., visible to the observer. Moreover, if it were only in one terminal set, it could not be necessary presence for any of the output sets at the non-terminal vertices. In an unrooted binary branching tree, each non-terminal vertex subtends three subtrees, as in Fig. 1. 1. input n sets of characters 2. i = 1 1. input n sets of characters 2. i = 1 3. all n sets are “eligible” vertices 3. all n sets are “eligible” vertices 4. while i ≤n −3 4. while i ≤n −3 (a) *for each pair (G, H) of the m = n −i + 1 eligible vertices calculate a potential ancestral vertex A containing all phylogenetically validated characters (in both G and H, or in one of G or H plus any other eligible vertex). ) (b) pick the pair (G′, H′) with maximum total number of the characters in their potential ancestor A. (b) pick the pair (G′, H′) with maximum total number of the characters in their potential ancestor A. (c) ancestor vertex A becomes eligible, and the other two, G′ and H′, become ineligible (c) ancestor vertex A becomes eligible, and the other two, G′ and H′, become ineligible (d) two edges of the tree are deﬁned: AG′ and AH′ ( ) g (e) i = i + 1 (e) i = i + 1 ( ) 5. Now i = n−3, so that there are three eligible vertices G′, H′, K′. Deﬁne three edges of the tree: AG′, AH′ and AK′ 5. Now i = n−3, so that there are three eligible vertices G′, H′, K′. Deﬁne three edges of the tree: AG′, AH′ and AK′ g , 6. calculate ancestral vertex A containing all phylogenetically validated charac- ters (in any two or all three of G′, H′ and K′) 6. calculate ancestral vertex A containing all phylogenetically validated charac- ters (in any two or all three of G′, H′ and K′) ( ) 7. output all 2n−3 tree edges and all n−2 selected output (ancestral, or HTU) sets. 7. output all 2n−3 tree edges and all n−2 selected output (ancestral, or HTU) sets. The asterisked step 4a may be interpreted in at least two diﬀerent ways. In requiring that a character be present in a subtree, we may mean that character must be present in some input set associated with a terminal vertex of that subtree, or we may require something stronger, that the character be present in the eligible (ancestral) vertex of that subtree. 2 Dollo’s Law in the Context of Unrooted Binary Trees For a data set to be used in constructing a phyloge- netic tree and the output sets, clearly each character must be present in at least two of the three subtrees, as illustrated in Fig. 1. More precisely, each character must be present at least in one terminal vertex set in at least two of the three subtrees. We call these characters “phylogenetically validated”. 3 Generic Algorithm 1. input n sets of characters 2. i = 1 1. input n sets of characters 2. i = 1 In constructing a potential ancestor A of two input sets G and H, we deﬁne a Dollo character to be one in either G or H, but not both, as well as in some eligible vertex other than G or H. Our sketch of the generic algorithm is meant to illustrate how the Dollo principle allows a kind of look-ahead in the hierarchical construction of the Gene Order Phylogeny via Ancestral Genome Reconstruction Under Dollo Gene Order Phylogeny via Ancestral Genome Reconstruction Under Dollo 103 Fig. 2. The enclosed area contains input vertices. Outside the enclosed area are ances- tors. Five of the input vertices have already been joined to an ancestor. These are shaded, as is one ancestor that itself been joined to an earlier ancestor. Eligible ver- tices, either input or ancestral, are not shaded. The dotted lines show an attempt to join an input vertex to an ancestor, with one dotted line potentially linking a Dollo character. Fig. 2. The enclosed area contains input vertices. Outside the enclosed area are ances- tors. Five of the input vertices have already been joined to an ancestor. These are shaded, as is one ancestor that itself been joined to an earlier ancestor. Eligible ver- tices, either input or ancestral, are not shaded. The dotted lines show an attempt to join an input vertex to an ancestor, with one dotted line potentially linking a Dollo character. phylogeny (Fig. 2). Without some modiﬁcation, however, the algorithm can result in some counter-intuitive results. Consider the input sets (1, 2), (3, 4), (1, 3), (2, 4), where all four characters are also in other sets. The largest potential ancestor (1, 2, 3, 4) is constructed by pairing (1, 2) with (3, 4), or pairing (1, 3) with (2, 4). Note that this ancestor is based entirely on Dollo characters. y The grouping of (1, 2) with (3, 4) in this construction is not intuitively sat- isfying from the phylogenetic viewpoint since these two input sets have nothing in common. This suggests down-weighting the Dollo characters. For example, if we assign weight 1/2 to Dollo characters compared to weight 2 to a character in both G and H, the potential ancestor of (1, 2) and (3, 4) only has weight 2, while the ancestor of (1, 2) and (1, 3) has weight 3. 1. input n sets of characters 2. i = 1 It is the latter that will be chosen by the algorithm if we replace “maximum number” by “maximum weight”. A principled way of assigning weights might be 2 + αq/m for characters in both G and H plus any other eligible vertex, and 1 + βq/m for a character in one of G or H plus any other eligible vertex, where q is the number of other eligible vertices out of m containing the character, and 2β < α < 1. 4 Genomics Case: Generalized Adjacencies of Genes 4.1 Algorithm: Ancestor via Maximum Weight Matching (MWM) .1 Algorithm: Ancestor via Maximum Weight Matching (MWM) In this version of the algorithm, the content of the sets corresponding to the given extant genomes are all the generalized gene adjacencies in the genome. This includes pairs of adjacent genes, taking their orientations or DNA strand- edness into account, but also includes pairs of gene that are not immediate neighbours, but that are separated by at most 7 other genes in the gene order on a chromosome. 104 Q. Xu and D. Sankoﬀ The ancestor genomes also contain adjacencies, not the simple union of the contents of the two daughter genomes, but only the best set of consistent adja- cencies, namely the output of a MWM from among all the adjacencies in the two daughters plus certain adjacencies from other “eligible” genomes. The matching criterion ensures that all the adjacencies in a set associated with an ancestor are compatible with a linear ordering, as in a chromosome, although we are not concerned with actually building chromosomes here. The deﬁnition of orthologous genes in the various input genomes, the con- struction of the sets of adjacency, and the use of MWM are taken from the ﬁrst steps of the Raccroche method [1–5], which is concerned with actual chromo- some reconstruction of the ancestors genomes in a small phylogeny context. This is not our concern here, which is the branching structure of a phylogeny, namely the large phylogeny problem. The MWM analysis was carried out using Joris van Rantwijk’s implementa- tion [6] of Galil’s algorithm for maximum matching in general (not just bipartite) graphs [7], based on Edmonds “blossom” method including the search for aug- menting paths and his“primal-dual” method for ﬁnding a matching of maximum weight [8]. 1. input n extant genomes 2. i = 1 3. all n genomes are “eligible” 4. while i ≤n −3 ( ) f ll i 4. while i ≤n −3 ( ) f ll ( (a) for all pairs (G, H) of the m −i + 1 eligible vertices ﬁnd potential ances- tral genome A as the Maximum Weight Matching of phylogenetically validated generalized adjacencies (in both G and H - weight 2, or in one of G or H plus any other eligible vertex - weight 2, or in both G and H plus any other eligible vertex - weight 3). 5 Application to Phylogenies of Three Plant Orders Detailed references to all the genomes mentioned here are given in reference [5], including access codes for the CDS ﬁles of the genomes we use in the CoGe platform [9,10]. [ ] The phylogenies that serve as validation of our constructs are by and large uncontroversial, based on up-to-date sources, mainly [11–13]. .1 Algorithm: Ancestor via Maximum Weight Matching (MWM) ) (b) pick the pair (G′, H′) with the highest Maximum Weight Matching score ( ) p p ( , ) g g g (c) ancestor genome A becomes eligible, and the other two, G′ and H′ become ineligible. g (d) two edges of the tree are deﬁned: AG′ and AH′ g (d) two edges of the tree are deﬁned: AG′ and AH′ (d) two edges of the tree are deﬁned: AG and AH (e) i = i + 1 ′ ′ ′ ( ) Now i = n−3, so that there are three eligible vertices G′, H′, K′. Deﬁne three edges of the tree : AG′, AH′ and AK′ calculate ancestral vertex A containing all phylogenetically validated adja cencies (in any two or all three of G′, H′ and K′) A down-weighting scheme for Dollo characters may also be adopted here, although we do not consider that further here. The use of MWM is what distinguishes our method from Dollo parsimony and other adjacency methods. MWM ensures that the adjacencies that appear in a vertex set are consistent, that they can all be part of a chromosome, whereas all other phylogenetic methods based on adjacencies do not care that some sets of adjacencies are not consistent with a chromosome-based genome. These methods may base their results on a gene that is adjacent to three or four genes, while MWM will only retain adjacencies that allow a gene to be adjacent to only two other genes. Gene Order Phylogeny via Ancestral Genome Reconstruction Under Dollo 105 5.1 Asterales From the family Asteraceae, we used the published genomes of saﬄower (Carthamus tinctorius) and artichoke (Cynara cardunculu) from the subfamily Carduoideae, lettuce (Lactuca sativa) and dandelion (Taraxacum mongolicum) from the subfamily Cichorioideae, and Mikania micrantha and Stevia rebaudian from the subfamily Asteroideae. Figure 3 and Table 1 show that our method partitioned the six genomes cor- rectly into three groups. Fig 3 Partial phylogeny of the order Asterales (family Asteraceae) as correctly recon- Fig. 3. Partial phylogeny of the order Asterales (family Asteraceae) as correctly recon- structed by our algorithm, with haploid numbers of chromosomes. Labels on interior nodes indicate the algorithm steps at which they were created (cf Table 1). Fig. 3. Partial phylogeny of the order Asterales (family Asteraceae) as correctly recon- structed by our algorithm, with haploid numbers of chromosomes. Labels on interior nodes indicate the algorithm steps at which they were created (cf Table 1). Fig. 3. Partial phylogeny of the order Asterales (family Asteraceae) as correctly recon- structed by our algorithm, with haploid numbers of chromosomes. Labels on interior nodes indicate the algorithm steps at which they were created (cf Table 1). Table 1. Steps in searching for Asterales ancestors Table 1. Steps in searching for Asterales ancestors potential sisters MWM for MWM for MWM for anc 4 potential anc 1 pot. anc 2 potential. anc 3 Saﬄower Artichoke 1844325 Lettuce Dandelion 1767015 1711720 Stevia Mikania 1677330 1648609 1537870 anc3 anc2 anc1 1322071 Lettuce Saﬄower 1752682 Lettuce Stevia 1663257 1646505 Lettuce Artichoke 1814764 Lettuce Mikania 1588424 1576347 (continued) 106 Q. Xu and D. Sankoﬀ Table 1. (continued) Table 1. (continued) potential sisters MWM for MWM for MWM for anc 4 potential anc 1 pot. anc 2 potential. anc 3 Saﬄower Stevia 1631338 Saﬄower Dandelion 1683778 Saﬄower Mikania 1565038 Stevia Dandelion 1590905 1581354 Stevia Artichoke 1685538 Dandelion Artichoke 1741613 Dandelion Mikania 1525367 1518558 Artichoke Mikania 1612480 anc1 Lettuce 1644223 anc1 Stevia 1494957 1467125 anc1 Dandelion 1573684 anc1 Mikania 1414983 1399206 anc2 Stevia 1468319 anc2 Mikania 1395159 anc2 anc1 1374650 5.3 Sapindales While the results on the Asterales and Fagales are heartening, we cannot expect the method to always perform as well. This is illustrated by an analysis of the order Sapindales. From this order, we used the published genomes of cashew (Anacardium occidentale), mango (Mangifera indica) and pistachio (Pistacia vera) from the family Anacardiaceae and maple (Acer catalpifolium), longan (Dimocarpus longan), lychee (Litchi chinensis) and yellowhorn (Xanthoceras sorbifoli) from the family Sapindaceae. As seen in Fig. 5, except for the incorrect placement of yellowthorn, the method separates the two families as expected. Were this genome to be removed, the reconstructed tree would be identical to the known tree, aside from a permu- tation of the species within the Sapindaceae. Indeed, Table 3 shows that at the fourth step in the algorithm, yellowthorn was almost assigned to join the other Sapindaceae. 5.2 Fagales From the order Fagales, we used the published genomes of oak (Quercus robur) and beech (Fagus sylvatica) from the family Fagaceae, birch (Betula platyphylla) and hazelnut (Corylus mandshurica) from the family Betulaceae, and walnut (Juglans regia), pecan (Carya illinoinensis) and hickory (Carya cathayensis) from the family Juglandaceae. Fig. 4. Partial phylogeny of the order Fagales, as correctly reconstructed by our algo- rithm, with haploid numbers of chromosomes. Labels on interior nodes indicate the algorithm steps at which they were created (cf Table 2). Fig. 4. Partial phylogeny of the order Fagales, as correctly reconstructed by our algo- rithm, with haploid numbers of chromosomes. Labels on interior nodes indicate the algorithm steps at which they were created (cf Table 2). Gene Order Phylogeny via Ancestral Genome Reconstruction Under Dollo 107 Gene Order Phylogeny via Ancestral Genome Reconstruction Under Dollo 107 Table 2. Steps in searching for Fagales ancestors potential sisters MWM for MWM for MWM for MWM for ancestor 5 potential anc 1 pot. anc 2 pot. anc 3 pot.anc 4 Hickory Pecan 2280458 Birch Hazelnut 2133010 2130194 anc1 Walnut 1938190 1929418 anc3 anc2 1744607 anc4 Oak Beech 1502714 Walnut Birch 1780904 1777540 Walnut Oak 1432598 1414818 1399566 Walnut Hickory 2138055 Walnut Pecan 2048931 Walnut Beech 1662931 1654141 1623485 Walnut Hazelnut 2026641 2013614 Birch Oak 1474374 1481635 Birch Hickory 1980129 Birch Pecan 1892718 Birch Beech 1735741 1744305 Oak Hickory 1612423 Oak Pecan 1534218 Oak Beech 1390113 1394735 1366291 1332320 Oak Hazelnut 1675695 1677963 Hickory Beech 1851957 Hickory Hazelnut 2244383 Pecan Beech 1772551 Pecan Hazelnut 2148460 Beech Hazelnut 1951082 1952202 anc1 Oak 1443626 1460361 anc1 Beech 1704015 1711078 anc1 Hazelnut 2108715 anc1 Birch 1844003 anc2 Walnut 1817364 anc2 Oak 1473713 1487666 anc2 Beech 1743668 1737292 anc2 anc1 1849853 anc3 Oak 1410577 anc3 Beech 1650427 Table 2. Steps in searching for Fagales ancestors 108 Q. Xu and D. Sankoﬀ Fig. 5. Partial phylogeny of the order Sapindaless with haploid numbers of chromo- somes. Left is the known phylogeny and right is the reconstructed one. Labels on interior nodes at right indicate the algorithm steps at which they were created. Fig. 5. Partial phylogeny of the order Sapindaless with haploid numbers of chromo- somes. Left is the known phylogeny and right is the reconstructed one. Labels on interior nodes at right indicate the algorithm steps at which they were created. 5.2 Fagales Figure 4 and Table 2 show that our method partitioned the seven genomes correctly into three families. 6 Discussion and Conclusions The algorithmic reconstruction of ancient gene orders, and associated phyloge- nies, has a history of over three decades (e.g., [14–17]). The present work diﬀers from all these in that it is situated in a paradigm [1–5] where only the strictly monoploid ancestors in a phylogeny are reconstructed, or strictly linear chro- mosomal fragments, whether or not such genomes actually existed or are just inherent in the basic chromosomal organization within the possibly re-occurring polyploid history of the group. Although our approach builds a hierarchy by combining pairs of OTUs or already constructed subtrees, it is unusual that it does make use, via Dollo, of a Gene Order Phylogeny via Ancestral Genome Reconstruction Under Dollo 109 Gene Order Phylogeny via Ancestral Genome Reconstruction Under Dollo 109 Table 3. Steps in searching for Sapindales ancestors potential sisters MWM for MWM for MWM for MWM for ancestor 5 potential anc 1 pot. anc 2 pot. anc 3 pot.anc 4 Cashew Mango 2114258 Lychee Maple 1909424 1915717 anc2 Longan 1881830 anc3 anc1 1812017 anc4 Yellowhorn Pistachio 177012 Lychee Longan 1839623 1847180 Lychee Yellowhorn 1792086 1718022 Lychee Cashew 1715727 Lychee Pistachio 1881466 1885124 Lychee Mango 1899616 Longan Maple 1741981 1750100 Longan Yellowhorn 1775908 1568607 1561594 Longan Cashew 1566047 Longan Pistachio 1707939 1713967 1699756 Longan Mango 1729214 Maple Yellowhorn 1892491 1652072 Maple Cashew 1652307 Maple Pistachio 1821498 1820282 Maple Mango 1847530 Yellowhorn Cashew 1840431 Yellowhorn Pistachio 1966295 1689715 1694495 1511822 Yellowhorn Mango 1945493 Cashew Pistachio 1702324 Pistachio Mango 1920295 anc1 Lychee 1795812 anc1 Longan 1621378 1634641 anc1 Maple 1734318 anc1 Yellowhorn 1651337 1678486 1636597 anc1 Pistachio 1814383 1830387 1656675 anc2 Yellowhorn 1696079 anc2 Pistachio 1803140 anc2 anc1 1739109 anc3 Yellowhorn 1802826 anc3 Pistachio 1754820 Table 3. Steps in searching for Sapindales ancestors limited amount of information from outside these pairs. This is in eﬀect a very partial look-ahead. limited amount of information from outside these pairs. This is in eﬀect a very partial look-ahead. The goal of this paper was to present evidence that our method can construct accurate or plausible phylogenies, based entirely on comparative gene order. We were not preoccupied with questions of computing time. Indeed, since our Q. Xu and D. Sankoﬀ 110 reconstruction is based on maximum weight matching (MWM) software on very large graphs, it is bound to be computationally expensive. 6 Discussion and Conclusions Moreover, since the current experimental version uses MWM to exhaustively evaluate all possibilities separately at each step in building a hierarchy, only a moderate number of OTUs can be input. There are, however, many possibilities to improving the eﬃciency, by constraining the search space, by branch and bound techniques, by saving a certain number of partial solutions in parallel, and other techniques. References 1. Xu, Q., Jin, L., Zheng, C., Leebens Mack, J.H., Sankoﬀ, D.: RACCROCHE: ancestral ﬂowering plant chromosomes and gene orders based on generalized adjacencies and chromosomal gene co-occurrences. In: Jha, S.K., M˘andoiu, I., Rajasekaran, S., Skums, P., Zelikovsky, A. (eds.) ICCABS 2020. LNCS, vol. 12686, pp. 97–115. Springer, Cham (2021). https://doi.org/10.1007/978-3-030-79290-9 9 2. Xu, Q., Jin, L., Leebens-Mack, J.H., Sankoﬀ, D.: Validation of automated chro- mosome recovery in the reconstruction of ancestral gene order. Algorithms 14(6), 160 (2021) 3. Chanderbali, A.S., Jin, L., Xu, Q., et al.: Buxus and Tetracentron genomes help resolve eudicot genome history. Nat. Communun. 13, 643 (2022). https://doi.org/ 10.1038/s41467-022-28312-w 4. Xu, Q., et al.: Ancestral ﬂowering plant chromosomes and gene orders based on generalized adjacencies and chromosomal gene co-occurrences. J. Comput. Biol. 28(11), 1156–79 (2021) ( ) ( ) 5. Xu, Q., Jin, L., Zheng, C., Zhang, X., Leebens-Mack, J., Sankoﬀ, D.: From com- parative gene content and gene order to ancestral contigs, chromosomes and kary- otypes. Sci. Rep. 13, 6095 (2023) ( ) 6. van Rantwijk, J.: Maximum Weighted Matching (2008). http://jorisvr.nl/article/ maximum-matching 7. Galil, Z.: Eﬃcient algorithms for ﬁnding maximum matching in graphs. ACM Comput. Surv. 18, 23–38 (1986) p , ( ) 8. Edmonds, J.: Paths, trees, and ﬂowers. Can. J. Math. 17, 449–67 (1965) 9. Lyons, E., Freeling, M.: How to usefully compare homologous plant genes and chromosomes as DNA sequences. Plant J. 53, 661–673 (2008) 10. Lyons, E., et al.: Finding and comparing syntenic regions among Arabidopsis and the outgroups papaya, poplar and grape: CoGe with rosids. Plant Physiol. 148, 1772–1781 (2008) ( ) 11. Published Plant Genomes. Usadel lab, Forschungszentrum J¨ulich, Heinrich Heine University., D¨usseldorf (2022). https://www.plabipd.de/ ( ) // / 12. Stevens, P.F.: Angiosperm Phylogeny Website. Version 14 (2017). http://www. mobot.org/MOBOT/research/APweb/ 13. Chase, M.W., et al.: An update of the Angiosperm Phylogeny Group classiﬁcation for the orders and families of ﬂowering plants: APG IV. Bot. J. Linn. Soc. 181, 1–20 (2016) ( ) 14. Sankoﬀ, D., Leduc, G., Antoine, N., Paquin, B., Lang, B.F., Cedergren, R.: Gene order comparisons for phylogenetic inference: evolution of the mitochondrial genome. Proc. Natl. Acad. Sci. 15, 6575–9 (1992) ( ) 15. Moret, B.M., Warnow, T.: Advances in phylogeny reconstruction from gene order and content data. Methods Enzymol. Gene Order Phylogeny via Ancestral Genome Reconstruction Under Dollo 16. Hu, F., Lin, Y., Tang, J.: MLGO: phylogeny reconstruction and ancestral inference from gene-order data. BMC Bioinf. 15, 1–6 (2014) 17. Perrin, A., Varr´e, J.S., Blanquart, S., Ouangraoua, A.: ProCARs: progressive reconstruction of ancestral gene orders BMC genomics 16 S5:S6 (2015) References 395, 673–700 (2005) 111 Open Access This chapter is licensed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license and indicate if changes were made. The images or other third party material in this chapter are included in the chapter’s Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the chapter’s Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder.
https://openalex.org/W3153968310	https://pure.uva.nl/ws/files/68944255/s40337_021_00402_y.pdf	English	null	Changes in brain and behavior during food-based decision-making following treatment of anorexia nervosa	Journal of eating disorders	2,021	cc-by	10,624	Citation for published version (APA): Foerde, K., Walsh, B. T., Dalack, M., Daw, N., Shohamy, D., & Steinglass, J. E. (2021). Changes in brain and behavior during food-based decision-making following treatment of anorexia nervosa. Journal of Eating Disorders, 9(1), Article 48. https://doi.org/10.1186/s40337-021-00402-y UvA-DARE (Digital Academic Repository) UvA-DARE (Digital Academic Repository) Changes in brain and behavior during food-based decision-making following treatment of anorexia nervosa Foerde, K.; Walsh, B.T.; Dalack, M.; Daw, N.; Shohamy, D.; Steinglass, J.E. Citation for published version (APA): Foerde, K., Walsh, B. T., Dalack, M., Daw, N., Shohamy, D., & Steinglass, J. E. (2021). Changes in brain and behavior during food-based decision-making following treatment of anorexia nervosa. Journal of Eating Disorders, 9(1), Article 48. https://doi.org/10.1186/s40337-021-00402-y Citation for published version (APA): Foerde, K., Walsh, B. T., Dalack, M., Daw, N., Shohamy, D., & Steinglass, J. E. (2021). Changes in brain and behavior during food-based decision-making following treatment of anorexia nervosa. Journal of Eating Disorders, 9(1), Article 48. https://doi.org/10.1186/s40337-021-00402-y Changes in brain and behavior during food-based decision-making following treatment of anorexia nervosa Karin Foerde1,2, B. Timothy Walsh1,2, Maya Dalack1, Nathaniel Daw3, Daphna Shohamy4 and J E St i l 1,2* Karin Foerde1,2, B. Timothy Walsh1,2, Maya Dalack1, Nathaniel Daw3, Daphna Shohamy4 and Joanna E. Steinglass1,2* Joanna E. Steinglass1,2 © The Author(s). 2021 Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated in a credit line to the data. Abstract Background: Anorexia nervosa is a severe illness with a high mortality rate, driven in large part by severe and persistent restriction of food intake. A critical challenge is to identify brain mechanisms associated with maladaptive eating behavior and whether they change with treatment. This study tested whether food choice-related caudate activation in anorexia nervosa changes with treatment. Methods: Healthy women (n = 29) and women hospitalized with anorexia nervosa (n = 24), ages 18 to 40 years, completed a Food Choice Task during fMRI scanning at two timepoints. Among patients, procedures occurred upon hospital admission (Time 1) and again after patients had gained to normal weight (Time 2). Healthy controls were tested twice at an interval group-matched to patients. Choice-related caudate activation was assessed at each timepoint, using parametric analyses in an a priori region of interest. Results: Among patients, the proportion of high-fat foods selected did not change over time (p’s > 0.47), but decreased neural activity in the caudate after treatment was associated with increased selection of high-fat foods (r23 = −0.43, p = 0.037). Choice-related caudate activation differed among women with anorexia nervosa vs healthy control women at Time 1 (healthy control: M = 0.15 ± 0.87, anorexia nervosa: M = 0.70 ± 1.1, t51 = −2.05, p = 0.045), but not at Time 2 (healthy control: M = 0.18 ± 1.0, anorexia nervosa: M = 0.37 ± 0.99, t51 = −0.694, p = 0.49). (Continued on next page) * Correspondence: js1124@cumc.columbia.edu 2Psychiatry Department, Columbia University Irving Medical Center, 1051 Riverside Drive, Unit 98, New York, NY 10032, USA Full list of author information is available at the end of the article 2Psychiatry Department, Columbia University Irving Medical Center, 1051 Riverside Drive, Unit 98, New York, NY 10032, USA Full list of author information is available at the end of the article General rights I i i General rights It is not permitted to download or to forward/distribute the text or part of it without the consent of the author(s) and/or copyright holder(s), other than for strictly personal, individual use, unless the work is under an open content license (like Creative Commons). Disclaimer/Complaints regulations If you believe that digital publication of certain material infringes any of your rights or (privacy) interests, please let the Library know, stating your reasons. In case of a legitimate complaint, the Library will make the material inaccessible and/or remove it from the website. Please Ask the Library: https://uba.uva.nl/en/contact, or a letter to: Library of the University of Amsterdam, Secretariat, Singel 425, 1012 WP Amsterdam, The Netherlands. You will be contacted as soon as possible. Disclaimer/Complaints regulations UvA-DARE is a service provided by the library of the University of Amsterdam (https://dare.uva.nl) Download date:24 Oct 2024 Foerde et al. Journal of Eating Disorders (2021) 9:48 https://doi.org/10.1186/s40337-021-00402-y (2021) 9:48 Foerde et al. Journal of Eating Disorders https://doi.org/10.1186/s40337-021-00402-y © The Author(s). 2021 Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated in a credit line to the data. Page 2 of 12 Foerde et al. Journal of Eating Disorders (2021) 9:48 (2021) 9:48 Foerde et al. Journal of Eating Disorders (Continued from previous page) p p g Conclusions: Caudate activity was more strongly associated with decisions about food among individuals with anorexia nervosa relative to healthy comparison individuals prior to treatment, and decreases in caudate engagement among individuals with anorexia nervosa undergoing treatment were associated with increases in high-fat food choices. The findings underscore the need for treatment development that more successfully alters both eating behavior and the neural mechanisms that guide it. Keywords: Anorexia nervosa, Treatment, fMRI, Neuroscience, Longitudinal, Eating behavior and anxiety [5]. In these structured settings, individuals with AN are able to consume the ~ 4000 kcal per day that are required to gain weight at an appropriate pace. And yet, as soon as treatment incentives are removed after hospital discharge, these same individuals struggle to consume adequate nutrition and often lose weight [6, 7]. In two laboratory meal studies, hospitalized patients with AN showed significant improvement in psycho- logical measures after successful weight restoration, but still consumed significantly fewer calories and less fat than healthy volunteers at both timepoints [8, 9]. The persistence of this maladaptive eating pattern indicates that it is important to understand this behavior. Plain English summary Treatment that leads to full weight restoration in an- orexia nervosa (AN) is accompanied by improvements in psychological symptoms. Yet, restrictive eating does not improve. Prior studies have shown that different neural circuits are engaged when deciding what to eat for pa- tients with AN, compared with healthy controls. Here, we tested whether there were changes in brain activation associated with eating after weight restoration. Overall, food choices did not change with weight restoration, again showing that restrictive eating is resistant to change. However, changes among patients with AN showed a significant association between increased selec- tion of high-fat foods and decreased activity in the caud- ate, with treatment. These findings highlight the importance of specific brain systems related to eating in the treatment of AN. Decision science has recently yielded advances in un- derstanding the behavioral, computational, and neural mechanisms of choices [10–13]. Much of this research has involved food choices, partly for convenience of experimental design, leading to insights of particular relevance for eating disorders. Choice is a complex phenomenon that incorporates multiple cognitive and neural processes, such as attention, valuation, and action selection [14]. In decisions about what to eat, healthiness and tastiness have repeatedly been shown to be distin- guishable values that are integrated into choice selection [12, 15–17]. The neurobiology of food choice has been shown to involve regions associated with value represen- tations ventromedial prefrontal cortex (VMPFC) and self-control dorsolateral prefrontal cortex (DLPFC) [12]. The neural mechanisms of decision-making, and deci- sions about what foods to eat, have received relatively little attention in AN. Yet understanding the neural bases of maladaptive eating behavior may be important in order to improve treatment. Background The first step in the treatment of anorexia nervosa (AN) is weight restoration [1]. Behaviorally-based structured programs incentivizing patients with AN to consume ad- equate calories and normative macronutrient compos- ition (i.e., at least 30% kcal from dietary fat) have high success rates [2]. Yet, relapse and rehospitalization of adults with AN are common, even after full weight res- toration [3]. A central behavioral disturbance that con- tributes to relapse and chronicity of illness is the persistent restriction of food intake [4]. The neural mechanisms that underlie decisions about eating in AN have become a focus of recent research, providing an opportunity to test empirically how such mechanisms are affected by treatment. Increased understanding of the effects of treatment on neural mechanisms under- lying the decision of what to eat may be useful in identi- fying biomarkers and new treatment targets. In two studies from our group, food choices made by adult women with AN were associated with neural activ- ity in the dorsal striatum (specifically, the anterior caud- ate) [18, 19], which differed from healthy peers. This research used a food choice task that has been demon- strated to capture the maladaptive restrictive choices of patients with AN [20, 21] with both the restricting and the binge-eating/purging subtypes [22]. Task-based choices have been related to actual food intake in a la- boratory meal [19], and test-retest reliability of the task has been demonstrated among healthy individuals [23]. In standard treatment programs for AN, renourish- ment is accomplished by providing appropriate nutri- tion, supervision around meals, and behavioral and psychological therapy to promote healthy eating and weight restoration [2]. Success rates for inpatient treat- ment are high in terms of weight normalization, reso- lution of medical instability, and improvement in associated psychological symptoms including depression Page 3 of 12 Foerde et al. Journal of Eating Disorders (2021) 9:48 Foerde et al. Journal of Eating Disorders However, changes in the neural underpinnings of decision-making about food have not been examined after weight restoration treatment. participants were right-handed, between the ages 18–40 years, and had an estimated IQ > 80. Participants were enrolled between November 2015 and August 2018. g Here, we examined whether food choice-related acti- vation in regions of interest identified in previous studies [12, 19, 24], primarily the anterior caudate, VMPFC, and DLPFC changed, with treatment. Participants Participants were female HC and female inpatients with AN at the New York State Psychiatric Institute (NYSPI). Eligible patients met DSM-5 [25] criteria for AN at the time of admission assessed via Eating Disorders Assess- ment for DSM-5 [26]. HC were included if they had no current or lifetime psychiatric illness (including an eat- ing disorder) and had a body mass index (BMI) between 18.5 and 25.0 kg/m2. In both groups, individuals were excluded if they were taking psychotropic medication, had a known history of a neurological disorder or injury, had any contraindication to MRI, or were pregnant. All This study was approved by the NYSPI Institutional Review Board and all participants provided written in- formed consent. Background Patients with AN par- ticipated in a Food Choice Task during fMRI scanning twice, once at admission for inpatient weight restoration treatment, and again after treatment. Change with treat- ment was compared with a group of healthy comparison (HC) women, who were studied twice, at an interval matched to the AN group. We hypothesized that indi- viduals with AN would show greater choice-related acti- vation of the anterior caudate as compared with HC and that this difference between groups would diminish after patients received weight restoration treatment, whereas groups would not differ in choice-related activation in the VMPFC and DLPFC. We additionally explored dif- ferences between groups and changes with treatment in choice-related, as well as tastiness and healthiness rating-related, activation in whole-brain analyses. Procedures occurred twice, once upon admission (Time 1) and once again after weight restoration (Time 2) for individuals with AN, and within 1–2 months for HC, to match the treatment time frame. Inpatient treat- ment at NYSPI is provided at no cost and consists of a behaviorally-based program aimed at full weight restor- ation [2] with a BMI goal between 19.5–20.5 kg/m2. HC were compensated $150 for participation at each time- point ($300 for both). Only data from individuals who completed the Choice phase of the Food Choice Task during neuroimaging at both timepoints were analyzed. Time 1 data from these participants were included in one previous publication of food choice, focused on individuals across a spectrum of restrictive eating [18]. Of 31 HC, two were excluded from analyses due to loss of neuroimaging data (1 scan- ner error, 1 excessive motion) leading to a study sample of 29 HC. Of 26 AN, two were excluded due to a loss of task responses during neuroimaging, leaving a study sample of 24 AN (9 restricting subtype and 15 binge- eating/purging subtype). Two HC were additionally excluded from the Healthiness rating phase (1 due to excessive motion, 1 due to image artifact), and one AN was additionally excluded from the Healthiness and Tastiness rating phase due to excessive motion. Procedure Height was measured on a Seca 240 wall-mounted stadi- ometer. Weight was measured in the morning at each timepoint on a Detecto balance-beam scale. Fig. 1 Food Choice Task. The task consisted of two rating phases and a choice phase. In each phase participants made decisions about 76 food items. On each trial, the food stimulus was presented for 4 s, during which the participant made her response. a Healthiness rating phase. Participants rated the healthiness of each food item on five-point Likert scale from Bad to Good (or Good to Bad, counterbalanced across participants). b Tastiness rating phase. Participants rated the tastiness of each food item on five-point Likert scale from Unhealthy to Healthy (or Healthy to Unhealthy, counterbalanced across participants). c Choice phase. On each trial, participants indicated their preference for a changing food item (shown on the right) relative to a repeated reference item (previously rated neutral on healthiness and tastiness; shown on the left). After task completion, one Choice trial was randomly selected and the participant served a snack-sized portion of the food selected on that trial Fig. 1 Food Choice Task. The task consisted of two rating phases and a choice phase. In each phase participants made decisions about 76 food items. On each trial, the food stimulus was presented for 4 s, during which the participant made her response. a Healthiness rating phase. Participants rated the healthiness of each food item on five-point Likert scale from Bad to Good (or Good to Bad, counterbalanced across participants). b Tastiness rating phase. Participants rated the tastiness of each food item on five-point Likert scale from Unhealthy to Healthy (or Healthy to Unhealthy, counterbalanced across participants). c Choice phase. On each trial, participants indicated their preference for a changing food item (shown on the right) relative to a repeated reference item (previously rated neutral on healthiness and tastiness; shown on the left). After task completion, one Choice trial was randomly selected and the participant served a snack-sized portion of the food selected on that trial Page 4 of 12 Page 4 of 12 Foerde et al. Journal of Eating Disorders (2021) 9:48 Foerde et al. Journal of Eating Disorders (2021) 9:48 Foerde et al. Journal of Eating Disorders harvard.edu/optseq/). Each learning run lasted 480 s. Mean ITI = 2.3 s, median = 2 s, and range = 1–10 s, across all three phases. Data analysis Demographic and clinical characteristics were compared using Student’s t-tests and chi-square, as appropriate. Repeated measures ANOVA were used to test for sig- nificance of change over time within each group. Tests were two-tailed unless otherwise specified. Analyses were conducted in JASP [28] (https://jasp-stats.org), and alpha was set at p = 0.05. Procedure All task phases were presented using Matlab and the Psychophysics toolbox [27]. Participants were given a standardized lunch (4 oz. turkey breast, 2 slices whole wheat bread, 1 packet may- onnaise, Nutrigrain bar, and 8 oz. water, ~ 500 kcal) at noon. At Time 2, for patients with AN, the post- treatment standardized lunch was increased in calories, such that the percent of the daily caloric prescription was the same at both timepoints. At approximately 2 pm, participants were brought to the NYSPI MR Unit for the Food Choice Task with fMRI scanning. At the end of the task, the participant was served a snack based on one of her choices in the task (see below), observed by staff. Procedures at Time 2 were identical. y p y Whole-brain imaging data were acquired on a GE 3 T MR750 scanner with a 32-channel phased-array head coil. Structural images were collected using a high- resolution T1-weighted BRAVO pulse sequence (1 × 1 × 1 mm voxel size) for image registration. Functional im- ages were collected using a gradient echo T2-weighted echoplanar (EPI) sequence with blood oxygenation level- dependent (BOLD) contrast (TR = 2000 ms, TE = 19 ms, flipangle = 77, 3 × 3 × 3 mm voxel size, 45 contiguous axial slices, FOV = 19.2, interleaved acquisition). Each Food Choice Task run consisted of 240 volumes. Food choice task with fMRI (Fig. 1) Food Choice Task procedures have been described pre- viously [19]. The task consisted of three phases: Healthi- ness rating, Tastiness rating, and a Choice block (Fig. 1). In each block, participants rated 76 foods on a 5-point scale. Half of the foods were classified as “high-fat” (> 30% of the calories from fat; images and macronutrient information available at Columbia Academic Commons: https://doi.org/10.7916/d8-497c-2724 [17]. The rating scale appeared at the bottom of the screen for each item and participants were instructed that they could rate it as “neutral” or along the scale. For the Healthiness block, the anchors were “Unhealthy” to “Healthy.” For the Tastiness block, the anchors were “Bad” to “Good.” For the Tastiness block, participants were additionally instructed to rate it only for tastiness. All task parame- ters (including order of the rating scale) were counterba- lanced and randomized across participants. After completion of the rating scales, a “Reference” item was selected for each participant that had been rated by her as “Neutral” on both Healthiness and Tastiness. If no item neutral on both scales was identified, an item was selected that was rated neutral on Healthiness and 1 point in the positive direction on Tastiness, in order to minimize biasing choices based on taste value. In the Choice block, participants made a selection by indicating whether they ‘Strongly Preferred’ or ‘Preferred’ the Ref- erence item (which was constant) or the other food (which varied on each trial) using a 5-point scale. To en- sure that responses reflected true preferences, partici- pants were told that they would be served one of their choices at the conclusion of the task. One Choice trial was randomly selected, and the participant received a snack-sized portion of the item chosen on that trial. Food choice task Rating and choice data were analyzed using multilevel regression models lme4 linear mixed effects package for R [29]. For the Choice phase, responses on the 5-point scale were converted to binary ‘Yes’ (1) or ‘No’ (0) pref- erences for the trial-unique food versus the Reference item, and neutral responses were omitted from analyses. Binomial choice data were modelled with multilevel lo- gistic regression, in which participant choice (selection of the trial-unique food item over the reference food) was the dependent variable. Continuous outcome rating data from the Healthiness and Tastiness phases were modelled using multilevel linear regression. To assess re- lationships between ratings and choices, participant choice (selection of the trial-unique food item over the reference food) was the dependent variable and (z- scored) healthiness and tastiness ratings entered as inde- pendent variables. The relationship between Healthiness and Tastiness was assessed with Healthiness ratings as the dependent variable and (z-scored) Tastiness ratings as the independent variable. In all analyses, Group (HC/ AN, coded as −1/1), Food type (low-fat/high-fat, coded as −1/1), and Time (Time 1/Time 2, coded as −1/1) were entered as independent variables, and models in- cluded by-subject random intercepts and slopes and by- item (food images) random intercepts [30, 31]. In all blocks, the food stimulus was presented for 4 s on each trial, during which the participant made her re- sponse. Each trial was followed by a fixation cross inter- trial interval (ITI). The duration of ITIs was jittered for optimization of event-related fMRI design. Stimulus presentation sequence and timing were optimized using the optseq2 algorithm (http://surfer.nmr.mgh. Correlations between brain, behavior, and demographic measures were assessed using Pearson correlation. Robust regression was implemented as needed due to presence of Page 5 of 12 Page 5 of 12 Foerde et al. Journal of Eating Disorders Foerde et al. Journal of Eating Disorders (2021) 9:48 outlier data points using the rlm function in the MASS package with bisquare weighting [32]. made (i), onsets for each trial on which a response was made parametrically modulated by the (demeaned) rat- ing on that trial (ii) and the (demeaned) response time on that trial (iii), and onsets for missed trials (iv). Re- gressors i-iii were modeled with duration equal to the response time on each trial, and regressor iv with dur- ation equal to the trial length (4 s). Motion and con- found regressors were included as outlined above. fMRI analyses Imaging data were converted from DICOM to NIFTI format and preprocessed and analyzed using the FSL (http://fsl.fmrib.ox.ac.uk/fsl/) package version 6 (FMRIB’s Software Library; Oxford Centre for Functional Resonance Imaging of the Brain, FMRIB) [33]. Image pre-processing Functional images were aligned using the MCFLIRT tool [34] and the six scan-to-scan head motion parameters estimated during motion cor- rection obtained. The skull was removed from functional images using the brain extraction tool (BET) [35] and from structural images using Freesurfer [36, 37]. Spatial smoothing was applied with a Gaussian kernel of 5 mm (FWHM). Data and design matrix were high-pass fil- tered with a cutoff period of 100 s. After analysis at the individual level, the results were normalized to a stand- ard template: Functional images were first aligned to the T1-weighted image using a boundary-based registration method implemented in FSL6 (BBR) and then the struc- tural image to the standard MNI152 2-mm template using FLIRT (12 degrees of freedom) and FNIRT (10- mm warp resolution) [34, 38]. ROI analyses y Based on a previous study indicating involvement of the right anterior caudate in food choices [19], an anatomical region of interest (ROI) was obtained from the Harvard- Oxford probabilistic atlas included in FSL, thresholded at 25% probability and anterior to y = 0 (Fig. 3a). Additional ROIs were included based on their implication in food choices in healthy individuals, with the VMPFC involved in value-based decisions and the DLPFC implicated in self-control [12]. Moreover, VMPFC regions (e.g., OFC) have been implicated in studies of decision making in AN [41] and DLPFC has been used as a target for neuromodu- lation (rTMS) intervention in AN [20]. Six-mm spheres were created centered on MNI coordinates taken from a previous study using the Food Choice task [12]: VMPFC (MNI = [3 51 3]; Fig. 3b) and DLPFC (MNI = [−48 15 24]; Fig. 3c). Analyses At the level of individual participants, each event was convolved with a canonical hemodynamic re- sponse function (except added confound regressors, see below) and entered into a general linear model (GLM). The temporal derivative of each regressor (except added confound regressors) was also included in the model. To account for any residual effects of subject movement, we included the six scan-to-scan head motion parameters estimated during motion correction as well as framewise displacement (FD) and root mean square intensity differ- ence from one volume to the next (DVARS) [39] as con- found regressors. In addition, volumes with FD and DVARS exceeding a threshold of 0.5 were modeled out by adding a single time point regressor for each volume to be ‘dropped’ from analysis [40]. Runs for which more than 25% of volumes were dropped were excluded from analysis: one Tastiness rating run at Time 1 (AN), two Healthiness rating runs at Time 2 (1 HC, 1 AN). The number of dropped volumes did not differ between groups (ps > 0.05, Supplemental Table S1). Whole-brain analyses Whole-brain higher-level analyses were thresholded using clusters determined by Z > 3.1 and a whole-brain corrected, FWE cluster significance threshold of p = 0.05 [42]. For exploratory purposes, we additionally consid- ered analyses thresholded at Z > 2.3, FWE cluster signifi- cance threshold of p = 0.05 [43]. Results are displayed on a study-specific mean ana- tomical image resulting from averaging all participants’ normalized high-resolution structural images. Food choice task Linear contrasts were performed on specific comparisons of interest. These contrasts were used for mixed-effects group analyses using FSL’s FLAME 1 (FMRIB’s local analysis of mixed effects) tool, using two-sample un- paired t-tests to compare groups and paired t-tests to compare changes between Time 1 and Time 2. Results Clinical characteristics of participants are presented in Table 1. Psychological measures differed between indi- viduals with AN and HC at both timepoints, and the AN group showed improvement in all psychological mea- sures and BMI from Time 1 to Time 2. At Time 2, AN and HC groups were well matched on BMI (t51 = 1.29, p = 0.20). The average time between scans was similar in both groups (meanHC: 49 ± 21 days; meanAN: 58 ± 21 days, t51 = −1.4, p = 0.17). Parametric analysis of food choices and ratings were conducted (rating-related results in Supplemental Figure S1) [19]. Each person’s choices were normalized to their own response range; analyses were therefore not biased by overall differences in choice preferences. The GLMs for the choice and rating phases included the following regressors: onsets for each trial on which a response was Foerde et al. Journal of Eating Disorders (2021) 9:48 Foerde et al. Food choice task behavior Across both timepoints Healthiness influenced Choice more among individuals with AN relative to HC (Est = 0.49, z = 3.94, p = 8.2 × 10−5) and Tastiness influenced Choice more among HC relative to individuals with AN (Est = −0.46, z = −3.69, p = 0.0002). There was a Time by Tastiness interaction (Est = 0.17, z = 2.08, p = 0.037), with Tastiness influencing Choice more at Time 2 in both groups (Fig. 2d, Supplemental Table S5). Healthi- ness was more strongly associated with Tastiness among AN relative to HC (Est = 0.13, z = 2.19, p = 0.03; Fig. 2e, Supplemental Table S6) and the relationship between Healthiness and Tastiness ratings did not change over time (ps > 0.25). All participants were able to respond within the re- sponse window, indicated by response rates exceeding 96% in all task phases and in both groups. Response rates did not differ between groups (ps > 0.25). g p p Choices of high-fat and low-fat foods (proportion of trials chosen over the Reference food) did not change significantly between Time 1 and Time 2 (ps > 0.16) (Fig. 2a). On average, the AN group chose high-fat foods less often than the HC group at both time points (Group X Food type interaction: Est = −0.42, z = −4.72, p < 0.00001; Supplemental Table S2; significant effect of Group for high-fat foods: Est = −0.14, z = −4.61, p < 0.0001, but not for low-fat foods: Est = −0.0008, z = − 0.03, p = 0.97). Healthiness ratings did not change significantly with treatment (ps > 0.05, Fig. 2b, Supplemental Table S3). High-fat foods were rated lower in healthiness than low- fat foods overall (Est: −0.75, t80.3 = −6.70, p = 2.6 × 10−9) and significantly more so among individuals with AN (Group X Food type interaction: Est = −0.05, t50.8 = − 2.02, p = 0.048; significant effect of Group for high-fat foods: Est = −0.15, t51 = −2.84, p = 0.006, but not for low-fat foods: Est = −0.04, t51 = −1.15, p = 0.26). Results Journal of Eating Disorders Page 6 of 12 Table 1 Demographics and clinical characteristics of participants Time 1 Time 2 Time 1 versus Time 2 HC (n = 29) AN (n = 24) HC AN HC AN M SD M SD t p M SD M SD t p t p t p Age (years) 25.8 5.2 26.9 6.5 −0.7 0.50 Caucasian (n,%) 22 76% 16 67% 0.55a 0.46 Estimated IQ 117.3 13.4 112.0 9.5 1.7 0.11 LNS 12.0 3.1 11.7 2.4 0.4 0.67 BMI (kg/m2) 21.0 1.4 16.3 1.9 10.3 < .001 21.0 1.6 20.5 0.9 1.4 0.17 −0.33 0.75 −13.4 < .001 Duration of Illness (years) 8.6 6.9 EDE-Q Global 0.4 0.5 4.3 1.7 −12.2 < .001 0.4 0.4 2.8 1.3 −9.1 < .001 −0.54 0.6 6.6 < .001 TFEQ-Restraint 5.9 4.2 17.4 3.6 −10.6 < .001 6.1 4.8 14.2 5.3 −5.9 < .001 −0.43 0.67 3.9 < .001 BDI 2.1 2.3 31.1 13.4 −11.5 < .001 3.1 2.5 17.4 14.4 −5.2 < .001 −1.94 0.06 5.3 < .001 STAI(T) 32.6 6.5 62.9 11.4 −12.2 < .001 32.9 8.2 56.0 12.2 −8.2 < .001 −0.26 0.8 3.6 0.003 aChi square statistic Missing data: BDI is missing from 1 individual with AN AN anorexia Nervosa, BDI beck depression index [44], BMI body mass index, EDE-Q eating disorder examination-questionnaire version [45], HC healthy control, LNS letter number sequence from the Weschler Adult Intelligence Scale [46], STAI(T) Spielberger trait anxiety inventory [47], TFEQ-Restraint Three Factor Eating Questionnaire Restraint subscale [48] Estimated IQ was assessed with the Wechsler Abbreviated Scale of Intelligence [49] Table 1 Demographics and clinical characteristics of participants q Missing data: BDI is missing from 1 individual with AN Missing data: BDI is missing from 1 individual with AN AN anorexia Nervosa, BDI beck depression index [44], BMI body mass index, EDE-Q eating disorder examination-questionnaire version [45], HC healthy control, LNS letter number sequence from the Weschler Adult Intelligence Scale [46], STAI(T) Spielberger trait anxiety inventory [47], TFEQ-Restraint Three Factor Eating Questionnaire Restraint subscale [48] Estimated IQ was assessed with the Wechsler Abbreviated Scale of Intelligence [49] Neuroimaging Anterior caudate During the Choice phase there was no significant inter- action between Group and Time in a rmANOVA (F [1, 51]=1.19, p = 0.28) nor a main effect of Time (F [1, 51]= 0.87, p = 0.36). The main effect of Group approached statistical significance (F [1, 51]=2.997, p = 0.089) and groups differed significantly during Choice at Time 1 (t51 = −2.05, p = 0.045). In contrast, group differences at Time 2 were not significant (t51 = −0.694, p = 0.49; Fig. 3a. During Healthiness and Tastiness ratings there were no effects of Group or Time (ps > 0.05; Supplemen- tal Figure S1A). Tastiness ratings did not change with treatment (Fig. 2c, Supplemental Table S4). Individuals with AN rated high-fat foods as less tasty than did HC (Food type X Group: Est = −0.14, t = −3.23, p = 0.002; Group: Est = −0.13, t = −2.51, p = 0.015; significant effect of Group for high-fat foods: Est = −0.28, t51 = −3.22, p = 0.002, but not for low-fat foods: Est = 0.007, t51 = 0.14, p = 0.89), and this pattern did not change over time (all other ef- fects ps > 0.27). We next examined whether change in caudate activa- tion over time was related to changes in choice behavior in AN. Among individuals with AN, increased choice of high-fat foods was associated with decreased activation of the caudate during the Choice block (r22 = −0.427, Foerde et al. Journal of Eating Disorders (2021) 9:48 Page 7 of 12 Fig. 2 Food choice task behavior at Time 1 and Time 2. a High-fat and low-fat food choices did not change significantly from Time 1 to Time 2 (ps > 0.37). The AN group made fewer high-fat, but not low-fat, food choices before and after treatment (Supplemental Table S2). b Overall low-fat items were rated higher on healthiness than high-fat items. High-fat foods were rated as lower in healthiness than low-fat foods overall and the groups differed significantly (Supplemental Table S3). c The AN group rated high-fat foods specifically lower in tastiness than did the HC group and this did not change with treatment (Supplemental Table S4). d Logistic regression of Healthiness and Tastiness ratings on choice. Choice was influenced more by tastiness among HC than AN, whereas choice was influenced more by healthiness among AN relative to HC (Supplemental Table S5). Neuroimaging Anterior caudate e Regression of Healthiness on Tastiness ratings. Ratings were more strongly associated among AN than HC (Supplemental Table S6). indicates p < 0.05. ^ indicates significant Group difference and significant difference between Time 1 and Time 2 in both groups Fig. 2 Food choice task behavior at Time 1 and Time 2. a High-fat and low-fat food choices did not change significantly from Time 1 to Time 2 (ps > 0.37). The AN group made fewer high-fat, but not low-fat, food choices before and after treatment (Supplemental Table S2). b Overall low-fat items were rated higher on healthiness than high-fat items. High-fat foods were rated as lower in healthiness than low-fat foods overall and the groups differed significantly (Supplemental Table S3). c The AN group rated high-fat foods specifically lower in tastiness than did the HC group and this did not change with treatment (Supplemental Table S4). d Logistic regression of Healthiness and Tastiness ratings on choice. Choice was influenced more by tastiness among HC than AN, whereas choice was influenced more by healthiness among AN relative to HC (Supplemental Table S5). e Regression of Healthiness on Tastiness ratings. Ratings were more strongly associated among AN than HC (Supplemental Table S6). * indicates p < 0.05. ^ indicates significant Group difference and significant difference between Time 1 and Time 2 in both groups were no effects of Group or Time during Healthiness and Tastiness ratings (ps > 0.05; Supplemental Figure S1BC). p = 0.037, robust regression: p = 0.036; Fig. 3a). This re- lationship between neural activation and behavior was not found among the HC (r27 = −0.11, p = 0.57; Fig. 3a). Changes in choices of low-fat foods were not related to caudate engagement in either group (AN: r22 = −0.08, p = 0.70; HC: r27 = −0.02, p = 0.90). Whole-brain analyses to explore areas outside a priori ROIs revealed a group difference at Time 1 in the super- ior parietal lobule/precuneus, with greater choice related engagement in the AN than the HC group (Supplemen- tal Figure 2). To examine changes over time in the par- ietal region, we extracted values at Time 2 from the cluster identified at Time 1. There were no group differ- ences at Time 2 (t51 = −0.15, p = 0.88; Group X Time Interaction: F [1, 51]=6.67, p = 0.013; Fig. 3d). Neuroimaging Anterior caudate For com- pleteness, as for the anterior caudate, we explored the relationship between changes from Time 1 to Time 2 in the parietal region and high-fat food choice changes and found a significant correlation (r22 = −0.476, p = 0.019; robust regression: p = 0.022; Fig. 3d). Changes over time in the anterior caudate and parietal lobe were signifi- cantly correlated in the AN group (r = 0.745, p < 0.001) Prefrontal cortex activity and food choice As in previous studies [12, 19], there was significant choice-related activation in the VMPFC ROI. That acti- vation did not differ between groups (F [1, 51]=0.007, p = 0.93) or across time (F [1, 51]=0.57, p = 0.45) nor was there a significant interaction (F [1, 51]=0.19, p = 0.67; Fig. 3b). We additionally explored whether the DLPFC, previously implicated in the use of self-control for food choices [12], was differentially engaged across groups and time, but found no significant differences (Time: F [1, 51]=0.16, p = 0.69; Group: F [1, 51]=1.32, p = 0.26; Group x Time: F [1, 51]=1.00, p = 0.32; Fig. 3c). There Foerde et al. Journal of Eating Disorders (2021) 9:48 Foerde et al. Journal of Eating Disorders Page 8 of 12 Fig. 3 Choice-related engagement in regions of interest before and after treatment and relationship with eating behavior. a Values extracted from the parametric choice analysis in our a priori anatomical ROI in the right anterior caudate (top panel). Choice-related activation in the caudate differed significantly between HC and AN at Time 1, but not at Time 2 (middle panel). Change in activity in the caudate from Time 1 to Time 2 was significantly correlated with the change in proportion high-fat food choices on the food choice task among individuals with AN, but not HC. Robust regression was used due to the presence of outliers. b Same analyses as in a for the a priori VMPFC ROI (MNI = [3 51 3]; top panel). c Same analyses as in a for the a priori DLPFC ROI (MNI = [−48 15 24]; top panel). d For illustration purposes the same analyses as in a are presented for the parietal region identified in exploratory whole-brain analyses h l d f b f d f d l h h b h l d Fig. 3 Choice-related engagement in regions of interest before and after treatment and relationship with eating behavior. a Values extracted from the parametric choice analysis in our a priori anatomical ROI in the right anterior caudate (top panel). Choice-related activation in the caudate differed significantly between HC and AN at Time 1, but not at Time 2 (middle panel). Change in activity in the caudate from Time 1 to Time 2 was significantly correlated with the change in proportion high-fat food choices on the food choice task among individuals with AN, but not HC. Prefrontal cortex activity and food choice Robust regression was used due to the presence of outliers. b Same analyses as in a for the a priori VMPFC ROI (MNI = [3 51 3]; top panel). c Same analyses as in a for the a priori DLPFC ROI (MNI = [−48 15 24]; top panel). d For illustration purposes the same analyses as in a are presented for the parietal region identified in exploratory whole-brain analyses but not in the HC group (r = 0.338, p = 0.073); the differ- ence in correlation between groups was significant (p = 0.038). restoration. Change over time in this region was also asso- ciated with food-choice change and with caudate change. These findings underscore the association between re- strictive eating and choice-related activation of the caud- ate in AN and suggest this brain-behavior link may be a target for treatment development. Whole-brain analyses of the Choice phase (Supple- mental Figure S2), Healthiness rating phase (Supplemen- tal Figure S3), and Tastiness rating phase (Supplemental Figure S4) did not reveal any other regions that changed between Time 1 and Time 2 or differed between groups. The lack of change in choice of high-fat foods for a snack after ~ 2.5 months of hospital care is disappoint- ing, though not surprising. The challenge of changing eating behavior, even during inpatient treatment, has been demonstrated in laboratory meal studies [8, 9] and is consistent with the high rate of relapse following hos- pital discharge. The ability to select a diet higher in energy density at the end of hospital care is associated with better treatment outcomes 1 year later [4, 50], suggesting that in- terventions that effect change in dietary patterns are likely necessary to decrease relapse. The caudate region impli- cated here is important for action control and choice, in general [51]. Change in maladaptive behavior in AN may depend on therapeutic techniques that focus directly on the difficult problem of changing food choices. A habit re- versal based approach that targeted cues for maladaptive eating routines was successful in changing behavior in AN in a pilot study [52]. Similarly, exposure therapies that emphasize finer-grained focus on maladaptive eating have Discussion This study tested whether neural mechanisms associated with maladaptive food choice in AN change with acute weight restoration. The fMRI findings during a Food Choice Task immediately following hospital admission were consistent with prior results, with greater choice- related engagement of the anterior caudate among indi- viduals with AN. Increased selection of high-fat foods following weight restoration was associated with de- creased choice-related caudate activation. On average, however, patients with AN did not substantially increase their selection of high-fat foods from Time 1 to Time 2, consistent with a prior study demonstrating the persist- ence of dietary restriction after treatment [9]. Additionally, whole brain analyses identified choice-related activation in a parietal/precuneus region that normalized with weight Foerde et al. Journal of Eating Disorders (2021) 9:48 Page 9 of 12 Foerde et al. Journal of Eating Disorders abnormalities have not consistently reported normalization with weight restoration [69, 70]. One task-based fMRI study indicated that learning related re- sponse abnormalities in the striatum normalized after weight restoration [71]. Two studies of monetary decision-making found increases in neural activity in the striatum [72] and the dorsal anterior cingulate after weight restoration [73]. These studies together indicate that the brain is altered by starvation (and by renourish- ment) and that fronto-striatal circuits are among regions that undergo change across multiple studies and modal- ities. The stability of the neural response in the HC group in the present study supports the utility of the longitudinal assessments and strengthens the ability to draw inferences from the changes seen in the AN group. And, unlike prior longitudinal research, the present study directly examined treatment-associated changes in neural mechanisms of restrictive eating - the behavioral disturbance most directly connected to relapse, morbid- ity, and mortality. These findings highlight the value of explicitly linking brain and behavior by using a task demonstrated to capture maladaptive eating behavior in AN. also shown promise [53]. Neuromodulation of the dorsal frontostriatal system using transcranial magnetic stimula- tion (TMS) showed some impact in decreasing excessive self-control over food choices among individuals with chronic AN [20]. Targeting the brain systems identified here may amplify the effects of existing treatments for AN. In addition to the a priori ROI analyses that largely confirmed hypotheses about the caudate in food-based decision making in AN, exploratory whole-brain analyses suggested that the brain-behavior link also includes par- ietal/precuneus regions. Discussion The parietal lobe has been im- plicated in decision making and choice [54, 55]. In addition, parietal cortex is often considered part of a cognitive control network (fronto-cingulo-parietal sys- tem) [56–58], with functions that could be relevant for food choice (e.g., selective attention, response inhibition, etc.). The precuneus has been shown to play a role in a variety of higher cognitive functions, including self- referential processing and episodic memory retrieval [59]. Abnormalities in precuneus activation among indi- viduals with AN have been identified several studies that include food related tasks [60–62] and, notably, is the most highly implicated region associated with the term ‘anorexia nervosa’ on the Neuroquery meta-analysis tool [63]. While the present study design does not allow for inferences about the specific contribution to decision- making, the findings suggest that further investigation of these cognitive processes is warranted. Although, reverse inference of cognitive processes based on brain activity is of limited utility, one hypothesis po- tentially worth investigating based on differential en- gagement of the precuneus among AN and HC, is that individuals with AN rely on memory retrieval differently than do HC; the role of episodic memories in decision making among individuals with AN has not been carefully examined. These cortical regions might also be targets for neuromodulation. Limitations The current results should be viewed in light of several limitations. First, the sample size was modest, and it is possible that there was insufficient power to detect the full effect of treatment on neural circuits. Small sample size is a common limitation in studies of AN and sug- gests a need for standardization of methods such that studies can be pooled. Second, patients were studied at the end of acute inpatient treatment, shortly after full weight restoration. While weight restoration is a neces- sary first step in treatment of AN, it is not associated with full resolution of psychological and behavioral symptoms and may not be the optimal moment to assess brain changes in AN. On the other hand, it has been suggested that neural changes should precede behavioral effects, hopefully permitting understanding of the neural mechanisms of treatment response [74]. Third, examin- ation of brain regions that may become engaged only after weight restoration was limited by a priori regional hypotheses. It may be that the ROIs selected (VMPFC and DLPFC) were not the most relevant or sensitive to change and, in this small study, we were unable to detect other regions of increased activation in a whole brain analysis. There are limited data about brain systems that con- tribute to the persistence of AN, especially as few studies have examined neural systems before and after treat- ment. The mean age and duration of illness in this study is representative of a general sample of treatment seek- ing adults, and is within the range seen in across neuro- imaging research [64]. However, duration of illness has not been systematically considered in neuroimaging studies in AN, and may influence brain findings [65]. Existing longitudinal data, across different imaging mo- dalities, have indicated that renourishment is associated with some brain changes. For example, structural brain abnormalities, such as gray matter volume decreases and cortical thinning, are seen among individuals with AN at low weight and tend to normalize with weight restor- ation [66–68]. However, studies of white matter Acknowledgments We would like to thank Akram Bakkour, PhD for technical assistance with the data analyses, and thoughtful contributions to discussions of these data. The authors would like to thank the patients and staff of the New York State Psychiatric Institute for their contributions to this research. 7. Uniacke B, Wang Y, Biezonski D, Sussman T, Lee S, Posner J, et al. Resting- state connectivity within and across neural circuits in anorexia nervosa. Brain Behav. 2019;9(1):e01205. https://doi.org/10.1002/brb3.1205. Funding This work was supported by National Institute of Mental Health R01 MH105452; R01 MH105452 S1; K24 MH113737. This work was supported by National Institute of Mental Health R01 MH105452; R01 MH105452 S1; K24 MH113737. 11. Rangel A. Regulation of dietary choice by the decision-making circuitry. Nat Neurosci. 2013;16(12):1717–24. https://doi.org/10.1038/nn.3561. Availability of data and materials 12. Hare TA, Camerer CF, Rangel A. Self-control in decision-making involves modulation of the vmPFC valuation system. Science. 2009;324(5927):646–8. https://doi.org/10.1126/science.1168450. The datasets analyzed during the current study are available from the corresponding author on reasonable request. The food choice task stimuli and nutritional content are publicly available at Columbia Academic Commons (https://doi.org/10.7916/d8-497c-2724) and Open Science Framework (https:// osf.io/483mx/); information included in Lloyd et al., 2020 [17]. 13. Hare TA, Malmaud J, Rangel A. Focusing attention on the health aspects of foods changes value signals in vmPFC and improves dietary choice. J Neurosci. 2011;31(30):11077–87. https://doi.org/10.1523/JNEUROSCI.6383-1 0.2011. Abbreviations 1. American Psychiatric Association. Treatment of patients with eating disorders,third edition. Am J Psychiatry. 2006;163(7 Suppl):4–54. 1. American Psychiatric Association. Treatment of patients with eating disorders,third edition. Am J Psychiatry. 2006;163(7 Suppl):4–54. AN: Anorexia nervosa; BMI: Body mass index; BOLD: Blood-oxygen level dependent; DLPFC: Dorsolateral prefrontal cortex; GLM: General linear model; HC: Healthy control volunteer; ITI: Intertrial interval; NYSPI: New York State Psychiatric Institute; VMPFC: Ventromedial prefrontal cortex; IQ: Intelligence quotient; DICOM: Digital Imaging and Communications in Medicine; NIFTI: Neuroimaging Informatics Technology Initiative; FSL: FMRIB Software Library; MNI: Montreal Neurological Institute; FD: Framewise Displacement 2. Attia E, Walsh BT. Behavioral management for anorexia nervosa. N Engl J Med. 2009;360(5):500–6. https://doi.org/10.1056/NEJMct0805569. 2. Attia E, Walsh BT. Behavioral management for anorexia nervosa. N Engl J Med. 2009;360(5):500–6. https://doi.org/10.1056/NEJMct0805569. 3. Khalsa SS, Portnoff LC, McCurdy-McKinnon D, Feusner JD. What happens after treatment? A systematic review of relapse, remission, and recovery in anorexia nervosa. J Eat Disord. 2017;5(1):20. https://doi.org/10.1186/s40337- 017-0145-3. 4. Schebendach J, Mayer LE, Devlin MJ, Attia E, Walsh BT. Dietary energy density and diet variety as risk factors for relapse in anorexia nervosa: a replication. Int J Eat Disord. 2012;45(1):79–84. https://doi.org/10.1002/eat.2 0922. 4. Schebendach J, Mayer LE, Devlin MJ, Attia E, Walsh BT. Dietary energy density and diet variety as risk factors for relapse in anorexia nervosa: a replication. Int J Eat Disord. 2012;45(1):79–84. https://doi.org/10.1002/eat.2 0922. Consent for publication Not applicable. 17. Lloyd EC, Shehzad Z, Schebendach J, Bakkour A, Xue AM, Assaf NF, et al. Food Folio by Columbia Center for Eating Disorders: A Freely Available Food Image Database. Front Psychol. 2020;11(3556). https://doi.org/10.3389/ fpsyg.2020.585044. Received: 26 January 2021 Accepted: 31 March 2021 Received: 26 January 2021 Accepted: 31 March 2021 Received: 26 January 2021 Accepted: 31 March 2021 Additional file 1. 6. Kaplan AS, Walsh BT, Olmsted M, Attia E, Carter JC, Devlin MJ, et al. The slippery slope: prediction of successful weight maintenance in anorexia nervosa. Psychol Med. 2009;39(6):1037–45. https://doi.org/10.1017/S0033291 70800442X. 6. Kaplan AS, Walsh BT, Olmsted M, Attia E, Carter JC, Devlin MJ, et al. The slippery slope: prediction of successful weight maintenance in anorexia nervosa. Psychol Med. 2009;39(6):1037–45. https://doi.org/10.1017/S0033291 70800442X. Declarations 14. Krajbich I, Armel C, Rangel A. Visual fixations and the computation and comparison of value in simple choice. Nat Neurosci. 2010;13(10):1292–8. https://doi.org/10.1038/nn.2635. Competing interests Dr. Steinglass receives royalties from UpToDate. Dr. Walsh reports receiving royalties and honoraria from Guilford Press, McGraw-Hill, Oxford University Press, UpToDate, British Medical Journal, Johns Hopkins Press and Guidepoint Global. Drs. Daw, Foerde, and Shohamy and Ms. Dalack report no financial re- lationships with commercial interests. 18. Foerde K, Schebendach JE, Davis L, Daw N, Walsh BT, Shohamy D, et al. Restrictive eating across a spectrum from healthy to unhealthy: behavioral and neural mechanisms. Psychol Med. 2020:1–10. https://doi.org/10.1017/ S0033291720003542. Authors’ contributions 8. Sysko R, Walsh BT, Schebendach J, Wilson GT. Eating behavior among women with anorexia nervosa. Am J Clin Nutr. 2005;82(2):296–301. https://doi.org/10.1093/ajcn/82.2.296. JS, DS, KF, and BTW conceptualized and designed the study. KF conducted analyses and DS and ND contributed to analytic design and interpretation. MD contributed to data collection and management, and manuscript review. JS, KF, and BTW wrote and edited the manuscript. All authors reviewed the manuscript. The author(s) read and approved the final manuscript. g j 9. Mayer LE, Schebendach J, Bodell LP, Shingleton RM, Walsh BT. Eating behavior in anorexia nervosa: before and after treatment. Int J Eat Disord. 2012;45(2):290–3. https://doi.org/10.1002/eat.20924. 10. Rangel A, Camerer C, Montague PR. A framework for studying the neurobiology of value-based decision making. Nat Rev Neurosci. 2008;9(7): 545–56. https://doi.org/10.1038/nrn2357. Supplementary Information The online version contains supplementary material available at https://doi. org/10.1186/s40337-021-00402-y. 5. Attia E, Haiman C, Walsh BT, Flater SR. Does fluoxetine augment the inpatient treatment of anorexia nervosa? Am J Psychiatry. 1998;155(4):548– 51. https://doi.org/10.1176/ajp.155.4.548. 5. Attia E, Haiman C, Walsh BT, Flater SR. Does fluoxetine augment the inpatient treatment of anorexia nervosa? Am J Psychiatry. 1998;155(4):548– 51. https://doi.org/10.1176/ajp.155.4.548. Ethics approval and consent to participate 15. Maier SU, Raja Beharelle A, Polania R, Ruff CC, Hare TA. Dissociable mechanisms govern when and how strongly reward attributes affect decisions. Nat Hum Behav. 2020;4(9):949–63. https://doi.org/10.1038/s41562-020-0893-y. This study was approved by the New York State Psychiatric Institute Institutional Review Board (Protocol #6717) and all participants signed informed consent. 16. Suzuki S, Cross L, O'Doherty JP. Elucidating the underlying components of food valuation in the human orbitofrontal cortex. Nat Neurosci. 2017;20(12): 1780–6. https://doi.org/10.1038/s41593-017-0008-x. Conclusions By linking brain and behavior, this study confirmed pre- vious findings that during active decisions about what to eat, acutely ill individuals with AN differed from HC in engagement of the anterior caudate. The change Page 10 of 12 Page 10 of 12 Page 10 of 12 Foerde et al. Journal of Eating Disorders (2021) 9:48 Foerde et al. Journal of Eating Disorders (2021) 9:48 Foerde et al. Journal of Eating Disorders (decrease) in caudate activation after treatment was as- sociated with an increased tendency to choose high-fat foods, and similar activation patterns occurred in the parietal/precuneus region. Additional research is needed to determine whether food choice behavior—and associ- ated neural mechanisms—change with longer-term remission. Center, 1051 Riverside Drive, Unit 98, New York, NY 10032, USA. 3Princeton Neuroscience Institute, Department of Psychology, Princeton University, Princeton, NJ 08540, USA. 4Psychology Department and Zuckerman Mind Brain and Behavior Institute, Columbia University, 3227 Broadway, New York, NY 10027, USA. Author details 1 19. Foerde K, Steinglass JE, Shohamy D, Walsh BT. Neural mechanisms supporting maladaptive food choices in anorexia nervosa. Nat Neurosci. 2015;18(11):1571–3. https://doi.org/10.1038/nn.4136. 1New York State Psychiatric Institute, 1051 Riverside Drive, Unit 98, New York, NY 10032, USA. 2Psychiatry Department, Columbia University Irving Medical Page 11 of 12 Page 11 of 12 Page 11 of 12 Foerde et al. Journal of Eating Disorders (2021) 9:48 Foerde et al. Journal of Eating Disorders Foerde et al. Journal of Eating Disorders (2021) 9:48 20. Dalton B, Foerde K, Bartholdy S, McClelland J, Kekic M, Grycuk L, et al. The effect of repetitive transcranial magnetic stimulation on food choice-related self-control in patients with severe, enduring anorexia nervosa. Int J Eat Disord. 2020;53(8):1326–36. https://doi.org/10.1002/eat.23267. 43. Bakkour A, Palombo DJ, Zylberberg A, Kang YH, Reid A, Verfaellie M, Shadlen MN, Shohamy D. The hippocampus supports deliberation during value-based decisions. Elife. 2019;8:e46080. https://doi.org/10.7554/eLife.4 6080. Disord. 2020;53(8):1326–36. https://doi.org/10.1002/eat.23267. 44. Beck AT, Steer RA. Depression inventory manual. San Antonio: Psychological Corporation, Harcourt, Brace; 1993. 21. Steinglass J, Foerde K, Kostro K, Shohamy D, Walsh BT. Restrictive food intake as a choice--a paradigm for study. Int J Eat Disord. 2015;48(1):59–66. https://doi.org/10.1002/eat.22345. 45. Fairburn CG, Beglin SJ. Assessment of eating disorders: interview or self- report questionnaire? Int J Eat Disord. 1994;16(4):363–70. 22. Uniacke B, Slattery R, Walsh BT, Shohamy D, Foerde K, Steinglass JE. A comparison of food-based decision making between restricting and binge- eating/purging subtypes of anorexia nervosa. In J Eat Disord. 2020;53(10): 1751–6. https://doi.org/10.1002/eat23359. report questionnaire? Int J Eat Disord. 1994;16(4):363–70. 46. Wechsler D. Wechsler Adult Intelligence Scale - III. San Antonio: The Psychological Corporation; 1997. 47. Spielberger CD, Gorsuch RL, Lushene R, Vagg PR, Jacobs AG. Manual for the state-trait anxiety inventory. Palo Alto: Consulting Psychologists Press, Inc.; 1983. 23. Foerde K, Gianini L, Wang Y, Wu P, Shohamy D, Walsh BT, et al. Assessment of test-retest reliability of a food choice task among healthy individuals. Appetite. 2018;123:352–6. https://doi.org/10.1016/j.appet.2018.01.010. 48. Stunkard AJ, Messick S. The three-factor eating questionnaire to measure dietary restraint, disinhibition and hunger. J Psychosom Res. 1985;29(1):71– 83. https://doi.org/10.1016/0022-3999(85)90010-8. 24. Bartra O, McGuire JT, Kable JW. The valuation system: a coordinate-based meta-analysis of BOLD fMRI experiments examining neural correlates of subjective value. Neuroimage. 2013;76:412–27. https://doi.org/10.1016/j. neuroimage.2013.02.063. 49. Wechsler D. Wechsler abbreviated scale of intelligence manual. San Antonio: The Psychological Corporation; 1999. 25. American Psychiatric Association. Diagnostic and statistical manual of mental disorders. 5th ed. Author details 1 Arlington: American Psychiatric Association; 2013. https://doi.org/10.1176/appi.books.9780890425596. 50. Schebendach JE, Mayer LE, Devlin MJ, Attia E, Contento IR, Wolf RL, et al. Dietary energy density and diet variety as predictors of outcome in anorexia nervosa. Am J Clin Nutr. 2008;87(4):810–6. https://doi.org/1 0.1093/ajcn/87.4.810. 26. Sysko R, Glasofer DR, Hildebrandt T, Klimek P, Mitchell JE, Berg KC, et al. The eating disorder assessment for DSM-5 (EDA-5): development and validation of a structured interview for feeding and eating disorders. Int J Eat Disord. 2015;48(5):452–63. https://doi.org/10.1002/eat.22388. 51. Liljeholm M, O'Doherty JP. Contributions of the striatum to learning, motivation, and performance: an associative account. Trends Cogn Sci. 2012;16(9):467–75. https://doi.org/10.1016/j.tics.2012.07.007. 27. Brainard DH. The psychophysics toolbox. Spat Vis. 1997;10(4):433–6. https:// doi.org/10.1163/156856897X00357. 52. Steinglass JE, Glasofer DR, Walsh E, Guzman G, Peterson CB, Walsh BT, et al. Targeting habits in anorexia nervosa: a proof-of-concept randomized trial. Psychol Med. 2018;48(15):2584–91. https://doi.org/10.1017/S00332917180002 0X. 28. JASP Team (2020). JASP (Version 0.14.1)[Computer software]. 29. Bates D, Maechler M, Bolker B, Walker S. lme4: linear mixed-effects models using Eigen and S4. R package version. 2014;1(7):1–23. 53. Butler RM, Heimberg RG. Exposure therapy for eating disorders: a systematic review. Clin Psychol Rev. 2020;78:101851. https://doi.org/10.1016/j.cpr.2020.1 01851. 30. Barr DJ, Levy R, Scheepers C, Tily HJ. Random effects structure for confirmatory hypothesis testing: Keep it maximal. J Mem Lang. 2013;68(3): 255–78. https://doi.org/10.1016/j.jml.2012.11.001. 54. Platt ML, Glimcher PW. Neural correlates of decision variables in parietal cortex. Nature. 1999;400(6741):233–8. https://doi.org/10.1038/22268. 31. Barr DJ. Random effects structure for testing interactions in linear mixed-effects models. Front Psychol. 2013;4:328. https://doi.org/10.3389/fpsyg.2013.00328. 55. Pesaran B, Nelson MJ, Andersen RA. Free choice activates a decision circuit between frontal and parietal cortex. Nature. 2008;453(7193):406–9. https:// doi.org/10.1038/nature06849. 32. Venables WN, Ripley BD. Modern applied statistics with S. Fourth ed. New York: Springer; 2002. https://doi.org/10.1007/978-0-387-21706-2. 33. Smith SM, Jenkinson M, Woolrich MW, Beckmann CF, Behrens TE, Johansen- Berg H, et al. Advances in functional and structural MR image analysis and implementation as FSL. Neuroimage. 2004;23(Suppl 1):S208–19. https://doi. org/10.1016/j.neuroimage.2004.07.051. 56. Niendam TA, Laird AR, Ray KL, Dean YM, Glahn DC, Carter CS. Meta-analytic evidence for a superordinate cognitive control network subserving diverse executive functions. Cogn Affect Behav Neurosci. 2012;12(2):241–68. https:// doi.org/10.3758/s13415-011-0083-5. 34. Jenkinson M, Bannister P, Brady M, Smith S. Improved optimization for the robust and accurate linear registration and motion correction of brain images. Neuroimage. 2002;17(2):825–41. https://doi.org/10.1006/nimg.2002.1132. 57. Derrfuss J, Brass M, Neumann J, von Cramon DY. Author details 1 Involvement of the inferior frontal junction in cognitive control: meta-analyses of switching and Stroop studies. Hum Brain Mapp. 2005;25(1):22–34. https://doi.org/10.1002/hbm.2 0127. 35. Smith SM. Fast robust automated brain extraction. Hum Brain Mapp. 2002; 17(3):143–55. https://doi.org/10.1002/hbm.10062. 58. Rae CL, Hughes LE, Weaver C, Anderson MC, Rowe JB. Selection and stopping in voluntary action: a meta-analysis and combined fMRI study. Neuroimage. 2014;86:381–91. https://doi.org/10.1016/j.neuroimage.2013.10. 012. 36. Fischl B, Sereno MI, Tootell RB, Dale AM. High-resolution intersubject averaging and a coordinate system for the cortical surface. Hum Brain Mapp. 1999;8(4):272–84. https://doi.org/10.1002/(SICI)1097-0193(1999)8:4< 272::AID-HBM10>3.0.CO;2-4. 59. Cavanna AE, Trimble MR. The precuneus: a review of its functional anatomy and behavioural correlates. Brain. 2006;129(Pt 3):564–83. https://doi.org/10.1 093/brain/awl004. 37. Segonne F, Dale AM, Busa E, Glessner M, Salat D, Hahn HK, et al. A hybrid approach to the skull stripping problem in MRI. Neuroimage. 2004;22(3): 1060–75. https://doi.org/10.1016/j.neuroimage.2004.03.032. 1060–75. https://doi.org/10.1016/j.neuroimage.2004.03.032. 60. Jiang T, Soussignan R, Carrier E, Royet JP. Dysfunction of the mesolimbic circuit to food odors in women with anorexia and bulimia nervosa: a fMRI study. Front Hum Neurosci. 2019;13:117. https://doi.org/10.3389/fnhum.2019. 00117. 38. Andersson JL, Jenkinson M, Smith S. Non-linear registration aka spatial normalisation FMRIB Technial report TR07JA2. Oxford: FMRIB Analysis Group of the University of Oxford; 2007. 61. Brooks SJ, O'Daly OG, Uher R, Friederich HC, Giampietro V, Brammer M, et al. Differential neural responses to food images in women with bulimia versus anorexia nervosa. Plos One. 2011;6(7):e22259. https://doi.org/10.1371/journal. pone.0022259. 39. Power JD, Barnes KA, Snyder AZ, Schlaggar BL, Petersen SE. Spurious but systematic correlations in functional connectivity MRI networks arise from subject motion. Neuroimage. 2012;59(3):2142–54. https://doi.org/10.1016/j. neuroimage.2011.10.018. 62. Nikendei C, Friederich HC, Weisbrod M, Walther S, Sharma A, Herzog W, et al. Event-related potentials during recognition of semantic and pictorial food stimuli in patients with anorexia nervosa and healthy controls with varying internal states of hunger. Psychosom Med. 2012;74(2):136–45. https://doi.org/10.1097/PSY.0b013e318242496a. 40. Siegel JS, Power JD, Dubis JW, Vogel AC, Church JA, Schlaggar BL, et al. Statistical improvements in functional magnetic resonance imaging analyses produced by censoring high-motion data points. Hum Brain Mapp. 2014;35(5):1981–96. https://doi.org/10.1002/hbm.22307. 41. Bodell LP, Keel PK, Brumm MC, Akubuiro A, Caballero J, Tranel D, et al. Longitudinal examination of decision-making performance in anorexia nervosa: before and after weight restoration. J Psychiatr Res. 2014;56:150–7. https://doi.org/10.1016/j.jpsychires.2014.05.015. 63. Dockes J, Poldrack RA, Primet R, Gozukan H, Yarkoni T, Suchanek F, et al. NeuroQuery, comprehensive meta-analysis of human brain mapping. Elife. 2020;9. https://doi.org/10.7554/eLife.53385. 63. Author details 1 https://doi.org/10.11 76/appi.ajp.2016.16060671. 72. Decker JH, Figner B, Steinglass JE. On weight and waiting: delay discounting in anorexia nervosa pretreatment and posttreatment. Biol Psychiatry. 2015; 78(9):606–14. https://doi.org/10.1016/j.biopsych.2014.12.016. 72. Decker JH, Figner B, Steinglass JE. On weight and waiting: delay discounting in anorexia nervosa pretreatment and posttreatment. Biol Psychiatry. 2015; 78(9):606–14. https://doi.org/10.1016/j.biopsych.2014.12.016. 73. Doose A, King JA, Bernardoni F, Geisler D, Hellerhoff I, Weinert T, et al. Strengthened Default Mode Network Activation During Delay Discounting in Adolescents with Anorexia Nervosa After Partial Weight Restoration: A Longitudinal fMRI Study. J Clin Med. 2020;9(4). https://doi.org/10.3390/ jcm9040900. Epub date 2020/03/29. 74. Chung T, Noronha A, Carroll KM, Potenza MN, Hutchison K, Calhoun VD, et al. Brain mechanisms of change in addictions treatment: models, methods, and emerging findings. Curr Addict Rep. 2016;3(3):332–42. https:// doi.org/10.1007/s40429-016-0113-z. Author details 1 Dockes J, Poldrack RA, Primet R, Gozukan H, Yarkoni T, Suchanek F, et al. NeuroQuery, comprehensive meta-analysis of human brain mapping. Elife. 2020;9. https://doi.org/10.7554/eLife.53385. 42. Eklund A, Nichols TE, Knutsson H. Cluster failure: why fMRI inferences for spatial extent have inflated false-positive rates. Proc Natl Acad Sci U S A. 2016;113(28):7900–5. https://doi.org/10.1073/pnas.1602413113. 64. Alfano V, Mele G, Cotugno A, Longarzo M. Multimodal neuroimaging in anorexia nervosa. J Neurosci Res. 2020;98(11):2178–207. https://doi.org/10.1 002/jnr.24674. 64. Alfano V, Mele G, Cotugno A, Longarzo M. Multimodal neuroimaging in anorexia nervosa. J Neurosci Res. 2020;98(11):2178–207. https://doi.org/10.1 002/jnr.24674. Page 12 of 12 Foerde et al. Journal of Eating Disorders (2021) 9:48 Foerde et al. Journal of Eating Disorders Foerde et al. Journal of Eating Disorders (2021) 9:48 65. Steinglass J, Glasofer D, Dalack M, Attia E. Between wellness, relapse and remission: Stages of illness in anorexia nervosa. Int J Eat Disord. 2020; 53(7): 1088–96. https://doi.org/10.1002/eat.23237. 66. Nickel K, Joos A, Tebartz van Elst L, Matthis J, Holovics L, Endres D, et al. Recovery of cortical volume and thickness after remission from acute anorexia nervosa. Int J Eat Disord. 2018;51(9):1056–69. https://doi.org/10.1 002/eat.22918. 67. von Schwanenflug N, Muller DK, King JA, Ritschel F, Bernardoni F, Mohammadi S, et al. Dynamic changes in white matter microstructure in anorexia nervosa: findings from a longitudinal study. Psychol Med. 2019; 49(9):1555–64. https://doi.org/10.1017/S003329171800212X. 68. Bernardoni F, King JA, Geisler D, Stein E, Jaite C, Natsch D, et al. Weight restoration therapy rapidly reverses cortical thinning in anorexia nervosa: a longitudinal study. Neuroimage. 2016;130:214–22. https://doi.org/10.1016/j. neuroimage.2016.02.003. 69. Martin Monzon B, Hay P, Foroughi N, Touyz S. White matter alterations in anorexia nervosa: a systematic review of diffusion tensor imaging studies. World J Psychiatry. 2016;6(1):177–86. https://doi.org/10.5498/wjp.v6.i1.177. 70. Cha J, Ide JS, Bowman FD, Simpson HB, Posner J, Steinglass JE. Abnormal reward circuitry in anorexia nervosa: a longitudinal, multimodal MRI study. Hum Brain Mapp. 2016;37(11):3835–46. https://doi.org/10.1002/hbm.23279. 70. Cha J, Ide JS, Bowman FD, Simpson HB, Posner J, Steinglass JE. Abnormal reward circuitry in anorexia nervosa: a longitudinal, multimodal MRI study. Hum Brain Mapp. 2016;37(11):3835–46. https://doi.org/10.1002/hbm.23279. 71. DeGuzman M, Shott ME, Yang TT, Riederer J, Frank GKW. Association of elevated reward prediction error response with weight gain in adolescent anorexia nervosa. Am J Psychiatry. 2017;174(6):557–65. https://doi.org/10.11 76/appi.ajp.2016.16060671. 71. DeGuzman M, Shott ME, Yang TT, Riederer J, Frank GKW. Association of elevated reward prediction error response with weight gain in adolescent anorexia nervosa. Am J Psychiatry. 2017;174(6):557–65. Publisher’s Note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
https://openalex.org/W1888012092	https://cadmus.eui.eu/bitstream/1814/9588/1/LAW_2008_28.pdf	English	null	The Concept of State Aid in Liberalised Sectors	Social Science Research Network	2,008	public-domain	7,772	EUI Working Papers LAW 2008/28 The Concept of State Aid in Liberalised Sectors Thomas von Danwitz Thomas von Danwitz Thomas von Danwitz Abstract The author highlights the balancing act both on the regulatory as well as on the institutional level between state aid control and the liberalization of public services. He focuses on partially liberalized markets and tackling cross-subsidisation where Member States infringe the competitive neutrality of the privatisation process by various funding schemes. These are all subject to three criteria linked to the private investor test. Once partially liberalized, sectors traditionally shaped by public service obligations are prone to state intervention owing to the special needs they fulfil. Starting from the premise that the concept of universal services is designed to combine public policy objectives with a fully competitive market, the author allocates the role of state aid control as both a specific mandate avoiding selective distortions through the granting of state resources imputable to the State and as a regulatory mandate to maintain a level playing field for all undertakings in the Internal Market. The jurisprudence of Community Courts – e.g. UFEX, Chronopost and Laboratoires Boiron - is faced with the demarcation of the European Commission's powers and the determination of the nature and extent of judicial review. Its analysis focuses on cost calculation and allocation in search of cross- subsidisation of liberalised market segments by using state resources originally designed to compensate for public service obligations. He closes with the assumption that, because of the narrow confines of aleatory references made to the Courts, preference should be given to a best practice approach to cost allocation standards EUROPEAN UNIVERSITY INSTITUTE DEPARTMENT OF LAW The Concept of State Aid in Liberalised Sectors THOMAS VON DANWITZ EUI Working Paper LAW No. 2008/28 This text may be downloaded for personal research purposes only. Any additional reproduction for other purposes, whether in hard copy or electronically, requires the consent of the author(s), editor(s). If cited or quoted, reference should be made to the full name of the author(s), editor(s), the title, the working paper or other series, the year, and the publisher. The author(s)/editor(s) should inform the Law Department of the EUI if the paper is to be published elsewhere, and should also assume responsibility for any consequent obligation(s). ISSN 1725-6739 © 2008 Thomas von Danwitz Printed in Italy European University Institute Badia Fiesolana I – 50014 San Domenico di Fiesole (FI) Italy http://www.eui.eu/ http://cadmus.eui.eu/ ∗ Prof. Dr. Thomas von Danwitz, D.I.A.P. (ENA, Paris), Judge of the Court of Justice of the European Communities, Luxembourg/Cologne 1 See treaty of Lisbon, protocol no. 26. 2 COM (2007) 725 final. The Concept of State Aid in Liberalised Sectors Thomas von Danwitz∗ Keywords Regulatory politics – sectoral governance – European law – judicial review – state aids – liberalization – European Court of Justice – universal services – cross-subsidisation I. Introduction The following reflections on the concept of state aid in liberalised sectors have, of course, been inspired by the particular intellectual appeal of this subject resulting from the indistinguishable combination of legal and economic elements which, like all crossbreeds, led to particularly original features and the need for a high degree of commitment to fully understand the nature of those developments. Less of intellectual but more of practical concern are the far reaching economic, financial and even political consequences with which we are faced when dealing with the subject of state aid in liberalised sectors. Already a first glance at the relevant jurisprudence tells us that it might not be easy to establish a clear cut concept, since the Community Courts have only delivered a rather limited number of judgements in recent times of state aid features in liberalised sectors. Beyond this interest in our subject lies a second dimension, which is of no lesser importance. After long years of discussion over the role of public services within the creation of the single market, the framework for public services has become more and more precisely defined after the elaboration of Article 16 in the treaty of Amsterdam, and particularly after its amendment by the treaty of Lisbon. The new paragraph added to Article 16, now re-numbered as Article 14, empowers the Community legislator to establish principles and conditions in order to enable the Member States and the Community to provide, commission and to fund such services. This specific rulemaking provision on the functioning of public services inserted in the treaty, to which a specific protocol on services of general interest1 has been added and which is accompanied by a Communication of the Commission2, constitutes a major change in the existing legal framework and, so far, the culmination point in the fight for a specific legal status of public services. It goes without saying that this rulemaking provision is designed to bear legislation specifically designed to meet the particular mission of public services and thereby, at least to a certain extend, replacing the predominant focus of the existing Thomas von Danwitz legal framework on competition rules and their implementation. Such an evolution could of course lead to a significant change of the existing framework, going far beyond the well balanced solution adopted in the Constitutional treaty3. But it will hopefully not limit the completion of the internal market in this field. 6 Joined Cases C-83/01 P, C-93/01 P, 94/01 P Chronopost [2003] ECR I-6993. 5 Case C-482/99 French Republic v Commission – Stardust Marine [2002] ECR I-4397. I. Introduction It is self evident that these changes in the legal framework will become the predominant source of inspiration for legal scholars and will, once secondary legislation has been enacted, become more and more relevant for the status of public services and the legal appreciation of the conditions under which they operate. Given this perspective, it remains to be seen, to what extend the relevant jurisprudence of the Court, in particular in Preußen Elektra4, Stardust Marine5, Chronopost6 and Altmark Trans7, will continue to play a crucial role in this field. As you are certainly expecting a view from the Court on these matters and given the wide range of legal aspects having an impact on the mission and the functioning of public services, I will focus on the role of state aid control and its major challenges in this field. On the basis of some general remarks on the specific nature and difficulties of state aid control over public services, the major part of my presentation will be dedicated to the main elements of the established jurisprudence in respect to the problems arising from partially liberalised markets and in particular to the crucial question of cross-subsidisation. 4 Case C-379/98 PreussenElektra [2001] ECR I-2099. p 7 Case C-280/00 Altmark Trans [2003] ECR I-7747. 3 See von Danwitz, Die Rolle der Unternehmen der Daseinsvorsorge im Verfassungsentwurf, i Schwarze (Hrsg.), Der Verfassungsentwurf des Europäischen Konvents, Baden-Baden 2004, S. 251 f. p 7 Case C-280/00 Altmark Trans [2003] ECR I-7747. 4 Case C-379/98 PreussenElektra [2001] ECR I-2099. ned Cases C-83/01 P, C-93/01 P, 94/01 P Chronopost [2003] ECR I-6993. II. The role of state aid control in the process of liberalisation and privatisatio Undoubtedly, the purpose of privatisation is to deprive state aid control of its principal object, a public undertaking. But, in order to achieve this goal, privatisation has to be accomplished under conditions ensuring the competitive neutrality of the conducted privatisation procedure. Typically, state aid control in the field of privatisation gives rise to three different issues, all closely linked to the famous private investor test: − firstly, the well-known phenomenon of "dressing the bride" in order to make an undertaking more attractive to private investors; − firstly, the well-known phenomenon of "dressing the bride" in order to make an undertaking more attractive to private investors; − secondly, the famous dilemma of stranded costs: contributions or preferential treatments in order to compensate for investments or other engagements of formerly public undertakings which have become void under liberalised market conditions. − thirdly, the old problem of how to assess whether the sale of a public undertaking is at market value or entails state aid. − thirdly, the old problem of how to assess whether the sale of a public undertaking is at market value or entails state aid. Each of these problems would undoubtedly justify me concentrating my remarks on them individually. Instead, let me confine my remarks to the general statement that the EUI WP LAW 2008/28 © 2008 Thomas von Danwitz 2 The Concept of State Aid in Liberalised Sectors The Concept of State Aid in Liberalised Sectors settled case-law on contributions aimed at making an undertaking more attractive to private investors8 gives full guidance for the application of the private investor test in that respect. Nevertheless, the central question of whether a comparable private investor pursuing a structural policy which is guided by prospects of long term viability leaves considerable scope for the assessment of each particular case. In contrast to that, there is only rather limited jurisprudence on the problem of stranded costs9 and on the question of whether the sale of a public undertaking entails state aid10. Finally, it should be noted that the general experience of state aid law to witness an incredible inventiveness of Member States when it comes to original funding schemes and methods, is quite impressively confirmed in the field of privatisation and liberalisation. 8 For ex: Case T-157/01 Dansk Busvognamænd [2004] ECR II-917; joined Cases T-129/95, T-2/96, T- 97/96 Maxhütte [1999] ECR II-17; joined Cases C-278/92, C-279/92 and C-280/92 Spain/COM – Hytasa [1994] ECR I-4103. 9 Joined Cases C-128/03 and C-129/03 AEM [2005] ECR I-2861; C-206/06 Essent Netwerk Noord – pending. 10 Case C-334/99 Germany/COM – Gröditzer Stahlwerke [2003] ECR I-1139; C-280/92 Spain/COM – Hytasa [1994] ECR I-4103. 11 Joined Cases C-128/03 and C-129/03 AEM [2005] ECR I-2861. 12 See Case 107/84 [1985] ECR 02655 (2666, point 11). II. The role of state aid control in the process of liberalisation and privatisatio In the AEM case11 the Court has already been asked to juge whether an increased charge for access to the national transmission grid only levied on certain undertakings in order to offset a particular advantage which they presumably might have gained from liberalisation constitutes state aid. Unfortunately, the preliminary reference did not contain sufficient factual information to conclude that the existence of state aid was established. I am convinced that Member States will also in the future ensure that the task of the Court will not end up in routine, but will continue to be faced with innovative systems of preferential treatment which have to be monitored strictly under state aid rules. 13 For the latter see Koenig/Kiefer, Public Funding of Infrastructure Projects under EC-State Aid Law, EStAL 3/2005, p. 415 seq. 14 See v. Danwitz, Der europäische Elektrizitätsbinnenmarkt zwischen Wettbewerbsorientierung und Wettbewerbsbeschränkung, 2007, under A, in: Recht und Ordnung der Wirtschaft, Festsschrift für Peter J. Tettinger (Jörg Ennuschat, Jörg Gerling, Thomas Mann and Johann Christian Pielow, eds.). 15 Services of General Economic Interest – Opinion Prepared by the State Aid Group of EAGCP, June 29, 2006. III. The competitive environment Despite the focus of my following remarks on state aid law, let me bring to your attention that state aid is by far not the only tool for Member States to favour a particular undertaking or a certain type of public services. In the preliminary ruling C- 357/07 TNT Post UK, the Court is faced with the question of VAT exoneration in favour of services carried out by public postal service operators under the circumstances of liberalisation. The notion figuring in Article 13 A. paragraph 1 under a) of the Sixth VAT Directive received a self evident interpretation in the good old days of the delivery monopole for state owned postal undertakings. Accordingly, the Court has held in 198512 under those circumstances that the public service operator was exempted from VAT including all offered services. Based on the assumption that the existence of a reserved market would constitute a prerequisite for such a VAT exemption, advocate general Geelhoed proposed in 2003 under conditions of a partially liberalised market that the VAT exemption should only cover reserved market shares and not go beyond. EUI WP LAW 2008/28 © 2008 Thomas von Danwitz 3 3 Thomas von Danwitz Thomas von Danwitz Unfortunately a withdrawal of the preliminary request made it impossible for the Court to render its judgement. Unfortunately a withdrawal of the preliminary request made it impossible for the Court to render its judgement. Under conditions of entirely liberalised markets the question is today asked from a different angle. On the one hand it is quite evident that public postal services do merit a VAT exemption, even if those services are accomplished in a market environment by one or more operators. On the other hand, it is equally evident that such a VAT exemption of relevant services is of sufficient importance for the evolution of a competitive environment in postal markets to possibly produce distortions of competition. Therefore the decision of the Court will, in any event, have to take the importance of public postal services and the maintenance of a high level of universal services into consideration as possible justification. Given these diverging goals, it will be revealing to see how the Court might strike the balance. 13 For the latter see Koenig/Kiefer, Public Funding of Infrastructure Projects under EC-State Aid La EStAL 3/2005, p. 415 seq. 14 IV. State aid control in partially liberalised markets and cross-subsidisation Once a sector has been liberalised, it seems to be, at a first glance at least, rather peculiar to look for a major role of state aid control under circumstances of liberalisation. But in fact, state aid control is faced with quite a significant set of questions which go far beyond the traditional patterns of regional and sectorial state aid schemes or infrastructural support measures13. Particularly in sectors which have traditionally been shaped by public service obligations, it is not uncommon to see a member state devoting continuous efforts to the development of this market according to priorities inspired by the general interest. The process of liberalisation, which we have experienced in the European Communities in sectors such as telecommunications, postal services, energy supply and public transportation since the 90s, has shown quite strikingly that Member States are in general not willing to pursue a strict “hands off” policy once public undertakings have been privatised and market conditions in these sectors have been liberalised14. The basic reason is simply to be found in the mere fact that those sectors continue to be perceived under a different angle than most other industrial or commercial markets, both by consumers and by public authorities. Those sectors are and will, for the foreseeable future, remain to be marked by the particular interest of consumers in services of general economic interest in terms of continuity, periodicity and affordability15. In consequence, public authorities continue to seek ways and means of ensuring that those consumer needs are met and, depending on traditional policy orientations, consider state subsidies and the maintenance of exclusive rights quite a legitimate way of ensuring that public policy priorities in delivering services of general economic interest are achieved. In that respect it seems that Member States have not yet fully EUI WP LAW 2008/28 © 2008 Thomas von Danwitz 4 The Concept of State Aid in Liberalised Sectors embraced the deeper logic of the concept of universal services which is designed to combine the continuous pursuit of public policy objectives with the conditions of a fully competitive market environment. This might explain why the liberalisation process has in general been enacted on a step by step basis with a persistent tendency of maintaining certain exclusive rights of limited scope in postal services16, energy supply17 and public transport18. 1. The role of state aid control in partially liberalised markets The undoubtedly highly technical nature of these examples, in particular when it comes to the allocation of costs, should not lead to the false conclusion that this subject-matter is too detailed to be deserving of particular attention. A closer look into the relevant jurisprudence and the academic research quite strikingly shows the considerable importance of these questions and the far-reaching consequences in economic and even in political terms23. The principal challenge facing Community Courts in the environment of liberalised and partially liberalised markets is to ask whether there is a specific role for state aid control, to define the scope of the powers of the European Commission in the assessment of state aid measures and to determine the nature and the extent of judicial review in this field. Fundamental questions altogether. IV. State aid control in partially liberalised markets and cross-subsidisation Beyond direct state subsidies designed to fulfil public service objectives, in particular the phenomenon of partially liberalised markets, be it a constant or a transitional one, most naturally gives rise to possible distortions of competition through cross-subsidisation. The different manifestations, in which cross-subsidisation might occur are as numerous and possibly as diverse as the granting of state aid itself. The Court’s practise already shows a quite significant range of different schemes, be it direct payments to compensate for public service obligations19 or the levy of increased charges to finance public service missions20, logistical and commercial support to a subsidiary operating in a competitive market21 and the calculation of charges for the access and use of a transmission or a transport system22 to a subsidiary. 16 Article 7 paragraph 1 of Directive 97/67/EC as amended by Directive 2002/39/EC. 17 Article 3 paragraph 8 of Directive 2003/54/EC. 18 Art. 10 of Directive 91/440/EEC. 19 Case C-280/00 Altmark Trans [2003] ECR I-7747. 20 Case C-340/90 TNT Traco [2001] ECR I-4109; Case C-526/04 Laboratoires Boiron [2006] ECR I- 7529. 21 Joined Cases C-83/01 P, C-93/01 P, 94/01 P Chronopost [2003] ECR I-6993; Case T-613/97 UFEX [2006] ECR II-1531. 22 Joined Cases C-128/03 and C-129/03 AEM [2005] ECR I-2861; Case T-266/02 Deutsche Post v Commission - pending. 23 The ongoing discussion on Community and national level instigated by the Court's decisions in Cases like Stardust Marine and Altmark Trans and the resulting legislative initiatives may serve as examples. 16 Article 7 paragraph 1 of Directive 97/67/EC as amended by Directive 2002/39/EC. 17 Article 3 paragraph 8 of Directive 2003/54/EC. 18 Art. 10 of Directive 91/440/EEC. 19 Case C-280/00 Altmark Trans [2003] ECR I-7747. a) A “specific” role for state aid control in liberalised sectors? The starting-point of our reflections on the scope of state aid control in liberalised sectors is the basic question of whether state aid control has a specific role to play in liberalised areas. To put it in quite simple terms: My suggestion is that state aid control has no different role to play in liberalised sectors than in any other field. State aid 23 The ongoing discussion on Community and national level instigated by the Court's decisions in Cases like Stardust Marine and Altmark Trans and the resulting legislative initiatives may serve as examples. EUI WP LAW 2008/28 © 2008 Thomas von Danwitz 5 5 Thomas von Danwitz control has to effectively ensure that Member States cannot create distortions of competition by aids which, firstly, are granted directly or indirectly through state resources and, secondly, are imputable to the State24. In view of the wide spectrum of ways in which aid can be granted, state aid control has to cover the entire range of possible configurations. In that respect, let me recall the settled case-law according to which Community law cannot permit the rules on state aid to be circumvented by the creation of autonomous institutions or procedures charged with allocating state aid25. But the purpose of state aid control of avoiding distortions of competition is not a self- sufficient objective which is on its own capable of justifying any given measure designed to improve conditions of competition in general. In that respect, it has to be stressed that a clear distinction must be drawn between the specific mandate of state aid control of avoiding selective distortions of competition by specific state intervention in granting state resources and a general, far-reaching regulatory mandate to maintain a level playing field for all undertakings active on the Internal Market. In particular, the improvement of competitive conditions as such between former public undertakings with enduring strong market positions in their home market and new competitors, be it new private undertakings or former public undertakings from other Member States, lies, as urgent as it might be for the realisation of the Internal Market, beyond the Commission’s mandate in matters of state aid control26. Cross-subsidisation in general is subject to Article 82 EC27 and specific regulatory rules for each sector come within the ambit of state aid rules only if state resources have been granted in a way which is imputable to the state. 28 Decision on the application of Article 86 (2) of the EC Treaty to State aid in the form of public service compensation granted to certain undertakings entrusted with the operation of services of general economic interest, O.J. L 312, p. 67. 27 See v. Danwitz, Europarechtliche Grundlagen, in: Beck’scher PostG-Kommentar, 2. ed. 2002, para. 113 seq. 26 Vesterdorf, A Further Comment on the New State Aid Concept as this Concept Continues to be Reshaped, EstAQ 2005 p. 393, 398. 25 See Case C-482/99 French Republic v Commission – Stardust Marine [2002] ECR I-4397 (4435, pa 23). 24 Case C-482/99 French Republic v Commission – Stardust Marine [2002] ECR I-4397 (4436, para 24). 24 Case C-482/99 French Republic v Commission – Stardust Marine [2002] ECR I-4397 (4436, para 24). 25 See Case C-482/99 French Republic v Commission – Stardust Marine [2002] ECR I-4397 (4435, para 23). b) Use and abuse of state resources – in search of imputability State aid control in liberalised sectors requires the granting of state resources in a manner which is imputable to the State. Quite evidently, both conditions must be met. All advantages financed by state resources constitute aid, irrespective of whether they are granted directly by the State or indirectly by a public or private body designated or established by the State29. By contrast, a selective measure that does not involve any direct or indirect transfer of State resources to undertakings may not be qualified an aid, even if it confers an undeniable advantage on certain undertakings30. In quite a number of cases, the use of state resources has not been in question31, but the focus was rather on the question of whether the decision to grant those state resources can be attributed to the State and the measure therefore constitutes state aid. The conditions for such an imputability have been developed in the Stardust Marine judgement32 on the basis of a realistic assumption. The Court states clearly that relations between the State and public undertakings are in general close and that it will therefore be very difficult for a third party to demonstrate in a particular case that aid measures taken by such an undertaking were in fact adopted on instructions of public authorities33. The reference made in Stardust Marine to the circumstances of the case and the context in which the measure was taken, has been criticised for being a rather vague notion34. But in fact it seems hard to imagine how one could set up criteria once and for all relevant cases that could be significantly more precise. In that respect it seems to be more important to note that the Court has deployed great efforts on identifying indicators which might be relevant to establish that an aid measure adopted by a public undertaking is imputable to the State. Seen from this angle, it does not seem unreasonable to conclude that a merely formalistic approach fixing on the fact that an undertaking is legally placed under public control, would not be sufficient for drawing conclusions as regards the imputability of its decisions to the State. 34 See Gahlleithner, Die Zurechenbarkeit von Handlungen öffentlicher Unternehmen an den Staat, 2002, European Law Reporter 271, 276; Lübbig/von Merveldt, Stardust Marine: Introducing Imputability into State Aid Rules - Plain Sailing into Calm Seas or Rowing Back into Shallow Waters?, (2003) EC.L.R. 629, 632. 35 Case C-482/99 French Republic v Commission – Stardust Marine [2002] ECR I-4397 (4445, para. 56). 36 Ibid. Para. 57. 29 Case C-379/98 PreussenElektra [2001] ECR I-2099 (2181, para. 58) with reference to the settled case- law of the Court. 30 Case C-379/98 PreussenElektra [2001] ECR I-2099 (2181, para. 61). 31 Case C-39/94 SFEI [1996] ECR I-3549 (3595, point 56); Case C-482/99 French Republic v Commission – Stardust Marine [2002] ECR I-4397 (4438, para. 32). 32 Case C-482/99 French Republic v Commission – Stardust Marine [2002] ECR I-4397 (4444, para. 52 seq.). 33 Ibid 53 St d t M i 29 Case C-379/98 PreussenElektra [2001] ECR I-2099 (2181, para. 58) with reference to the settled case- law of the Court. 30 Case C-379/98 PreussenElektra [2001] ECR I-2099 (2181, para. 61). 31 Case C-39/94 SFEI [1996] ECR I-3549 (3595, point 56); Case C-482/99 French Republic v Commission – Stardust Marine [2002] ECR I-4397 (4438, para. 32). 32 Case C-482/99 French Republic v Commission – Stardust Marine [2002] ECR I-4397 (4444, para. 52 seq.). 33 Ibid. para 53 seq. – Stardust Marine. 34 See Gahlleithner, Die Zurechenbarkeit von Handlungen öffentlicher Unternehmen an den Staat, 2002, European Law Reporter 271, 276; Lübbig/von Merveldt, Stardust Marine: Introducing Imputability into State Aid Rules - Plain Sailing into Calm Seas or Rowing Back into Shallow Waters?, (2003) EC.L.R. 629, 632. 35 Case C-482/99 French Republic v Commission – Stardust Marine [2002] ECR I-4397 (4445, para. 56). 36 Ibid. Para. 57. 37 See the reasoning of the CFI, Case T-613/97 UFEX [2006] ECR II-1531 (para. 166). 32 Case C-482/99 French Republic v Commission – Stardust Marine [2002] ECR I-4397 (4444, para. 52 seq.). 29 Case C-379/98 PreussenElektra [2001] ECR I-2099 (2181, para. 58) with reference to the settled case- law of the Court. 30 a) A “specific” role for state aid control in liberalised sectors? Although the distinction between the general regulatory need for the realisation of a level playing field in liberalised markets and the rather “limited” function of state aid control is certainly less evident in practice, an important conclusion should be drawn from this differentiation. The regulatory standards contained in regulations and directives on the liberalisation of a particular sector can go well beyond or stay behind the requirements of state aid control. Therefore, those standards as such cannot, it seems to me, be taken as valid criteria for decisions in state aid law without particular justification, which might of course be possible under given circumstances. This is why the Commission has rightly elaborated specific compensation standards in Article 5 paragraph 2 of the decision taken on the application of Article 86 paragraph 2 EC28 concerning the compensation granted to undertakings entrusted with the operation of services of general economic interest. 28 Decision on the application of Article 86 (2) of the EC Treaty to State aid in the form of public service compensation granted to certain undertakings entrusted with the operation of services of general economic interest, O.J. L 312, p. 67. EUI WP LAW 2008/28 © 2008 Thomas von Danwitz 6 The Concept of State Aid in Liberalised Sectors 38 Faulhaber, Cross-Subsidization: Pricing in Public Enterprises, American Economy Review 1975, p. 966; idem. Cross-Subsidization: Analysis with more than two Services, Journal of Competition Law and Economics, 2005, p. 441-448. 2. Cost allocation - in search of the “right” standard Ever since the ground-breaking study of Faulhaber on the phenomenon of cross- subsidisation in public enterprises,38 the pricing behaviour of public undertakings and their internal allocation of costs have become central features in proceedings on the abuse of a dominant market position. It should however be noted that Faulhaber's studies concerned wholly regulated industries.39 Therefore it might be worthwhile to do some research on the question to what extent his approach is transposable to state aid control over undertakings offering services both on reserved and on competitive markets. In the early Corbeau-judgement the Court concluded that cross-subsidisation can be considered a generally justified behaviour in accordance with Article 86 paragraph 2 EC, but added quite strongly that the exclusion of competition is not justified as regards specific services which are dissociable from services of general interest and meet special needs of economic operators for which the traditional postal services make no adequate offer40. The perception of cross-subsidisation as a subject to state aid rules is a more recent evolution, rather closely linked to the process of step by step liberalisation giving rise to partially liberalised markets in which the cross-subsidisation of liberalised market segments by using state resources originally designed to compensate for public service obligations has become a real danger for the creation of a competitive environment41. b) Use and abuse of state resources – in search of imputability In making reference to the intensity of the supervision exercised by public authorities35 and to the real possibilities of exercising a dominant influence36, Stardust Marine aims to achieve a realistic assessment of the particular legal and factual situation in concreto in which a public undertaking adopts a certain aid measure37. In the end it is for the Commission and the Member States to demonstrate to the Court for what legal and factual reasons an 34 See Gahlleithner, Die Zurechenbarkeit von Handlungen öffentlicher Unternehmen an den Staat, 2002, European Law Reporter 271, 276; Lübbig/von Merveldt, Stardust Marine: Introducing Imputability into State Aid Rules - Plain Sailing into Calm Seas or Rowing Back into Shallow Waters?, (2003) EC.L.R. 629, 632. 35 Case C-482/99 French Republic v Commission – Stardust Marine [2002] ECR I-4397 (4445, para. 56). 36 Ibid. Para. 57. 37 See the reasoning of the CFI, Case T-613/97 UFEX [2006] ECR II-1531 (para. 166). EUI WP LAW 2008/28 © 2008 Thomas von Danwitz 7 Thomas von Danwitz aid measure should be considered imputable to the State or not. The jurisprudence will evolve in the light of this experience. aid measure should be considered imputable to the State or not. The jurisprudence will evolve in the light of this experience. 39 Faulhaber, (Footnote 35), p. 966 (967). 40 Case C-320/91 Corbeau [1993] ECR I- 2533 (2569, para.16-19). 41 See Hancher/Buendia Sierra, Cross-Subsidization and EC Law, CMLRev. 35 (1998), 901 ff.; Dimitris Triantafyllou, L’encadrement communautaire du service public, RTDEur 1999 p. 21. 42 Case C-340/90 TNT Traco [2001] ECR I-4109. 43 Joined Cases C-83/01 P, C-93/01 P, 94/01 P Chronopost [2003] ECR I-6993. 44 Case C-475/99 Ambulanz Glöckner [2001] ECR I-8089. 45 Joined Cases C-83/01 P, C-93/01 P, 94/01 P Chronopost [2003] ECR I-6993 (7042, para. 38). 45 Joined Cases C-83/01 P, C-93/01 P, 94/01 P Chronopost [2003] ECR I-6993 (7042, para. 38). 39 Faulhaber, (Footnote 35), p. 966 (967). 40 Case C-320/91 Corbeau [1993] ECR I- 2533 (2569, para.16-19). 41 See Hancher/Buendia Sierra, Cross-Subsidization and EC Law, CMLRev. 35 (1998), 901 ff.; Dimitris Triantafyllou, L’encadrement communautaire du service public, RTDEur 1999 p. 21. a) Scope and intensity of judicial review Ever since the fundamental judgement in Corbeau, the Court has had quite a number of opportunities of developing its jurisprudence with regards to Article 86 paragraph 2 EC. The general message of the judgements in TNT Traco42, Chronopost43 and Ambulanz Glöckner44 seems to be quite clear. In particular the Court pointed out clearly in Chronopost that the public network made available to Chronopost was clearly not a market network and would never have been created by a private undertaking. Accordingly, cost calculation and allocation in search of possible cross-subsidisation have to be based on objective and verifiable elements45 on the factual basis of the EUI WP LAW 2008/28 © 2008 Thomas von Danwitz 8 The Concept of State Aid in Liberalised Sectors existing network46. The factual and legal difficulties for the Commission of meeting this standard may not be neglected, as the follow-on UFEX-judgment of the CFI shows, in which the Tribunal annulled the contested decision of the Commission for defective reasoning. For the future, the rather detailed accounting standards of the transparency directive 2006/111/EC47 and the cost calculation standards of the Commission decision on the application of Article 86 paragraph 2 EC to state aid in the form of public service compensation48 should improve the factual conditions quite considerably to identify objective and verifiable elements for the required assessments relating to cross- subsidisation. In the UFEX-judgement, the Tribunal addressed a rather fundamental question in stating that the Commission’s calculation of the cost incurred in providing logistical and commercial assistance to a subsidiary involves, in the absence of analytical accounts, a complex economic appraisal for which the Commission enjoys a wide discretion. In consequence, the judicial review of that measure, even though comprehensive in principle as the CFI adds, is limited to manifest errors of assessment as well as the respect of procedural rules and the statement of reasons49. In the light of the complexities of cost calculation, this reasoning is fully understandable from any lawyer’s and not only from a judge’s point of view. In addition to that, we all know since the Court's reasoning in TU München50 that a rather strict monitoring of the obligation to state reasons may in practice lead to a judicial review which is as strict as a comprehensive review in substantive terms. 46 See CFI, Case T-613/97 UFEX [2006] ECR II-1531 para. 131 and with respect to Article 82 EC see Case 27/76 United Brands [1978] ECR 207 (305; para. 248/257); Case C-323/93 Centre d’insémination de la Crespelle [1994] I-5077 (5106, para. 26 f.). p p 47 Commission Directive 2006/111/EC of 16 November 2006 on the transparency of financial relations between Member States and public undertakings as well as on financial transparency within certain undertakings, O.J. 2006, L 318, p. 17. 48 Commission decision of 28 November 2005 on the application of Article 86 (2) of the EC Treaty to State aid in the form of public service compensation granted to certain undertakings entrusted with the operation of services of general economic interest, O.J. 2005, L 312, p. 67. 50 Case C-269/90 Technische Universität München [1991] ECR I-5469 (5501, para. 26). 49 See CFI, Case T-613/97 UFEX [2006] ECR II-1531 (para. 128). 51 See CFI, Case T-613/97 UFEX [2006] ECR II-1531 (1581, para. 138). p g , , , p 49 See CFI, Case T-613/97 UFEX [2006] ECR II-1531 (para. 128). 50 Case C-269/90 Technische Universität München [1991] ECR I-5469 (5501, para. 26). 51 b) The cost-standard for cross-subsidisation b) The cost-standard for cross-subsidisation When it comes to the central question of the “right” cost-standard for monitoring and detecting cross-subsidisation, the interrelation of cost-accounting rules and the legal requirements of Community law deserve our full attention. Even if it might be in practice a quite self-evident starting point, it appears important to note that the allocation of costs has to be made on the basis of generally accepted cost accounting principles, as the Commission decision of November 28, 200552 states in Article 5 paragraph 2. This ensures that state aid control of cross-subsidisation relies in the very first place on a “best practice” approach, which is in particular completed by the requirements of the transparency directive. On that basis, the opposing aims of the prohibition of any state aid in Article 87 paragraph 1 EC and the necessity of ensuring workable conditions for public services according to Article 86 paragraph 2 and Article 16 EC have to guide our reflections. The basic idea is to keep the prohibition of state aid effective, even if aid is granted by way of cross-subsidisation. This purpose may, of course, not endanger the good functioning of public service operations, which enjoy a particular recognition in Community law. To me it seems quite inappropriate to strive for a hierarchical understanding of both values in terms of Community law. The challenge is rather to achieve practical solutions designed for an equivalent realisation of both objectives. To my mind, the rationale of the jurisprudence in Chronopost seems to be rightly based on this balancing approach. The formula for cost-allocation used in Chronopost reflects quite clearly the commitment to a strict application of state aid rules without interfering with the conditions necessary for the operation of services of general economic interest, although this standard might indeed not have been written with the precision of a text book in business administration. Nonetheless, it becomes quite clear that the focus of the judgement is on the additional cost-concept. Therefore, as the judgement states, prices charged to a subsidiary need to cover all the additional variable costs incurred in providing the assistance for a service in liberalized markets, an appropriate contribution to the fixed costs arising from the use of the existing postal network and an adequate return on capital investment used for competitive activities53. 52 Commission decision of 28 November 2005 on the application of Article 86 (2) of the EC Treaty to State aid in the form of public service compensation granted to certain undertakings entrusted with the operation of services of general economic interest, O.J. 2005, L 312, p. 67 (Footnote 45). p g , , , p ( ) 53 Joined Cases C-83/01 P, C-93/01 P, 94/01 P Chronopost [2003] ECR I-6993 (7042, para. 40). a) Scope and intensity of judicial review But still, the question remains whether it would be compatible with Article 87 paragraph 1 EC to assume that the Commission could lawfully, by a certainly non-arbitrary and most likely reasonable choice between different methods of calculation, come in one particular case to opposite conclusions as regards the existence of state aid? One could, of course, consider in line with the UFEX-judgement that this would constitute a manifest error of assessment51, if that could be demonstrated to the Court. But again, would the same logic apply if not the existence of an aid, but only the amount were to differ considerably according to the choice of the calculation method? Where could one draw the line? In any event, we will follow with particular interest how the Commission will eventually justify for what reasons a particular cost-allocation standard has been chosen in one case while using a different cost-allocation standard in another case. 49 See CFI, Case T-613/97 UFEX [2006] ECR II-1531 (para. 128). 50 Case C-269/90 Technische Universität München [1991] ECR I-5469 (5501, para. 26). 51 See CFI, Case T-613/97 UFEX [2006] ECR II-1531 (1581, para. 138). EUI WP LAW 2008/28 © 2008 Thomas von Danwitz 9 Thomas von Danwitz Thomas von Danwitz 52 Commission decision of 28 November 2005 on the application of Article 86 (2) of the EC Treaty to State aid in the form of public service compensation granted to certain undertakings entrusted with the operation of services of general economic interest, O.J. 2005, L 312, p. 67 (Footnote 45). 53 Joined Cases C-83/01 P, C-93/01 P, 94/01 P Chronopost [2003] ECR I-6993 (7042, para. 40). c) Remaining uncertainties Nevertheless it is quite evident, that the problem of cost allocation between services of general economic interest and competitive services has not fully been resolved by the existing jurisprudence. Details of considerable economic importance have so far not been brought to the attention of the Court and, therefore, have remained undecided yet. Questions relating to an eventual difference in calculating additional costs or incremental costs, to the degree of admissible standardisation, to the timeframe for calculation which is important for shifting the demarcation line between variable and fixed costs and, finally, to the consequences of a top-down as opposed to a bottom-up approach do not find an answer in the rules on state aid law. Consequently, they have not been reflected in the jurisprudence of the Court. It should however be noted that the Commission has made an effort to cope with these problems in a discussion paper on Article 82 EC54. Here again, it remains to be seen whether those standards are fully transposable to state aid law. When it comes to the mission of Community Courts in that respect, it seems to be quite clear that enacting cost-allocation standards cannot be considered a genuine task of the judiciary and, beyond that, defining common rules for cost-accounting and cost- allocation in essence does not require a specifically legal reasoning. This only comes into play when the suitability and proportionality of such standards or the requirements of a homogeneous application respecting an equal treatment of all undertakings concerned give rise to legal doubts. But, in my perception, the elaboration of such standards in itself ought to be regarded as a non-judicial question which should be answered in accordance with generally recognized cost-accounting and cost-allocation standards. In that perspective, preference should be given to a best practice approach. 54 DG Competition discussion paper on the application of Article 82 of the Treaty to exclusionary abuses, para. 64 seq. 55 Case C-526/04 Laboratoires Boiron [2006] ECR I-7529 (7565, para. 51 seq.) 56 Ibid., para. 55. b) The cost-standard for cross-subsidisation In my understanding only those fixed costs can be taken into account which result from the actual use of the postal network for a particular competitive activity. Otherwise, this competitive activity would have to bear costs of services of general economic interest. Admittedly, the Commission’s decision on compensation of public services is not crystal clear in that crucial respect, when it states in Article 5 paragraph 2 under c) that costs allocated to a service of general economic interest may cover the variable costs and “a proportionate contribution to fixed costs common to both services of general economic interest and other activities”. But quite evidently, this provision can and has to be read in conformity with the requirements established in Chronopost, since they constitute an interpretation of primary law in Article 87 paragraph 1 EC. EUI WP LAW 2008/28 © 2008 Thomas von Danwitz 10 The Concept of State Aid in Liberalised Sectors The Concept of State Aid in Liberalised Sectors 54 DG Competition discussion paper on the application of Article 82 of the Treaty to exclusionary abuse para. 64 seq. 55 55 Case C-526/04 Laboratoires Boiron [2006] ECR I-7529 (7565, para. 51 seq.) 56 Ibid., para. 55. 57 See BVerfGE 115, 205; v. Danwitz, Der Schutz von Geschäfts- und Betriebsgeheimnissen im Recht der Regulierungsverwaltung, DVBl. 2005, p. 597. Thomas von Danwitz Thomas von Danwitz circumstances in this specific case, where a national court had the power to order a party to produce a certain document. In particular, this judgement has not dealt with the protection of business secrets and the possibility of rejecting claims for the production of documents on such grounds57. circumstances in this specific case, where a national court had the power to order a party to produce a certain document. In particular, this judgement has not dealt with the protection of business secrets and the possibility of rejecting claims for the production of documents on such grounds57. V. Conclusion This overview over the jurisprudence of Community Courts on the concept of aid in liberalised sectors has been an occasion to identify the main challenges for state aid control in this field and to reflect the state of Community law jurisprudence. In essence, it seems to me that the jurisprudence is well established on a balanced approach which takes the necessities of a stringent state aid control seriously without neglecting the importance of maintaining the good functioning of services of general economic interest. In particular, let me draw three conclusions from the foregoing: Firstly, we have to acknowledge that state aid control is not a generally usable, unconditioned instrument of regulatory policy for realising a level playing field in liberalised markets. State aid control is rather focused on the use and abuse of state resources in a competitive environment. In that respect, the role of state aid control in liberalised markets does not differ from the traditional mandate to render the prohibition of state aids in Article 87 paragraph 1 EC effective. Secondly, the discussion of the issue of imputability has shown that Community Courts will have to live up to the basic standards established in PreußenElektra and in Stardust Marine. Quite evidently, these judgments have not provided us with a standard of review which is so easy to handle that most proceedings before the Commission and the Community Courts have or will become superfluous. But, as you well know, these judgements were not adopted to fulfil such unrealistic expectations and no future judgment will do so. The standards of review in state aid control need to be designed in a way which takes two aspects into account. A certain degree of flexibility is necessary to ensure a reasonable application in the Commission’s practice and in particular in order to cover future evolutions. But a solid concept with a visible profile is indispensable to give Member States and public undertakings the orientation which they clearly need. Thirdly, the jurisprudence of the Court has developed a realistic approach to cross- subsidisation, which might still need some refinement and future evolution. In particular, judicial review will have to ensure a homogenous application of the criteria for cost allocation as they have been established in Chronopost. On that basis, procedural consequences, as they have been discussed in Laboratoires Boiron for the very first time, will certainly require more attention in the future. 3. Procedural consequences What remains is the question of procedural consequences and in particular, how a party can eventually prove features of overcompensation. Since Member States enjoy, in general, procedural autonomy, the requirements of Community law are limited to the principles of equivalence and effectiveness. On this basis, in Laboratoires Boiron the Court has dealt with the delicate question of the burden of proof55. According to the findings of the Court, in a situation in which it is likely to be impossible or excessively difficult for the required evidence to be produced by a party, a national Court is required to use all procedures available under national law for this purpose, including the production of a particular document56. By this finding, the Court evidently responds to the need arising from a particular situation to make sure that a national court can verify a potential overcompensation. There is indeed an undeniable need for such a procedural possibility in order to assess whether an alleged overcompensation in fact took place. But still, this judgement should not be taken as a general rule without limitations giving rise to all sorts of investigatory actions. It should rather be noted that the Court refers to the factual and legal EUI WP LAW 2008/28 © 2008 Thomas von Danwitz 11 Thomas von Danwitz V. Conclusion EUI WP LAW 2008/28 © 2008 Thomas von Danwitz 12
https://openalex.org/W4388703810	https://link.springer.com/content/pdf/10.1007/s00414-023-03116-9.pdf	English	null	Awareness of forensic anthropology in Switzerland: a survey among forensic practitioners, police, and prosecutors	International journal of legal medicine	2,023	cc-by	7,974	Abstract Forensic anthropology (FA) as a specialized discipline has been practised in multi-lingual Switzerland for over a decade. A variety of expertise regarding osteological assessments as well as facial image comparison (FIC) is provided by different centres. Nevertheless, information is lacking about the awareness of FA and its benefits for forensic investigations among forensic stakeholders. Therefore, a survey was sent to Swiss anthropologists (AN) and related professions (police officers, prosecutors, and forensic pathologists) to assess three main aspects: (1) the experience of working (biological/forensic) anthropologists within FA; (2) how FA is perceived by other professions within the legal system; and (3) identify gaps (if any) in understanding of FA with the aim to suggest avenues for improvement if necessary. The results show that awareness of FA varies by occupation and cantonal regions. In areas where close collaborations between forensic anthropologists (FAs) and other stakeholders have been formally established, be it with focus on osteological analyses or FIC, the awareness of FA competencies was superior to areas where this was not the case. An overwhelming majority of forensic actors expressed interest in continuing education related to the role of FA. These findings indicate that facilitation of communication and col- laboration leads to improvement in the awareness of the competencies of FAs and their contribution to forensic investigations. Keywords Forensic anthropology · Human remains · Questionnaire · Forensic collaboration · Legal medicine · Facial image comparison Keywords Forensic anthropology · Human remains · Questionnaire · Forensic collaboration · Legal medicine · Facial image comparison Awareness of forensic anthropology in Switzerland: a survey among forensic practitioners, police, and prosecutors Received: 14 August 2023 / Accepted: 29 October 2023 / Published online: 15 November 2023 © The Author(s) 2023 Received: 14 August 2023 / Accepted: 29 October 2023 / Published online: 15 November 2023 © The Author(s) 2023 * Negahnaz Moghaddam negahnaz.moghaddam@chuv.ch 6 Centre for Forensic Anthropology, School of Social Sciences, The University of Western Australia, Perth, Australia International Journal of Legal Medicine (2024) 138:1067–1077 https://doi.org/10.1007/s00414-023-03116-9 International Journal of Legal Medicine (2024) 138:1067–1077 https://doi.org/10.1007/s00414-023-03116-9 ORIGINAL ARTICLE ORIGINAL ARTICLE 5 Swiss Human Institute of Forensic Taphonomy, University Centre of Legal Medicine Lausanne-Geneva, Lausanne University Hospital and University of Lausanne, Chemin de la Vulliette 4, CH – 1000, 25 Lausanne, Switzerland Negahnaz Moghaddam and Zuzana Obertová contributed equally to this work. Introduction The anthropologists are based at the Unit of Visual Identification of Persons (VIP), with their casework encompassing mainly facial image compari- son (FIC) and more recently osteological analyses (collabo- rating with the Institute of Forensic Medicine at the Univer- sity of Zurich). educational pathways for FAs; as anatomists and forensic pathologists have been the traditional professions dealing with forensic requests concerning skeletal remains [1]. A similar diversity in the involvement of FAs in forensic casework has been reported in a survey by Obertová et al. [5] from 2019. However, there has been some progress since among 28 studied countries, 20 countries reported that anthropologists (AN) were involved to at least some extent in cases of skeletonized remains, while in Austria and Finland only FAs, not forensic pathologists, were han- dling such cases. Switzerland was not included in the [1] survey, but at that time, in contrast to many other Euro- pean countries, AN had already been working on forensic cases, holding positions at the Unit of Forensic Imaging and Anthropology at the University Centre of Legal Medi- cine in Lausanne/Geneva (CURML), and at the Depart- ment of Physical Anthropology (DPA) at the Institute of Legal Medicine in Bern. Apart from the FAs working at the above-mentioned institutions, there are freelance AN and AN employed within archaeological excavation firms. The primary case- work of these two groups concerns archaeological remains, with occasional involvement in forensic casework. There is no specific higher degree in forensic anthropology offered at Swiss universities, although AN/FAs at uni- versities in Basel, Bern, Lausanne and Geneva, and those from the FOR in Zurich, do provide individual lectures for under- and post-graduate courses in medicine, law, and archaeology. Switzerland is divided into 26 cantons. Besides the fed- eration, each canton has its own cantonal laws, judiciary, and police corps within a federal system. Although Switzerland has a population of only 8.7 million [6], there are seven insti- tutes of legal medicine (in alphabetical order: Aarau, Basel, Bern, Chur, CURML in Lausanne and Geneva, St. Gallen, Zurich). As mentioned previously, forensic anthropology can be found in Lausanne/Geneva and in Bern, while the Institute of Forensic Medicine in Zurich has a collaboration agreement with the AN based at the Zurich Forensic Science Institute since 2020. Prior studies have shown that acceptance and knowledge of FA on the national level varies widely [1, 5, 8–14]. Introduction Negahnaz Moghaddam and Zuzana Obertová contributed equally to this work. Forensic anthropologists (FAs) take on numerous roles (Fig. 1) since the tasks allocated to them are commonly dic- tated by the needs of local law enforcement agencies, as well as forensic and judiciary institutions. These tasks can vary from the analysis of human skeletal remains to facilitate identification, skeletal trauma analysis, visual identification of persons, or being part of Disaster Victim Identification (DVI) teams [1–4]. * Negahnaz Moghaddam negahnaz.moghaddam@chuv.ch 1 Zurich Forensic Science Institute, Zurich, Switzerland 2 Unit of Forensic Imaging and Anthropology, University Centre of Legal Medicine Lausanne-Geneva, Lausanne University Hospital and University of Lausanne, Lausanne, Switzerland Kranioti and Paine [1], in their overview of forensic anthropologists in Europe from 2011, have mentioned diverse employment situations for FAs within Europe, mainly freelance actives, and some employment in aca- demia or governmental institutes. While pointing out that among the 18 European countries evaluated in their study, most of the cases involving skeletal remains were handled by forensic pathologists and/or anatomists, and only in two countries physical anthropologists are reported to be involved in skeletal evaluation [1]. They argue that this division of tasks may be due to a lack of specific 3 Unit of Forensic Medicine, University Centre of Legal Medicine Lausanne-Geneva, Lausanne University Hospital and University of Lausanne, Lausanne, Switzerland 4 Unit of Forensic Medicine, University Centre of Legal Medicine Lausanne-Geneva, Geneva University Hospital and University of Geneva, Geneva, Switzerland 5 Swiss Human Institute of Forensic Taphonomy, University Centre of Legal Medicine Lausanne-Geneva, Lausanne University Hospital and University of Lausanne, Chemin de la Vulliette 4, CH – 1000, 25 Lausanne, Switzerland (0121 3456789) 3 3456789) 3 1068 International Journal of Legal Medicine (2024) 138:1067–1077 Fig. 1 Main areas of work and responsibilities of forensic anthropologists FA at the CURML lies mostly with forensic casework and research. The Department of Physical Anthropology at the Institute of Forensic Medicine in Bern is also involved in forensic casework. However, as described by Indra and Lösch [7] the main investigations concern archaeological skeletal remains and only a smaller proportion relates to forensic contexts. The Zurich Forensic Science Institute (FOR), an independent governmental police institute, is responsible for crime scene investigations and provides forensic specialist consultations to the police of mainly German-speaking cantons. Results At the beginning, each participant was asked to name the canton of their primary employment. No identifying information beyond the main canton of employment was collected. The questionnaires consisted of eight to eleven questions and were organised into three main parts: expertise of FA, awareness of the role of FA, and education and train- ing. For each profession, the survey questions were adapted Introduction For example, just recent and very limited involvement of FAs in relevant forensic investigations are reported from Finland [14]; for Sweden, only some FAs involvement in casework The CURML covers the forensic needs of French- speaking Switzerland (mainly the cantons of Vaud, Valais, Geneva, Fribourg, Neuchâtel, and the Jura). The focus of 1 3 1069 International Journal of Legal Medicine (2024) 138:1067–1077 mainly depending on the knowledge/personal contacts of the crime scene investigator or forensic pathologists in charge are stated by Alfsdotter [8] and Alfsdotter et al. [13], whereas clear collaboration structures for FAs in Denmark exist and different aspects of FA are practised at three dif- ferent departments of forensic pathology [12]. Compared to these three countries, the Swiss situation of forensic anthropology and forensic anthropologists seems to be most comparable with the Swedish situation. Despite the varied employment opportunities for AN in Switzerland, little is known about the knowledge of their role regarding human identification and other aspects of forensic anthropological expertise among other actors involved in forensic casework, such as police officers (PO), prosecutors (PR), and forensic pathologists (FP). to the field of expertise; however, the content of the ques- tions was similar to ensure comparability across the different professions. Methods To reach the members of the various professions targeted by the questionnaires, invitations to distribute the survey were sent by email to various national working groups (WGs) in 2022. The WGs included the Swiss Anthropological Society (SGA/SSA), police WGs in the individual cantons and at the FedPol (the Swiss Federal Police), as well as the prosecu- tors’ WG. Forensic pathologists were approached through their respective institutions with emails sent to all institutes of legal medicine in Switzerland. The aim of this study was therefore to survey the state of FA in Switzerland and more specifically the perception of the role of FA both within the discipline of anthropology, and among other actors of legal proceedings likely to engage with FAs, namely police officers, prosecutors, and foren- sic pathologists. The objectives were to acquire knowledge about (1) the experience of practitioners of FA in Switzer- land and the methods used in forensic analyses; (2) how FA is perceived by other professions within the legal system; and (3) identify gaps (if any) in the understanding of the role of FA and suggest potential avenues of improvement in areas of need. The questionnaires were available online for two months, after which time all results were exported as Microsoft Excel files. Descriptive statistics were performed using Microsoft Excel using percentages of responses compared among the professions and, where relevant, among cantons of employ- ment grouped by official language. The cantons of employ- ment or canton groups were classified as follows I. DEC—cantons with German as the official language (Argovia, Appenzell Outer-Rhodes, Appenzell Inner- Rhodes, Basel, Basel District, Glarus, Lucerne, Nid- walden, Obwalden, Schaffhausen, Schwyz, Solo- thurn, St Gallen, Thurgau, Uri, Zug, Zurich);fi II. FRC—cantons with French as the official language (Geneva, Vaud, Neuchâtel, Jura);fi Materials and methods III. MLC—cantons with two or more official languages (Bern, Fribourg, Valais, Grisons), the Canton of Ticino, where Italian is the official language, was also included in this group. Materials Four anonymous questionnaires were designed for this sur- vey, using the LimeSurvey platform (LimeSurvey Commu- nity Edition, Version 3.28.19+220712), access to which was provided through the University of Geneva, Switzerland. Each questionnaire targeted a specific profession: anthro- pologists (AN), police officers (PO), prosecutors (PR), and forensic pathologists (FP). The questionnaires were designed in English, and then translated into three of the official Swiss languages [15]: German, French, and Italian. The English version of the questionnaires can be found in Supplementary Material 1. The responses of police officers (PO), prosecutors (PR), and forensic pathologists (FP), regarding their knowledge of different methodologies potentially associated with FA (question 7) were grouped into four main categories: general (morphological, metric), specialized (FIC, facial approxima- tion), technical (imaging, histology), and analytical (dating, isotopes). Response rate Overall, there were 188 responses, 16 responses from anthropologists (AN), 65 from police officers (PO), 80 from prosecutors (PR), and 17 from forensic pathologists (FP). Table 1 shows the distribution of responses by canton group 1 3 1070 International Journal of Legal Medicine (2024) 138:1067–1077 studying FA as a major degree in Switzerland (question 7), 14 (88%) responded positively. including the proportion of the Swiss population living within these cantons, which was used as a proxy to assess the response rates. According to this proxy comparison the responses from AN and PO in DEC and from FP in MLC (a single response) were underrepresented. Responses by AN regarding their employment and education (questions 1–5, 7) When PO, PR, and FP were asked if they are familiar with the role of FAs (question 1), 14% of PO, 25% of PR, and 12% of FP responded negatively, while 32% of PO, 56% of PR and 6% of FP responded they are not sure. Question 2 asked whether PO, PR and FP are aware of FAs working in Switzerland, with 95% of PO, 66% of PR and 94% of FP responding ‘yes’, while question 3 queried whether they know how to contact a FAs, with, 62% of PO, 45% of PR and 76% of FP responding positively. When FP were asked if they employed AN in their institution or whether they had any cooperation with AN (question 4, specific to FP), 35% responded that they have AN employed in their institution, 30% that they cooperate with AN, while 35% replied that they do not cooperate with AN (this response was only given by FP in DEC). Regarding their current employer (question 1) AN answered as follows: four (25%) were employed by a university, five (31%) by a forensic institute, four (25%) by an archaeologi- cal company, and three (19%) were self-employed. The frequency of involvement in forensic casework was queried in question 2. Here, one participant reported no interest in being involved in forensic work, four each (25% each or 75% in total) responded that they have so far not been involved in any forensic casework, they have rarely (<once a year) worked on forensic cases or they have worked on forensic cases occasionally (more than once a year), while three (19%) reported that they were involved regu- larly in forensic casework, which represents their main work. Responses to the educational background (question 3) showed that the highest degree of 9 (56%) responders was PhD or equivalent, with the remaining 6 (44%) having a master’s degree or equivalent. One-quarter of the 16 AN responders had a forensic degree (question 4), while none had acquired their forensic degree in Switzerland (question 5). When asked if they would welcome the possibility of The frequency of involvement in forensic casework was queried in question 2. Table 1 Distribution of responses by cantons grouped according to official languages; AN, anthropologists; PO, police officers; PR, prosecutors; FP, forensic pathologists Responses by AN regarding their employment and education (questions 1–5, 7) Here, one participant reported no interest in being involved in forensic work, four each (25% each or 75% in total) responded that they have so far not been involved in any forensic casework, they have rarely (<once a year) worked on forensic cases or they have worked on forensic cases occasionally (more than once a year), while three (19%) reported that they were involved regu- larly in forensic casework, which represents their main work. Responses to the educational background (question 3) showed that the highest degree of 9 (56%) responders was PhD or equivalent, with the remaining 6 (44%) having a master’s degree or equivalent. One-quarter of the 16 AN responders had a forensic degree (question 4), while none had acquired their forensic degree in Switzerland (question 5). When asked if they would welcome the possibility of Table 2 shows the distribution of responses to questions 4 (PO, PR) and 5 (FP), which asked how often PO, PR and FP work with FAs. The most common responses were ‘never’ or ‘rarely (<1 case annually), accounting for 81% of responses from PO, 89% from PR and 70% from FP. While two-thirds of FP in FRC reported working with FA occasionally or regularly, this was the case for 10% of DEC FP. Table 1 Distribution of responses by cantons grouped according to official languages; AN, anthropologists; PO, police officers; PR, prosecutors; FP, forensic pathologists Official language In canton/ Profession AN PO PR FP Percentage population living in canton group Canton group n (%) n (%) n (%) n (%) German (DEC) 5 (31) 19 (29) 52 (65) 10 (59) 56% French (FRC) 5 (31) 26 (40) 12 (15) 6 (35) 18% Multilingual (MLC) 6 (38) 20 (31) 16 (20) 1 (6) 26% Total 16 65 80 17 Table 2 Responses of police officers, prosecutors, and forensic pathologists about how frequently they work with a FAs How often do you work with forensic anthropologists? Profession Never Rarely (<1/yr) Occasionally Regularly (several times/yr) n (%) n (%) n (%) n (%) Police officers 15 (23) 38 (58) 9 (14) 3 (5) Prosecutors 38 (48) 33 (41) 6 (8) 3 (4) Forensic pathologists 6 (35) 6 (35) 2 (12) 3 (18) Table 2 Responses of police officers, prosecutors, and forensic pathologists about how frequently they work with a FAs How often do you work with forensic anthropologists? Responses by AN regarding their employment and education (questions 1–5, 7) Profession Never Rarely (<1/yr) Occasionally Regularly (several times/yr) n (%) n (%) n (%) n (%) Police officers 15 (23) 38 (58) 9 (14) 3 (5) Prosecutors 38 (48) 33 (41) 6 (8) 3 (4) Forensic pathologists 6 (35) 6 (35) 2 (12) 3 (18) 1 3 3 1071 International Journal of Legal Medicine (2024) 138:1067–1077 In question 5, PO and PR were asked if they usually con- tact FAs directly, or if they had a different contact person in cases of skeletal remains. Of the 80 PR, 28% replied they would contact FAs directly, 20% had a different contact person and 53% had as of the time of enquiry no interac- tion with such cases. Among the 16 PR, who would con- tact someone else, 88% would contact an institute of legal medicine and 6% would contact a forensic institute. Among PO, 17% would contact FAs directly, 52% would contact someone else, and 31% had so far, no interaction with skel- etal cases. Among the 34 PO, who would turn to a differ- ent contact person, 79% would contact an institute of legal medicine and 9% would contact a forensic institute. (The responses mentioning a forensic institute were all from DEC and pertained to the Zurich Forensic Science Institute.)i 94% establishment of biological profile, 88% assessment of skeletal trauma and pathology, 56% assistance with iden- tification of human remains, 38% estimation of age in the living, 38% assistance with humanitarian action, and 69% research. The differentiation between human and nonhuman remains was considered a main FA task by 92% of PO, 58% of PR and 82% of FP, the establishment of biological profile by 85% of PO, 65% of PR, and 76% of FP, the assessment of skeletal trauma and pathology by 54% of PO, 46% of PR, and 59% of FP, the assistance with identification by 45% of PO, 59% of PR, and 24% of FP, the assessment of the post- mortem interval by 63% of PO, 48% of PR, and 71% of FP, the age estimation in the living by 46% of PO, 54% of PR, and 12% of FP, and dental assessment by 31% of PO, 44% of PR, and 12% of FP (Fig. 2). Question 8, specific to PO, asked what they usually do if they encounter a case concerning skeletal remains. Responses by PO, PR, and FP regarding their knowledge of different methodologies associated with FA (question 7) The responses for question 7 were grouped into four main categories: general (morphological, metric), specialized (FIC, facial approximation), technical (imaging, histology), and analytical (dating, isotopes). In this question, PO, PR, and FP were asked whether they are aware of certain meth- odological approaches being applied by FAs. Figure 3 shows responses by PO, PR and FP concerning the individual meth- ods. Overall, FP were more likely to be aware of FAs using general (morphological 65% and metric assessment 53%), and technical methods (imaging 94% and histology 41%) compared with PO (51%, 15%, 40%, and 15%, respectively) and PR (39%, 9%, 66%, and 19%, respectively). In contrast, PO and PR were more aware of FIC being performed by FAs (31% and 38%, respectively) versus 18% of FP. i Questions 8–9 (PR, FP) and 9–10 (PO) queried whether PO, PR and FP have ever attended a presentation on the role of FA, and whether they would be interested in such a pres- entation, respectively. Overall, 25% of PO, 18% of PR, and 76% of FP responded positively (Table 3). In comparison, all FP in FRC and MLC attended a presentation on the role of FA versus 60% in DEC. In total, 88% of PO, 96% of PR and 95% of FP would be interested in a presentation on the role of FA (Table 3). Responses by AN regarding their employment and education (questions 1–5, 7) Among the 65 responders, 12% said they do not work such cases, 2% would contact a FAs, 29% would collect the remains and bring them to FP, and 57% responded that their actions would depend on the specific case. Responses by PO, PR, FP, and AN regarding the main tasks of FAs (question 6) In question 6 the AN were asked what they consider as their main tasks, while question 6 posed to PO, PR and FP queried their perception of the main tasks of FAs. The AN answered positively concerning the following main tasks: Figure 4 shows the variation of responses by canton group and profession regarding the awareness of FIC being a method used by FAs. This method was selected because it is performed by forensic anthropologists solely at the Zurich Forensic Science Institute located in DEC. Both PO and PR in DEC were more frequently (53% and 42%, respectively) aware of facial image comparison being performed by FAs compared with PO and PR in FRC (27% and 17%, respec- tively) and MLC (15% and 38%, respectively). Table 3 Responses of police officers, prosecutors and forensic pathol- ogists about whether they have attended, or they would be interested to attend a presentation on the role of FA Have you ever attended a presentation on the role of FA? Would you be interested in a presentation on the role of FA? Profession Yes Yes Yes, but only if it would take less than 2 hours n (%) n (%) n (%) Police officers 16 (25) 44 (68) 13 (20) Prosecutors 14 (18) 50 (63) 26 (33) Forensic pathologists 13 (76) 12 (71) 4 (24) Table 3 Responses of police officers, prosecutors and forensic pathol- ogists about whether they have attended, or they would be interested to attend a presentation on the role of FA Table 3 Responses of police officers, prosecutors and forensic pathol- ogists about whether they have attended, or they would be interested to attend a presentation on the role of FA Education of Swiss (forensic) anthropologists or regularly, compared with 10% of FP in DEC. This may be explained by the employment strategies for FAs differing between these canton groups; with all FP in FRC respond- ing that they employ FAs in their institution and some FP in DEC replying that they do not cooperate with FAs at all. Due to a lack of responses from FP in MLC, it was not pos- sible to make a valid conclusion. This difference may also partly reflect the duration of established FA positions in the various regions. While FA positions in FRC have been estab- lished for more than ten years, in DEC the FAs at the FOR have taken over the responsibility for the analysis of skeletal remains from FP in 2020. Responses to questions about the educational and profes- sional background of AN revealed that more than half held a PhD or equivalent, but only a quarter of the responders had forensic-specific degrees. Those degrees were acquired at foreign universities as there is currently no possibility of gaining a FA degree in Switzerland, although almost 90% of the AN would welcome the possibility of studying a major degree in the discipline in Switzerland. The current employment status of responding AN was almost equally split among universities, forensic institutes, archaeolog- ical excavation companies and freelancing. This finding is in accordance with previous surveys, which also reported a range of different employment opportunities for FAs [1, 5, 16]. Overall, 62% of PO, 45% of PR, and 76% of FP reported that they know how to contact FAs. However, only a fraction of police officers (17%) and prosecutors (28%) reported that they would contact FAs directly should the need arise. More than half of PO and one in five PR (with more than 50% having not worked on a case, which required an interaction with FAs) would prefer contacting someone else, mostly an institute of legal medicine. When PO were asked about how they usually handle a case concerning skeletal remains, only 2% would contact FAs, more than half would tailor their action to the specific case and 29% would collect the remains and bring them to FP. Discussion 1 3 3 International Journal of Legal Medicine (2024) 138:1067–1077 1073 Fig. 4 Responses of police officers, prosecutors and foren- sic pathologists considering awareness of facial image com- parison (FIC) being performed by forensic anthropologists (FAs) by canton group Discussion Forensic anthropologists have had positions at Swiss univer- sities and governmental institutions for more than a decade, but little is known about how the role of FA is understood by different actors (police officers, prosecutors, and forensic pathologists) involved in legal proceedings regarding human identification and other aspects of forensic anthropological 1 3 1 3 1072 International Journal of Legal Medicine (2024) 138:1067–1077 0% 20% 40% 60% 80% 100% Human/Nonhuman diﬀeren a on Biological proﬁle Pathology/Trauma Iden ﬁca on Postmortem interval Age es ma on Living Dental assessment Police oﬃcers Prosecutors Forensic pathologists Percentage of forensic actors considering this a FAs task FAs tasks 0% 20% 40% 60% 80% 100% Human/Nonhuman diﬀeren a on Biological proﬁle Pathology/Trauma Iden ﬁca on Postmortem interval Age es ma on Living Dental assessment Police oﬃcers Prosecutors Forensic pathologists Percentage of forensic actors considering this a FAs task FAs tasks Police oﬃcers Prosecutors Forensic pathologists Fig. 2 Responses of police officers, prosecutors and forensic pathologists considering certain tasks typical for forensic anthropologists (FAs) Fig. 2 Responses of police officers, prosecutors and forensic pathologists considering certain tasks typical for forensic anthropologists (FAs) esponses of police officers, prosecutors and forensic pathologists considering certain tasks typical for forensic anthro Fig. 2 Responses of police officers, prosecutors and forensic pathologists considering certain tasks typical for forensic anthropologists (FAs) 0% 20% 40% 60% 80% 100% Morphological Metric Facial image comparison Facial approximaon Imaging (CT/RTG) Histology Dang C14 Isotopes Police oﬃcers Prosecutors Forensic pathologists Percentage of forensic actors aware of a method being used by FAs Methods Fig. 3 Responses of police officers, prosecutors, and forensic pathologists considering awareness about certain methods being applied by foren- sic anthropologists (FAs); CT: computed tomography, RTG: radiography Fig. 3 Responses of police officers, prosecutors, and forensic pathologists considering awareness about certain methods being applied by foren- sic anthropologists (FAs); CT: computed tomography, RTG: radiography expertise. To address this question, a survey was designed and distributed among anthropologists and the above-men- tioned forensic professionals in Switzerland. The results of this survey help identify potential knowledge gaps, which can in turn be translated into evidence-based education pathways. In addition, information about different employ- ment and cooperation structures involving FA and the other forensic actors can guide future working environments and collaboration strategies in cases of skeletal remains and human identification not just in Switzerland but worldwide. Involvement of AN/FAs in forensic casework and collaboration among forensic actors Only three (19%) forensic anthropologists reported working on forensic cases regularly. These forensic anthropologists worked at a forensic institute (meaning an institute of foren- sic sciences or institute of legal medicine) and four more AN worked on forensic cases occasionally (>1 case/annually); two reported working for a forensic institute and two for a univer- sity. The case numbers can vary greatly per year but overall, an increase in FAs’ involvement in skeletal investigations is recorded in Switzerland according to the authors and Indra and Lösch [7]. Considering this, the reported collaboration of other forensic stakeholders with FAs occasionally or regularly (19% of PO, 11% of PR and 30% of FP) is not surprising. Ideally, the police would directly contact FAs when skel- etal remains are found or FP would forward the request to FAs, seeing that they seem to be the primary contact so far (Fig. 5). However, as the results of this survey indicate, there is an overall knowledge gap about FAs working in Swit- zerland, how to contact them, and what their competencies are across the different forensic actors. Moreover, there are currently no standardised national guidelines on how to deal with skeletal human remains, which results in PO, who are The cooperation patterns varied by canton group, where 67% of FP in FRC reported working with FAs occasionally 1 3 1 1074 International Journal of Legal Medicine (2024) 138:1067–1077 International Journal of Legal Medicine (2024) 138:1067–1077 Fig. 5 Depicting the current pathway of handling skeletal remains according to the present finding for Switzerland and an advised improved pathway optimizing the inclusion of FA knowledge in relevant casework Fig. 5 Depicting the current pathway of handling skeletal remains according to the present finding for Switzerland and an advised improved pathway optimizing the inclusion of FA knowledge in relevant casework Fig. 5 Depicting the current pathway of handling skeletal remains according to the present finding for Switzerland and an advised improved pathway optimizing the inclusion of FA knowledge in relevant casework future. These findings highlight that although the knowledge about FAs competencies is limited, the willingness to learn more about the discipline and how it can contribute to foren- sic investigations is present and should be built upon by FAs providing lectures, seminars and other educational opportu- nities about their field of work. Involvement of AN/FAs in forensic casework and collaboration among forensic actors Moreover, close cooperation (especially between FAs and FP) clearly fosters an under- standing of the respective roles. usually first responders in such cases, acting on a case-by- case basis rather than having a standardised procedure when skeletal remains are concerned. The solution to strengthen collaborations, streamline workflows and provide the most appropriate expertise in cases of skeletal remains would be to first establish national guidelines on how to deal with skeletal human remains, and second, a structured dissemina- tion of information about FA competencies to other foren- sic stakeholders. Similar recommendations were previously emphasized in international studies on FAs involvement in casework [1, 5, 8, 13, 16]. The responses regarding the main tasks of FA can be interpreted in light of FA-specific literature, assigning an expected value of 100% to tasks, which are typically addressed in FA textbooks (e.g. [17–20]) or have been identified as main tasks in other FA surveys/studies [4, 5, 9, 12, 14, 21, 22]. These tasks include the evaluation of human vs. non-human bones, the assessment of skeletal features leading to biological profile, the descriptive evalu- ation of skeletal trauma and pathology, and the assistance in the identification process of human remains. Other tasks that have been associated with FA but are either country-/ laboratory-specific or are considered emerging areas in FA were assigned partially arbitrary expected percentages fol- lowing the results of recent surveys [1, 5]. These tasks, with expected percentages noted in brackets, were the evaluation of the post-mortem interval (75%), age estimation of the liv- ing (50%), and dental assessment (20%). It should be noted that these percentages represent a reflection of mixed, differ- ent wholes, such as how often they are performed in different countries by FAs (in some this may be 100%, while in others 0%), how often they are performed by forensic anthropolo- gists (some may perform these tasks on a daily basis, while others rarely or never), or whether these are perceived as Familiarity with the role of FAs, their competencies and methods among police officers, prosecutors and forensic pathologists This survey found that more than 80% of Swiss FP reported being familiar with the role of FA, compared with only 54% of PO and 19% of PR. While the majority of PO and FP were aware of FAs working in Switzerland, only two-thirds of PR reported such knowledge. The different responses about the familiarity with the role of FA closely reflected the variation in positive replies regarding the attendance of a presenta- tion on FA with 76% FP but only 18% of PR reporting ever attending a presentation on FA. Similarly, the difference in the cooperation patterns of FP and FAs found between DEC and FRC are mirrored in the differences regarding previous attendance of a presentation on FA, with all FP in FRC but only 60% of FP in DEC reporting presentation attendance. Most respondents (88% of PO, 96% of PR and 95% of FP) would welcome a presentation on the role of FA in the 1 3 1075 International Journal of Legal Medicine (2024) 138:1067–1077 FA tasks (some may be perceived as such but may never be performed by FAs). of FA diverted into specialized and distinct requests closely related to forensic sciences. Table 4 shows the comparison between the expected and the survey responses regarding FA competencies. The responses of AN reflected well the expected values regard- ing the establishment of biological profile, and the assess- ment of skeletal trauma and pathology (note: they were not asked about human/nonhuman differentiation, post-mortem interval and dental assessment). However, only 56% of AN were considered assistance with the identification of human remains as their task. This surprising result may be explained by the fact that only a quarter of the responding AN had a forensic-specific degree and a quarter have not been involved in forensic casework at all. Although similar methods are used in biological and forensic anthropology, the objectives of the subdisciplines differ as Passalacqua et al. [10] recently argued that forensic anthropology should no longer be considered a sub-discipline of bioarchaeology (biological anthropology) as the responsibilities and tasks The response of PO aligned well with the expected val- ues for human/nonhuman differentiation, biological profile, estimation of post-mortem interval and age estimation of the living, while they exceeded the expected value in dental assessment and only half of the PO viewed skeletal trauma/ pathology assessment and identification as FA tasks. -15% from the expected value, e.g., the initial value of 50% has the following increments: 42.5%, 35%, 27.5%, 20%, 12.5% and 5%. No fill is assigned to values that exceed the expected value by more than 5% Familiarity with the role of FAs, their competencies and methods among police officers, prosecutors and forensic pathologists Pros- ecutors have only considered age estimation of the living in agreement with the expected value, while their response exceeded the expected value for dental assessment and was lower than 30% of the initial value for human/nonhuman differentiation, biological profile, assessment of skeletal trauma/pathology, post-mortem interval, and assistance with identification. These values seem to reflect the lack of aware- ness of PR about the role of FA as shown in their reported responses throughout this survey. Further education for PR, which they reported to be welcome, should therefore be a priority, considering that they commission forensic experts. Table 4 The comparison of the percentage of positive responses and expected values concerning the perception of main FA tasks among the various professions. The darker the fill, the closer the response to the expected value. The increment of greyscale difference represents le 4 The comparison of the percentage of positive responses and ected values concerning the perception of main FA tasks among various professions. The darker the fill, the closer the response to expected value. The increment of greyscale difference represents 15% from the expected value, e.g., the initial value of 50% has t following increments: 42.5%, 35%, 27.5%, 20%, 12.5% and 5%. N fill is assigned to values that exceed the expected value by more th 5% Positive responses Forensic anthropologists Police officers Prosecutors Forensic pathologists Expected values Human/nonhuman differentiation NA 92% 58% 82% 100% Biological profile 94% 85% 65% 76% 100% Assessment of skeletal trauma/pathology 88% 54% 46% 59% 100% Assistance with identification 56% 45% 59% 24% 100% Assessment of post- mortem interval NA 63% 48% 71% 75% Age estimation of the living 38% 46% 54% 12% 50% Dental assessment NA 31% 44% 12% 20% 1 3 International Journal of Legal Medicine (2024) 138:1067–1077 1076 The anthropologists were additionally asked if they considered assistance with humanitarian action and research as one of the main tasks, with positive responses of about one-third and two-thirds, respectively. Since FA in humanitarian actions and disaster victim identification is an emerging field [23–25] it is not surprising that a relatively low number accounted for this as a main task of FAs (38%). Even though conducting research is not requested at every FA employment, research is signifi- cant for the discipline and positive responses would have been expected to be higher (69%). Familiarity with the role of FAs, their competencies and methods among police officers, prosecutors and forensic pathologists This discrepancy may be explained by the varied employment situations of the responding AN (e.g., a quarter were freelancers) and pos- sibly the fact that research might not be part of employ- ment contracts, as well as limited time, money and/or access to specimens, which might hinder research in the field to be considered. Conclusion To our knowledge, this is the first study addressing the perception of forensic anthropology among related profes- sions, including police officers and prosecutors. This sur- vey showed inherent gaps in the knowledge of the role and competencies of FAs among police, prosecutors, and foren- sic pathologists in Switzerland. However, an overwhelming majority of the participants indicated willingness to learn more about the discipline of FA and therefore it is recom- mended that forensic anthropologists promote their field of expertise by disseminating information about FA tasks and competencies in lectures or seminars for other forensic actors. While FAs have been working in Switzerland for over a decade, it seems that establishing such a special- ized discipline in a country with multiple official languages and a variety of local laws and regulations may be a slow process, where close collaboration networks (e.g., forensic anthropologists being employed at institutions with foren- sic pathologists or the police) lead to improved understand- ing of the benefits of FA in forensic investigations. i The survey concluded by asking about the awareness of PO, PR and FP regarding methodological approaches used by FAs. The knowledge of general methods (morphologi- cal and metric), which are routinely used by FAs for the assessment of human/nonhuman origin, biological profile, and identification, was surprisingly limited, ranging between 9% of PO for metrics and 65% in FP for morphology. Over- all, FP were more likely to be familiar with the methodo- logical approaches of FA than PO and PR, except for FIC. However, FIC was registered by about one-third of PO and PR, although this increased to almost 50% in DEC, where this specialized methodology is applied routinely by FAs and trained experts employed at the Zurich Forensic Science Institute. It is not surprising that FP are less aware of FIC than PO and PR as the latter two professionals are mainly give mandates in such cases. i By exploring the perception of the role of FA among other forensic actors in Switzerland, a country with strong cantonal division in laws and language, the findings of this survey may be extrapolated to European or even global level in relation to the optimization of employment struc- tures and educational pathways in forensic disciplines. Supplementary Information The online version contains supplemen- tary material available at https://doi.org/10.1007/s00414-023-03116-9. Supplementary Information The online version contains supplemen- tary material available at https://doi.org/10.1007/s00414-023-03116-9. Acknowledgements The authors thank all participants for their time and insights. We are also grateful to our colleagues who helped to distribute the survey within the various working groups. We take this opportunity to thank the anonymous reviewer for their constructive comments. A limitation of this study is the underrepresentation of responses from DEC AN and PO, as well as MLC FP. The reasons for the observed response rate are difficult to evaluate as the survey was distributed via e-mail to the different working groups with the request to pass the link to the relevant personnel. This, however, relies on the goodwill of the first in line to pass the survey on as well as on the participants to spare time to respond to a survey that relates to the highly specialized discipline of forensic anthropology. Nevertheless, the high participa- tion rate of PR of the various canton shows the interest of the important representatives of the cantonal government in criminal cases. However, the results of this question- naire showed that the competencies and skills of FAs are often not well understood among forensic stakeholders. This survey indicates that in the future, FAs in Switzer- land should establish structured and close collaboration with PO, PR and FP, provide more insights into FA to the various professions, and finally continue their own profes- sional training. Author contribution All authors contributed to the study conception and design. The first draft of the manuscript was written by Inga Siebke and all authors commented on previous versions of the manuscript. All authors read and approved the final manuscript. i Inga Siebke: conceptualization, methodology, project administra- tion, visualization, writing - original draft, writing—review and editing. Claudine Abegg: data curation, methodology, project administration, software. Tony Fracasso: supervision, review and editing. Negahnaz Moghaddam: conceptualization, methodology, project administration, writing—review and editing. Zuzana Obertová: formal analysis, meth- odology, project administration, visualization, writing—original draft, writing—review and editing. Funding Open access funding provided by University of Lausanne References 1. Kranioti E, Paine R (2011) Forensic anthropology in Europe: an assessment of current status and application. J Anthropol Sci 89:71–92. https://doi.org/10.4436/jass.89002 20. Hackman L, Delabarde T, Roberts J, de Boerr H (2022) Forensic anthropology: a primer for courts. The Royal Society. 21. Ubelaker DH, Shamlou A, Kunkle A (2019) Contributions of forensic anthropology to positive scientific identification: a criti- cal review. Forensic Sci Res 4(1):45–50. https://doi.org/10.1080/ 20961790.2018.1523704i 2. Park DK, Park KH, Ko JS, Kim YS, Chung NE, Ahn YW et al (2009) The role of forensic anthropology in the examination of the Daegu subway disaster (2003, Korea). J Forensic Sci 54(3):513– 518. https://doi.org/10.1111/j.1556-4029.2009.01027.x 22. Marquez-Grant N, Roberts J (2021) Redefining forensic anthropol- ogy in the 21st century and its role in mass fatality investigations. Eur J Anat 25:19–34 3. Ubelaker DH, Shamlou A, Kunkle AE (2019) Forensic anthro- pology in the global investigation of humanitarian and human rights abuse: perspective from the published record. Sci Justice 59(2):203–209. https://doi.org/10.1016/j.scijus.2018.10.008 3. Ubelaker DH, Shamlou A, Kunkle AE (2019) Forensic anthro- pology in the global investigation of humanitarian and human rights abuse: perspective from the published record. Sci Justice 59(2):203–209. https://doi.org/10.1016/j.scijus.2018.10.008 23. de Boer HH, Blau S, Delabarde T, Hackman L (2019) The role of forensic anthropology in disaster victim identification (DVI): recent developments and future prospects. Forensic Sci Res 4(4):303–315. https://doi.org/10.1080/20961790.2018.1480460 p g j j 4. Cattaneo C (2007) Forensic anthropology: developments of a clas- sical discipline in the new millennium. Forensic Sci Int 165(2– 3):185–193. https://doi.org/10.1016/j.forsciint.2006.05.018 4. Cattaneo C (2007) Forensic anthropology: developments of a clas- sical discipline in the new millennium. Forensic Sci Int 165(2– 3):185–193. https://doi.org/10.1016/j.forsciint.2006.05.018 24. Blau S, Briggs CA (2011) The role of forensic anthropology in disaster victim identification (DVI). Forensic Sci Int 205(1–3):29– 35. https://doi.org/10.1016/j.forsciint.2010.07.038 5. Obertová Z, Adalian P, Baccino E, Cunha E, De Boer HH, Fra- casso T et al (2019) The status of forensic anthropology in Europe and South Africa: results of the 2016 FASE Questionnaire on Forensic Anthropology. J Forensic Sci 64(4):1017–1025. https:// doi.org/10.1111/1556-4029.14016 5. Obertová Z, Adalian P, Baccino E, Cunha E, De Boer HH, Fra- casso T et al (2019) The status of forensic anthropology in Europe and South Africa: results of the 2016 FASE Questionnaire on Forensic Anthropology. J Forensic Sci 64(4):1017–1025. https:// doi.org/10.1111/1556-4029.14016 25. Cordner S, Tidball-Binz MV (2022) Humanitarian forensic action: a new forensic discipline helping to implement international law and construct peace. Funding Open access funding provided by University of Lausanne Data availability The data presented in this study are available on request from the corresponding author. The data are not publicly avail- able due to ethical restrictions. 1 3 3 3 International Journal of Legal Medicine (2024) 138:1067–1077 1077 Conflict of interest The authors declare no competing interests. Open Access This article is licensed under a Creative Commons Attri- bution 4.0 International License, which permits use, sharing, adapta- tion, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/. 14. Maekinen T, Maijanen H, Seitsonen O (2022) The status and future of forensic archaeology and anthropology in Finland. Scand J For Res 28:32–39. https://doi.org/10.2478/sjfs-2022-0013fi 15. Federal Statistical Office (2022) Presence Switzerland 2022 - Lan- guage - facts and figures. https://www.eda.admin.ch/aboutswitz erland/en/home/gesellschaft/sprachen/die-sprachen---fakten-und- zahlen.html. Accessed 3 Nov 2023 16. Passalacqua NV, Pilloud MA (2020) Education and training in forensic anthropology. Forensic Anthropol 3(2):65–74. https://doi. org/10.5744/fa.2020.1013 17. Byers SN (2017) Introduction to forensic anthropology, 5th edn. Routledge, New York g 18. Blau S, Ubelaker DH (2016) Handbook of forensic anthropology and archaeology, Second edn. Routledge, New York 19. Langley NR, Tersigni-Tarrant MT (2017) Forensic anthropology: a comprehensive introduction, Second edn. CRC Press, Boca Raton Declarations suggestions for future developments. Forensic Sci Int Rep 3:100178. https://doi.org/10.1016/j.fsir.2021.100178 9. Baryah N, Krishan K, Kanchan T (2019) The development and status of forensic anthropology in India: a review of the literature and future directions. Med Sci Law 59(1):61–69. https://doi.org/ 10.1177/0025802418824834 Research involving human participants and/or animals No participa- tion of animals. This study is a voluntary survey of professional experi- ences in Forensic Anthropology in Switzerland. All participants were informed about the details of the study and the publication of the results before starting the questionnaire. By submitting the questionnaire, the participants agreed to participate. Participation in the questionnaire was anonymized upon entry by the participant. 10. Passalacqua NV, Pilloud MA, Congram D (2021) Forensic anthro- pology as a discipline. Biology (Basel) 10(8):691. https://doi.org/ 10.3390/biology10080691 11. Traithepchanapai P, Mahakkanukrauh P, Kranioti EF (2016) His- tory, research and practice of forensic anthropology in Thailand. Forensic Sci Int 261(167):e1–e6. https://doi.org/10.1016/j.forsc iint.2016.02.025 Informed consent All participants were informed about the publica- tion of the results before starting the questionnaire. Participation in the questionnaire was anonymous. 12. Villa C, Lynnerup N, Boel LWT, Boldsen JL, Weise S, Bjarno C et al (2022) Forensic anthropology and archaeology in Denmark. Scand J For Res 28:3–9. https://doi.org/10.2478/sjfs-2022-0016 Ethical approval Not applicable, as no medical data were collected. Participation was totally voluntary and anonymous. 13. Alfsdotter C, Petaros A, Güvencel A, Molnar P, Teglind R, Alkass K (2022) Development and implementation of forensic anthro- pology in Swedish forensic practice. Scand J For Res 28:10–19. https://doi.org/10.2478/sjfs-2022-0015 Conflict of interest The authors declare no competing interests. Conflict of interest The authors declare no competing interests. References Wiley interdisciplinary reviews: Forensic Sci 4(1):e1438. https://doi.org/10.1016/j.forsciint.2017.08.011 g 6. Bundesamt für Statistik BFS (2022) Die Bevölkerung der Schweiz im Jahr 2021. Neuchâtel: BFS; 2022 7. Indra L, Lösch S (2021) Forensic anthropology casework from Switzerland (Bern): taphonomic implications for the future. Foren- sic Sci Int Rep 4:100222. https://doi.org/10.1016/j.fsir.2021.100222 Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. 8. Alfsdotter C (2021) Forensic archaeology and forensic anthro- pology within Swedish law enforcement: current state and 1 3
https://openalex.org/W2897306447	https://pure.mpg.de/pubman/item/item_3006554_1/component/file_3006555/Deakin_18.pdf	English	null	Altered hippocampal gene expression and structure in transgenic mice overexpressing neuregulin 1 (Nrg1) type I	Translational psychiatry	2,018	cc-by	7,572	A R T I C L E O p e n A c c e s s © The Author(s) 2018 OpenAccessThisarticleislicensedunderaCreativeCommonsAttribution4.0InternationalLicense,whichpermitsuse,sharing,adaptation,distributionandreproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this license, visit http://creativecommons.org/licenses/by/4.0/. Abstract Transgenic mice overexpressing the type I isoform of neuregulin 1 (Nrg1; NRG1) have alterations in hippocampal gamma oscillations and an age-emergent deﬁcit in hippocampus-dependent spatial working memory. Here, we examined the molecular and morphological correlates of these ﬁndings. Microarrays showed over 100 hippocampal transcripts differentially expressed in Nrg1tg-type I mice, with enrichment of genes related to neuromodulation and, in older mice, of genes involved in inﬂammation and immunity. Nrg1tg-type I mice had an enlarged hippocampus with a widened dentate gyrus. The results show that Nrg1 type I impacts on hippocampal gene expression and structure in a multifaceted and partly age-related way, complementing the evidence implicating Nrg1 signaling in aspects of hippocampal function. The ﬁndings are also relevant to the possible role of NRG1 signaling in the pathophysiology of schizophrenia or other disorders affecting this brain region. multifaceted and partly age-related way, complementing the evidence implicating Nrg1 signaling in aspects of hippocampal function. The ﬁndings are also relevant to the possible role of NRG1 signaling in the pathophysiology of schizophrenia or other disorders affecting this brain region. Altered hippocampal gene expression and structure in transgenic mice overexpressing neuregulin 1 (Nrg1) type I Inga H. Deakin1, Beata R. Godlewska1, Mary A. Walker1, Guo-Jen Huang2, Markus H. Schwab3,4, Klaus-Armin Nave3, Amanda J. Law1,5 and Paul J. Harrison 1,6 Deakin et al. Translational Psychiatry (2018) 8:229 DOI 10.1038/s41398-018-0288-2 Deakin et al. Translational Psychiatry (2018) 8:229 DOI 10.1038/s41398-018-0288-2 Translational Psychiatry Introduction selectively overexpressing this isoform (Nrg1tg-type I) was created19, and shown to have alterations in myelination20 and some aspects of behavior21, including an age- emergent deﬁcit in hippocampus-dependent spatial working memory22. Nrg1tg-type I mice also exhibit a reduced frequency of carbachol-induced hippocampal gamma oscillations22. These ﬁndings complement a broad body of evidence linking Nrg1 to hippocampal function and plasticity23–28. Differing phenotypic proﬁles are seen in other genetic mouse models of Nrg1, highlighting the existence of isoform-speciﬁc properties29–34. Neuregulin 1 (Nrg1; NRG1) is a growth factor, signaling via Erbb3 and Erbb4 receptor tyrosine kinases. Nrg1 plays diverse roles in the development, plasticity, and diseases of the nervous system1–5. Its pleiotropy arises, in part, from a family of structurally and functionally distinct isoforms (types I to VI), transcribed from different 5’ exons6. In humans, allelic variation can affect NRG1 iso- form expression7–9, and polymorphisms in NRG1 may be a risk gene for schizophrenia10–12, although this has not been conﬁrmed in genome-wide association studies13. The type I isoform is affected in schizophrenia, with increased expression in hippocampus7 and prefrontal cortex14 compared with controls, and representing one of the abnormalities of NRG1-ErbB4 signaling observed in the disorder15–18. Reﬂecting the interest in the functional and pathological roles of NRG1 type I, a transgenic mouse In the current study we examined two other aspects of the hippocampus in Nrg1tg-type I mice: its transcriptome, and its structure, in comparison with wildtype (wt) lit- termates. Because of the age-dependent features of the hippocampal phenotype observed previously, the tran- scriptomic analysis was performed at two ages. Correspondence: Paul J. Harrison (paul.harrison@psych.ox.ac.uk) 1Department of Psychiatry, University of Oxford, Oxford, UK 2Wellcome Trust Centre for Human Genetics, University of Oxford, Oxford, UK Full list of author information is available at the end of the article. © The Author(s) 2018 OpenAccessThisarticleislicensedunderaCreativeCommonsAttribution4.0InternationalLicense,whichpermitsuse,sharing,adaptation,distributionandreproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. Correspondence: Paul J. Harrison (paul.harrison@psych.ox.ac.uk) 1Department of Psychiatry, University of Oxford, Oxford, UK 2Wellcome Trust Centre for Human Genetics, University of Oxford, Oxford, UK Full list of author information is available at the end of the article. Introduction To view a copy of this license, visit http://creativecommons.org/licenses/by/4.0/. Page 2 of 9 Page 2 of 9 Deakin et al. Translational Psychiatry (2018) 8:229 RNA extraction and preparation for microarrays Hippocampal RNA was treated with 1 μl (1 unit) RQ1 RNase-free DNAse (Promega) and 0.6 μl (24 units) RNa- sin ribonuclease inhibitor (Promega) at 37 °C for 30 min and then heated to 72 °C for 10 min. DNAse-treated RNA was reverse-transcribed using 1 μl (200 units) MMLV reverse transcriptase, Promega), 1xMMLV buffer, 0.8 μl of 10 mM dNTPs, 0.6 μl (24 units) RNasin, 0.6 μl of 10 mM oligoDTs. The reaction mix was incubated at 42 °C for 1 h and then heated to 72 °C for 10 min. The reverse- transcribed RNA (complementary DNA; cDNA) was diluted with nuclease-free water and stored at −80 °C. Brains were frozen in isopentane cooled on dry ice. The left hemisphere was placed into RNAlater®-ICE Frozen Tissue Transition Solution (Ambion) at −20 °C for 18 h, after which the hippocampus was dissected and homo- genized in Qiazol in a TissueLyser (Qiagen). Total RNA was extracted and puriﬁed with RNeasy lipid tissue Mini kits (Qiagen) according to manufacturer’s protocol. In total, 300 ng RNA was used for ampliﬁcation and labeling with Illumina® TotalPrep™ RNA Ampliﬁcation Kit (Ambion). Complementary RNA (cRNA) quality was determined with an Agilent 2100 Bioanalyser (Agilent Technologies, Palo Alto, CA). TaqMan assays were ordered from Applied Biosystems (Npy, Mm00445771_m1; Gfap, Mm01253034_m1; Inhba, Mm00434339_m1; Cntfr, Mm00516697_m1; C1qa, Mm00432142_m1; Rbbp4, Mm00771401_g1). Bdnf was detected using pan-BDNF primers and a TaqMan probe (Forward, 5′-GGGTCACAG CGGCAGATAAA-3′, Reverse 5′-GCCTTTGGATACCGGGACTT-3′; Probe, T CTGGCGGGACGGTCACAGTCCTA)38. Bdnf v1- speciﬁc probes39 were: Forward, 5′-CACATTACCTTCC TGCATCTGTTG-3′, reverse 5′-ACCATAGTAAGGAA AAGGATGGTCAT-3′, probe AAGCCACAATGTTCC ACCAG. The PCR reaction mix included 15 ng cDNA, 1xTaqMan Universal PCR Master Mix (Applied Biosys- tems) and the assay in a ﬁnal volume (with nuclease-free water) of 15 μl in 384-well plates. Nrg1 mRNA was quantiﬁed using a SyBr green assay, with the primers designed so that the PCR product would span the exon boundary between the type I-speciﬁc exon and the immunoglobulin-like domain, and checked in the BLAST database for binding speciﬁcity (Forward, 5′- A total of 1.5 µg cRNA from each brain was hybridized to an Illumina Mouse WG-6 v1.1 (young mice) or v2 (old mice) chip according to manufacturers’ protocol and scanned with a BeadStation 500 machine. Materials and methods A linear model with an empirical Bayes t-statistic was ﬁtted to the data to generate lists of signiﬁcant effects of genotype, using the Limma program35. We corrected for multiple testing with the Benjamini–Hochberg false dis- covery rate to give an adjusted p value36. The criteria for differential expression were an adjusted p value < 0.05 and an absolute (unlogged RMA) fold change in expression (FC) > 1.5. All experiments were conducted in accordance with the United Kingdom Animals (Scientiﬁc Procedures) Act, 1986, and had local ethical approval. Microarrays and quantitative real-time PCR (RT-qPCR) Two microarray experiments were performed, one in ‘young adult’ mice (2.5–4 months), the other in ‘old’ mice (14–15 months). Each comprised 24 animals, 6 of each genotype and sex. Generation and genotyping of Nrg1tg-type I mice The generation and genotyping of the Nrg1tg-type I mice has been described19. The mice overexpress Nrg1 type I (β1a-isoform) under a Thy-1 promoter, with robust overexpression in multiple brain regions, including the hippocampus, with no alteration in Nrg1 types II or III20,21. The experiments reported here were performed in F6-F9 generations of backcross of heterozygous Nrg1tg-type I males with wt C57BL/6 J females, comparing Nrg1tg-type I mice with their wt littermates. Differentially expressed probe lists were condensed into lists of differentially expressed transcripts, including known splice variants, identiﬁed with PubMed basic local alignment search tool (BLAST). Probes with no accession number were identiﬁed by nucleotide sequence with PubMed Nucleotide BLAST. Results from each array were compared with obtain lists of genes that were differen- tially expressed in NRG1tg-type I mice at both ages, or only at one of the two ages. These three lists of genes were used for further investigation with Ingenuity Pathways Analysis (IPA; Ingenuity Systems), which generated net- works of up to 35 genes based on their known functional links37. Microarray analysis Standard quality control measures were performed with the BeadStudio program (Illumina, CA), including sub- tracting the background from each array from the raw signal intensity of each probe type. The raw signal intensity data underwent variance stabilizing normal- ization, which is a generalized log2 transformation of the signal. A quantile standardization procedure was used to centralize the mean signal in the distribution (i.e., of all the probes) and to equalize the variance between mice. Further quality control was performed such as hier- archical clustering and box plots of normalized intensity (robust multichip average; RMA) values for each chip. Page 3 of 9 Deakin et al. Translational Psychiatry (2018) 8:229 was known about their functions and interactions with Nrg1 (Fig. 1b–d; and Table 1). AAGGGGAAGGGCAAGAAGAA-3′, Reverse 5′- TCTTTCAATCTGGGAGGCAAT-3′; Eurogentech). The reaction mix for Nrg1 type I was 1xSyBr Green Mix, 15 ng cDNA, 200 nM of each primer and nuclease-free water up to a ﬁnal volume of 15 μl. AAGGGGAAGGGCAAGAAGAA-3′, Reverse 5′- TCTTTCAATCTGGGAGGCAAT-3′; Eurogentech). The reaction mix for Nrg1 type I was 1xSyBr Green Mix, 15 ng cDNA, 200 nM of each primer and nuclease-free water up to a ﬁnal volume of 15 μl. Genes upregulated at both ages in the NRG1tg-type I mice included neuropeptide Y (Npy), brain-derived neuro- trophic factor (Bdnf), and glial ﬁbrillary acidic protein (Gfap). Consistent with the Npy mRNA data, Npy- immunoreactive hippocampal interneurons were mark- edly more prominent in the Nrg1tg-type I mice (not shown). The Bdnf mRNA increase affected the V1 isoform selec- tively. IPA generated two networks with scores corre- sponding to signiﬁcance values of p = 10−31 and p = 10−25 (Supplementary Figs. 1 and 2). Network 1 included, as well as Npy and Bdnf, several transcripts involved in neurotransmission and implicated in schizophrenia, such as dopamine D1 and D4 receptors (Drd1 and Drd4). μ Standard curves of pooled cDNA from all samples were set up in triplicate with the starting amount of cDNA ranging from 100 ng to 0.54 pg (Nrg1), 5 pg (Npy), 0.01 ng (Gfap), or 0.39 ng (Inhba, Cntfr, C1qa, Rbbp4). The R2 of all standard curves was > 0.99. All experimental samples fell within the standard curve. Samples were run in tri- plicate on the same plate as the standard curve. No- template controls and RT-negative controls were also run in triplicate to test for any contamination of the reaction mix or cDNA, respectively. Discussion Mice selectively overexpressing the type I isoform of Nrg1 show differences in hippocampal function and age-emergent deﬁcits in hippocampus-dependent behavior22. Here, we report that these changes are accompanied by an altered proﬁle of gene expression which differs between young adult (2.5–4 month) and old (14–15 month) Nrg1tg-type I mice, and by an increase in hippocampal volume. The hippocampal transcriptome of Nrg1tg-type I mice: age- related effects on genes involved in myelination, neurotransmission, and immunity Nrg1 impacts upon the expression of many individual genes1,3. Here we used microarrays as an unbiased method to identify transcripts and networks altered in the hippocampus of Nrg1tg-type I mice. Adopting a stringent statistical approach, and with a 1.5-fold change threshold, we identiﬁed over 100 differentially expressed genes, of which ~ 80% were increased in Nrg1tg-type I mice com- pared with wt (Fig. 1a, and Supplementary Tables 1–3). Thirty-eight transcripts were altered in the same direction at both ages (Supplementary Table 1), 20 were differen- tially expressed only in the young adult (2.5–4 month) Nrg1tg-type I mice (Supplementary Table 2) and 54 only in old (14–15 month) Nrg1tg-type I mice (Supplementary Table 3). Eight transcripts were selected for qPCR vali- dation, based on the microarray results as well as what Microarray analysis Cycling conditions for all qRT-PCR reactions were 50 °C for 2 min, 95 °C for 10 min and then 40 repeats of 95 °C for 15 seconds to denature and 60 °C for 1 min. Of the genes differentially expressed in young but not old Nrg1tg-type I mice, three were involved in myelination (myelin basic protein (Mbp), myelin oligodendrocyte glycoprotein (Mog), and myelin-associated oligoden- drocytic basic protein). IPA generated a network that included Mog and Mbp (p = 10−35; Supplementary Fig. 3). Genes overexpressed in old but not young Nrg1tg-type I mice (Supplementary Table 3) included many with immune and inﬂammatory functions, and IPA generated two highly signiﬁcant networks (p = 10−37 and p = 10−34; one is shown in Fig. 1e, the other in Supplementary Fig. 4). In particular, the network shown in Fig. 1e includes many upregulated HLA and major histocompatibility complex (MHC) genes (Supplementary Table 4). Of the genes differentially expressed in young but not old Nrg1tg-type I mice, three were involved in myelination (myelin basic protein (Mbp), myelin oligodendrocyte glycoprotein (Mog), and myelin-associated oligoden- drocytic basic protein). IPA generated a network that included Mog and Mbp (p = 10−35; Supplementary Fig. 3). I Nrg1tg-type I mice have altered hippocampal morphology Nrg1 y mice have altered hippocampal morphology As shown in Fig. 2a, the hippocampus was enlarged (by ~ 25%) in Nrg1tg-type I mice, with brain volume unchan- ged. In terms of individual subﬁelds, the only difference observed in Nrg1tg-type I mice was that the dentate gyrus granule cell layer was wider than in wt mice; this was primarily in the infra-pyramidal (external) blade rather than the supra-pyramidal (internal) blade (Fig. 2b, c). The density of PV + interneurons was unchanged in each subﬁeld measured (Fig. 2d). Morphology and histology Snap-frozen brains from 10-month-old wt and Nrg1tg- type I mice were coronally cryosectioned at 20 μm thick- ness. Hippocampal area was measured by point counting on multiple cresyl violet-stained sections throughout the dorsoventral extent, and hippocampal volume estimated using Cavalieri’s theorem as described40. Whole brain volume was estimated in the same way. We also measured the cross-sectional area of subﬁelds (dentate gyrus, CA3, CA1) by manual tracing using a Nikon Eclipse 3600 microscope coupled to an MCID Elite image analysis system (Interfocus, Haverhill, UK). Other sections were immunostained for parvalbumin (PV27, 1:100; Swant, Switzerland), detected with diaminobenzidine, by standard methods. All measurements were made blind to genotype. Genes overexpressed in old but not young Nrg1tg-type I mice (Supplementary Table 3) included many with immune and inﬂammatory functions, and IPA generated two highly signiﬁcant networks (p = 10−37 and p = 10−34; one is shown in Fig. 1e, the other in Supplementary Fig. 4). In particular, the network shown in Fig. 1e includes many upregulated HLA and major histocompatibility complex (MHC) genes (Supplementary Table 4). The hippocampal molecular and morphological proﬁle of Nrg1tg-type I mice The statistical approach to the microarray data is described in text; p values fo tests (two-tailed) comparing transgenic and wt mice of each age group . NC: no signiﬁcant change. The statistical approach to the microarray data is described in text; p values for qRT-PCR are from unpaired t ansgenic and wt mice of each age group 1–4; Supplementary Tables 1–3). Of the genes over- expressed in Nrg1tg-type I mice at both ages, several are noteworthy. In particular, ﬁve transcripts (Npy, Gfap, Bdnf, Drd1, and Drd4) were part of the most signiﬁcant gene network and all had been linked previously to Nrg1. The fact they were upregulated in both age cohorts, which were studied separately and with different versions of the microarray chip, strengthens the robustness of the ﬁndings. an increase of hilar cells immunoreactive for doublecortin, a marker of newly formed neurons, in the Nrg1tg-type I mice (I.H.D. and P.J.H., unpublished observations), sug- gesting that the enlarged dentate gyrus might reﬂect increased adult neurogenesis—a process stimulated by Npy48–50, and inﬂuenced by Nrg151. Whether the per- sisting upregulation of Gfap mRNA in Nrg1tg-type I mice (Table 1) is also indicative of enhanced cell proliferation— as many neural precursors express Gfap52—remains to be seen; it might also be a remnant of the developmental role of Nrg1-ErbB signaling in neuron–astrocyte differentia- tion53,54. Bdnf is a regulator of hippocampal plasticity and function55, and its elevated expression may have many manifestations in the Nrg1tg-type I mice, including a con- tribution to their spatial working memory deﬁcit22,56. Finally, the upregulation of Drd1 and Drd4 mRNAs complements evidence that hippocampal Drd4 mediates Nrg1-induced reversal of LTP24, and that Nrg1 applica- tion produces acute57 and sustained58 increases in dopa- mine release and dopamine neuron ﬁring59. Another Nrg genetic mouse model also shows dopamine receptor alterations60. 1–4; Supplementary Tables 1–3). Of the genes over- expressed in Nrg1tg-type I mice at both ages, several are noteworthy. In particular, ﬁve transcripts (Npy, Gfap, Bdnf, Drd1, and Drd4) were part of the most signiﬁcant gene network and all had been linked previously to Nrg1. The fact they were upregulated in both age cohorts, which were studied separately and with different versions of the microarray chip, strengthens the robustness of the ﬁndings. Npy is expressed by a subpopulation of hippocampal interneurons, bistratiﬁed and ivy cells, which impose a strong inhibitory inﬂuence on pyramidal cell den- drites41,42. The hippocampal molecular and morphological proﬁle of Nrg1tg-type I mice Nrg1tg-type I mice exhibited differential expression of a number of genes (Fig. 1; Table 1; Supplementary Figures Deakin et al. Translational Psychiatry (2018) 8:229 Page 4 of 9 Fig. 1 (See legend on next page.) Fig. 1 (See legend on next page.) Fig. 1 (See legend on next page.) Page 5 of 9 Deakin et al. Translational Psychiatry (2018) 8:229 (see ﬁgure on previous page) Fig. 1 Hippocampal gene expression in Nrg1tg-type I mice. a Genes meeting the criteria described in text for differential expression in young (yellow) or old (blue) adult Nrg1tg-type I mice, or Nrg1tg-type I mice of both ages (orange). b–d examples of RT-qPCR validation of transcripts from each group. b Npy mRNA, increased in Nrg1tg-type I mice of both ages; c Cntfr mRNA, decreased in young but not old adult Nrg1tg-type I mice; d C1q mRNA, increased in old but not young adult Nrg1tg-type I mice. Statistics for the data shown in panels b–d are given in Table 1. e An IPA network of transcripts differentially expressed in old but not young adult Nrg1tg-type I mice. The network comprises nodes (genes) and their biological relationships shown by interconnecting lines. Red nodes are transcripts with increased expression, and the green nodes are transcripts with lower expression, in the old Nrg1tg-type I mice, compared with their age-matched wt controls. Increasing color intensity indicates a greater fold change. White nodes show genes that are functionally related to the other differentially expressed genes in the network and added by IPA. Solid lines between nodes indicate a direct interaction between them and dashed lines indicate indirect relationships. A continuous line denotes “binding only”; pointed line, “acts upon” and blunt ended line, “inhibits”. For gene symbols and names, see Supplementary Table 4. For additional IPA networks identiﬁed in one or both age groups of Nrg1tg-type I mice, see Supplementary Tables 1–3 and Supplementary Figures 1–4 ( g p p g ) Fig. 1 Hippocampal gene expression in Nrg1tg-type I mice. a Genes meeting the criteria described in text for differential expression in young (yellow) or old (blue) adult Nrg1tg-type I mice, or Nrg1tg-type I mice of both ages (orange). b–d examples of RT-qPCR validation of transcripts from each group. The hippocampal molecular and morphological proﬁle of Nrg1tg-type I mice b Npy mRNA, increased in Nrg1tg-type I mice of both ages; c Cntfr mRNA, decreased in young but not old adult Nrg1tg-type I mice; d C1q mRNA, increased in old but not young adult Nrg1tg-type I mice. Statistics for the data shown in panels b–d are given in Table 1. e An IPA network of transcripts differentially expressed in old but not young adult Nrg1tg-type I mice. The network comprises nodes (genes) and their biological relationships shown by interconnecting lines. Red nodes are transcripts with increased expression, and the green nodes are transcripts with lower expression, in the old Nrg1tg-type I mice, compared with their age-matched wt controls. Increasing color intensity indicates a greater fold change. White nodes show genes that are functionally related to the other differentially expressed genes in the network and added by IPA. Solid lines between nodes indicate a direct interaction between them and dashed lines indicate indirect relationships. A continuous line denotes “binding only”; pointed line, “acts upon” and blunt ended line, “inhibits”. For gene symbols and names, see Supplementary Table 4. For additional IPA networks identiﬁed in one or both age groups of Nrg1tg-type I mice, see Supplementary Tables 1–3 and Supplementary Figures 1–4 Table 1 Quantitative RT-PCR validation of differentially expressed genes in Nrg1tg-type I mice Young adult (2.5–4 months) Old (4–15 months) Microarray qRT-PCR Microarray qRT-PCR Accession Gene FC p FC p FC p FC p NRG1 type I 8.60 2.62E-19 801 0.001 NA NA 466 0.001 NM_023456.2 NPY 2.22 1.01E-05 2.0 0.005 2.46 9.93E-06 2.6 0.001 NM_010277 GFAP 1.59 1.69E-03 NC NC 1.62 1.42E-02 1.5 0.036 NM_007540.3 BDNF v1 1.65 4.12E-04 3.5 0.001 2.76 8.00E-07 NC NC NM_008380.1 Inhba 2.28 4.83E-06 NC NC NC NC 2.7 0.001 NM_016673.1 Cntfr −1.86 5.66E-03 −3.4 0.068 NC NC NC NC NM_007572 C1qa NC NC NC NC 1.60 6.14E-04 1.4 0.032 NM_009030 Rbbp4 NC NC NC NC −1.89 6.14E-04 NC NC NA: probe not present on array. NC: no signiﬁcant change. The statistical approach to the microarray data is described in text; p values for qRT-PCR are from unpaired t tests (two-tailed) comparing transgenic and wt mice of each age group 1 Quantitative RT-PCR validation of differentially expressed genes in Nrg1tg-type I mice NA: probe not present on array. NC: no signiﬁcant change. The hippocampal molecular and morphological proﬁle of Nrg1tg-type I mice d The density of parvalbumin (PV)-immunoreactive cells did not differ in Nrg1tg-type I mice (n = 10) compared with wt (n = 12). Bars in a, b, and d show mean and standard deviation. All data in this Figure come from 10- month-old mice The gene expression differences that occurred in the old but not young Nrg1tg-type I were striking, comprising many immune and inﬂammatory genes, such as HLA-DR, MHC class II CD74, and complement C1q. Similar changes have been reported in old vs. young wt rodents and in this respect there may be an ‘accelerated aging’ phenotype in Nrg1tg-type I mice61–64. C1q, like other complement fac- tors, is also involved in neuronal and synaptic function65– 67 and dysfunction68–70 and brain aging71. As such, the altered expression of the genes in the older Nrg1tg-type I mice may represent molecular correlates of, and might contribute to, their age-emergent memory impairment. hippocampus; any resulting hypermyelination may in turn contribute to the hippocampal volume increase. However, this remains speculative; indeed, more generally, the processes that link the morphological and molecular alterations reported here remain unknown. The altered transcriptomic proﬁle of the Nrg1tg-type I mice highlights an issue that pertains broadly to geneti- cally modiﬁed animals: their phenotypes need not arise solely from the targeted gene(s) but also from the cascade of molecular changes which the manipulation induces. Moreover, these effects are not static but vary with age, and illustrate the value of going beyond the 3–6 month time-point at which characterization is often completed, even though capturing the temporal dynamic and long- itudinal proﬁle is demanding of resources and time. Fewer genes were differentially expressed in the young but not old Nrg1tg-type I mice, and they were primarily myelin-related. This is not unexpected, in that Nrg1 is a critical player in myelination72,73. Although the type I isoform has hitherto been implicated primarily in per- ipheral myelination19,74,75, these mice do show hyper- myelination of small diameter axons in the central nervous system20. The upregulation of myelin-related transcripts in the Nrg1tg-type I mice may be a molecular indication that this process is also occurring in the The hippocampal molecular and morphological proﬁle of Nrg1tg-type I mice Its marked upregulation in the Nrg1tg-type I mice (Fig. 1b) is of interest for several reasons. First, it provides another hint that interneurons are affected and may contribute to the oscillatory and circuitry alterations that underlie the phenotype of the mice41–44. Second, Npy is anti-epileptic45–47, and enhanced Npy expression may help prevent the epileptiform predisposition of Nrg1tg-type I mice22 progressing to overt seizure activity. Third, the Npy mRNA elevation may be related to the morphological ﬁnding of a widened dentate gyrus. Preliminary data show Deakin et al. Translational Psychiatry (2018) 8:229 Page 6 of 9 Fig. 2 Hippocampal morphology in Nrg1tg-type I mice. a Hippocampal volume is increased in Nrg1tg-type I mice (n = 10) compared with wt (n = 9; two-tailed unpaired t test, t = 3.249, df = 17, p = 0.006), but whole brain volume is unchanged. b The width of the dentate gyrus granule cell layer is increased in Nrg1tg-type I mice in the infra-pyramidal blade (n = 6 in each group; two-tailed unpaired t test, t = 4.126, df = 10, p = 0.002) but not in the supra-pyramidal blade. c Illustration of the data in b, showing the wider infra-pyramidal blade (IPB) in a Nrg1tg-type I mouse compared with a wt mouse. The supra-pyramidal blade (SPB) is also shown, with CA3 in between. d The density of parvalbumin (PV)-immunoreactive cells did not differ in Nrg1tg-type I mice (n = 10) compared with wt (n = 12). Bars in a, b, and d show mean and standard deviation. All data in this Figure come from 10- month-old mice Fig. 2 Hippocampal morphology in Nrg1tg-type I mice. a Hippocampal volume is increased in Nrg1tg-type I mice (n = 10) compared with wt (n = 9; two-tailed unpaired t test, t = 3.249, df = 17, p = 0.006), but whole brain volume is unchanged. b The width of the dentate gyrus granule cell layer is increased in Nrg1tg-type I mice in the infra-pyramidal blade (n = 6 in each group; two-tailed unpaired t test, t = 4.126, df = 10, p = 0.002) but not in the supra-pyramidal blade. c Illustration of the data in b, showing the wider infra-pyramidal blade (IPB) in a Nrg1tg-type I mouse compared with a wt mouse. The supra-pyramidal blade (SPB) is also shown, with CA3 in between. References 1. Corfas, G., Roy, K. & Buxbaum, J. D. Neuregulin 1-erbB signaling and the molecular/cellular basis of schizophrenia. Nat. Neurosci. 7, 575–580 (2004). 2. Harrison, P. J. & Law, A. J. Neuregulin 1 and schizophrenia: genetics, gene expression, and neurobiology. Biol. Psychiatry 60, 132–140 (2006). 3. Mei, L. & Xiong, W. C. Neuregulin 1 in neural development, synaptic plasticity and schizophrenia. Nat. Rev. Neurosci. 9, 437–452 (2008). 4. Buonanno, A. The neuregulin signaling pathway and schizophrenia: from genes to synapses and neural circuits. Brain Res. Bull. 83, 122–131 (2010). 5. Mei, L. & Nave, K. A. Neuregulin-ERBB signaling in the nervous system and neuropsychiatric diseases. Neuron 83, 27–49 (2014). p However, there is less congruence when other ﬁndings are considered. Thus, in schizophrenia, in contrast to ﬁndings in the Nrg1tg-type I mice, hippocampal myelin- related transcripts are decreased90, hippocampal volume is unchanged or reduced91, and there is a lower density of hippocampal PV + neurons92. Thus, although the phe- notype of Nrg1tg-type I mice may be seen as overlapping to an extent with that of schizophrenia, the differences are at least as striking as the similarities. One speciﬁc factor to consider is that the magnitude of overexpression in the Nrg1tg-type I mouse is far greater than the increased NRG1 type I expression reported in schizophrenia. More gen- erally, these considerations draw attention to the need for cautious interpretation and extrapolation from any genetic mouse model to the human syndrome. Never- theless, the results extend the evidence that Nrg1 type I is functional in the hippocampus, and hence may play a role in any disease in which its expression in this brain region is altered. 6. Steinthorsdottir, V. et al. Multiple novel transcription initiation sites for NRG1. Gene 342, 97–105 (2004). 7. Law, A. J. et al. Neuregulin 1 transcripts are differentially expressed in schi- zophrenia and regulated by 5’ SNPs associated with the disease. Proc. Natl. Acad. Sci. USA 103, 6747–6752 (2006). 8. Nicodemus, K. K. et al. A 5 ‘promoter region SNP in NRG1 is associated with schizophrenia risk and type III isoform expression. Mol. Psychiatry 14, 741–743 (2009). 9. Weickert, C. S., Tiwari, Y., Schoﬁeld, P. R., Mowry, B. J. & Fullerton, J. M. Schizophrenia-associated HapICE haplotype is associated with increased NRG1 type III expression and high nucleotide diversity. Transl. Psychiatry 2, e104 (2012). 10. Mostaid, M. S. et al. References Neuregulin-1 and schizophrenia in the genome-wide association study era. Neurosci. Biobehav Rev. 68, 387–409 (2016). 11. Mostaid, M. S. et al. Elevated peripheral expression of neuregulin-1 (NRG1) mRNA isoforms in clozapine-treated schizophrenia patients. Transl. Psychiatry 7, 1280 (2017). 12. Jagannath, V., Gerstenberg, M., Correll, C. U., Walitza, S. & Grünblatt, E. A systematic meta-analysis of the association of Neuregulin 1 (NRG1), D-amino acid oxidase (DAO), and DAO activator (DAOA)/G72 polymorphisms with schizophrenia. J. Neural Transm. (Vienna) 125, 89–102 (2018). 13. Farrell, M. S. et al. Evaluating historical candidate genes for schizophrenia. Mol. Psychiatry 20, 555–562 (2015). 14. Hashimoto, R. et al. Expression analysis of neuregulin-1 in the dorsolateral prefrontal cortex in schizophrenia. Mol. Psychiatry 9, 299–307 (2004). Author details 1 1Department of Psychiatry, University of Oxford, Oxford, UK. 2Wellcome Trust Centre for Human Genetics, University of Oxford, Oxford, UK. 3Department of Neurogenetics, Max Planck Institute of Experimental Medicine, Goettingen, Germany. 4Center of Physiology, Hannover Medical School, Hannover, Germany. 5Department of Psychiatry, University of Colorado, Denver, USA. 6Oxford Health NHS Foundation Trust, Oxford, UK 19. Michailov, G. V. et al. Axonal neuregulin-1 regulates myelin sheath thickness. Science 304, 700–703 (2004). 20. Brinkmann, B. G. et al. Neuregulin-1/ErbB signaling serves distinct functions in myelination of the peripheral and central nervous system. Neuron 59, 581–595 (2008). 21. Deakin, I. H. et al. Behavioural characterization of neuregulin 1 type I over- expressing transgenic mice. Neuroreport 20, 1523–1528 (2009). Publisher’s note S i N Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional afﬁliations. 24. Kwon, O. B. et al. Neuregulin-1 regulates LTP at CA1 hippocampal synapses through activation of dopamine D4 receptors. Proc. Natl. Acad. Sci. USA 105, 15587–15592 (2008). Acknowledgements h k 15. Hahn, C. G. et al. Altered neuregulin 1-erbB4 signaling contributes to NMDA receptor hypofunction in schizophrenia. Nat. Med. 12, 824–828 (2006). We thank P. Burnet, L. Chen, S. Eastwood, J. Flint K. Lamsa, H. Lockstone, W. Nissen, and J. Taylor for their contributions. I.H.D. was supported by a Wellcome Trust doctoral studentship. A.J.L. was a Medical Research Council (UK) Career Development Fellow. K.L. was a Wellcome Trust Career Development Fellow. Additional support from Stanley Medical Research Institute and Medical Research Council (UK) grants to P.J.H. M.H.S. holds a Heisenberg Fellowship from the Deutsche Forschungsgemeinschaft (DFG) and acknowleges funding by a DFG research grant (SCHW741/4-1). 16. Law, A. J., Kleinman, J. E., Weinberger, D. R. & Weickert, C. S. Disease-associated intronic variants in the ErbB4 gene are related to altered ErbB4 splice- variant expression in the brain in schizophrenia. Hum. Mol. Genet. 16, 129–141 (2007). 17. Law, A. J. et al. Neuregulin 1-ErbB4-PI3K signaling in schizophrenia and phosphoinositide 3-kinase-p110δ inhibition as a potential therapeutic strategy. Proc. Natl. Acad. Sci. USA 109, 12165–12170 (2012). 18. Parlapani, E. et al. Gene expression of neuregulin-1 isoforms in different brain regions of elderly schizophrenia patients. World J. Biol. Psychiatry 11, 243–250 (2010). Conﬂict of interest 22. Deakin, I. H. et al. Transgenic overexpression of the type I isoform of neur- egulin 1 affects working memory and hippocampal oscillations but not long- term potentiation. Cereb. Cortex. 22, 1520–1529 (2012). The authors declare no conﬂicts of interest. The authors declare no conﬂicts of interest. 23. Kwon, O. B., Longart, M., Vullhorst, D., Hoffman, D. A. & Buonanno, A. Neuregulin-1 reverses long-term potentiation at CA1 hippocampal synapses. J. Neurosci. 25, 9378–9383 (2005). Implications for NRG1 in schizophrenia The evidence mentioned earlier showing Nrg1-dopa- mine interactions, and the increased D1 and D4 receptor expression seen here is notable, given that dopaminergic abnormalities are a ﬁnal common pathway in schizo- phrenia pathophysiology76, and alterations in both Deakin et al. Translational Psychiatry (2018) 8:229 Page 7 of 9 Page 7 of 9 receptors have been reported in the disorder77,78. The MHC complex is strongly implicated in schizophrenia79, in part through the complement C4 gene80, and several81– 84 though not all85 studies report elevated expression of many immune and inﬂammatory genes, including C1q, especially in older, chronically ill patients86, reminiscent of the gene expression changes being restricted to the old Nrg1tg-type I mice. There is also some evidence for direct links between NRG1, immune function, and schizo- phrenia87–89. Supplementary Information accompanies this paper at (https://doi.org/ 10.1038/s41398-018-0288-2). Neuregulin 1 transgenic mice display reduced mis- match negativity, contextual fear conditioning and social interactions. Brain Res. 1294, 116–127 (2009). 56. Cunha, C. et al. Brain-derived neurotrophic factor (BDNF) overexpression in the forebrain results in learning and memory impairments. Neurobiol. Dis. 33, 358–368 (2009). 31. Duffy, L., Cappas, E., Lai, D., Boucher, A. A. & Karl, T. Cognition in transmem- brane domaine neuregulin 1 mutant mice. Neurosci 170, 800–807 (2010). 57. Yurek, D. M., Zhang, L., Fletcher-Turner, A. & Seroogy, K. B. Supranigral injection of neuregulin1-[beta] induces striatal dopamine overﬂow. Brain Res. 1028, 116–119 (2004). 32. Yin, D. M. et al. Reversal of behavioral deﬁcits and synaptic dysfunction in mice overexpressing neuregulin 1. Neuron 78, 644–657 (2013). 33. Luo, X., He, W., Hu, X. & Yan, R. Reversible overexpression of bace1-cleaved neuregulin-1 N-terminal fragment induces schizophrenia-like phenotypes in mice. Biol. Psychiatry 76, 120–127 (2014). 58. Kato, T. et al. Transient exposure of neonatal mice to neuregulin-1 results in hyperdopaminergic states in adulthood: implication in neurodevelopment hypothesis for schizophrenia. Mol. Psychiatry 16, 307–320 (2011). 59. Namba, H., Okubo, T. & Nawa, H. Perinatal exposure to neuregulin-1 results in disinhibition of adult midbrain dopaminergic neurons: implication in schizo- phrenia modeling. Sci. Rep. 6, 22606 (2016). 34. Papaleo, F. et al. Behavioral, neurophysiological, and synaptic impairment in a transgenic neuregulin1 (NRG1-IV) murine schizophrenia model. J. Neurosci. 36, 4859–4875 (2016). 35. Smyth, G. K. Limma: Linear models for microarray data. In: R. Gentleman, V. J. Carey, W. Huber, R. A. Irizarry, Dudoit S editors. Bioinformatics and Computa- tional Biology Solutions Using R and Bioconductor. (pp. 397–420. Springer, New York, 2005). 60. Newell, K. A., Karl, T. & Huang, X. F. A neuregulin 1 transmembrane domain mutation causes imbalanced glutamatergic and dopaminergic receptor expression in mice. Neuroscience 248, 670–680 (2013). p 61. Verbitsky, M. et al. Altered hippocampal transcript proﬁle accompanies an age- related spatial memory deﬁcit in mice. Learn Mem. 11, 253–260 (2004). 36. Benjamini, Y. & Hochberg, Y. Controlling the false discovery rate - a practical and powerful approach to multiple testing. J. Roy. Stat. Soc. B 57, 289–300 (1995). 62. Frank, M. G. et al. mRNA upregulation of MHC II and pivotal pro-inﬂammatory genes in normal brain aging. Neurobiol. Aging 27, 717–722 (2006). genes in normal brain aging. Neurobiol. Aging 27, 717–7 37. Calvano, S. E. et al. A network-based analysis of systemic inﬂammation in humans. Nature 437, 1032–1037 (2005). 63. Supplementary Information accompanies this paper at (https://doi.org/ 10.1038/s41398-018-0288-2). Kadish, I. et al. Hippocampal and cognitive aging across the lifespan: a bioenergetic shift precedes and increased cholesterol trafﬁcking parallels memory impairment. J. Neurosci. 29, 1805–1816 (2009). 38. Burbach, G. J., Dehn, D., Del Turco, D. & Deller, T. Quantiﬁcation of layer-speciﬁc gene expression in the hippocampus: effective use of laser microdissection in combination with quantitative RT-PCR. J. Neurosci. Methods 131, 83–91 (2003). y 64. Pawlowski, T. L. et al. Hippocampal gene expression changes during age- related cognitive decline. Brain Res. 1256, 101–110 (2009). 65. Stevens, B. et al. The classical complement cascade mediates CNS synapse elimination. Cell 131, 1164–1178 (2007). 39. Liu, Q. R. et al. Rodent BDNF genes, novel promoters, novel splice variants, and regulation by cocaine. Brain Res. 1067, 1–12 (2006). 66. Boulanger, L. M. Immune proteins in brain development and synaptic plas- ticity. Neuron 64, 93–109 (2009). 40. Walker, M. A. et al. Estimated neuronal populations and volumes of the hip- pocampus and its subﬁelds in schizophrenia. Am. J. Psychiatry 159, 821–828 (2002). 67. Bialas, A. R. & Stevens, B. TGF-beta signaling regulates neuronal C1q expression and developmental synaptic reﬁnement. Nat. Neurosci. 16, 1773–1782 (2013). 41. Klausberger, T. et al. Spike timing of dendrite-targeting bistratiﬁed cells during hippocampal network oscillations in vivo. Nat. Neurosci. 7, 41–47 (2004). 68. Di Filippo, M., Sarchielli, P., Picconi, B. & Calabresi, P. Neuroinﬂammation and synaptic plasticity: theoretical basis for a novel, immune-centred, therapeutic approach to neurological disorders. Trends Pharmacol. Sci. 29, 402–412 (2008). ppocampal network oscillations in vivo. Nat. Neurosci. 7, 41–47 (2004) 42. Fuentealba, P. et al. Ivy cells: a population of nitric-oxide-producing, slow- spiking GABAergic neurons and their involvement in hippocampal network activity. Neuron 57, 917–929 (2008). 69. Chu, Y. X. et al. Enhanced synaptic connectivity and epilepsy in C1q knockout mice. Proc. Natl. Acad. Sci. USA 107, 7975–7980 (2010). y 43. Tukker, J. J., Fuentealba, P., Hartwich, K., Somogyi, P. & Klausberger, T. Cell type- speciﬁc tuning of hippocampal interneuron ﬁring during gamma oscillations in vivo. J. Neurosci. 27, 8184–8189 (2007). 70. Michailidou, I. et al. Complement C1q-C3-associated synaptic changes in multiple sclerosis hippocampus. Ann. Neurol. 77, 1007–1026 (2015). 71. Ziv, Y. et al. Immune cells contribute to the maintenance of neurogenesis and spatial learning abilities in adulthood. Nat. Neurosci. 9, 268–275 (2006). 44. Gulyas, A. I. et al. Parvalbumin-containing fast-spiking basket cells generate the ﬁeld potential oscillations induced by cholinergic receptor activation in the hippocampus. J. Supplementary Information accompanies this paper at (https://doi.org/ 10.1038/s41398-018-0288-2). Supplementary Information accompanies this paper at (https://doi.org/ 10.1038/s41398-018-0288-2). Supplementary Information accompanies this paper at (https://doi.org/ 10.1038/s41398-018-0288-2). 25. Neddens, J., Vullhorst, D., Paredes, D. & Buonanno, A. Neuregulin links dopa- minergic and glutamatergic neurotransmission to control hippocampal synaptic plasticity. Commun. Integr. Biol. 2, 261–264 (2009). 25. Neddens, J., Vullhorst, D., Paredes, D. & Buonanno, A. Neuregulin links dopa- minergic and glutamatergic neurotransmission to control hippocampal synaptic plasticity. Commun. Integr. Biol. 2, 261–264 (2009). Received: 19 June 2018 Revised: 24 August 2018 Accepted: 26 September 2018 Received: 19 June 2018 Revised: 24 August 2018 Accepted: 26 September 2018 Deakin et al. Translational Psychiatry (2018) 8:229 Page 8 of 9 Page 8 of 9 26. Nason, M. W. Jr, Adhikari, A., Bozinoski, M., Gordon, J. A. & Role, L. W. Disrupted activity in the hippocampal-accumbens circuit of type III neuregulin 1 mutant mice. Neuropsychopharmacology 36, 488–496 (2011). 51. Ghashghaei, H. T. et al. The role of neuregulin-ErbB4 interactions on the proliferation and organization of cells in the subventricular zone. Proc. Natl. Acad. Sci. USA 103, 1930–1935 (2006). y gy 27. Tamura, H., Kawata, M., Hamaguchi, S., Ishikawa, Y. & Shiosaka, S. Processing of neuregulin-1 by neuropsin regulates GABAergic neuron to control neural plasticity of the mouse hippocampus. J. Neurosci. 32, 12657–12672 (2012). 52. Kriegstein, A. & Alvarez-Buylla, A. The glial nature of embryonic and adult neural stem cells. Annu Rev. Neurosci. 32, 149–184 (2009). 53. Schmid, R. S. et al. Neuregulin 1-erbB2 signaling is required for the estab- lishment of radial glia and their transformation into astrocytes in cerebral cortex. Proc. Natl. Acad. Sci. USA 100, 4251–4256 (2003). 28. Kawata, M., Morikawa, S., Shiosaka, S. & Tamura, H. Ablation of neuropsin- neuregulin 1 signaling imbalances ErbB4 inhibitory networks and disrupts hippocampal gamma oscillation. Transl. Psychiatry 7, e1052 (2017). cortex. Proc. Natl. Acad. Sci. USA 100, 4251–4256 (2003). 54. Sardi, S. P., Murtie, J., Koirala, S., Patten, B. A. & Corfas, G. Presenilin-dependent ErbB4 nuclear signaling regulates the timing of astrogenesis in the developing brain. Cell 127, 185–197 (2006). 29. O’Tuathaigh, C. M. P. et al. Phenotypic characterization of spatial cognition and social behavior in mice with ‘knockout’ of the schizophrenia risk gene neur- egulin 1. Neuroscience 147, 18–27 (2007). 55. Waterhouse, E. G. & Xu, B. J. New insights into the role of brain-derived neurotrophic factor in synaptic plasticity. Mol. Cell Neurosci. 42, 81–89 (2009). 30. Ehrlichman, R. S. et al. Deakin et al. Translational Psychiatry (2018) 8:229 Deakin et al. Translational Psychiatry (2018) 8:229 Supplementary Information accompanies this paper at (https://doi.org/ 10.1038/s41398-018-0288-2). Neurosci. 30, 15134–15145 (2010). 72. Chen, S. et al. Neuregulin 1-erbB signaling is necessary for normal myelination and sensory function. J. Neurosci. 26, 3079–3086 (2006). 45. Colmers, W. F., Lukowiak, K. & Pittman, Q. J. Neuropeptide Y reduces ortho- dromically evoked population spike in rat hippocampal CA1 by a possibly presynaptic mechanism. Brain Res. 346, 404–408 (1985). 73. Nave, K. A. & Salzer, J. L. Axonal regulation of myelination by neuregulin 1. Curr. Opin. Neurobiol. 16, 492–500 (2006). p 74. Taveggia, C. et al. Neuregulin-1 type III determines the ensheathment fate of axons. Neuron 47, 681–694 (2005). 46. Sorensen, A. T. et al. Hippocampal NPY gene transfer attenuates seizures without affecting epilepsy-induced impairment of LTP. Exp. Neurol. 215, 328–333 (2009). 75. Krishnan, A. Neuregulin-1 type I: a hidden power within Schwann cells for triggering peripheral nerve remyelination. Sci. Signal. 6, 270 (2013). 47. Gøtzsche, C. R. et al. Combined gene overexpression of neuropeptide Y and its receptor Y5 in the hippocampus suppresses seizures. Neurobiol. Dis. 45, 288–296 (2012). 6. Howes, O. D. & Kapur, S. The dopamine hypothesis of schizophre 76. Howes, O. D. & Kapur, S. The dopamine hypothesis of schizophrenia: version III--the ﬁnal common pathway. Schizophr. Bull. 35, 549–562 (2009). III--the ﬁnal common pathway. Schizophr. Bull. 35, 549–562 77. Harrison, P. J. Neurochemical alterations in schizophrenia affecting the puta- tive receptor targets of atypical antipsychotics: Focus on dopamine (D1, D3, D4) and 5-HT2A receptors. Br. J. Psychiatry 174, 12–22 (1999). 48. Gray, W. P. Neuropeptide Y signaling on hippocampal stem cells in health and disease. Mol. Cell Endocrinol. 288, 52–62 (2008). 49. Decressac, M. et al. Neuropeptide Y stimulates proliferation, migration and differentiation of neural precursors from the subventricular zone in adult mice. Neurobiol. Dis. 34, 441–449 (2009). 78. Abi-Dargham, A. et al. Increased prefrontal cortical D-1 receptors in drug-naïve patients with schizophrenia: a PET study with [C-11]NNC112. J. Psycho- pharmacol. 26, 794–805 (2012). 50. Geloso, M. C., Corvino, V., Di Maria, V., Marchese, E. & Michetti, F. Cellular targets for neuropeptide Y-mediated control of adult neurogenesis. Front. Cell Neu- rosci. 9, 85 (2015). 79. Schizophrenia Working Group of the Psychiatric Genomics Consortium.. Bio- logical insights from 108 schizophrenia-associated genetic loci. Nature 511, 421–427 (2014). 79. Schizophrenia Working Group of the Psychiatric Genomics Consortium.. Bio- logical insights from 108 schizophrenia-associated genetic loci. Nature 511, 421–427 (2014). Page 9 of 9 Page 9 of 9 Deakin et al. Supplementary Information accompanies this paper at (https://doi.org/ 10.1038/s41398-018-0288-2). Translational Psychiatry (2018) 8:229 87. Kanakry, C. G., Li, Z., Nakai, Y., Sei, Y. & Weinberger, D. R. Neuregulin-1 regulates cell adhesion via an ErbB2/phosphoinositide-3-kinase/Akt-dependent path- way: potential implications for schizophrenia and cancer. PLoS ONE 2, e1369 (2007). 80. Sekar, A. et al. Schizophrenia risk from complex variation of complement component 4. Nature 530, 177–183 (2016). 81. Arion, D., Unger, T., Lewis, D. A., Levitt, P. & Mirnics, K. Molecular evidence for increased expression of genes related to immune and chaperone function in the prefrontal cortex in schizophrenia. Biol. Psychiatry 62, 711–721 (2007). 88. Marballi, K. et al. In vivo and in vitro genetic evidence of involvement of neuregulin 1 in immune system dysregulation. J. Mol. Med. 88, 1133–1141 (2010). 82. Saetre, P. et al. Inﬂammation-related genes up-regulated in schizophrenia brains. BMC Psychiatry 7, 46 (2007). 89. O’Leary, C. et al. Phenotypic effects of maternal immune activation and early postnatal milieu in mice mutant for the schizophrenia risk gene neuregulin-1. Neuroscience 277, 294–305 (2014). 83. Kim, S., Hwang, Y., Webster, M. J. & Lee, D. Differential activation of immune/inﬂammatory response-related co-expression modules in the hippocampus across the major psychiatric disorders. Mol. Psychiatry 21, 376–385 (2016). 90. Dracheva, S. et al. Myelin-associated mRNA and protein expression deﬁcits in the anterior cingulate cortex and hippocampus in elderly schizophrenia patients. Neurobiol. Dis. 21, 531–540 (2006). 84. Sanders, A. R. et al. Transcriptome sequencing study implicates immune- related genes differentially expressed in schizophrenia: new data and a meta- analysis. Transl. Psychiatry 7, e1093 (2017). 91. Harrison, P. J. The hippocampus in schizophrenia: a review of the neuro- pathological evidence and its pathophysiological implications. Psycho- pharmacology 174, 151–162 (2004). 85. Birnbaum, R. et al. Investigating the immunogenetic architecture of schizo- phrenia. Mol. Psychiatry 23, 1251–1260 (2017). 92. Zhang, Z. J. & Reynolds, G. P. A selective decrease in the relative density of parvalbumin-immunoreactive neurons in the hippocampus in schizophrenia. Schizophr. Res. 55, 1–10 (2002). 86. Narayan, S. et al. Molecular proﬁles of schizophrenia in the CNS at different stages of illness. Brain Res. 1239, 235–248 (2008).
https://openalex.org/W4225880908	https://zenodo.org/records/5824596/files/299.pdf	English	null	Reducing Consequences of Car Collision Using Inspiration From Nature	Zenodo (CERN European Organization for Nuclear Research)	2,022	cc-by	5,778	/////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////// /////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////// STATE OF THE ART In biomimetics, engineering problems are solved by biological analogies (Benyus 1997, Vogel 1998, Mak & Shu 2004). Scientific knowledge, gained in biological research, is used for technical applications (Speck & Erb 2009). The degree of the abstraction of biomimetic solutions during the transfer from biology into engineering ranges from an exact copy of the biological system up to an analogy on the level of principles. A higher level of abstraction results in fewer boundaries for the technical realization, but also additional insights about the adopted biologic phenomena are needed (Mak & Shu 2004). ABSTRACT Biomimetic design stimulates the creative idea gen- eration within product development by searching nature’s enormous database of fine tuned solutions. The challenge for the designer is to find efficiently the solution analogies and principles that will solve the design problems. The present paper is a result of a project where researchers from two universities (Technical University of Denmark, DTU, and Tech- nische Universität München, TUM) explored three different biomimetic design approaches namely the transfer checklist approach, the inspiration card ap- proach and the interdisciplinary team approach. The design problem was to reduce the consequences of car collision. All three approaches led to a large number of solution analogies that were then analysed amongst others regarding the structuredness and detail level of solution. The work also addresses a critique earlier raised in the discussions on another biomimetic paper at the International Design Conference in Dubrovnik 2010 (Lenau et al 2010), namely if biomimetic design was best suited for products with relation to humans, like medical devises. The question was if natural inspira- tion could be a valuable inspiration to mechanical problems as those found in the car industry. For this reason the car collision problem was selected. Keywords: Biomimetic design, idea generation, solution space. REDUCING CONSEQUENCES OF CAR COLLISION USING INSPIRATION FROM NATURE Torben Lenau1, Katharina Helten2, Clemens Hepperle2, Sebastian Schenkl2, Udo Lindemann2 (1) Technical University of Denmark, (2) Technische Universität München. Lenau@man.dtu.dk Torben Lenau1, Katharina Helten2, Clemens Hepperle2, Sebastian Schenkl2, Udo Lindemann2 (1) Technical University of Denmark, (2) Technische Universität München. Lenau@man.dtu.dk The paper describes how a problem from the car industry – how to reduce the consequences of car collisions – can be solved using inspiration from na- ture. Three different approaches have been investi- gated, namely the transfer checklist approach, the inspiration card approach and the interdisciplinary team approach. The problem was focused to the handling of the mechanical impact when a car col- lides with solid objects, like another car, a wall or a tree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iomimetic design is an interesting design process that stimulates the creative idea generation within product development. Nature represents an enor- mous database of fine tuned solutions that has been developed through billions of years of evolution. The challenge for the designer is to search in nature and find the solutions and principles that will solve the design problems. Literature reports about several different approaches including procedures and meth- ods for how to work with biomimetics. It was our aim to explore some of the approaches using student projects to get experiences from working with the approaches. In the last years, a lot of research has been con- ducted in the development of biomimetic technolo- gies and products as well as in methods, methodolo- gies and procedural models to support biomimetic DIVERSITY AND UNITY ledge sources like books, scientific papers, the World Wide Web and by contacting biologists. development projects (Speck & Erb 2011). It has been demonstrated that there are many factors that favor the attractiveness of biologically inspired de- sign but also that there are factors that might lead to problems. METHOD AND PROBLEM DESCRIPTION One of the challenges is, that a biological model has to be identified that has the potential to be success- fully transferred into a technology to solve a techni- cal problem. To overcome that, a range of different approaches has been developed in the last years. Three classes of approaches can be distinguished: searching in databases, specially edited for biomi- metics, searching in technical literature and using personal knowledge (Helten et al. 2011). Databases, mainly accessed by a comparison of a functional description of the technical problem, have been developed as papers based catalogues (e.g. Hill 1997, Lindemann & Gramann 2004) or as a software (Löffler 2009). These catalogues list natural systems or earlier developed biologically inspired concepts (Bruck et al. 2007). They may contain only biological solutions, but there are also catalogues with biologi- cal as well as technical solutions (Sarkar et al. 2008). These hybrid catalogues have a broader applicability since there exist technical solutions that are not found in nature like the wheel. The applied research method is a form of action research (Whyte 1991) where the researchers them- selves actively participate in the research process. Two research groups from different universities solved the same design problem using three different approaches, namely a transfer checklist approach (Lindemann & Gramann 2004), the inspiration card approach (Lenau 2009, Lenau et al 2010) and an in- terdisciplinary team approach (Helten et al. 2011). A basic research question for the present work is if the creative idea generation in design work can be stimulated using inspiration from nature. Secondary questions include how this is done using the three different approaches and what the similarities and differences were. THE APPROACHES AND RESULTS TRANSFER CHECKLIST APPROACH Gramann (2004) describes a procedural model involv- ing so-called transfer checklists and Stricker (2006) reports a pre-selection procedure. In this paper we have chosen to name the approach involving the 2 procedures ‘transfer checklist approach’. The check- list approach is an iterative approach, focusing the level of abstraction of the problem formulation (cp. figure 1). The approach was executed by an engi- neering student, without support by a person with a significant expertise in biology. In the first step, the goal of the search for biologic solutions is defined. The goal is described as a function. Vincent & Mann (2002) discuss an approach called BioTriz, enhancing the TRIZ-database with natural systems. An approach using technical literature, not especially edited for biomimetic product development as a source is presented by Shu (2010): Within textbooks of biology, Shu suggest searching a function as well as synonyms in textbooks. The text book gives also a further concretization of the biological solution. Using fundamental literature such as Purves et al. (2003) has the benefit that engineers have less diffi- culties to understand these texts. In the next step, biological systems are allocated to the problem by using the checklist for biological associations according to Gramann (2004). This checklist contains of 177 specific functions that are attributed with associations in nature (cp. figure 2). In the next step, biological systems are allocated to the problem by using the checklist for biological associations according to Gramann (2004). This checklist contains of 177 specific functions that are attributed with associations in nature (cp. figure 2). Subsequent, the allocated systems are analyzed re- garding their suitability and possible implementa- tion. Approches using personal knowledge as a source for biological solutions are discussed e.g. 1H#L5Uer (2009) or Helten et al. (2011). Löffler suggests iden- tifying natural solutions in a brainstorming session. There are also hybrid approaches that incorporate several classes. One of them is the inspiration card method, discussed by Lenau et al. (2010). Within this methodology, the solution search is performed as following: Based on functional keywords searches for biological analogies are performed in different know- Subsequent, the allocated systems are analyzed re- garding their suitability and possible implementa- tion. This evaluation was performed based on the method called “characterizing description” by Stricker (2006) to early pre-select biological effects appropriate for a technical problem. Based on a list of parameters 2 PRODEEDINGS IASDR2011 (e.g. THE APPROACHES AND RESULTS speed, force, etc.), applicable to describe both biological and technical systems, first the technical systems is characterized by describing the pursued specification of the parameters (e.g. high speed, high force). If the biological phenomenon is also characterized by a high speed and high force, the biological phenomenon is taken into account for fur- ther development. If only some parameters show parallel characteristics both for the technical and the biological system, it depends which level of con- gruence the two characterized systems show. The pre-selection method is illustrated in figure 3. In the case study, the problem formulation for "shock absorbance / replacement for car bumpers" was ab- stracted as "attenuating collision impacts". Suitable functions and biological associations were identified by applying the transfer check list (cp. figure 2). The associations are listed below: In the case study, the problem formulation for "shock absorbance / replacement for car bumpers" was ab- stracted as "attenuating collision impacts". Suitable functions and biological associations were identified by applying the transfer check list (cp. figure 2). The associations are listed below: intervertebral disk meniscus / collagen fibres meniscus / collagen fibres During this process, the engineering student had the challenge that only the name of the associations is given. Lacking an appropriate knowledge of biology, the student had problems gathering the working principle of the biological solutions. It was necessary to consult a textbook in biology to get more informa- tion and execute the subsequent steps. Figure 1. Procedural model (Lindemann & Gramann 2004) Figure 2. Excerpt from the transfer checklist between technical functions and terms in biology (Gramann 2004) Figure 2. Excerpt from the transfer checklist between technical functions and terms in biology (Gramann 2004) Based on the more detailed descriptions the “charac- terizing description” method (Stricker 2006) was applied to the five identified biological phenomena (cp. figure 3.). Therefore, first the technical prob- lem was characterized by four suitable parameters chosen from a parameter catalogue provided by Stricker (2006). The parameters with the respective pursued specifications for the technical system are the following: Figure 1. Procedural model (Lindemann & Gramann 2004) If a technical analogy was deducible at a satisfying level of congruence, the system is implemented technically according to Gramann's procedural model (cp. figure 1). If not, the degree of abstraction of the biological system is evaluated. THE APPROACHES AND RESULTS If the degree of abstraction was appropriate, it is checked if the goal definition is adequate as well as if the search was deep enough. low Structure Weight/size high Flexibility low Structure Weight/size high Impact Force/ Structure Weight high Degree of transformation of kinetic en- ergy into deformations 3 DIVERSITY AND UNITY After also describing the above mentioned biological phenomena based on the four parameters, all of the five phenomena show a high congruence with the technical system. The stem, horny skin, combs and meniscus show congruence for three of the four pa- rameters. The intervertebral disk even shows con- gruence in the specifications for all four considered parameters and therefore should be taken into ac- count for the further development. Figure 4: Five step procedure from the inspiration card approach. Figure 4: Five step procedure from the inspiration card approach. Second step is the search for biological analogies and a number of different sources can be used: Intuitive brainstorm based on own knowledge, general world wide-web-search, the website asknature.org, dia- logue with an expert (a biologist), library search based on textbooks and library search in on scientific journals. In the third step the biological analogies are analysed which is done through a literature study or by consulting relevant biologists. The fourth step is to extract the relevant information about the analogies and describe them on inspiration cards (Figure 5). Figure 3. Application of Stricker’s method to preselect promising biological phenomena for the technical challenge Figure 5. Example of inspiration card for ‘woodpecker pecking’ Figure 3. Application of Stricker’s method to preselect promising biological phenomena for the technical challenge Figure 3. Application of Stricker’s method to preselect promising biological phenomena for the technical challenge As a conclusion, the student applying the methods by Gramann and Stricker detected reasonable associa- tions for the problem of “attenuating collision im- pacts”. As Stricker’s method is only a first step to compare biological and technical systems on an ab- stract level, the biological phenomena have to be researched in depth in the next steps before going into detailed design. In order to detect even further appropriate biological phenomena, it is also possible to take other sources for inspiration, such as asknature.org, besides the association list of Gra- mann into consideration. INSPIRATION CARD APPROACH The website asknature.org gave a number of relevant search results for example the energy absorbing properties of the nuchal ligament in grazing animals and the exoskeletons of dragonflies. In total was 15 relevant biological analogies found and for each of those an inspiration card was made. Figure 5 shows one of the cards illustrating the woodpecker shock absorption principle. Three of the cards were then used to sketch pro- posed solutions on a conceptual level for a car bumper: The hedgehog spine, the goat butting and the landing bird bone. The website asknature.org gave a number of relevant search results for example the energy absorbing properties of the nuchal ligament in grazing animals and the exoskeletons of dragonflies. The functional problems were found to be ’Collision shock absorption, absorb energy and control decel- eration’. Further keywords were found by brain- storming based on the student’s own knowledge about how nature solved this type of problems. The brainstorm resulted in biological phenomena like the woodpecker pecking, the fighting of rams (male sheep and goats) and the capability of cats falling from great heights without injury. Based on the mechanisms that animals and plants use to ensure shock absorption more keyword could be formulated like ‘impact AND shell, protection layer, protection mechanism, energy absorption and impact AND pro- tection’. The functional problems were found to be ’Collision shock absorption, absorb energy and control decel- eration’. Further keywords were found by brain- storming based on the student’s own knowledge about how nature solved this type of problems. The brainstorm resulted in biological phenomena like the woodpecker pecking, the fighting of rams (male sheep and goats) and the capability of cats falling from great heights without injury. Based on the mechanisms that animals and plants use to ensure shock absorption more keyword could be formulated like ‘impact AND shell, protection layer, protection mechanism, energy absorption and impact AND pro- tection’. In total was 15 relevant biological analogies found and for each of those an inspiration card was made. Figure 5 shows one of the cards illustrating the woodpecker shock absorption principle. INSPIRATION CARD APPROACH The inspiration card approach is a five step proced- ure that is used to find and describe relevant biolo- gical solutions to functional problems as described in figure 4 (Lenau 2009, Lenau et al. 2010). First step is to identify the relevant functional problems and to rephrase them to make a search possible. In general this is done by formulating the functional problem in terms taken from the specific product and then gen- eralise these formulations, for example as keywords. Figure 5. Example of inspiration card for ‘woodpecker pecking’ Such a card should include the following information: 4 PRODEEDINGS IASDR2011 A title describing the biological organism and the desired phenomena, e.g. Woodpecker pecking. It is an advantage to also mention the Latin name of the organism to allow for easy further information search The chicks that fall out of nests can manage to drop many meters down without getting injured thanks to their feathers and their elastic body. Searching biology books like Life (Purves 2003) showed to be a little difficult. A reason could be that Life does not cover much of the macro biology where the keywords are likely to apply. A more successful search would probably require expanded considera- tions on the keyword formulation e.g. as described by (Shu 2010). However one book showed to be use- ful namely Biomechanics by Vogel (2003). It gave insight into the basic mechanisms of energy trans- formation in animals. A picture or drawing of the organism - pref- erably in action. A description of the phenomena using biology terminology. A description of the interesting principle ap- plied in the phenomena formulated in func- tional engineering terms Journal articles were searched using the Biosis and Zoological records databases and gave relevant re- sults like the surprising information that hedgehog spines also act as shock absorbers for falling animals and the energy absorption in spider webs. A simplified drawing describing the principle Each card represents a conceptual idea and is used in the fifth step to generate design proposals. The method was used on the collision problem by a single design-engineering student in collaboration with a biology trained librarian. The method was used on the collision problem by a single design-engineering student in collaboration with a biology trained librarian. THE FOUND BIOLOGICAL ANALOGIES All together was 42 biological analogies found while working with the three approaches. The transfer checklist approach produced 20 phenomena of which 11 were considered to particular relevant after using the preselection procedure. The inspiration card approach resulted in 15 analogies and the interdisci- plinary approach found 7 analogies. Nevertheless the sheer number of analogies is in general no indication for their quality and applicability. Table 1 lists all the found analogies. Figure 6: Procedure of interdisciplinary team approach (adapted from Helten et al. 2011) As main requirements of a car bumper the engineer- ing student defined the following ones: to protect the car from scratches and dents when bumping into another car, to absorb kinetic energy in a frontal crash and to improve the bumper as the first contact point with a pedestrian. Based on these require- ments main functionalities were derived: INTERDISCIPLINARY TEAM APPROACH Cushion the leg of a pedestrian Finally they came up with the idea to use a liquid with a shear-thinning viscosity for energy absorbance as a principal solution. The fluid remains tenacious at low forces and get fluid at high forces. The starting point of phase B was the generally for- mulated question how nature cares for shock absor- bance. The biology student searched by the use of the following keywords for solutions in sources such as Ask Nature, libraries, internet and eol.org: bio- logical shock absorbance/absorption, natural shock absorption, shock absorption of mam- mals/insects/birds/reptiles/amphibians, cats and human shock absorbance/absorption, molecular ab- sorbance/absorption, shock absorbance in plants, and spinal disc/spinal column fluid. Related findings were the invertebral disc, ventricles of the brain, footpads bush-cricket, exoskeleton dragonfly, foot sole (humans), articular cartilage, and skeleton of cats, see table 1. Figure 6: Procedure of interdisciplinary team approach (adapted from Helten et al. 2011) In phase C the interdisciplinary team looked for dif- ferent kinds of viscosity in natural liquids. The liquor (the liquid between brain and skull) is of a similar viscosity as water, blood is a shear-thinning liquid, and sea-anemones use shear-thickening body liquids to enable extreme shape change. They furthermore seeked to understand the underlying structure of the intervertebral disc. The disc consists of a jelly core that absorbs the shock, surrounded by several layers of fibrocartilage. The final concept then had a simi- lar structure. INTERDISCIPLINARY TEAM APPROACH In contrast to the two approaches mentioned before, this approach is mainly characterized by the inter- disciplinary team. An engineering as well as a biology student worked together to find new biomimetic solutions. Their approach can be described as a four phase procedure, see figure 6. In phase A the engi- neering student defined the problem. The desired functionalities were described, followed by a first draft of principal ways to realize the functionality. In addition, search questions for the biologist were formulated referring to the main principles with the purpose to enhance the biomimetic solution finding process. In phase B the biology student looked in nature for examples and phenomena that fulfilled the required functionality - mainly through literature research, based on the student´s broad knowledge These keywords were then used to search the library database and asknature.com. It was found that some of the keywords gave confusing results since they had more meanings. For example do protection mechanisms like shells have additional functions (protection from predators) so the keyword could be combined with additional keywords like ‘impact’ or ‘shock’. A synonym dictionary could also increase the number of keywords, e.g. ‘collision’ have synonyms like clash, conflict, impact, crash, smash, encounter and shock. Talks with the biology librarian expanded the results further with chicks that fall out of nests, eggs that can withstand relatively large forces and flying birds that resist large impact when landing. 5 DIVERSITY AND UNITY about the phenomena and the biology taxonomy. The phenomena were then presented to the engineering student with focus on the natural structure and the functionality of the objects to make them easylier understandable. In the following phase C both stu- dents discussed together iteratively about the find- ings. The main question was how to adapt the find- ings from nature into the technical problem. Often the discussion was so inspiring that they ran the cy- cle again. They redefined the research question and repeated the search. In some cases a redefinition was needed to go more in detail of the identified natural organism or the like. In the final phase D both students decided on the main principle and generated a concept that incorporated the same. In the following the main questions and findings are presented according to the four phases. DISCUSSION Meniscus / collagen fibers Inspiration card team 1.Woodpecker pecking, 2.Fighting rams, 3.Falling cats, 4.Chicks that fall out of nest, 5.Egg shells, 6.Mussle shells, 7.Leg bones in landing birds, 8.Falling hedge- hogs, 9.Spiderweb, 10.Nuchal ligament in grazing animals, 11.Butting dragonflies (exoskelleton), 12.Albumen – the fluid in bird eggs, 13.Mussle thread attachement to stones, 14.Swaying palm trees, 15.Wood cylinders Interdisciplinary team 1.Invertebral disk, 2.Ventricles with liquor, 3.Bush-cricket foodpads (partially fluid), 4.Dragonfly exoskeleton, 5.Foot sole, 6.Articular cartilage, 7.Skelleton of cats Table 1. Overview of found biological analogies for the collision problem. A ‘’ indicate the phenomena is selected as more relevant for the desired properties. Table 1. Overview of found biological analogies for the collision problem. A ‘’ indicate the phenomena is selected as more relevan for the desired properties. ciples to the basic design problem, which is a desir- able characteristic for good idea generation work. However, since a comparison study of idea genera- tion without using biomimetic methods is missing, it is not possible to qualify the assumption that more or better result are possible using biomimetics. Extent of information: The transfer checklist supply only names of solutions and requires a further con- cretization by using literature. The two other ap- proaches deliver first details already during the first search. Evaluation: The transfer checklist incorporates a systematic evaluation of solutions. The interdiscipli- nary team approach includes an “intuitive” evalua- tion of solutions since the biologist has a deeper background knowledge about phenomena. Thus, a continuous pre-evaluation is done during the genera- tion of ideas by the biologist. The Inspiration card approach does not include a formal evaluation of the found biological phenomena, but their value is indi- rectly measured since the designer have to formulate the functional principles when making the inspiration cards. If it is not possible to formulate a reasonable functional principle for an otherwise interesting bio- logical phenomena, this is probably a good indication of less relevance. The secondary research question refers to the simi- larities and differences between the three different approaches. In the following the main findings re- lated to the overall approach and the solutions are presented. DISCUSSION Being stiff Being compressable The basic research question for the present work was if the creative idea generation in design work could be stimulated using inspiration from nature. For all Being stiff Being compressable 6 6 PRODEEDINGS IASDR2011 approaches, the transfer checklist approach is strongly formalized. three approaches biological inspiration was found which favor a positive answer to this question. Re- markable is that the initial problem of car collision is taken from the automotive industry. The fact that so many ideas could be found emphasizes the usability of biomimetics in all industrial areas, not only the medical industry (see motivation). Impressively the students came up with rather different solution prin- Description of phenomenon: The inspiration card approach focuses on a structured description of the biologic phenomena. In contrary, the interdiscipli- nary team approach shows up a less structured but due to involvement of a biologist a broader descrip- tion of the phenomena. Transfer checklist team 1.Musk ox, 2.Duck, 3.Cat, 4.Flea, 5.Woodpecker, 6.Cassowary, 7.Dragonfly, 8.Birds’ bones, 9.Hedgehog, 10.Pied kingfisker, 11.Pads cusion (mammals), 12.Toucan beaks, 13.Flea, 14.Seeds from brazil nut trees, 15.Stem, 16.Horny skin, 17.Fiber alignment, 18.Combs, 19. Invertebral disk, 20. Meniscus / collagen fibers Inspiration card team 1.Woodpecker pecking, 2.Fighting rams, 3.Falling cats, 4.Chicks that fall out of nest, 5.Egg shells, 6.Mussle shells, 7.Leg bones in landing birds, 8.Falling hedge- hogs, 9.Spiderweb, 10.Nuchal ligament in grazing animals, 11.Butting dragonflies (exoskelleton), 12.Albumen – the fluid in bird eggs, 13.Mussle thread attachement to stones, 14.Swaying palm trees, 15.Wood cylinders Interdisciplinary team 1.Invertebral disk, 2.Ventricles with liquor, 3.Bush-cricket foodpads (partially fluid), 4.Dragonfly exoskeleton, 5.Foot sole, 6.Articular cartilage, 7.Skelleton of cats Table 1. Overview of found biological analogies for the collision problem. A ‘’ indicate the phenomena is selected as more relevant for the desired properties. Transfer checklist team 1.Musk ox, 2.Duck, 3.Cat, 4.Flea, 5.Woodpecker, 6.Cassowary, 7.Dragonfly, 8.Birds’ bones, 9.Hedgehog, 10.Pied kingfisker, 11.Pads cusion (mammals), 12.Toucan beaks, 13.Flea, 14.Seeds from brazil nut trees, 15.Stem, 16.Horny skin, 17.Fiber alignment, 18.Combs, 19. Invertebral disk, 20*. FINDINGS CONCERNING THE IDENTIFIED SOLUTIONS: FINDINGS CONCERNING THE IDENTIFIED SOLUTIONS: All three approaches resulted in a number of useful biological analogies that represented principles for design mechanisms that solve the problem, and some of them were used to propose conceptual design solutions. The learning curve for the students was satisfactory. They all got sufficiently skilled to use the methods in about two weeks, which is of espe- cially of interest for discussions about further indus- trial implementation. One group included a biology student. The collaboration between the two disci- plines worked well. The students expressed their experience of a better understanding of the biologi- cal phenomena. Level of abstraction: Comparing the solutions it can be seen that the transfer checklist approach provides solutions on the subsystems level (e.g. horny skin), whereas the description of the inspiration card ap- proach contains the situation when the ability of the technical system is relevant (e.g. fighting rams, fal- ling cats). The transfer checklist only names the biological phenomenon without explaining the use context (e.g. horny skin, stem). The interdisciplinary team approach delivers a description of the function owners (e.g. skeleton of cats). Biomimetic design faces at least five difficulties or challenges that require increased research focus namely 1) problem delimitation, 2) biology search, 3) understanding biology phenomena, 4) formulate de- sign principles and 5) communication of the biomimetic findings. The first and fourth difficulty are broader design challenges found in most design work and are addressed by the existing design re- search. To overcome the second and third difficulty biology/engineering collaboration and improved knowledge access is required. Improved collabora- tion is probably best approached at the education level where the different competences are brought together in mutual courses and projects. The im- proved knowledge access can be addressed through new types of dedicated knowledge sources or through more intelligent interfaces to the huge exist- ing body of biology literature, such as the transfer checklist approach described in this paper, the asknature website (Asknature 2011) and the biotriz approach (Vincent and Mann 2002) are examples of the former. The inspiration cards described for one of biomimetic approaches in this paper represent one way of overcoming the communication difficulty. Congruence of solutions: Interestingly, no identified solution can be assigned to all three approaches, but at least some solutions can be assigned to two of them. The solutions are complementary, and the solution space shows only a small overlap. FINDINGS CONCERNING THE IDENTIFIED SOLUTIONS: Thus, the approaches can be accomplished complementary. The solutions of the card approach mostly show a zoologic background, whereas the interdisciplinary team approach shows more variation within the solu- tion space – solutions come from human biology and zoology. The transfer list approach shows mainly human and zoology-oriented biological phenomena, still also providing botany-oriented (stem) and bot- any-associated (honeycombs) solutions. PERSPECTIVES After the joined effort with students using biomimetic methods for generating design proposals for a specific problem we are more confident that biomimetics represent a promising way of improving systematic design work. Biomimetics can be ap- proached on different knowledge levels and we find FINDINGS CONCERNING THE PROCEDURAL STEPS: The first and fourth difficulty are broader design challenges found in most design work and are addressed by the existing design re- search. To overcome the second and third difficulty biology/engineering collaboration and improved knowledge access is required. Improved collabora- tion is probably best approached at the education level where the different competences are brought together in mutual courses and projects. The im- proved knowledge access can be addressed through new types of dedicated knowledge sources or through more intelligent interfaces to the huge exist- ing body of biology literature, such as the transfer checklist approach described in this paper, the asknature website (Asknature 2011) and the biotriz approach (Vincent and Mann 2002) are examples of the former. The inspiration cards described for one of biomimetic approaches in this paper represent one way of overcoming the communication difficulty. ACKNOWLEDGEMENT The authors thank Meiken Hansen, Dirk Engelbrecht, FINDINGS CONCERNING THE PROCEDURAL STEPS: Procedural structure: All three approaches present a top-down approach that starts with an explicit problem definition for the technical system and they all provide the possibility of a cyclic procedure in- corporating several iterations. The interdisciplinary team approach includes micro-cycles during the dis- cussion of possible solutions, whereas the transfer list approach shows a macro-cycle (after searching solutions if necessary). In contrast to the other two 7 DIVERSITY AND UNITY that valuable results can emerge even for the de- signer with only a modest knowledge of biology. In the present work 43 design principles were identi- fied. However, large areas of the biology like cell biology, microbiology and biochemistry are difficult to access for the untrained and significant biomimetic search result from these areas require a certain level of knowledge typically found among biologists. A very positive finding from our work therefore was the fruitful collaboration between the engineering and the biology student in the interdis- ciplinary team. Beforehand we were afraid that bi- ologists in a collaborative biomimetic design team only would see themselves as knowledge providers and therefore only have limited interest. The biology student emphasized that the collaboration increases the collaboration competences that would improve the student´s value on the labor market and give access to a broader variety of jobs. that valuable results can emerge even for the de- signer with only a modest knowledge of biology. In the present work 43 design principles were identi- fied. However, large areas of the biology like cell biology, microbiology and biochemistry are difficult to access for the untrained and significant biomimetic search result from these areas require a certain level of knowledge typically found among biologists. A very positive finding from our work therefore was the fruitful collaboration between the engineering and the biology student in the interdis- ciplinary team. Beforehand we were afraid that bi- ologists in a collaborative biomimetic design team only would see themselves as knowledge providers and therefore only have limited interest. The biology student emphasized that the collaboration increases the collaboration competences that would improve the student´s value on the labor market and give access to a broader variety of jobs. Biomimetic design faces at least five difficulties or challenges that require increased research focus namely 1) problem delimitation, 2) biology search, 3) understanding biology phenomena, 4) formulate de- sign principles and 5) communication of the biomimetic findings. REFERENCES Asknature, www.asknature.org. Benyus, J.M. (1997) : Biomimicry: Innovation Inspired by Nature, William Morrow. Benyus, J.M. (1997) : Biomimicry: Innovation Inspired by Nature, William Morrow. Bruck, H. A., Gershon, A. L., Golden, I., Gupta, S. K., Gyger Jr., L. S., Magrab, E. B., Spranklin, B. W. (2007) "Training mechanical engineering students to utilize biological inspiration during prod- uct development.", Bioinspiration & Biomimetics 2(2), pp.198- 209. Gramann, J. (2004) Problemmodelle und Bionik als Methode, Doctoral thesis, Technische Universität München. Helten, K., Schenkl, S. and Lindemann, U. (2011) Biologizing product development - results from a student project. In: Chakra- barti, A. (Ed.): International Conference in Research into Design ICoRD2011, Bangalore, pp27-34. Hill, B.(2007) Innovationsquelle Natur, Shaker, Aachen. Hill, B. (1993) Bionik – notwendiges Element im Konstruktionspro- zess, Konstruktion, Vol. 45, 1993, pp.283-287 Lindemann, U. and Gramann, J. (2004) Engineering Design using Biological Principles. 8th International Design Conference - DE- SIGN 2004. Zagreb. Lenau, T. (2009) Biomimetics as a design methodology – possibili- ties and challenges, International Conference on Engineering Design, ICED'09 Stanford University, CA, USA. Lenau, T., A., Dentel, Ingvarsdóttir, Þ., Guðlaugsson T. (2010) Engineering Design of an Adaptive Leg Prosthesis Using Biological Principles”, International Design Conference - Design 2010, pp. 331-340. Löffler, S. (2009) Anwenden bionischer Konstruktionsprinzipe in der Produktentwicklung, Doctiral Thesis TU Braunschweig Mak, T.W. & Shu, L.H. (2004) Abstraction of Biological Analogies for Design. CIRP Annals - Manufacturing Technology, 53(1), pp117- 120. Purves, W.K., Sadava, D., Orians, G.H., Heller, H.C. (2003) Life: The Science of Biology, W.H. Freeman Co, 7th Edition. Sarkar, P., Phaneendra, S., Chakrabarti, A. (2008) Developing Engineering Products Using Inspiration From Nature, Journal of Computer Science and Information Engineering 8(3). Shu, L. (2010) A Natural-language Approach to Biomimetic Design, Journal on Artificial Intelligence for Engineering Design, Analysis and Manufacturing, 24(4), 2010, pp. 507-519. Speck, T. and Erb, R. (2009) Prozessketten in Natur und Wirt- schaft. In Otto, K.-S. and Speck, T. (eds) Darwin meets Business, Gabler, Wiesbaden, pp95-112. Speck, T. and Erb, R. (2011) Biomimetics International Industrial Convention, BIOKON Berlin. Stricker, H.(2006) Bionik in der Produktentwicklung unter der Berücksichtigung menschlichen Verhaltens. Doctoral thesis, Tech- nische Universität München. Vincent, J. F. V. and Mann, D. L. (2002) Systematic technology transfer from biology to engineering, Philosophical Transactions of the Royal Society A: Mathematical, Physical & Engineering Sci- ences 360(2), pp. 159-173. Vogel, S. (1998) Cats' Paws and Catapults: Mechanical Worlds of Nature and People, Norton. Vogel S. ACKNOWLEDGEMENT The authors thank Meiken Hansen, Dirk Engelbrecht, Moritz Heber and Verena Huber for their highly moti- vated engagement in this study. 8 PRODEEDINGS IASDR2011 REFERENCES (2003) Comparative Biomechanics, Princeton University Press. Whyte, W.F. (ed) (1991) Participatory action research, Sage Publi- cations, Newbury Park, California. 9
https://openalex.org/W2100022944	https://pure.aber.ac.uk/portal/files/63613016/bioinformatics_26_12_1513.pdf	English	null	LCE: a link-based cluster ensemble method for improved gene expression data analysis	Bioinformatics	2,010	cc-by	7,340	LCE Citation for published version (APA): Iam-On, N., Boongoen, T., & Garrett, S. (2010). LCE: A link-based cluster ensemble method for improved gene expression data analysis. Bioinformatics, 26(12), 1513-1519. Article btq226. https://doi.org/10.1093/bioinformatics/btq226 Citation for published version (APA): Iam-On, N., Boongoen, T., & Garrett, S. (2010). LCE: A link-based cluster ensemble method for improved gene expression data analysis. Bioinformatics, 26(12), 1513-1519. Article btq226. https://doi.org/10.1093/bioinformatics/btq226 Document License CC BY Document License CC BY Aberystwyth University LCE Iam-On, Natthakan; Boongoen, Tossapon; Garrett, Simon LCE Iam-On, Natthakan; Boongoen, Tossapon; Garrett, Simon LCE Citation for published version (APA): Iam-On, N., Boongoen, T., & Garrett, S. (2010). LCE: A link-based cluster ensemble method for improved gene expression data analysis. Bioinformatics, 26(12), 1513-1519. Article btq226. https://doi.org/10.1093/bioinformatics/btq226 General rights General rights Copyright and moral rights for the publications made accessible in the Aberystwyth Research Portal (the Institutional Repository) are retained by the authors and/or other copyright owners and it is a condition of accessing publications that users recognise and abide by the legal requirements associated with these rights. • Users may download and print one copy of any publication from the Aberystwyth Research Portal for the purpose of private study or research. • Users may download and print one copy of any publication from the Aberystwyth Research Portal for the purp research esearch. • You may not further distribute the material or use it for any profit-making activity or commercial gain • You may freely distribute the URL identifying the publication in the Aberystwyth Research Portal • You may not further distribute the material or use it for any profit-making activity or commercial gain • You may freely distribute the URL identifying the publication in the Aberystwyth Research Portal ABSTRACT Motivation: It is far from trivial to select the most effective clustering method and its parameterization, for a particular set of gene expression data, because there are a very large number of possibilities. Although many researchers still prefer to use hierarchical clustering in one form or another, this is often sub-optimal. Cluster ensemble research solves this problem by automatically combining multiple data partitions from different clusterings to improve both the robustness and quality of the clustering result. However, many existing ensemble techniques use an association matrix to summarize sample-cluster co-occurrence statistics, and relations within an ensemble are encapsulated only at coarse level, while those existing among clusters are completely neglected. Discovering these missing associations may greatly extend the capability of the ensemble methodology for microarray data clustering. Clinical researchers commonly use simple clustering methods, such as agglomerative hierarchical and k-means (Bredel et al., 2005; Sorlie et al., 2003) to cluster cancer microarray samples, despite the advent of several new techniques that capitalize on the inherent characteristics of gene expression data (noise and high dimensionality) to improve clustering quality (e.g. Brunet et al., 2004; Liu et al., 2003; McLachlan et al., 2002). de Souto et al. (2008) says, this is because the use of such methods is difﬁcult for non-expert users. Recently, cluster ensembles or consensus clusterings have emerged as simple, effective, one-stop methods for improving the robustness and quality of clustering results. Cluster ensembles combine multiple clustering decisions (referred to as ‘base clusterings’ or ‘ensemble members’) where the base clusterings contain diversity in their choice of clusters by: (i) using a single clustering algorithm with random parameter initializations (Kim et al., 2009; Monti et al., 2003; Yu et al., 2007); (ii) employing multiple clustering algorithms (Swift et al., 2004); (iii) selecting a random number of clusters (Fred and Jain, 2005; Kuncheva and Vetrov, 2006); (iv) using different subsets of gene (Avogadri and Valentini, 2009; Yu et al., 2007); or (v) using data sampling techniques (Dudoit and Fridyand, 2003; Monti et al., 2003). Most existing methods compare cluster associations between each of the N samples in the dataset to produce an N ×N pairwise similarity matrix [i.e. consensus (Monti et al., 2003), agreement (Swift et al., 2004) and co-association (Fred and Jain, 2005) matrices], to which a consensus function (e.g. agglomerative hierarchical clustering) is applied to acquire the ﬁnal data partition. ABSTRACT With the ensemble of two base clusterings ={π1,π2} and ﬁve samples (x1,...x5) that is given in Figure 1a, the corresponding similarity matrix is shown in Figure 1b. Results: The link-based cluster ensemble (LCE) method, presented here, implements these ideas and demonstrates outstanding performance. Experiment results on real gene expression and synthetic datasets indicate that LCE: (i) usually outperforms the existing cluster ensemble algorithms in individual tests and, overall, is clearly class-leading; (ii) generates excellent, robust performance across different types of data, especially with the presence of noise and imbalanced data clusters; (iii) provides a high-level data matrix that is applicable to many numerical clustering techniques; and (iv) is computationally efﬁcient for large datasets and gene clustering. Availability: Online supplementary and implementation are available at: http://users.aber.ac.uk/nii07/bioinformatics2010 Availability: Online supplementary and implementation are available at: http://users.aber.ac.uk/nii07/bioinformatics2010 Contact: nii07@aber.ac.uk; natthakan@mfu.ac.th Supplementary information: Supplementary data are available at Bioinformatics online. Supplementary information: Supplementary data are available at Bioinformatics online. Supplementary information: Supplementary data are available at Bioinformatics online. Received on December 14, 2009; revised on March 18, 2010; accepted on April 20, 2010 An alternative approach (Fern and Brodley, 2004; Strehl and Ghosh, 2002) to pairwise similarity methods makes use of an N ×P binary cluster-association matrix (BM) (where P denotes the number of clusters in an ensemble). Figure 1c shows the example of such matrix that is generated from the ensemble of Figure 1a. Despite reported success and efﬁciency, these methods generate the ultimate clustering result based on incomplete information of a cluster ensemble. The underlying association matrix presents sample–cluster relations at a coarse level and completely ignores the relations among clusters (Iam-on et al., 2008). As a result, the ∗To whom correspondence should be addressed. LCE: a link-based cluster ensemble method for improved gene expression data analysis expression data analysis Natthakan Iam-on1,∗, Tossapon Boongoen1,2 and Simon Garrett1 1Department of Computer Science, Aberystwyth University, Aberystwyth, Ceredigion, UK and 2Department of Mathematics and Computer Science, Royal Thai Air Force Academy, Thailand Natthakan Iam-on1,∗, Tossapon Boongoen1,2 and Simon Garrett1 1Department of Computer Science, Aberystwyth University, Aberystwyth, Ceredigion, UK an Mathematics and Computer Science, Royal Thai Air Force Academy, Thailand Associate Editor: Olga Troyanskaya Downloaded from https://academic.oup.com/bioinformatics/article/26/12/1513/286962 by Aberystwyth University user on 09 March 2023 clustering algorithm can perform the best for all datasets (Kuncheva and Hadjitodorov, 2004), and discovering all types of cluster shapes and structures presented in data is impossible for any known clustering algorithm (Duda et al., 2000; Handl et al., 2005). ORIGINAL PAPER BIOINFORMATICS ORIGINAL PAPER Vol. 26 no. 12 2010, pages 1513–1519 doi:10.1093/bioinformatics/btq226 BIOINFORMATICS ORIGINAL PAPER ORIGINAL PAPER Oxford University Press. All rights reserved. For Permissions, please email: journals.permissions@oxfordjournals.org Take down policy Take down policy If you believe that this document breaches copyright please contact us providing details, and we will remove access to the work immediately and investigate your claim. tel: +44 1970 62 2400 email: is@aber.ac.uk tel: +44 1970 62 2400 email: is@aber.ac.uk Download date: 24. Oct. 2024 © The Author 2010. Published by Oxford University Press. All rights reserved. For Permissions, please email: journals.permiss 2 METHODS The proposed LCE methodology is illustrated in Figure 2. It includes three major steps: (i) creating M base clusterings to form a cluster ensemble; (ii) creating a reﬁned cluster-association matrix (RM) using a link-based similarity algorithm (Weighted Connected-Triples, WCT); and (iii) generating the ﬁnal data partition by exploiting the spectral graph partitioning (SPEC) technique as a consensus function. This framework is similar to that of HBGF (Fern and Brodley, 2004), except the second step that is introduced for constructing a reﬁned information matrix.As compared to HBGF that is based on the BM, LCE may enhance effectiveness of the former using a more informative RM. Based on this insight, the reﬁned cluster-association matrix (RM) is put forward as the enhanced variation of the original BM. Its aim is to approximate value of unknown associations (‘0’) from known ones (‘1’), whose association degrees are preserved within the RM (i.e. BM(xi,cl)=1→ RM(xi,cl)=1). For each clustering πt,t =1...M and their corresponding clusters Ct 1,...,Ct kt (where kt is the number of clusters in the clustering πt), the association degree RM(xi,cl)∈[0,1] that sample xi ∈X has with each cluster cl∈{Ct 1,...,Ct kt } is estimated as follows: 1 INTRODUCTION The use of clustering is vital both for visualizing and extracting useful information from microarray data. However, different algorithms (or even the same algorithm with different parameters) often provide distinct clusterings.As a result, it is extremely difﬁcult for users to decide which algorithm and parameters will be optimal for a given set of data—this is because no single-pass/simple 1513 shed by Oxford University Press. All rights reserved. For Permissions, please email: journals.permissions@oxfordjournals.org N.Iam-on et al. Fig. 2. The LCE framework: (i) a cluster ensemble ={π1,...,πM} created from M base clusterings; (ii) a reﬁned cluster-association matri (RM) is then generated from the ensemble using the WCT algorithm; an (iii) a ﬁnal clustering result (π∗) is produced by a consensus function of th spectral graph partitioning (SPEC) Fig. 1. An example of (a) cluster ensemble of samples {x1...x5} that consists of two base clusterings (π1 ={C1 1,C1 2,C1 3} and π2 ={C2 1,C2 2}), (b) the corresponding pairwise similarity matrix and (c) BM, respectively. Fig. 1. An example of (a) cluster ensemble of samples {x1...x5} that consists of two base clusterings (π1 ={C1 1,C1 2,C1 3} and π2 ={C2 1,C2 2}), (b) the corresponding pairwise similarity matrix and (c) BM, respectively. performance of such techniques may consequently be degraded as many matrix entries are left unknown, each presented with zero. (1) q=qmin +⌊α(qmax −qmin)⌋ (1) In response, we present a new method—the LCE—for clustering data. It signiﬁcantly extends the hybrid bipartite graph formulation (HBGF) technique (Fern and Brodley, 2004), by applying a graph- based consensus function to an improved cluster association matrix, instead of the conventional BM. This article explores its application to the problem of clustering cancer microarray samples, and is shown to reﬁne the cluster-association matrix, as well as reducing the number of unknown entries and, therefore, increasing accuracy; moreover, it can easily replace or augment a researcher’s existing clustering tools. here α∈[0,1] is a uniform random variable, qmin and qmax are the lower and upper bounds of the generated subspace, respectively. In particular, qmin and qmax are set to 0.75d and 0.85d. A gene is selected one by one from the pool of d genes, until the collection of q is obtained. The index of each selected gene is determined as follows, where h denotes the h-th gene in the pool of d genes and β∈[0,1) is a uniform random variable. h=⌊1+βd⌋ (2) Generating a reﬁned cluster-association matrix (RM) In particular to HBGF, BM has been used to summarized information presented in an ensemble . Each entry in this matrix BM(xi,Cj)∈{0,1} represents a crisp association degree between sample xi ∈X and cluster Cj ∈. According to Figure 1, which shows an example of cluster ensemble and the corresponding BM, a large number of entries in the BM are unknown, each presented with 0. Intuitively, this may limit the quality of a data partition generated by any consensus function. These conditions occur when relations between different clusters of a base clustering are originally assumed to be nil. It is important to note that each sample can associate (to a certain degree within [0,1]) to several clusters of any particular clustering, at the same time. These hidden or unknown associations can be estimated upon the similarity among clusters, discovered from a link network of clusters. 2.2 Applying a consensus function to RM Having obtained a reﬁned cluster-association matrix (RM) with the aforementioned link-based similarity algorithm, a graph-based partitioning method is exploited to obtain the ﬁnal clustering. This consensus function requires the underlying matrix to be initially transformed into a weighted bipartite graph. Formally, given an RM representing associations between N samples and P clusters in an ensemble , a weighted bipartite graph G=(V,W) can be constructed, where V =VX ∪VC is a set of vertices representing both samples VX and clusters VC, and W denotes a set of weighted edges that can be deﬁned as follows: 2.1.1 WCT: a link-based similarity algorithm Given a cluster ensemble of data samples X, a weighted graph G=(V,W) can be constructed, where V is the set of vertices each representing a cluster and W is a set of weighted edges between clusters. Formally, the weight assigned to the edge wxy ∈W, that connects clusters Cx,Cy ∈V, is estimated by wxy = \|Lx ∩Ly\| \|Lx ∪Ly\| (4) • wij =0 when vertices vi,vj ∈VX, i.e. correspond to samples. (4) • wij =0 when vertices vi,vj ∈VC, i.e. correspond to clusters. • wij =RM(vi,vj) when vertices vi ∈VX and vj ∈VC. The bipartite graph G is bi-directional such that wij is equivalent to wji. where Lz ⊂X denotes the set of samples belonging to cluster Cz ∈V. Figure 3a shows the network of clusters that is generated from the example given in Figure 1. Note that circle nodes represent clusters and edges exist only when the corresponding weights are non-zero. Given such graph, the spectral graph partitioning (SPEC) method similar to that of Ng et al. (2002) is applied to generate a ﬁnal data partition. This is a powerful method for decomposing an undirected graph, with good performance being exhibited in many application areas, including protein modelling, information retrieval and identiﬁcation of densely connected on-line hypertextual regions (Luxburg, 2007). Principally, given a graph G=(V,W), SPEC ﬁrst ﬁnds the K largest eigenvectors u1,...,uK of W, which are used to formed another matrix U (i.e. U =[u1,...,uK]), whose rows are then normalized to have unit length. By considering the row of U as K-dimensional embedding of the graph vertices, SPEC applies k-means to these embedded points in order to acquire the ﬁnal clustering result. Further details of SPEC can be found in Supplementary Section 1. 2.1 Creating a cluster ensemble Let X ={x1,...,xN} be a set of N samples and let ={π1,...,πM} be a cluster ensemble with M base clustering results. Each base clustering returns a set of clusters πi ={Ci 1,Ci 2,...,Ci ki}, such that ki j=1Ci j =X, where ki is the number of clusters in the i-th clustering. As in many previous studies (Fred and Jain, 2005; Kim et al., 2009), the k-means clustering algorithm is used to generate base clusterings, each with random initialization of cluster centers. Euclidean distance is used to measure the dissimilarity between two samples unless stated otherwise. For each base clustering, there are two schemes of selecting the number of clusters: Fixed-k (k = √ N, where N is the number of samples) and Random-k (k ∈ {2,..., √ N}). To create diversity in an ensemble, k should be greater than the expected number of clusters and the common rule-of-thumb is k = √ N (Fred and Jain, 2005; Hadjitodorov et al., 2006). Note that the quality of the RM(xi,cl)= 1 ifcl=Ct ∗(xi) sim(cl,Ct ∗(xi)) otherwise (3) (3) where Ct ∗(xi) is a cluster label (corresponding to a particular cluster of the clustering πt) to which the sample xi belongs. In addition, sim(Cx,Cy)∈ [0,1] denotes the similarity between any two clusters Cx,Cy, which can be discovered using the following link-based algorithm. Note that, for any clustering πt ∈, 1≤ ∀C∈πt RM(xi,C)≤kt. Unlike the measure of fuzzy membership, the typical constraint of ∀C∈πt RM(xi,C)=1 is 1514 LCE method for improved gene expression data analysis Fig. 4. Details of disclosed WCT similarities/relations with DC being 0.9, and the resulting RM. Fig. 3. Examples of (a) cluster network and (b) connected-triple between vertices C1 1 and C1 3, where wC1 1C1 3 =0. Fig. 3. Examples of (a) cluster network and (b) connected-triple between vertices C1 1 and C1 3, where wC1 1C1 3 =0. Fig. 4. Details of disclosed WCT similarities/relations with DC being 0.9, and the resulting RM. not appropriate for re-scaling associations within the RM. In fact, such local normalization will signiﬁcantly distort the true semantics of known associations (‘1’), such that their magnitudes become dissimilar, different from one clustering to another. According to our empirical investigation, the quality of RM is usually higher than other soft, fuzzy-like variations of the BM, which can be obtained from sample-to-cluster distances or a fuzzy cluster ensemble. 2.1 Creating a cluster ensemble See Supplementary Section 8.1 for details of such methods and associated experimental results. 2.3 Experiment design The experiments set out to investigate the performance of LCE compared to a number of different simple/standard clustering algorithms and state-of- the-art cluster ensemble methods, over real gene expression and synthetic datasets. The compared techniques include: (i) HBGF that is the baseline model of LCE; (ii) four simple clustering techniques that are usually used by clinical researchers to analyse microarray data [k-means (KM), single-linkage (SL), complete-linkage (CL) and average-linkage (AL)]; (iii) three pairwise similarity-based cluster ensemble algorithms that have been developed so far for gene expression data analysis [MULTI-K, consensus clustering with hierarchical clustering (CCHC) and graph-based consensus clustering (GCC)], and three graph-based cluster ensemble techniques that have been considered as benchmarks in the literature [Cluster- based Similarity Partitioning Algorithm (CSPA), Hyper-Graph Partitioning Algorithm (HGPA) and Meta-Clustering Algorithm (MCLA)]. Details of examined cluster ensemble techniques are given below. WCTk xy =min(wxk,wyk) (5) (5) where wxk,wyk ∈W are weights of the edges connecting clusters Cx and Ck, and clusters Cy and Ck, respectively. The count of all triples (1...q) between clusters Cx and Cy can be calculated as follows: where wxk,wyk ∈W are weights of the edges connecting clusters Cx and Ck, and clusters Cy and Ck, respectively. The count of all triples (1...q) between clusters Cx and Cy can be calculated as follows: WCTxy = q k=1 WCTk xy (6) (6) the similarity between clusters Cx and Cy can be estimated b Then, the similarity between clusters Cx and Cy can be estimated by sim(Cx,Cy)= WCTxy WCTmax × DC (7) (7) where WCTmax is the maximum WCTpq value of any two clusters Cp,Cq ∈V and DC ∈(0,1) is a constant decay factor (i.e. conﬁdence level of accepting two non-identical clusters as being similar). Following the example shown in Figs 1 and 3, the discovered link-based similarities/relations and the resulting RM are presented in Figure 4. where WCTmax is the maximum WCTpq value of any two clusters Cp,Cq ∈V and DC ∈(0,1) is a constant decay factor (i.e. conﬁdence level of accepting two non-identical clusters as being similar). Following the example shown in Figs 1 and 3, the discovered link-based similarities/relations and the resulting RM are presented in Figure 4. • Pairwise similarity-based cluster ensemble methods are based principally on the pairwise similarity among samples. 2.2 Applying a consensus function to RM Given this network formalism, the new WCT algorithm is introduced to disclose the similarity between any pair of clusters. It extends the Connected-Triple method (Reuther and Walter, 2006) that has been originally developed to identify ambiguous author names within publication databases. In particular, the similarity of any Cx,Cy ∈V can be estimated by counting the number of Connected-Triples (i.e. triples) they are part of. Formally, a triple, Triple = (VTriple,WTriple), is a subgraph of G containing three vertices VTriple ={Cx,Cy,Ck} ⊂V and two non-zero edges WTriple ={wxk,wyk}⊂W, with wxy =0. An example of triple within the network of Figure 3a is shown in Figure 3b. This simple counting might be sufﬁcient for any indivisible object, e.g. name or sample. However, to evaluate the similarity between clusters, it is important to realize and take into account the composite characteristic of a cluster (i.e. shared members). Inspired by this idea, the WCT measure of clusters Cx,Cy ∈V with respect to each triple Ck ∈V, is estimated as 2.3 Experiment design Note that Cm(xi) denotes the cluster label of the m-th clustering to which a sample xi ∈X belongs. 2.3.1 Real gene expression datasets This evaluation is based on real gene expression data, obtained from nine published microarray studies, and summarized in Table 1. The experiments were conducted over ﬁltered datasets as given in the empirical study of de Souto et al. (2008), where uninformative genes are removed for a better quality of clustering result. Details of the types of datasets, data preprocessing and the gene selection method are given in Supplementary Sections 2.1–2.2.To rigourously evaluate the robustness of LCE and its compared techniques, they are also assessed on both simulated gene expression data (with noise and imbalanced clusters) and geometrically complicated datasets (see Supplementary Sections 3–4 for data descriptions). Since co-association matrix (CO) is a similarity matrix, any similarity- based clustering algorithm (referred to as ‘consensus function’) can be applied to this matrix to yield the ﬁnal partition π∗(Fred and Jain, 2005). Among several existing similarity-based techniques, the most well-known is agglomerative hierarchical clustering algorithm. Speciﬁcally to the problem of clustering cancer samples, MULTI-K (Kim et al., 2009) and CCHC (Monti et al., 2003) methods make use the SL and AL agglomerative hierarchical clusterings as consensus functions, respectively. In addition, to obtain π∗, the GCC approach (Yu et al., 2007) transforms the CO matrix into a graph of samples to which the normalized cut algorithm (Shi and Malik, 2000) is applied. Graph-based cluster ensemble algorithms that are investigated herein include the methods of Fern and Brodley (2004) (HBGF) and Strehl and Ghosh (2002) (CSPA, HGPA and MCLA). Note that HBGF is included as the baseline model of LCE. It makes use of the bipartite graph that is generated from the BM. There is no edge connecting vertices of the same object type, and the weight of an edge between any data point and cluster is either 1 (when the sample belongs to the cluster) or 0 (otherwise). SPEC (Ng et al., 2002) is exploited to obtain the ﬁnal clustering result from this graph. This effectively allows the quality of the two cluster-association matrices (i.e. BM and RM) to be compared. CSPA creates a similarity graph, where vertices represent samples and edges’ weight represent similarity scores obtained from the CO matrix. 2.3 Experiment design Afterwards, a graph partitioning algorithm called METIS (Karypis and Kumar, 1998) is used to partition the similarity graph into K clusters. HGPA constructs a hyper-graph, where vertices represent samples and the same-weighted hyper-edges represent clusters in the ensemble. Then, HMETIS (Karypis et al., 1999) is applied to partition the underlying hyper-graph into K parts. MCLA creates a graph where each vertex corresponds to each cluster in the ensemble and each edge’s weight between any two cluster vertices is computed using the binary Jaccard measure. METIS is also employed to partition the meta-level graph into K meta-clusters. The ﬁnal clustering is produced by assigning each sample to the meta-cluster with which it is most frequently associated. 2.3.2 Experiment setting The proposed LCE method and its competitors are evaluated, using the experiment setting illustrated below. • Each cluster ensemble method is evaluated over four different types of ensemble: (i) Fixed-k with full-space data (with d genes), (ii) Fixed-k with subspace data (with q genes), (iii) Random-k with full-space data and (iv) Random-k with subspace data, respectively. • An ensemble size (M) of only 10 base clusterings was used. • To generate a reﬁned cluster-association matrix (RM), the constant decay factor (DC) of 0.9 is exploited with the underlying link-based similarity algorithm (i.e. WCT). • For a comparison purpose, as in Fern and Brodley (2004) and Kim et al. (2009), each clustering method divides data points into a partition of K (the number of true classes for each dataset, known as ‘gold standard’) clusters, which is then evaluated against the corresponding true partition using a set of well-known evaluation indices. Note that, true classes are known for all datasets but are absolutely not used in any way by the cluster ensemble process; they are only used to evaluate the quality of the clustering results after clustering is complete. This assessment framework has been successfully adopted in de Souto et al. (2008) to compare the performance of different simple clustering algorithms over a large number of gene expression datasets. • The current research follows several previous studies (Kim et al., 2009; Monti et al., 2003; Yu et al., 2007) that focus on clustering samples of a given microarray data into known groups, i.e. class prediction. 2.3 Experiment design Given a cluster ensemble ={π1,...,πM} of a dataset X ={x1,...,xN}, an N ×N • Pairwise similarity-based cluster ensemble methods are based principally on the pairwise similarity among samples. Given a cluster ensemble ={π1,...,πM} of a dataset X ={x1,...,xN}, an N ×N 1515 N.Iam-on et al. Table 1. Description of real gene expression datasets: tissue type, microarray chip type, number of samples (N), number of original genes (d∗), number of selected genes (d), number of classes (K) and class distribution sion datasets: tissue type, microarray chip type, number of samples (N), number of original genes (d∗), number of and class distribution Table 1. Description of real gene expression datasets: tissue type, microarray chip type, number of samples (N), number of original genes (d∗), number of selected genes (d), number of classes (K) and class distribution Dataset Tissue Chip Samples Original Selected Classes Class (N) genes (d∗) genes (d) (K) distribution Leukemia1 (Golub et al., 1999) Bone marrow Affy 72 7129 1877 2 47, 25 Leukemia2 (Golub et al., 1999) Bone marrow Affy 72 7129 1877 3 38, 9, 25 Leukemia3 (Armstrong et al., 2002) Blood Affy 72 12582 2194 3 20, 24, 28 Breast-Colon tumors (Chowdary et al., 2006) Breast and colon Affy 104 22283 182 2 62, 42 Brain Tumor (Nutt et al., 2003) Brain Affy 50 12625 1377 4 14, 14, 7, 15 Central nervous system (Pomeroy et al., 2002) Brain Affy 42 7129 1379 5 10, 8, 10, 10, 4 Multi-tissue1 (Ramaswamy et al., 2001) Multi-tissue Affy 190 16063 1363 14 11, 10, 11, 11, 22, 11, 10, 10, 30, 11, 11, 11, 11, 20 Multi-tissue2 (Su et al., 2001) Multi-tissue Affy 174 12533 1571 10 26, 8, 26, 23, 12, 11, 7, 27, 6, 28 Hepatocellular carcinoma (Chen et al., 2002) Liver cDNA 180 22699 85 2 104, 76 Small, round blue-cell tumors (Khan et al., 2001) Multi-tissue cDNA 83 6567 1069 4 29, 11, 18, 25 See Supplementary Material for further details. Downloaded from https://academic.oup.com/bioinformatics/article/26/12/1513/286962 by Aberystwyth University user on 09 March 2023 The effectiveness of LCE and other cluster ensemble methods with different ensemble sizes and types are also empirically examined. Details of gene expression datasets and experiment setting are presented below. similarity matrix (CO) is constructed as CO(xi,xj)= 1 M M m=1Sm(xi,xj), where CO(xi,xj)∈[0,1] represents the similarity measure between samples xi,xj ∈X. In addition, Sm(xi,xj)=1 if Cm(xi)=Cm(xj), and Sm(xi,xj)=0 otherwise. See Supplementary Material for further details. 2.3 Experiment design In particular to these methods, the quality of data partition π∗generated by a clustering technique is directly compared against the Note that the performance of SL, CL, AL and KM are always assessed over the original data, without using any information of cluster ensemble. 1516 LCE method for improved gene expression data analysis Fig. 6. The statistics of total performance, summarized across all evaluation indices, i.e. (B−W)i,∀i∈CM. Fig. 5. Average validity measures of different clustering methods, across all validity indices (CA, NMI, AR) and experimental settings. Fig. 5. Average validity measures of different clustering methods, across all validity indices (CA, NMI, AR) and experimental settings. Fig. 6. The statistics of total performance, summarized across all evaluation indices, i.e. (B−W)i,∀i∈CM. known partition ′ (i.e. class labels), using external validity indices such as Adjusted Rand (AR; Hubert and Arabie, 1985), Normalized Mutual Information (NMI; Strehl and Ghosh, 2002) and Classiﬁcation Accuracy (CA; Nguyen and Caruana, 2007). These speciﬁc indices are exploited for evaluating the performance of the proposed LCE method, against several other clustering techniques. setting β∈ST (ST is a set of 40 unique combination of four ensemble types and ten real gene expression datasets). The 95% CI, [LXC(i,β),UXC(i,β)], for the mean XC(i,β) of each validity criterion C is calculated by LXC(i,β) =XC(i,β)−1.96 SC(i,β) √n and UXC(i,β) =XC(i,β)+1.96 SC(i,β) √n . Note that SC(i,β) denotes the SD of the validity index C across n runs for a clustering method i and an experiment setting β. In addition, multiple runs of any setting β∈ST are different and independent—each with a unique ensemble that is generated by randomly selected parameters, and possibly dissimilar gene subsets. The limitation of this evaluation is that the capability of examined methods for ‘class discovery’ has not been reviewed. Unlike the task of class prediction, the quality of data partition is determined by a structural properties of clusters, e.g. a compactness of samples in a cluster and a distance between clusters. To this extent, an initial study regarding of LCE for the task of class discovery is provided in Supplementary Section 8.2. In addition, the analysis of gene domain is another prominent research, in which LCE may prove to be useful. In particular, a better quality assessment should make use of a validity index that takes into account known gene functions, instead of simple external or internal indices mentioned earlier. 2.3 Experiment design Here, the performance of a given clustering algorithm is justiﬁed in terms of its ability to produce biologically meaningful clusters using a reference set of functional classes, which can be obtained from prior biological knowledge speciﬁc to a microarray study or may be formed using the growing databases of Gene Ontologies. The number of times that one method i∈CM is signiﬁcantly better than its competitors, BC(i) (in accordance with the validity criterion C, across all experiment settings), can be deﬁned as BC(i)= ∀β∈ST ∀i∗∈CM,i∗̸=i betterβ C(i,i∗) (8) betterβ C(i,i∗)= 1 if LXC(i,β) >UXC(i∗,β) 0 otherwise (9) BC(i)= ∀β∈ST ∀i∗∈CM,i∗̸=i betterβ C(i,i∗) (8) (8) betterβ C(i,i∗)= 1 if LXC(i,β) >UXC(i∗,β) 0 otherwise (9) (9) • The quality of each cluster ensemble method with respect to a speciﬁc ensemble setting is generalized as the average of 50 runs. Similarly, the number of times that one method i∈CM is signiﬁcantly worse than its competitors, WC(i), in accordance with the validity criterion C, can be computed as 3 RESULTS WC(i)= ∀β∈ST ∀i∗∈CM,i∗̸=i worseβ C(i,i∗) (10) worseβ C(i,i∗)= 1 ifUXC(i,β) <LXC(i∗,β) 0 otherwise (11) (10) The results1 with real gene expression data are summarized in Figure 5, where each investigated clustering method is represented with its average validity measure across all validity indices, datasets and ensemble types. It is clear that LCE regularly performs better than any of these clustering methods. It also enhances the performance of KM, which is used as base clusterings. In particular, HBGF is apparently less effective than LCE. This information suggests that the quality of the reﬁned cluster-association matrix (RM) is superior than the original BM counterpart. See the full results with real gene expression datasets in Supplementary Section 2.3. worseβ C(i,i∗)= 1 ifUXC(i,β) <LXC(i∗,β) 0 otherwise (11) (11) Using the aforementioned assessment formalism, Figure 6 illustrates for each method i∈CM the statistics of total performance (B−W)i = ∀C∈{CA,NMI,AR}BC(i)−WC(i). The results shown in this ﬁgure indicate that LCE is more effective than other clustering techniques included in this experiment. Since SL and AL do not perform well over the examined datasets, MULTI-K and CCHC that use the former and latter as a consensus function, respectively, are less accurate than other cluster ensemble methods and KM. However, their performance may improve with an ensemble that is much larger than the one investigated herein (i.e. M >>10). In addition to this evaluation scheme, a further performance analysis with a paired t-test is discussed in Supplementary Section 2.4. Following the study of Kuncheva et al. (2006), to rigourously evaluate the quality of investigated clustering techniques, the number of times that one method is signiﬁcantly better and worse (to 95% conﬁdence level) than the others are assessed across all experiment settings. Let XC(i,β) be the average value of validity index C ∈{CA, NMI, AR} across n runs (n=50 in this evaluation) for a cluster ensemble method i∈CM (CM is a set of 12 experimented clustering methods), on a speciﬁc experiment Another important investigation is on the subject of relations between performance of experimented cluster ensemble methods and different types of ensemble being explored in the present evaluation. Figure 7 shows the average validity measures of different cluster ensemble methods across all validity indices and real gene expression datasets. For each method, its performance with 1This section only contains a summary of our empirical evaluation over real gene expression data. 3.1 Parameter analysis The parameter that has any effect on the results of LCE is DC [see Equation (7)]. With the ensemble size of 10, we varied this value from 0.1 through 0.9, in steps of 0.1, for three validity measures, and obtained the results in Figure 8. This ﬁgure clearly shows that the results are robust, and do not depend strongly on any particulary value of DC. This makes it easy for users to obtain high-quality, reliable results when using LCE, particularly since values of DC near 0.7 generally produce the best results. Although there is variation in response across the DC values, the performance of LCE is always better than any of the other cluster ensemble methods included in this assessment. 3 RESULTS For more detailed results and experiments with other data collections, please see Supplementary Sections 2-4. 1517 N.Iam-on et al. Fig. 9. Performance of different cluster ensemble methods in accordance with ensemble size (M ∈{10,20,...,100}), as the averages of validity measures (CA, NMI and AR) across all real gene expression datasets and ensemble types. Fig. 7. Average validity measures of different cluster ensemble methods across all validity indices and real gene expression datasets, categorized in accordance with four types of ensemble. Fig. 7. Average validity measures of different cluster ensemble methods across all validity indices and real gene expression datasets, categorized in accordance with four types of ensemble. Fig. 9. Performance of different cluster ensemble methods in accordance with ensemble size (M ∈{10,20,...,100}), as the averages of validity measures (CA, NMI and AR) across all real gene expression datasets and ensemble types. Fig. 8. The relations between DC ∈{0.1,0.2,...,0.9} and the performance of the LCE method (the averages of CA, NMI and AR over 10 real gene expression datasets and 4 ensemble types), whose values are presented in X-axis and Y-axis, respectively. Note that the averaged performance of other cluster ensemble methods are also included for a comparison purpose. According to Figure 9 in which DC=0.9, this heuristic is applicable to LCE, where its validity measures (averages of CA, NMI and AR across all experimental settings) gradually incline to the increasing value of M ∈{10,20,...,100}. Furthermore, LCE performs better than its competitors with all different ensemble sizes. Note that a bigger ensemble leads to an improved accuracy, but with the tradeoff of run time—but, again, even the worst results for LCE are better than the best results of the other methods. These ﬁndings regarding the relation between LCE and its parameters have also been observed when both DC and M are simultaneously analysed (see details in Supplementary Section 7). Fig. 8. The relations between DC ∈{0.1,0.2,...,0.9} and the performance of the LCE method (the averages of CA, NMI and AR over 10 real gene expression datasets and 4 ensemble types), whose values are presented in X-axis and Y-axis, respectively. Note that the averaged performance of other cluster ensemble methods are also included for a comparison purpose. 3.3 Additional utilization of RM with simple clusterings Besides its current utilization through the formation of a weighted bipartite graph, the RM can also be regarded as a ‘high-level’ data matrix to which any simple clustering algorithm can be directly applied. Promising results have been obtained from the exploitation of six simple clustering techniques with RM: RM + SL, RM + CL, RM +AL, RM + KM, RM + Partitioning Around Medoids and RM + spectral clustering, respectively (see detailed results in Supplementary Section 5). 3.2 Complexity analysis The space and time complexity of creating a reﬁned cluster- association matrix (RM) are O(P2+NP) and O(P2l+NP), where N is the number of samples, P denotes the number of all clusters in an ensemble and l represents the average number of neighbours connecting to one cluster in a link network of clusters. For each entry (corresponding to clusters Cx,Cy ∈) in the P×P matrix of cluster similarity, WCT searches through l neighbors of Cx (or Cy) to identify connected triples. Following this, the RM of size N ×P is created using the aforementioned similarity matrix. As a result, LCE is computationally efﬁcient with the time complexity generally converging to O(N). Please consult Supplementary Section 6 for details of the scalability test. four ensemble types (Full-space + Fixed-k, Full-space + Random-k, Subspace + Fixed-k, and Subspace + Random-k) are compared. It is clear that LCE is more effective than other cluster ensemble techniques over all ensemble types, with its best performance being generated from a ‘Full-space + Fixed-k’ ensemble. Most methods work better with Full-space ensembles, as compared to Subspace alternatives. Unlike LCE, GCC and other graph-based techniques that usually produce a superior performance with a Fixed-k ensemble type, MULTI-K and CCHC are best when coupled with Random-k ensembles. four ensemble types (Full-space + Fixed-k, Full-space + Random-k, Subspace + Fixed-k, and Subspace + Random-k) are compared. It is clear that LCE is more effective than other cluster ensemble techniques over all ensemble types, with its best performance being generated from a ‘Full-space + Fixed-k’ ensemble. Most methods work better with Full-space ensembles, as compared to Subspace alternatives. Unlike LCE, GCC and other graph-based techniques that usually produce a superior performance with a Fixed-k ensemble type, MULTI-K and CCHC are best when coupled with Random-k ensembles. nformatics/article/26/12/1513/286962 by Aberystwyth University user on 09 March 2023 Although the results are impressive, on several datasets, it is important to ensure they are obtainable in a wide range of conditions. To this end the LCE algorithm’s response was examined to perturbations in its parameters, and by investigating its time and space complexity. REFERENCES Liu,L. et al. (2003) Robust singular value decomposition analysis of microarray data. Proc. Natl Acad. Sci. USA, 100, 13167–13172. Armstrong,S. et al. (2002) MLL translocations specify a distinct gene expression proﬁle that distinguishes a unique leukemia. Nat. Genet., 30, 41–47. Luxburg,U. (2007) A tutorial on spectral clustering. Stat. Comput., 17, 395–416. Luxburg,U. (2007) A tutorial on spectral clustering. Avogadri,R. and Valentini,G. (2009) Fuzzy ensemble clustering based on random projections for DNA microarray data analysis. Artif. Intell. Med., 45, 173–183. McLachlan,G. et al. (2002) A mixture model-based approach to the clustering of microarray expression data. Bioinformatics, 18, 413–422. Monti,S. et al. (2003) Consensus clustering: a resampling-based method for class discovery and visualization of gene expression microarray data. Mach. Learn., 52, 91–118. Bredel,M. et al. (2005) Functional network analysis reveals extended gliomagenesis pathway maps and three novel MYC-interacting genes in human gliomas. Cancer Res., 65, 8679–8689. Ng,A. et al. (2002) On spectral clustering: analysis and an algorithm. NIPS, 14, 849–856. Brunet,J. et al. (2004) Metagenes and molecular pattern discovery using matrix factorization. Proc. Natl Acad. Sci. USA, 101, 4164–4169. Nutt,C. et al. (2003) Gene expressionbased classiﬁcation of malignant gliomas correlates better with survival than histological classiﬁcation. Cancer Res., 63, 1602–1607. Chen,X. et al. (2002) Gene expression patterns in human liver cancers. Mol. Biol. Cell., 13, 1929–1939. Pomeroy,S. et al. (2002) Prediction of central nervous system embryonal tumour outcome based on gene expression. Nature, 415, 436–442. Chowdary,D. et al. (2006) Prognostic gene expression signatures can be measured in tissues collected in RNAlater preservative. J. Mol. Diagn., 8, 31–39. Ramaswamy,S. et al. (2001) Multiclass cancer diagnosis using tumor gene expression signatures. Proc. Natl Acad. Sci. USA, 98, 15149–15154. Ramaswamy,S. et al. (2001) Multiclass cancer diagnosis using de Souto,M. et al. (2008) Clustering cancer gene expression data: a comparative study. BMC Bioinformatics, 9, 497. Reuther,P. and Walter,B. (2006) Survey on test collections and techniques for personal hi d S O l 1 89 99 Reuther,P. and Walter,B. (2006) Survey on test collections and techniques for personal name matching. Int. J. Metadata Semantics Ontologies, 1, 89–99. Duda,R.O. et al. (2000). Pattern Classiﬁcation. 2nd edn. Wiley-Interscience, New York. name matching. Int. J. Metadata Semantics Ontologies, 1, 89–99 Dudoit,S. and Fridyand,J. (2003) Bagging to improve the accuracy of a clustering procedure. Bioinformatics, 19, 1090–1099. Shi,J. and Malik,J. (2000) Normalized cuts and image segmentation. IEEE Trans. Pattern Anal. Mach. Intell., 22, 888–905. Fern,X.Z. and Brodley,C.E. 4 CONCLUSION A new LCE method has been introduced for clustering gene expression data samples that has greatly improved accuracy and efﬁciency. The performance of LCE is usually superior than existing graph-based ensemble techniques, and those that are particularly developed for gene data analysis. LCE is highly effective over real gene expression datasets and synthetic data collections (with Another important parameter that may determine the quality of a cluster ensemble technique, is the ensemble size (M). Intuitively, the larger an ensemble is, the better the performance becomes. 1518 LCE method for improved gene expression data analysis the presence of noise and non-equal-size clusters). Unlike existing pairwise similarity based counterparts, LCE is efﬁcient for clustering large-size datasets, including the clustering of genes. Speciﬁcally, the reﬁned cluster-association matrix (RM) used by LCE is able to recover and account for unknown entries in the original BM counterpart, and hence, delivers a superior clustering performance. With its consistent performance over settings of parameter, ensemble type and size, LCE also proves to be a user-friendly data analysis tool, especially for non-expert users. Iam-on,N. et al. (2008) Reﬁning pairwise similarity matrix for cluster ensemble problem with cluster relations. In Proceedings of Eleventh International Conference on Discovery Science, Springer, Budapest, Hungary, pp. 222–233. Karypis,G. and Kumar,V. (1998) Multilevel k-way partitioning scheme for irregular graphs. J. Parallel Distrib. Comput., 48, 96–129. Karypis,G. et al. (1999) Multilevel hypergraph partitioning: applications in VLSI domain. IEEE Trans. VLSI Syst., 7, 69–79. Khan,J. et al. (2001) Classiﬁcation and diagnostic prediction of cancers using gene expression proﬁling and artiﬁcial neural networks. Nat. Med., 7, 673–679. Kim,E. et al. (2009) MULTI-K: accurate classiﬁcation of microarray subtypes using ensemble k-means clustering. BMC Bioinformatics, 10, 260. Kim,E. et al. (2009) MULTI-K: accurate classiﬁcation of microarray Kuncheva,L.I. and Hadjitodorov,S.T. (2004) Using diversity in cluster ensembles. In Proceedings of the IEEE International Conference on Systems, Man & Cybernetics, IEEE, The Hague, The Netherlands, pp. 1214–1219. Funding: Scholarship of the Ministry of Science and Technology, Royal Thai Government (to N.I.-O.). Kuncheva,L.I. and Vetrov,D. (2006) Evaluation of stability of k-means cluster ensembles with respect to random initialization. IEEE Trans. Pattern Anal. Mach. Intell., 28, 1798–1808. Conﬂict of Interest: none declared. Conﬂict of Interest: none declared. Conﬂict of Interest: none declared. Kuncheva,L.I. et al. (2006) Experimental comparison of cluster ensemble methods. In Proceedings of International Conference on Fusion, International Society of Information Fusion, Florence, Italy, pp. 105–115. REFERENCES (2004) Solving cluster ensemble problems by bipartite graph partitioning. In Proceedings of International Conference on Machine Learning, ACM, Banff, Alberta, Canada, pp. 36–43. Sorlie,T. et al. (2003) Repeated observation of breast tumor subtypes in independent gene expression data sets. Proc. Natl Acad. Sci. USA, 100, 8418–8423. et al. (2003) Repeated observation of breast tumor subtypes in ind expression data sets. Proc. Natl Acad. Sci. USA, 100, 8418–8423 Strehl,A. and Ghosh,J. (2002) Cluster ensembles: a knowledge reuse framework for combining multiple partitions. J. Mach. Learn. Res., 3, 583–617. Fred,A.L.N. and Jain,A.K. (2005) Combining multiple clusterings using evidence accumulation. IEEE Trans. Pattern Anal. Mach. Intell., 27, 835–850. combining multiple partitions. J. Mach. Learn. Res., 3, 583–617. Su,A. et al. (2001) Molecular classiﬁcation of human carcinomas by use of gene expression signatures. Cancer Res., 61, 7388–7393. Golub,T. et al. (1999) Molecular classiﬁcation of cancer: class discovery and class prediction by gene expression monitoring. Science, 286, 531–537. Swift,S. et al. (2004) Consensus clustering and functional interpretation of gene- expression data. Genome Biol., 5, R94. Hadjitodorov,S.T. et al. (2006) Moderate diversity for better cluster ensembles. Inform. Fusion, 7, 264–275. Yu,Z. et al. (2007) Graph-based consensus clustering for class discovery from gene expression data. Bioinformatics, 23, 2888–2896. Handl,J. et al. (2005) Computational cluster validation in post-genomic data analysis. Bioinformatics, 21, 3201–3212. 1519
https://openalex.org/W2528280271	https://www.matec-conferences.org/articles/matecconf/pdf/2016/40/matecconf_icmmr2016_05003.pdf	English	null	Research of long-term properties of high-strength fiberglasses by means of the built-in FBG sensors	MATEC web of conferences	2,016	cc-by	3,346	2. Experimental Influence of atmospheric moisture is the most important of climatic factors by which the polymeric composite materials (PCM) at operation of products are affected in nature. As a result of diffusive penetration of water into a polymeric matrix and into the interface between the fiber and matrix in fibrous PCM the part of molecules of water is immobilized on polar groups of polymer, weakening thereby interchain interaction in a matrix and adhesive interaction on limit of the section polymer matrix - fibers. At the same time molecular mobility (effect of plasticization) is released and the relaxation processes leading to restructurings on molecular proceed and supramolecular level.[1,2] For production of a sample biaxial fiber glass fabric is used (specification TU 1916-033-61664530-2013, produced by JSC Prepreg-SKM with area density of 860 g/m2 and the direction of fibers 0 °/90 °. Multiaxial fabric has been chosen with the purpose to imitate the pipes made by winding. The quasiisotropic structure is chosen as the most applicable. Matrix epoxy - ED-I (specification TU B3-774-2010) consist of composition from ED-20 epoxy (100 m of h) of IMTGFA hardener (isomethyltetrahydrophthalic anhydride, 20 m of h) and the accelerator unitary enterprise 606/2 (2.4,6, Tris (dimethylaminomethyl) of phenol. This type of a matrix is used in the industry and differs in high resistance to moisture. Samples of binding have been made by method of filling of images in a form. And then their mechanical properties have been defined. Before filling binding was maintained under vacuum not to allow hit of vials of air in a sample. Highly sensitive fiber Bregg grating sensors (FBG- sensors) of deformations are successfully used to an assessment of external and internal mechanical and thermal tension in various materials and elements of composite detail [3-7]. Unlike metal FBG strain gages – sensors do not rust and therefore they can be applied in the damp hot environment. Besides their accuracy is higher than at wire strain gages. FGRP samples on the basis of biaxial fabric and epoxy resin ED-I produced by an infusion method. Specimen was cured at 120 °C. The purpose of work consisted in clarification of opportunities to use fiber-optical sensors of Bragg (FBG- sensors) for studying micromechanical moist arising in PCM at the accelerated heat-moist influence which is widely applied to modeling of climatic aging of PCM. Research of long-term properties of high-strength fiberglasses by means of the built-in FBG sensors Peter V.Mikheev , Andrey A. Dalinkevich 2 , Sergey A. Gusev 1, Timofey N. Igonin 2, Ludmila B. Maksaeva Nenasheva 2 1 ev , Andrey A. Dalinkevich 2 , Sergey A. Gusev 1, Timofey N. Igonin 2, Ludmila B. Maksaeva 2 and Tatyan 1 1 Bauman Moscow State Technical University: 2, bd. 1, 2nd Baumanskaya Street, 105005 Moscow, Russia, 2 Russian academy of sciences A.N. Frumkin Institute of Physical chemistry and Electrochemistry RAS (IPCE RAS) 119991, Leninsky blvd. 31. Moscow, Russia, Abstract. In this work the possibility of use of the built-in fiber-optical sensors (Bragg's sensors) of deformation for control of a condition of polymeric fiberglasses on epoxy polymer matrix ED-I cured at the increased temperature was investigated. Sensors were inserted in different layers of the layered plastic reinforced by biaxial fiber glass fabric. The package of layers had quasiisotropic structure. The sample made by method of vacuum impregnation was affected by saturated steam (90 °C) within 6 months. Operability of system of measurement has been as a result shown and swelling parameters in different layers of FGRP were recorded. For measurement measuring systems of two types — on a superbright light-emitting diode and on the tunable laser were used. © The Authors, published by EDP Sciences. This is an open access article distributed under the terms of the Creative Commons Attribution License 4.0 (http://creativecommons.org/licenses/by/4.0/). 0 (2016) MATEC Web of Conferences 7 ICMMR 2016 ,7 5003 0 (2016) MATEC Web of Conferences 7 ICMMR 2016 ,7 5003 DOI: 10.1051/matecconf/20167705003 2. Experimental Completeness of hardening polymer matrix was controlled by method of the differential scanning calorimetry, glass temperature of the sample cured binding (without reinforcing fibres ) has made 132 °C. (Figure 1) . The problem of influence of moisture on fibreglass is important, so this material is perspective for application in networks of water supply and oil pipelines. Fiberglass pipes of less steel are subject to corrosion. FBG sensors have been located between layers at a different depth of a sample. When laying layers, the 0 (2016) MATEC Web of Conferences 7 ICMMR 2016 ,7 5003 DOI: 10.1051/matecconf/20167705003 Figure 2. The scheme of an arrangement of sensors in sample layers second layer has been turned by 45 degrees to create structure close to isotropic structure. Previously the method for output of optical sensors out of plastic has been worked out. For this purpose samples of fiberglass plastic with optical fibre but without sensors were prepared. So we picked up a way to protect the optical fiber from being damaged at the edges of the plastic and at the junction of the measuring fiber with patch cords. Initially the stresses arising during cooling of the finished fiberglass plate either destroyed the sensor or bent it. This distortion leads to the loss of the optical signal below the level of sensitivity threshold of the system based on high-brightness LEDs (the method of Bauman Moscow State Technical University, BMSTU). Figure 2. The scheme of an arrangement of sensors in sample layers The sensors located in different layers have been carried on width not to create distortion of structure. For production of samples the following devices were used: Installation of vacuum impregnation Vacmobile INFUSION 28/2 (New Zealand); After picking up a set of heat shrinking tubing the problem was solved. At the same time on the plates with an optical fiber without sensors was selected the temperature for preserving patch cords connectors from damage or from resin casting. France Etuves XXL04.5 furnace (France). France Etuves XXL04.5 furnace (France). Figure 3. Process of laying of sensors between fiber glass fabric layers Figure 1. Chart of hardening of epoxy matrix Figure 1. Chart of hardening of epoxy matrix Figure 3. Process of laying of sensors between fiber glass fabric layers Figure 1. 2. Experimental Chart of hardening of epoxy matrix For determination of temperature of vitrification binding the differential scanning calorimeter Netzch DSC 204 F1 Phoenix (Germany) was used. Sensors have been integrated into a sample only when was obtained the attenuation of a light signal accepted for measurements. For determination of mechanical properties of a sample of the witness the universal test equipment Zwick- Roel Z-100 (Germany) was used. Mechanical properties of epoxy binding have been checked on standard samples "a dog bone". Results are given in Tab. 1. Binding has shown fragile linear behavior up to destruction. The endurance of samples was carried out in the certified thermo-moist camera Quantel the MHK 408 CL model (Singapore) Figure 4. Table 1 Properties of polymeric binding Table 1 Properties of polymeric binding Table 1 Properties of polymeric binding № Property Value Variation factor, % 1 Elasticity module, GPA 2.84 1.75 2 Strength, MPa 38.3 14.5 3 ultimate strain,% 1.4 15.7 4 Dencity, gm/cm 3 1.22 5.0 Parameters of FBG – sensors optical (produced by "Inversion Sensor") were the following: top layer (marking No. 1341) Wavelength second layer (marking No. 13801542,28 second layer (marking No. 1380) Wavelength 1550,23 center (marking No. 1340) Wavelength 1554,06 The arrangement of sensors is given in layers on rice.2. On Fig. 3 process of laying of optical fibers with sensors between layers of industrial fiber-glass fabric before impregnation of a sample binding is shown. Table 1 Properties of polymeric binding № Property Value Variation factor, % 1 Elasticity module, GPA 2.84 1.75 2 Strength, MPa 38.3 14.5 3 ultimate strain,% 1.4 15.7 4 Dencity, gm/cm 3 1.22 5.0 Figure 4. Thermo-moist camera Experimental measurements were made of the Parameters of FBG – sensors optical (produced by "Inversion Sensor") were the following: Parameters of FBG – sensors optical (produced by "Inversion Sensor") were the following: ( g , second layer (marking No. 1380) Wavelength 1550,23 center (marking No. 1340) Wavelength 1554,06 second layer (marking No. 1380) Wavelength 1550,23 center (marking No. 1340) Wavelength 1554,06 The arrangement of sensors is given in layers on rice.2. On Fig. 3 process of laying of optical fibers with sensors between layers of industrial fiber-glass fabric before impregnation of a sample binding is shown. Figure 4. 2. Experimental Thermo-moist camera Experimental measurements were made of the measuring system developed in NOС s "Photonics and Experimental measurements were made of the measuring system developed in NOС s "Photonics and 2 0 (2016) MATEC Web of Conferences 7 ICMMR 2016 ,7 5003 DOI: 10.1051/matecconf/20167705003 0 (2016) MATEC Web of Conferences 7 ICMMR 2016 ,7 5003 IK-equipment" of BMSTU and the Astro A312 device of production "Inversion Sensor". These measurements received by means of both systems corrections for their use in one measuring system are compared among themselves. weighing) and some physicomechanical characteristics (Tab. 2) have been cut out from FGRP plates. Measurements of mechanical properties of a doubling sample have shown that it possesses good mechanical properties. The doubling sample has been cutted near the sensor location. The photo on Fig.6 shows the end face of sensor. It is visible that there are no intervals between the sensor and matrix, and the structure of a composite is not broken. The measurement system constructed in BMSTU was based on source of the white light, technically more simply, but the Astro system based on the tunable laser has big sensitivity and bigger speed. One of optical spectra of the sensor is given in Figure 5 . Table 2. Properties of the made fibreglass № Property Value 1 Elasticity module at a bend, GPA 21,7 2 Elasticity module at stretching, GPA 20,6 3 Strength at bending , MPa 397 4 Strength at shift, MPa 38,3 5 Density of plastic, g/cm3 1,9 6 Volume fraction of the reinforcing fibers, % 55 7 Density of of the used polymer matrix, g/cm3 1,22 Table 2. Properties of the made fibreglass № Property Value 1 Elasticity module at a bend, GPA 21,7 2 Elasticity module at stretching, GPA 20,6 3 Strength at bending , MPa 397 4 Strength at shift, MPa 38,3 5 Density of plastic, g/cm3 1,9 6 Volume fraction of the reinforcing fibers, % 55 7 Density of of the used polymer matrix, g/cm3 1,22 Figure 5. The example of an optical signal from the FBG- sensor received on the equipment Astro. Figure 5. The example of an optical signal from the FBG- sensor received on the equipment Astro. The FGRP sample containing sensors has been subjected to heat-moist aging in a test climatic chamber at the temperature of 80 °C and relative humidity of 95%. 2. Experimental Accuracy of maintenance of parameters of tests made ±1°С for temperature and ±1,5% for humidity. The scheme of sensors arrangement is provided in on Fig 2 and in Table 1. To reveal spatial unevenness of penetration of water into a sample, sensors were put on different depth (layers 1, 2 and 3). The increased temperature of hardening of polymeric binding has demanded special protection of optical fibers in places of an exit from FGRP. p y Moisture content in the studied sample of FGRP was controlled by method of periodic weighing with an accuracy of 10-3 g. y g Figure 6. Sensor cut in a composite sample Table 1. Arrangement of sensors in layers of FGRP Material SМ-42019 (+0/90°) Layer № laying 1 SМ-42019 (+0/90°) FBG-sensors 1 Measuring layer 2 SМ-42019 (+45/-45°) FBG-sensors 2 Measuring layer 3 SМ-42019 (+0/90°) FBG-sensors 3 Measuring layer 4 SМ-42019 (+0/90°) 5 SМ-42019 (+45/-45°) 6 SМ-42019 (+0/90°) Table 1. Arrangement of sensors in layers of FGRP Material SМ-42019 (+0/90°) Layer № laying 1 SМ-42019 (+0/90°) FBG-sensors 1 Measuring layer 2 SМ-42019 (+45/-45°) FBG-sensors 2 Measuring layer 3 SМ-42019 (+0/90°) FBG-sensors 3 Measuring layer 4 SМ-42019 (+0/90°) 5 SМ-42019 (+45/-45°) 6 SМ-42019 (+0/90°) Figure 6. Sensor cut in a composite sample 3. Results and discussion From Fig. 7 it is visible that sorption of vapors of water FGRP has abnormal character and does not submit to Fick's law. Existence of a maximum and step on kinetic curve sorption demonstrates course in a model of relaxation processes, i.e. change of internal microstructural tension and according to deformations which, in turn, are caused by diffusive and sorption penetration and interaction of water with polymer [8-11]. High sensitivity to temperature and deformation of a FBG-sensors also allow to use it as a sensitive element of the measuring system. Measuring properties of a FBG- sensors depend on effective index of refraction and on the From Fig. 7 it is visible that sorption of vapors of water FGRP has abnormal character and does not submit to Fick's law. Existence of a maximum and step on kinetic curve sorption demonstrates course in a model of relaxation processes, i.e. change of internal microstructural tension and according to deformations which, in turn, are caused by diffusive and sorption penetration and interaction of water with polymer [8-11]. Dynamics of change of weight of a sample due to absorption of moisture is shown in Figure 7. In the same Figure the schedule of measurement of deformations is shown. Special attention has been paid to protection of connectors against the increased temperature and humidity. Influence of hot water was so strong that all factory marking of sensors has been washed away. High sensitivity to temperature and deformation of a FBG-sensors also allow to use it as a sensitive element of the measuring system. Measuring properties of a FBG- sensors depend on effective index of refraction and on the The doubling sample on the same epoxy matrix ED-I and fabrics, but without sensors has been in parallel made. Samples for determination of density (hydrostatic 3 DOI: 10.1051/matecconf/20167705003 0 (2016) MATEC Web of Conferences 7 ,7 5003 0 (2016) MATEC Web of Conferences 7 ICMMR 2016 ,7 5003 lattice period on a formula (1). And the period of a lattice is directly connected with its deformation: period (6 months), process becomes stationary and does not lead to destruction of a composite. Figure 7. Water sorption by fiberglass. Arrows showed the moments of measurement of deformations Figure 8. Deformations of layers of fibreglass at endurance in the damp atmosphere. Figure 7. Water sorption by fiberglass. Arrows showed the moments of measurement of deformations Figure 8. 3. Results and discussion Deformations of layers of fibreglass at endurance in the damp atmosphere. ∆ = 2 ∙( + ) ∙∆ + 2 ∙( + ) ∙ ∆ (1) 4. Conclusions The first part of equation gives value of shift of wavelength depending on deformation (lengthening), the second depending on temperature. Dependence of the central reflected wavelength on deformation can be also presented in the following form [5]: The possibility of fixing of thermomoist deformations in fiberglass of quasiisotropic structure is shown for a long time. Two types of measuring equipment are used: the BMSTU experimental system on the basis superbright light-emitting diode and the Astro A312 system on the basis of the tunable laser. g [ ] ∆ = ∙(1 −) ∙ (2) (2) Deformations of swelling in different layers of fiberglass are recorded. In general epoxy resin ED-I has shown high resistance at influence of temperature and moisture. where pe — the constant of deformation of optical fiber defined from the following expression where pe — the constant of deformation of optical fiber defined from the following expression = ∙[ − ∙( −)] (3) (3) The equipment of the built-in sensors can be used for a prediction of resource of the polymeric detail working in the conditions of long influence of water. Bragg's sensors are not subject to corrosion and therefore are suitable for such task. In our research ED-I epoxy has shown high resistance to hot water where p11 и p12 — Pokkels's coefficients in a tensor of optical tension, — Poisson's coefficient. For standard fiber p11=0,113, p12=0,252, =0,16 and nef =1,4682 [6] where p11 и p12 — Pokkels's coefficients in a tensor of optical tension, — Poisson's coefficient. For standard fiber p11=0,113, p12=0,252, =0,16 and nef =1,4682 [6] Proceeding from these values sensitivity for wavelength B =1550 nanometers make 12 nanometers/%. Certain results of this work were obtained as a part of research works upon Agreement with the Ministry of Science and Education of the Russian Federation concerning the provision of grant №14.577.21.0103 dated 16.09. 2014. The research project ID RFMEFI57714X0103 Deformation of sensors after endurance in water calculated on the basis of (3) is given in Table 3. Measurements were taken once a month and until optical measurement the sample was stored in the damp atmosphere. Thus we interfered with probable deformations when drying. References 1. A. Cusano, A. Cutolo, J. Albert. “Fiber Bragg Grating Sensors: Recent Advancements, Industrial Applications and Market Exploitation”, Editors, Bentham Science Publishers, 2010, PP. 35-52. Table 3. Deformation increment from the moment of GFRP producing Deformation in time, %, % 15.07.15 10.09.15 09.10.15 23.12.15 Upper layer 0,007 -0,017 -0,014 -0,074 Second layer 0,013 0,003 0,007 -0,001 Middle layer 0,015 0,003 0,009 -0,006 Table 3. Deformation increment from the moment of GFRP producing 2. A.A. Dalinkevich, K.Z. Gumargalieva, S.S. Marakhovsky, A.V. Soukhanov, Modern Basalt Fibrous Materials and Basalt Fiber-Based Polymeric// JOURNAL OF NATURAL FIBERS, 2009,VL 6, №3, P 248-271, 3. Yu.N Kulchin. The distributed fiber-optical sensors and measuring networks Vladivostok// Dalnauka, 1999, PP. 115-118.(In Russian) In Figure 8 changes of deformation of layers in the form of the histogram are presented. 4. M. Ramakrishnan, G. Rajan, Y. Semenovaand, G. Farrell, Overview of Fiber Optic Sensor Technologies forStrain/Temperature Sensing Applications in In drawing it is well visible that external layers absorb water stronger than internal. But during such a long 4 0 (2016) MATEC Web of Conferences 7 ICMMR 2016 ,7 5003 0 (2016) MATEC Web of Conferences 7 ,7 5003 DOI: 10.1051/matecconf/20167705003 Composite Materials, Sensors// 2016, № 16, V. 99; doi:10.3390/s16010099 designs of a football arena on the basis of fiber-optical sensors"// Foton-Express, 2013, №6, V.110, PP. 22-23 (2013) (In Russian) 5. S. Hannuscha, M. Stockmanna, J. Ihlemanna, Experimental method for residual stress analysis with fibre Bragg grating sensors, Materials Today: Proceedings 3 ( 2016 ), PP. 979 – 982 10. 10. V.V. Shishkin, A.E. Churin, D.S. Kharenko, M.A. Zheleznova, I.S.Shelemba, “Structural health monitoring system of soccer arena based on optical sensors”, Proc. SPIE 9157, 23rd International Conference on Optical Fibre Sensors, 9157D3 (2 June 2014); doi: 10.1117/12.2071269, (2014) g ( ) 6. K.Yükseln, D. Pala, Analytical investigation of an ovelinterrogation approach of fiber Bragg grating sensors using Optical Frequency Domain Reflectometry// Optics and Lasers in Engineering 2016, № 81, PP.119–124 11. B.S. Sarbayev, A.A. Smerdov, L.P. Tairova, V.A. Seleznyov, S.V. Sokolov, G.Ya. Buymistryuk, V.I. Izotov, A.M. Rogov, "Research of the deformed condition of designs from composite materials by means of fiber optical sensors"// Bulletin of Bauman Moscow State Technical University. 2011. N.SP, PP. 39-51. (In Russian) 7. D. Abbaneo and other, Fiber Bragg Grating(FBG) sensors as flatness and mechanical stretching sensors//, Nuclear Instrumentsand Method sin Physics Research, 2016, V.824, PP. 493–495 8. P.V. Mikheev, A.V. Artemyev, V.A. Lazarev, A.B. References Pnev, V.A. Nelyub, "Experience of creation of system of the built-in control of a condition of a design on the basis of fiber-optical the breggovskikh of sensors", the Collection of reports of the All-Russian conference on tests and researches of properties of the materials TEST MAT - 2013. Publishing house: All-Russian Research Institute of Aviation Materials (Moscow.) 2013, PP. 28-29. ( In Russian) 12. E. Bar, Constructional properties of plastic (physical and chemical bases of application). The lane with English M., "Chemistry", 1967. PP. 231-235. (In Russian 13. Yu.A. Mikhaylin Fibrous polymeric composite materials in equipment. – SPb.: "Profession", 2013. – 752 pages. (In Russian)Modern physical methods of research of polymers. Under the editorship of Slonimsky G. L. M, "Chemistry", 1982. PP. 198-208. (In Russian) 9. V.V. Shishkin, A.E. Churin, D.S. Kharenko, I.S. Shelemba, "System of monitoring of the bearing 5
https://openalex.org/W4361872713	https://figshare.com/articles/journal_contribution/Supplementary_Figure_1_from_Canonical_Wnt_Signaling_Is_Required_for_Pancreatic_Carcinogenesis/22394967/1/files/39840591.pdf	Haitian	null	Supplementary Figure 8 from Canonical Wnt Signaling Is Required for Pancreatic Carcinogenesis	null	2,023	cc-by	249	Cre! Ptf1a! Kras G12D! STOP! LoxP! LoxP! β-catenin! LoxP! LoxP! LoxP! LoxP! Kras G12D! Cre! Cre! A! Cre! Ptf1a! Kras G12D! STOP! LoxP! LoxP! β-catenin! LoxP! LoxP! LoxP! LoxP! Kras G12D! Cre! Cre! A! B! β-catenin/Alcian Blue! ! ! β-catenin! KBC! B! C! B! β-catenin/Alcian Blue! Cre! Ptf1a! Kras G12D! STOP! LoxP! LoxP! β-catenin! LoxP! LoxP! LoxP! LoxP! Kras G12D! Cre! Cre! A! HE! ! ! β-catenin! KBC! CK19/β-catenin/DAPI! Ki67/β-catenin/DAPI! pERK1/2/β-catenin/DAPI! E KBC! KC! KBC! KC! WT! KBC! D A! β-catenin/Alcian Blue! C! β-catenin! HE! β-catenin! G! CK19/β-catenin/DAPI! KC! KBC! Ki67/β-catenin/DAPI! pERK1/2/β-catenin/DAPI! E β-catenin/Amylase/CK19! F ! ! Ela-CreERT2; β-cateninf/f! ! KBC! Number of cell clusters! Ela-CreERT2; β-cateninf/+! Ela-CreERT2; β-cateninf/f! ! ! ! Supplemental Figure 1. β-catenin inactivation prevents PanIn formation. ! KC! KBC! KC! WT! KBC! D C! β HE! β-catenin! G! CK19/β-catenin/DAPI! KC! KBC! Ki67/β-catenin/DAPI! pERK1/2/β-catenin/DAPI! E β-catenin/Amylase/CK19! F ! ! Ela-CreERT2; β-cateninf/f! ! KBC! Number of cell clusters! Ela-CreERT2; β-cateninf/+! Ela-CreERT2; β-cateninf/f! ! ! ! Supplemental Figure 1. β-catenin inactivation prevents PanIn formation. ! KC! KBC! KC! WT! KBC! D C! HE! KBC! KC! D D C! KBC! G! CK19/β-catenin/DAPI! KC! KBC! Ki67/β-catenin/DAPI! pERK1/2/β-catenin/DAPI! E β-catenin/Amylase/CK19! F ! ! Ela-CreERT2; β-cateninf/f! ! Number of cell clusters! Ela-CreERT2; β-cateninf/+! Ela-CreERT2; β-cateninf/f! ! ! ! Supplemental Figure 1. β-catenin inactivation prevents PanIn formation. ! G CK19/β-catenin/DAPI! KC! KBC! Ki67/β-catenin/DAPI! pERK1/2/β-catenin/DAPI! E F El C ERT2 β t i f/f E F Ela-CreERT2; β-cateninf/+! F G! Ela-CreERT2; β-cateninf/f! ! ! *! Supplemental Figure 1. β-catenin inactivation prevents PanIn formation. !
https://openalex.org/W2259238261	https://dash.harvard.edu/bitstream/1/27013404/1/srep20359.pdf	English	null	Longitudinal Prediction of the Infant Gut Microbiome with Dynamic Bayesian Networks	Scientific reports	2,016	cc-by	10,262	Terms of Use This article was downloaded from Harvard University’s DASH repository, and is made available under the terms and conditions applicable to Other Posted Material, as set forth at http:// nrs.harvard.edu/urn-3:HUL.InstRepos:dash.current.terms-of-use#LAA Permanent link http://nrs.harvard.edu/urn-3:HUL.InstRepos:27013404 Citation Citation McGeachie, Michael J., Joanne E. Sordillo, Travis Gibson, George M. Weinstock, Yang-Yu Liu, Diane R. Gold, Scott T. Weiss, and Augusto Litonjua. 2016. “Longitudinal Prediction of the Infant Gut Microbiome with Dynamic Bayesian Networks.” Scientific Reports 6 (February 8): 20359. doi:10.1038/srep20359. Published Version doi:10.1038/srep20359 Share Your Story The Harvard community has made this article openly available. Please share how this access benefits you. Submit a story . The Harvard community has made this article openly available. Please share how this access benefits you. Submit a story . Accessibility Longitudinal Prediction of the Infant Gut Microbiome with Dynamic Bayesian Networks Citation McGeachie, Michael J., Joanne E. Sordillo, Travis Gibson, George M. Weinstock, Yang-Yu Liu, Diane R. Gold, Scott T. Weiss, and Augusto Litonjua. 2016. “Longitudinal Prediction of the Infant Gut Microbiome with Dynamic Bayesian Networks.” Scientific Reports 6 (February 8): 20359. doi:10.1038/srep20359. Longitudinal Prediction of the Infant Gut Microbiome with Dynamic Bayesian Networks Citation McGeachie, Michael J., Joanne E. Sordillo, Travis Gibson, George M. Weinstock, Yang-Yu Liu, Diane R. Gold, Scott T. Weiss, and Augusto Litonjua. 2016. “Longitudinal Prediction of the Infant Gut Microbiome with Dynamic Bayesian Networks.” Scientific Reports 6 (February 8): 20359. doi:10.1038/srep20359. Accessibility www.nature.com/scientificreports www.nature.com/scientificreports Longitudinal Prediction of the Infant Gut Microbiome with Dynamic Bayesian Networks Michael J. McGeachie1,, Joanne E. Sordillo1,, Travis Gibson1, George M. Weinstock Yang-Yu Liu1, Diane R.Gold1, ScottT.Weiss1 &Augusto Litonjua1 received: 05 June 2015 accepted: 31 December 2015 Published: 08 February 2016 Sequencing of the 16S rRNA gene allows comprehensive assessment of bacterial community composition from human body sites. Previously published and publicly accessible data on 58 preterm infants in the Neonatal Intensive Care Unit who underwent frequent stool collection was used. We constructed Dynamic Bayesian Networks from the data and analyzed predictive performance and network characteristics. We constructed a DBN model of the infant gut microbial ecosystem, which explicitly captured specific relationships and general trends in the data: increasing amounts of Clostridia, residual amounts of Bacilli, and increasing amounts of Gammaproteobacteria that then give way to Clostridia. Prediction performance of DBNs with fewer edges were overall more accurate, although less so on harder-to-predict subjects (p = 0.045). DBNs provided quantitative likelihood estimates for rare abruptions events. Iterative prediction was less accurate (p < 0.001), but showed remarkable insensitivity to initial conditions and predicted convergence to a mix of Clostridia, Gammaproteobacteria, and Bacilli. DBNs were able to identify important relationships between microbiome taxa and predict future changes in microbiome composition from measured or synthetic initial conditions. DBNs also provided likelihood estimates for sudden, dramatic shifts in microbiome composition, which may be useful in guiding further analysis of those samples. Scientific Reports \| 6:20359 \| DOI: 10.1038/srep20359 Microbiome Th b (email: michael.mcgeachie@channing.harvard.edu) Scientific Reports \| 6:20359 \| DOI: 10.1038/srep20359 1 www.nature.com/scientificreports/ Although classic longitudinal analysis captures changes in outcomes over time, this standard approach has many limitations. Often, individual taxa are treated as separate outcomes, and information on the connections between bacteria (i.e. how one bacterial population may influence another over time) is lost. Network-based methods are an alternative approach to longitudinal gut microbiome modeling. In general, network studies on gut microbiota have been limited thus far; most are comprised of correlation analyses, which others have noted have poor asymptotic prediction14. Two types of network methodologies specifically designed to capture complex interactions and dynamic change within the microbiome over time include the generalized Lotka-Volterra model and Dynamic Bayesian Networks.i y y Generalized Lotka-Volterra (GLV) model and other dynamic systems identification formalisms15 use longi- tudinal microbiota compositional data to identify parameters in ordinary differential equations that describe the dynamics of microbial ecosystems16,17. For instance, GLV has been used in a study of the murine gut microbiome to generate a network of interactions between bacterial taxa18; this network included almost all possible edges and was not used for prediction. In another microbiome study, a continuous GLV model was assumed and coefficients related to the individual microbe growth rates, the strengths of the microbe-microbe interactions, and suscepti- bility to antibiotics were learned using linear regression with regularization17. Discrete GLV, where coefficients were learned using a sparse linear regression technique, has also been employed16. Alternatively, other dynamic systems models have been used, in one case modeling two of possibly many interacting microbes in the gut19. These endeavors build upon a rich history of systems-identification literature, spanning the theoretical and prac- tical20, and these approaches have shown that data-derived models of microbiota dynamics can have significant analytic and predictive power17. However, the degree to which the microbiome datasets available meet the rig- orous requirements of exact parameter estimation in these models remains an outstanding question16. Methods that include explicit parameter estimation and allowances for noisy data may be more appropriate. One such method used Bayesian statistics to help inform dynamic models of a single independent bacteria taxon’s change in response to antibiotics21. p Bayesian networks (BNs) are an appropriate tool to model the interaction of many microbial taxa in the gut microbiome, since they include implicit parameter estimation techniques for inferring complex networks from noisy data and are intended to predict clinical outcomes of relevance. Microbiome Th b The microbiota living in the human gut performs a number of vital functions for homeostasis, including the harvest of essential nutrients1,2, synthesis of vitamins3, metabolism of xenobiotics4, and the development and maintenance of the immune system5,6. Alterations in the gut microbiome have been observed in a number of disease states7, and may be directly connected to pathogenesis. Microbes that populate an infant’s gut after birth serve as critical immune stimuli in the first days of life8, and could influence the composition of the “mature” gut microbiome, with subsequent implications for the health of the human host in both early life and adulthood. Studies on the initial colonization of the infant gut are very limited, with sparse, if any, longitudinal data1. Early studies have proposed an initial predominance of facultative anaerobes, followed by a progression to anaer- obic bacteria9,10. The “first colonizers” of the infant gut may derive from maternal sources (vaginal flora, skin flora, and gut flora) or from environmental microbes. Microbiome studies of meconium11, amniotic fluid and placenta12, suggest that infants encounter microbes even before birth. Microbial rDNA present in the intrauterine environment suggests that prenatal sources may also contribute to gut colonization12.h gg p y g The most comprehensive study on the progression of the infant gut microbiota thus far examined 58 preterm infants in a neonatal intensive care unit, with repeated measurements taken every few days on all study subjects starting within the first days of life, and ending at approximately one month of age13. In this study, La Rosa et al. used longitudinal analysis of time series data to demonstrate that the microbiota of the infant gut was initially dominated by Bacilli at birth, giving way to Gammaproteobacteria, then Clostridia at the end of the first month of life. Gestational age appeared to have the greatest influence on the pace (but not the pattern) of bacterial micro- biome progression, and the non-random assembly observed seems to suggest that host biology or consistent exposure sources (i.e. infant diet, maternal flora) play a key role in infant gut population (as compared to chance encounters with microbes in the environment). 1Channing Division of Network Medicine, Brigham and Women’s Hospital, Boston, MA. 2The Jackson Laboratory for Genomic Medicine, Farmington, CT. *These authors contributed equally to this work. Correspondence and requests for materials should be addressed to M.J.M. Methods Mi bi Microbiome Data and Clinical Characteristics. We obtained public microbiome data collected by La Rosa et al.13 for their paper Patterned progression of bacterial populations in the premature infant gut from the PNAS website. This data is described in greater detail in La Rosa et al.13; briefly, it contains 922 infant gut micro- biome measurements from a total of 58 pre-term infants, taken one or two days apart on average. Infants enrolled in the study were ≤ 1500 g at birth and were expected to live beyond the first week of life. Infants that experienced serious intra-abdominal pathology (i.e. necrotizing enterocolitis) were excluded from the study. Demographic and clinical data collected on the preterm infants included race, gender, gestational age at birth, post conceptional age when the stool sample was obtained, being housed in an open or closed room, dietary information (breast milk as a percentage of total enteral volume administered to the infant), mode of delivery (C-section vs. vaginal), and antibiotic use (percent of days of life on antibiotic). To determine the bacterial community composition of the infant stool samples, the 16S region (V3-V5) of bacterial rDNA was sequenced using the Roche 454 plat- form. The median number of 16S rDNA sequence reads across all infant stool specimens collected was 7,183 (IQR 5,243–9,314). Clinical and demographic data, along with counts of 16S bacterial ribosomal DNA sequences grouped by bacterial class, is contained in the file “pnas.1409497111.sd01.xlsx”. From that file, we retained all bac- teria, although two of the bacterial taxa (Cyanobacteria and Holophagae) are likely environmental contaminants, in the spirit of using all available data for model construction. Using all of the pre-term infant gut microbiome data, we constructed dynamic Bayesian networks to capture the influence of individual microbial classes on each other over time. In short, longitudinal data (repeated measures of the microbial classes) are accounted for within the dynamic Bayes net, and the connections between nodes show how one taxon may influence another over time. Our confidence in each individual network connection is reflected by the Bayes Factor; connections with the largest Bayes factors are more likely to represent a true causal association. Details of dynamic Bayes network methodology are shown below. Dynamic Bayesian Networks. We used the CGBayesNets package27 to build two-stage dynamic Bayesian networks of the microbiome population dynamics from the entire data set. Microbiome Th b In this work we model the infant gut micro- biome using Dynamic Bayesian Networks (DBN), which are related to (static) Bayesian Networks (BN), but are ideal for modeling biological systems with feedback loops, which traditional BNs cannot easily represent22. DBNs have been successfully used to model time-series data in other clinical and ecological applications, including bird populations in wildlife reserves23, cancer outcomes24, gene regulatory networks from transcriptome data in blood25 and from perturbation data26. In these applications DBNs were found to be superior to other techniques including LASSO regression23 and proportional hazards24 for outcome prediction. To our knowledge, DBNs have not yet been applied to human microbiome data. y pp In this paper, we employed DBNs to model the progression of microbiota in colonizing the infant gut, using microbiota abundance data from the previously published La Rosa et al. study13. Using DBNs, we developed a predictive model for the infant’s gut microbiome based on prior composition. We modeled relationships between multiple bacterial taxa (both dominant and rare), the compositional changes bacterial taxa exert on other com- munity members over time, and the influence of potential external characteristics (mode of delivery, antibiotic use, etc) on gut microbiome progression. We also investigated the likelihood of abruption events (rapid, extreme changes in microbiome composition), and determined how these abruptions influence future taxonomic com- position in the infant’s gut. Methods Mi bi We use “two-stage” here to refer to Scientific Reports \| 6:20359 \| DOI: 10.1038/srep20359 2 www.nature.com/scientificreports/ Figure 1. Dynamic Bayesian Network of Infant Gut Microbiota Dynamics. The DBN network included the clinical variables and bacteria taxa pictured above. Nodes indicate variables, with greater size and redder color indicating greater node degree. Direction of edges indicate temporal statistical influence: the source node’s prior value predicts the target node’s present value. Edge thickness indicates strength of statistical dependence. Figure 1. Dynamic Bayesian Network of Infant Gut Microbiota Dynamics. The DBN network included the clinical variables and bacteria taxa pictured above. Nodes indicate variables, with greater size and redder color indicating greater node degree. Direction of edges indicate temporal statistical influence: the source node’s prior value predicts the target node’s present value. Edge thickness indicates strength of statistical dependence. the use of an abstraction on time points: that the network model only considers “current time” samples and the immediately previous time samples. The CGBayesNets package performs both network learning and infer- ence with discrete and continuous (normal) nodes in a formalism known as a Conditional Gaussian Bayesian network28. We adapted this to construct two-stage DBNs. For each sample, bacterial ribosomal DNA sequence counts were converted to relative abundances (proportions). To satisfy requirements of a two-stage dynamic Bayesian network, each microbial sample was paired with the immediately preceding microbial sample from the same subject. This resulted in 864 paired microbial samples, used as the primary data for the DBN modeling.h jh p p p y g The two-stage DBN was constructed by using standard Bayesian network techniques, but limiting edges to only those going from prior time points to subsequent time points. In preliminary analysis, the variables “mode of birth” referring to cesarean or vaginal delivery, “milk” indicating the amount of breastfeeding the infant received, and “gender”, resulted in overfitting. Including these variables in the modeling step led to networks with these three variables linked to all of the bacteria (although not Gammaproteobacteria), which is a sign of overfitting. We assessed the degree of overfitting by five-fold cross validation, building the network on 4/5ths of the subjects and testing it on the remaining 1/5th, in the same manner as described below; where networks using these variables had on average greater per-person error (0.0325 vs. Methods Mi bi 0.0296, p = 0.009, two-tailed t-test), indicating a reduced abil- ity to make generalizations from the majority of subjects that would then predict bacteria concentrations of the held out minority. We removed these variables from further investigation. Clinical variables that remained in the DBN model were days of antibiotics, day of life of sample obtained, gestational age at birth, and post-conceptional age when sample obtained; and also including the binary indicator for the infant being cared for in an open (dou- ble) or closed (single) room . ) ( g ) While building our Bayesian networks, prior assumed distributions on each node are required for use with Bayes’ Rule to determine the posterior probability of the data; the parameters of these prior distributions were chosen to provide a strong complexity penalty (prior equivalent sample size: ν = 10; prior assumed standard deviation: σ 2 = 1; for more details, see McGeachie et al.28) and limited each node to a maximum of three pos- sible parent nodes (although with an unlimited number of child nodes). We used the CGBayesNet function FullBNLearn()27, which performs an exhaustive search through possible edges using a hill-climbing algorithm: starting with an empty network with no edges, it chooses the most likely edge addition (or deletion), adds that edge to the network, and then repeats this process until no more edges remain which will increase the posterior likelihood of the data given the network28. We predicted the progress of the microbiome of each subject from the DBN using the remaining subjects as training data to learn the parameters of the DBN, thus not biasing the network with the microbiome transitions unique to a particular subject. For comparison, we built a DBN using a strong prior (parameter weights ν = 50, σ 2 = 1), which biased toward Gaussians with large standard deviations compared to the data, and set the maximum number of parents per node to five; these settings result in more edges being included in the network. We compared predictive performance of this network (the “dense” network, supplemental Fig. 1) to our preferred network (the “sparse” network, Fig. 1) on each of the subjects. We compared performance differences in mean absolute error using a two-tailed t-test. www.nature.com/scientificreports/ www.nature.com/scientificreports/ independent observations, we chose the number of subjects (n = 58) as the Benjamini-Hochberg correction for multiple testing; this enabled us to identify subjects experiencing at least one statistically unlikely abruption event. independent observations, we chose the number of subjects (n = 58) as the Benjamini-Hochberg correction for multiple testing; this enabled us to identify subjects experiencing at least one statistically unlikely abruption event. Iterative Models. DBNs provide a way of simulating the action of perturbations or unusual initial conditions on the eventual microbiome composition. We performed such iterative predictions by starting with initial condi- tions similar to a particular subject, and then used those values to predict the bacteria abundances of the second sample for that subject. We then used those predictions to predict the third sample’s abundances, and then used that output to predict the fourth, and so on. This allowed us to project microbiome abundances far into the future and to simulate unusual initial conditions. Model Construction. We measured accuracy of the model in predicting each time point from the previous time point, for each subject. Using all available variables, we obtained predictions for the amount of each bacterial taxa included in the network. These values were then normalized to represent percent composition of bacteria out of the total population, in keeping with the original interpretation of these data. In further analysis, bacteria taxa were included if they either appeared in the DBN model or represented at least 1% of any one sample, we thus included Actinobacteria, Alphaproteobacteria, Bacilli, Bacteroidia, Betaproteobacteria, Clostridia, Cyanobacteria, Epsilonproteobacteria, Erysipelotrichi, Flavobacteria, Fusobacteria, Gammaproteobacteria, Holophagae, and unclassified. Average predictive accuracy was measured by mean absolute error across the included bacteria taxa. For clarity, figures are shown displaying the 12 most commonly-occurring bacteria taxa. Formal Bayesian Methodology. Our DBNs are a type of Conditional Gaussian Bayesian model; and as such it induces a joint distribution over the variables in the domain of the time series, a distribution which is a multidimensional normal mixture density. The DBN is composed of a directed acyclic graph G over the variables of the domain (discrete variables Δ in the present analysis, including open room; and continuous variables Ψ, including all of the bacteria concentrations and antibiotics, day of life sample obtained, gestational age at birth, post-conceptual age when sample obtained). We write π(X) for the set of parents of variable X in G. www.nature.com/scientificreports/ The DBN further specifies a set of conditional probability distributions P over Δ , and a set of conditional linear Gaussian density functions F over Ψ. We can then write the multivariate normal mixture density over all variables as: ∏ ∏ π π (∆) (Ψ ∆) = ( ( )) ( ( )) ( ) ∈∆ ∈Ψ P F p x x f y y 1 x y ( ) 1 Continuous variables are modeled as Gaussian regressions on their continuous parents with parameters depend- ent upon their discrete parents, so that we have: Continuous variables are modeled as Gaussian regressions on their continuous parents with parameters depend- ent upon their discrete parents, so that we have: α β σ , ( ( ) + ( ) , ( )), ( ) ~ y u v N u u v u 2 2 ( ) 2 α β σ , ( ( ) + ( ) , ( )), ~ y u v N u u v u 2 or u the discrete parents of y and v the continuous parents of y, according to the DBN, where α (u) is an intercep ependent upon u, and β(u) is a vector of regression coefficients of v dependent upon u.i for u the discrete parents of y and v the continuous parents of y, according to the DBN, where α (u) is an intercept dependent upon u, and β(u) is a vector of regression coefficients of v dependent upon u.i In the current work, we considered a simplified two-stage DBN (TS-DBN) which uses a Markov assumption that the values of variables at time t + 1 are independent of earlier time points (t − 1 and earlier) given the variable values at time t. We further assumed that all transitions of interest were from time t to time t + 1, and not within time t or t + 1. www.nature.com/scientificreports/ Given the DBN, we use the expectation of variable y at time t + 1, as the predicted outcome for yt+1, which is simply: α β \| , = ( ) + ( ) , ( ) + E y u v u u v [ ] 3 t t t t t t 1 ( ) 3 where ut is a vector of the assignments of discrete parents u at time t, and vt is a vector of assignments of continu- us parents v at time t. We modified the CGBayesNets package to compute this expectation in simple TS-DBNs. Given the DBN with distribution parameters θ , the predicted value for discrete variable x at time t+1 is given by the following, where z is a value of discrete variable x, ut is an assignment to the (necessarily discrete) parents u of x at time t, Γ () is the Gamma function, and the notation α[ut,z] refers to the cardinality of the prior assumed sample size having variables u = ut and x = z, similarly n[ut,z] refers to the number of data points in D such that u = ut and x = z. θ ( ( = , )), ( ) + P x z u arg max 4 z t t 1 θ ( ( = , )), ( ) + P x z u arg max 4 z t t 1 θ α α α α ( = , ) = Γ( ) Γ( + ) Γ( , + , ) Γ( , ) . ( ) + P x z u u u n u u z n u z u z [ ] [ ] [ ] [ ] [ ] [ ] 5 t t t t t t t t 1 ( ) 4 θ α α α α ( = , ) = Γ( ) Γ( + ) Γ( , + , ) Γ( , ) . ( ) + P x z u u u n u u z n u z u z [ ] [ ] [ ] [ ] [ ] [ ] 5 t t t t t t t t 1 ( ) 5 This equation is a simplification of Sebastiani et al.3, and its value can be directly obtained from Bayesian poste likelihood computations already included in CGBayesNets. Methods Mi bi j p pf g To identify unlikely abruption events, which signify departures from the underlying, expected model, we calculated the distribution of posterior log probabilities of each sample, on the –log10 scale, then converted to z-scores, and identified outliers that had a p-value less than 0.05/58. Since the dataset does not contain purely Scientific Reports \| 6:20359 \| DOI: 10.1038/srep20359 3 Results D i Dynamic Bayesian Network. We constructed a DBN model of the infant gut microbial ecosystem (see Fig. 1). This model encodes temporal conditional statistical dependence in the following way: each node is inde- pendent of all other node’s current time-point values, each node is dependent upon its immediate upstream neighbors’ values at the previous time point, and given those values, is independent of all other nodes’ previous values. This model included the most commonly observed microbiota classes: the top three being Bacilli, Clostridia, and Gammaproteobacteria. The model also included Actinobacteria, Betaproteobacteria, Bacteroidia, Holophagae, Fusobacteria, Flavobacteria, Epsilonproteobacteria, and the unclassified category of bacteria, all of which were absent in most samples and, when present, were detected at small concentrations. The model also included clin- ical variables, primarily several different measures of time: days of antibiotics, day of life of sample obtained, gestational age at birth, and post-conceptional age when sample obtained; and also including the binary indicator for the infant being cared for in an open (double) or closed (single) room. These measures of time were primarily indicators of Clostridia level, which increases with post-conceptional age. This is a result previously observed in this dataset by La Rosa et al. and identified here with the DBN. Rebuilding the model without Holophagae and Cyanobacteria, two possible environmental contaminants, did not alter results (data not shown). y p Average accuracy of the DBN in predicting each subject (n = 58) is shown in Fig. 2. Accuracy was measured in mean absolute error per sample, using the previous sample to predict the subsequent sample. Accuracy was averaged per subject across multiple samples. Some subjects were easier to predict than others. This is a general characteristic of this cohort: some of the subjects display unusual or drastic changes of microbiome sample com- position from one time point to the next; La Rosa et al. referred to these events as abruptions. Overall, two infants showed statistically significant abruption events (see Abruption Events, below). We further found that samples were easier to predict when the time between that sample and the previous one was smaller (Supplemental Fig. 3, p = 0.048, two-tailed t-test). Details about specific abruption events may be found in the discussion. p pi p y We further investigated the accuracy of DBN microbiota taxa concentration predictions by examining indi- vidual subjects. One subject (number 27) is shown in Fig. 3. www.nature.com/scientificreports/ p y y Typical BN models calculate the posterior probability of the data given the network to determine the network that best fits the data28. We use this posterior probability to provide concrete expectations of the data and deter- mine unlikely abruptions. We further adapted the software in CGBayesNets27 to explicitly output ∫ θ θ θ ( ) = ( , ) ( ) , p D G p D G p G d ∫ θ θ θ ( ) = ( , ) ( ) , ( ) p D G p D G p G d 6 ( ) 6 Scientific Reports \| 6:20359 \| DOI: 10.1038/srep20359 4 www.nature.com/scientificreports/ tificreports/ Figure 2. Comparative accuracy of DBN models. Accuracy in mean absolute error per subject is shown, where error is averaged across all bacteria taxa included in the Sparse network (Fig. 1). The Sparse DBN is simpler and performs better on subjects that are easier to predict (below empirical average of sparse networks’ mean absolute error, dashed blue line), while the Dense DBN (Supplemental Fig. 1) contains many more connections and performs better on subjects that are harder to predict (above the dashed blue line). This effect is significant when the two outliers (subjects #4 and #5) are removed (p = 0.045, t-test). Iterative prediction with the sparse network using only the first sample for each subject (blue circles) resulted in much larger average error per subject (p < 0.001, t-test). Figure 2. Comparative accuracy of DBN models. Accuracy in mean absolute error per subject is shown, where error is averaged across all bacteria taxa included in the Sparse network (Fig. 1). The Sparse DBN is simpler and performs better on subjects that are easier to predict (below empirical average of sparse networks’ mean absolute error, dashed blue line), while the Dense DBN (Supplemental Fig. 1) contains many more connections and performs better on subjects that are harder to predict (above the dashed blue line). This effect is significant when the two outliers (subjects #4 and #5) are removed (p = 0.045, t-test). Iterative prediction with the sparse network using only the first sample for each subject (blue circles) resulted in much larger average error per subject (p < 0.001, t-test). that is, the posterior probability of the data, D, given the DBN graph G, and DBN parameters θ . www.nature.com/scientificreports/ We rely on the CGBayesNets package to compute this according to standard Bayesian equations, further details of which are available in McGeachie, Chang, and Weiss28.i Our modifications that enable these DBN computations are available at www.cgbayesnets.com, where they are integrated into the latest release of the CGBayesNets package. Scientific Reports \| 6:20359 \| DOI: 10.1038/srep20359 Results D i The prediction for a sample is based on the data from the previous sample. The second axis shows the mean absolute error per sample (yellow line). predicted microbiome composition for Subject No 27. Subject No 27, panel (a) Figure 3. Actual and predicted microbiome composition for Subject No 27. Subject No 27, panel (a) microbiota taxa abundances by sample. Panel (b) predicted microbiota taxa abundances by DBN. This subject shows a somewhat typical distribution of microbiota, dominated by Gammaproteobacteria that gives way to Clostridia. Small amounts of Bacilli are present, along with trace amounts of Actinobacteria. On the second axis, panel (a) shows the posterior likelihood of the data conditioned on the previous sample (yellow line); shown in negative log scale. Panel (b) shows predictions of bacteria taxa abundances given the previous sample’s abundances, according to the DBN. The prediction for a sample is based on the data from the previous sample. The second axis shows the mean absolute error per sample (yellow line). are generally well-predicted. In general, the DBN predicts increasing amounts of Clostridia over time, residual amounts of Bacilli, and increasing amounts of Gammaproteobacteria that then give way to Clostridia. These trends are visible in Fig. 3 by comparing the true data (Fig. 3, panel a) to the predicted data (panel b). This subject also displays relatively minor changes from one time point to the next, resulting in both low prediction error and rea- sonable posterior probability of the true data. p p y We also observed connections between the three dominant bacterial classes and other relatively rare tax- onomic groupings. For instance, abundances of Bacilli were dependent upon prior Actinobacteria levels. Clostridia abundance depended upon prior abundance of Fusobacterium. Actinobacteria predicted abundance of Bacteroidia. Abruption Events. Abruption events, or large and unlikely departures from the previous sample’s abun- dances of bacteria, were visually evident in several of the subjects. We quantified the likelihood of these abrup- tions by calculating the posterior probability of the data from the network; this gives a quantitative estimate of the probability of seeing a more extreme combination of bacterial taxa, where this probability is based on the posterior density functions for each bacteria taxa (see Methods). We were unable to predict these abrup- tion events: the DBN has learned an essentially conservative model that predicts changes primarily between Clostridia, Bacilli, and Gammaproteobacteria. Results D i All 58 subjects are similarly shown in an Appendix. This subject is representative of many of the subjects in the cohort because Gammaproteobacteria, Clostridia, and Bacilli dominate; in general these three make up the majority of the microbial ecology of the infants studied. In the DBN, these three bacteria are connected; with prior levels of Clostridia influencing levels of Bacilli, and prior levels of Bacilli influencing levels of Gammaproteobacteria; in addition age (post-conceptional age when sample obtained) influences Clostridia. Our confidence in the dynamic relationship between two variables (nodes) is cap- tured in the natural log Bayes Factor (Supplementary Table S1). Two of these relationships are stronger than most inter-bacterial relationships (natural log Bayes Factor29 7.9 for age-> Clostridia, 0.3 for Clostridia-> Bacilli, and 2.5 for Bacilli-> Gammaproteobacteria; median log Bayes Factor is 0.65). Infants where these bacteria dominate Scientific Reports \| 6:20359 \| DOI: 10.1038/srep20359 5 www.nature.com/scientificreports/ Figure 3. Actual and predicted microbiome composition for Subject No 27. Subject No 27, panel (a) microbiota taxa abundances by sample. Panel (b) predicted microbiota taxa abundances by DBN. This subject shows a somewhat typical distribution of microbiota, dominated by Gammaproteobacteria that gives way to Clostridia. Small amounts of Bacilli are present, along with trace amounts of Actinobacteria. On the second axis, panel (a) shows the posterior likelihood of the data conditioned on the previous sample (yellow line); shown in negative log scale. Panel (b) shows predictions of bacteria taxa abundances given the previous sample’s abundances, according to the DBN. The prediction for a sample is based on the data from the previous sample. The second axis shows the mean absolute error per sample (yellow line). Figure 3. Actual and predicted microbiome composition for Subject No 27. Subject No 27, panel (a) h Figure 3. Actual and predicted microbiome composition for Subject No 27. Subject No 27, panel (a) microbiota taxa abundances by sample. Panel (b) predicted microbiota taxa abundances by DBN. This subject shows a somewhat typical distribution of microbiota, dominated by Gammaproteobacteria that gives way to Clostridia. Small amounts of Bacilli are present, along with trace amounts of Actinobacteria. On the second axis, panel (a) shows the posterior likelihood of the data conditioned on the previous sample (yellow line); shown in negative log scale. Panel (b) shows predictions of bacteria taxa abundances given the previous sample’s abundances, according to the DBN. Results D i We did quantify the likelihood of these abruptions as follows: we reported the posterior probability of a data sample based on the DBN model and the previous data sample for that subject. This resulted in probability estimates (e.g., Fig. 3) reported in negative log probabilities. We identified statistically significant outliers from the distribution of posterior sample likelihoods (below (0.05/58) on the z-scale of likelihoods, see Methods), identifying subjects experiencing at least one anomalous posterior data sample likelihood, and correspondingly, an unlikely abruption. This resulted in two subjects (numbers 16, log(p) = − 3661, and 55, log(p) = −2413) experiencing statistically significant abruptions (Fig. 4). These two sub- jects may represent natural anomalies, or in these particular cases, it may represent some type of processing error. We accordingly removed these two subjects, and repeated our measure of abruptions, obtaining three subjects that experience either unusually high amounts of Bacteroidia (number 17), or statistically unlikely transitions to majority-Bacteroidia samples (numbers 52 and 53), shown in Supplemental Fig. 4. The ability of our method to identify these events, which may represent some type of anomaly, whether in the sample processing or in the actual microbiome, is a strength of our DBN approach. Iterative Prediction. To investigate the DBN model’s ability to predict final outcomes from the initial con- ditions, we performed iterative prediction. In this scenario, we used the results from our prediction of one time point as input into the prediction of the next time point, and thus iteratively predicted a microbiome trajectory only from initial conditions. Average predictive accuracy in mean absolute error per subject was much higher than prediction using every sample (p < 0.001, two-tailed t-test) and is shown on Fig. 2. In some cases, the iter- ative prediction provides what appears to be a smoothed or idealized progression compared to the volatility in the measurements of the actual daily bacteria concentrations for a given subject (example, Fig. 5, subject number 27. Compare Fig. 3). In some subjects, the iterative prediction method had lower mean absolute error than the every-sample method; inspection showed these were either subjects with only 2–5 samples total, or subjects Scientific Reports \| 6:20359 \| DOI: 10.1038/srep20359 6 www.nature.com/scientificreports/ Figure 4. Subjects with unlikely abruptions. Subject 16 progresses through a typical volatile mix of Bacilli, Clostridia, and Gammaproteobacteria, but at day 29 this is replaced by 100% unclassified bacteria. Results D i Subject 55 has an unusual mix of rare bacteria taxa initially, but this is entirely replaced by Bacteroidia subsequently. On the second axis, the yellow line shows the posterior likelihood of the data conditioned on the previous sample, in negative log scale. Figure 4. Subjects with unlikely abruptions. Subject 16 progresses through a typical volatile mix of Bacilli, Clostridia, and Gammaproteobacteria, but at day 29 this is replaced by 100% unclassified bacteria. Subject 55 has an unusual mix of rare bacteria taxa initially, but this is entirely replaced by Bacteroidia subsequently. On the second axis, the yellow line shows the posterior likelihood of the data conditioned on the previous sample, in negative log scale. Figure 5. Iterative Prediction of subject number 27. Compare Fig. 3, Panel (a). This presents an idealized prediction of the trajectory of subject number 27 from day 15 iteratively simulated into the future. The second axis shows the mean absolute error per sample (yellow line). Figure 5. Iterative Prediction of subject number 27. Compare Fig. 3, Panel (a). This presents an idealized prediction of the trajectory of subject number 27 from day 15 iteratively simulated into the future. The second axis shows the mean absolute error per sample (yellow line). where one outlier sample biased prediction by the every-sample method (e.g., subject #14, Supplemental Fig. 2), however this may also be due to chance.h y Our iterative predictions displayed a remarkable insensitivity to initial conditions. This is demonstrated by our analysis using synthetic initial conditions of six varieties: 1) initial conditions composed of an even split between the three main bacteria taxa – Bacilli, Clostridia, and Gammaproteobacteria; 2) the same even split with small amounts of other rare taxa included; 3) similar to condition 2, but without any Clostridia; 4) similar to condition 2, but without any Gammaproteobacteria; 5) similar to condition 2, but without any Bacilli; and 6) initial conditions composed of an even split between each rare type of bacteria (Fig. 6). In each case, after 15 or 20 time points, the populations converge to the same pattern: increasing Clostridia, decreasing Gammaproteobacteria, residual Bacilli, and any rare bacteria taxa have decreased to very minor amounts. This is the general behavior of the DBN model; and it captures well the insight from the original La Rosa et al. paper13. Sparse Vs. Dense Networks. Results D i For comparison purposes, we also generated a DBN with many more edges and including more of the rare bacteria (see Methods). This resulted in a denser network (Supplemental Fig. 1). We hypothesized that this would be able to predict microbiome transitions involving rare bacterial taxa to a greater extent than our original sparser network model. In general, this model displayed the same charac- teristics as the sparse network model: capturing the essential relationship between Bacilli, Clostridia, and Gammaproteobacteria; while displaying little difference in the predictions involving rare bacteria concentrations. Although per-sample prediction accuracy differed little between the sparse and dense DBN models (Fig. 2), we Scientific Reports \| 6:20359 \| DOI: 10.1038/srep20359 7 www.nature.com/scientificreports/ Figure 6. Iterative predictions of infant gut microbiome composition with different initial conditions. From left to right and top to bottom, initial conditions are simulated starting with an even amount of the three predominate bacteria (Clostridia, Bacilli, and Gammaproteobacteria); those are then simulated with small amounts of rarer bacteria; next initial conditions with no Clostridia are simulated; then no Gammaproteobacteria; then no Bacilli; and finally initial conditions with very large amounts of rare bacteria. In all cases the differences in initial conditions are largely gone by 20 time steps. Figure 6. Iterative predictions of infant gut microbiome composition with different initial conditions. Figure 6. Iterative predictions of infant gut microbiome composition with different initial conditions. From left to right and top to bottom, initial conditions are simulated starting with an even amount of the three predominate bacteria (Clostridia, Bacilli, and Gammaproteobacteria); those are then simulated with small amounts of rarer bacteria; next initial conditions with no Clostridia are simulated; then no Gammaproteobacteria; then no Bacilli; and finally initial conditions with very large amounts of rare bacteria. In all cases the differences in initial conditions are largely gone by 20 time steps. did notice that the sparse network was better at predicting subjects which had lower total prediction error (below the empirical mean of the mean absolute error of the sparse model, blue line on Fig. 2); while the denser network was better at predicting subjects with greater total error (above the blue line on Fig. 2). We found this trend was statistically significant only after removing the two outliers (subjects number 4 and 5), at p = 0.045 (two-tailed t-test). Scientific Reports \| 6:20359 \| DOI: 10.1038/srep20359 Discussion Although the developing infant gut microbiome remains understudied, La Rosa and colleagues collected and analyzed extensive data on its progression in early life, in a cohort of preterm infants admitted to a NICU13. With 922 microbiome assessments, this is an attractive cohort for DBN analysis, of which other authors have suggested that 1000 samples is a good baseline for reconstructing an adequately accurate DBN25 La Rosa et al. used this data to demonstrate the succession of bacterial colonization (from Bacilli to Gammaproteobacteria to Clostridia), and showed the importance of post-conceptional age to the developing infant gut microbiome. They further con- cluded that (with rare exception) the trajectory of the infant gut microbiome was independent of the initial condi- tions. Our analysis of this data set, using DBNs, confirmed many of La Rosa et al.’s findings, with some important additions. With a DBN model, we were able to: (1) identify how the three dominant bacterial classes influence one another over time; (2) account for the role of relatively rare taxa in influencing these dominant groups; (3) deter- mine the importance of initial conditions (microbiome profile immediately after birth) for iterative prediction of the post-natal microbiome trajectory; and (4) identify samples that depart from the expected trajectory of infant gut microbiome development in the first month of life. i Given the choreographed sequence of progression consistently observed in La Rosa et al.13 (initial predomi- nance of Bacilli, followed by Gammaproteobacteria, then Clostridia), it may appear as though Bacilli are reduced as the result of increasing Gammaproteobacteria, and that subsequent decreases in Gammaproteobacteria are the direct result of increasing Clostridia. DBN analysis of these data show that Gammaproteobacteria is not nec- essarily an intermediary in the progression from predominance of Bacilli to Clostridia (Bacilli concentrations are independent of Gammaproteobacteria levels, but do depend on infiltrating Clostridia). Although the true nature of ecological change in infant gut microbiome after birth is difficult to know with certainty, our model may yield some insight into the dynamics of the well-known shift8–10 from micro-organisms that tolerate or utilize oxygen (facultative anaerobes) to populations of obligate anaerobes. Results from the DBN model suggest Scientific Reports \| 6:20359 \| DOI: 10.1038/srep20359 8 www.nature.com/scientificreports/ that as initial facultative anaerobe colonizers (Bacilli) are diminished by increasing abundance of obligate anaer- obes (Clostridia), a second class of facultative anaerobes (Gamma-proteobacteria) momentarily out-compete the Bacilli and flourish temporarily. Discussion Although direct comparisons between our class-level taxonomic model and Trosvik et al.’s phylum-level model are difficult to make, we did also capture interaction between Firmicutes (Bacilli) and Proteobacteria (Gammaproteobacteria). Our model of class-level taxonomic data also demonstrated interaction between two taxonomic classes within Firmicutes (the influence of developing Clostrida abundance on Bacilli). (l p g ) While we focused on a sparse DBN model in this work, we also constructed a dense DBN that included a greater number of nodes (between rare taxa) and edges. The majority of network connections in the sparse DBN were reiterated in the dense model, but new connections between dominant groupings and rare taxa also appeared. For instance, in the dense model, prior levels of both Flavobacteria and Bacilli relate to cur- rent Gammaproteobacteria composition (in the sparse network Flavobacteria is not included as a node). Prior Alphaproteobacteria composition (another rare taxa left out of the sparse network) is featured as a predictor of Bacilli and Clostridia levels over time. Microbiome prediction accuracy for subjects with more complex micro- biome ecology (due to the presence of various rare bacteria) improved when using the dense, as opposed to the sparse, DBN model. p For the majority of infants, the sparse DBN showed accurate predictions of the gut microbiome over time, based on prior composition from each previously measured time point. This type of model is ideal for data sets with multiple repeated assessments of the microbiome within a given time window. Alternatively, the iterative prediction model, a special case of the DBN method, makes sequential predictions over time using only one initial microbiome assessment. Using this model, the final prediction of the infant’s gut microbiome at about one month of age, based on assessment of the microbiome just days after birth, showed remarkable insensitivity to starting taxonomic composition. According to the DBN model, regardless of the initial conditions, Clostridia eventually predominate, with some remaining levels of Bacilli and Gammaproteobacteria. The lack of importance of initial conditions suggests extreme plasticity in colonization, and may indicate a larger role of host biology and/ or external factors in directing colonization. g While, overall, the DBN showed accurate predictions of the microbiome, subjects with rapid, extreme changes in microbiome composition were difficult to predict. In this context, abruptions represent the degree to which a measured microbiome departs from a predicted microbiome given the underlying Dynamic Bayes Net model. Discussion After this temporary increase, Gamma-proteobacteria ultimately gives way to obligate anaerobes (Clostrida). While Clostridia is a major predictor of decreasing Bacilli abundance over time, the DBN model also highlights Actinobacteria (many of which are obligate anaerobes), as having an equally important influence on declining Bacilli abundances. Actinobacteria are considered a rare taxon in this study, as they were present in relatively few infants. As was shown in La Rosa et al.13, post-conceptional age emerges in the DBN as the main factor associated with increasing abundance of Clostridia, which eventually predominates in the infant gut. In addition to post-conceptional age, the DBN model also demonstrates that prior levels of Fusobacteria (a relatively rare taxon) may influence Clostridia levels over time. Bayes Factors associated with the dynamic network connections between Clostridia, Bacilli and Gammaproteobacteria were highest, suggesting that these connections are more likely to represent true causal associations. While the DBN model allowed for the inclusion of relatively rare taxa, network connections involving these microbes (Actinobacteria, Fusobacteria and Betaproteobacteria) had lower Bayes Factors, indicating a lower level of confidence in these relationships (Supplementary Table S1). In general, infant characteristics, including antibiotics administered and single vs. double rooms, had little influence on the progression of the gut microbiota.i l p g g Most studies on development of the gut microbiome within the first months of life do not account for inter- actions between taxa, and instead consider changes in cluster profiles30,31 or employ time series analyses for community profiles as a whole32. Prior work on modeling microbe-microbe interactions within the infant gut microbiome over time has been conducted using targeted RNA microarrays, which limit investigation to a set of known bacteria33. In that work, nonlinear regression was used to identify significant interactions between gut microbiota at the phylum level33. Simulations using this model revealed a predominance of Proteobacteria at age four months, with Firmicutes occurring at mid-level abundance, and Bacteroides at lower levels (although concen- trations may have been underestimated). Proteobacteria, Firmicutes and Bacteroides interacted with one another, showing intra-phylum competition (including a competitive relationship between Bacteroides and Firmicutes). While our model of La Rosa et al.’s data featured Bacteroides as a node within the network, Bacteroides abundance did not appear to influence levels of other taxa in the time window between birth and one month of age. Scientific Reports \| 6:20359 \| DOI: 10.1038/srep20359 www.nature.com/scientificreports/ www.nature.com/scientificreports/ composition may reflect measurement error or sample swapping; such rapid change (occurring on the same day (day 5)) is an extremely unusual event, and more sudden than compositional changes following even the strongest external perturbations (administration of broad spectrum antibiotics)35. If these two subjects are removed, the remaining abruptions detected are large quantities of Bacteroidia, which we consider to be much more likely to represent true biological transition and abruption. In general, little is known about the implications of abrup- tions in gut microbiome progression during early infancy; theoretically those abruptions that initiate a persistent change in gut microbiota (over days or weeks, such as the abruption show in subject 55), may have greater physi- ological impact on the host, as compared to highly transient abruptions. g p p g y p While application of a DBN model to La Rosa et al.’s longitudinal microbiome data in preterm infants yielded new insights about microbiome prediction and interactions between microbes over time, the DBN also had some limitations. First, the observation that gender, mode of delivery, and breast feeding variables resulted in overfit- ting indicates that there was not enough data present in the dataset to accurately model their influence upon the microbiome with a DBN methodology. Second, the lack of data on rare taxa appeared to affect the prediction of those groupings. A third limitation is that while bacterial taxa may show statistical relationships within a network, the exact nature of the biological mechanisms (if any) underlying these relationships often remain unknown. Lastly, the results of the DBN in the population studied (preterm infants residing in the NICU), may not be gen- eralizable to full-term infants who reside in a home environment shortly after birth; these preterm infants may be subject to unusual exposures in a hospital, may be at risk for conditions or microbial abnormalities due to prematurity, or even be unusually insulated from exposures in a controlled NICU setting. p y y p g More accurate prediction could be achieved with more detailed datasets, ideally with greatly increased fre- quency of microbiome observation, at least daily. It is possible that existing datasets could be used as priors for future datasets, thus increasing the accuracy of the predictions. www.nature.com/scientificreports/ The extent to which the daily microbiome concen- trations in infants fluctuate is also a concern; with more data on the daily fluctuations, we may be able to estimate if a prediction is within the expected range of possible microbiomes for a particular infant on a particular day. Similarly, we don’t yet have a firm grasp of which environmental influences impact the microbiome concentration the most, and knowledge of those effects in future datasets would undoubtedly be helpful. Finally, our network methodology could be improved; perhaps modeling within time-slice interactions would lead to a greater under- standing of the microbial interactions within the microbiome. g Application of a Dynamic Bayesian Network model to longitudinal gut microbiome data from pre-term infants admitted to a NICU identified important relationships between microbiome taxa, and was used to simu- late future changes in microbiome composition from measured or synthetic initial conditions. Simulation studies revealed that initial composition of the gut microbiome mattered little for development of the infant gut microbi- ota at about one month of age. Likelihood estimates for sudden, dramatic departures from expected microbiome composition may be calculated using a DBN model, and may be useful for identifying laboratory error or true biological variation due to clinically relevant dysbiosis. Our analysis shows that DBNs can be applied to longitu- dinal microbiome data. Application of this method to longitudinal microbiome data from other cohorts of infants (i.e. non-diseased) may provide valuable insight into how the gut microbiome gets established. Discussion Given the lack of knowledge on the development of the infant gut microbiome, it is difficult to know if the observed abruptions represent truly unlikely events (even including errors in sample collection or processing), or just normal, albeit sudden, changes that deviate from the typical course of microbiome progression. Furthermore, since the DBN model is learned from the data, the model has little ability to predict events that occur rarely within that data; and consequently assigns small probability to those events. Nevertheless, the DBN does provide an objective output of data likelihood and thus could be used to identify samples for further scrutiny. Accordingly, we may examine the two least likely abruptions shown in Fig. 4. The first seems the most likely to represent some sort of error in collection or data processing: the abruption replaces all data from day 19 in subject 16 with unclas- sifiable bacteria; the preceding and subsequent samples are both composed of standard mixes of Actinobacteria, Bacilli, Clostridia, and Gammaproteobacteria. While our general ignorance of the processes governing infant gut microbiome development makes rendering judgments of this type tenuous, we would propose that this sample be retreated or reexamined in some way. Perhaps stranger is the abruption in subject 55, in which (an unusual) variety of bacteria (present only in the initial measurement on day 5), is suddenly replaced by 100% Bacteroidia. However, in this case, all subsequent samples from this patient are also dominated by Bacteroidia. It seems pos- sible that one, perhaps rare, outcome is for the infant gut microbiome to be replaced by Bacteroidia, which has been shown to be the major component of adult gut microbial populations34. Alternatively, the initial microbial Scientific Reports \| 6:20359 \| DOI: 10.1038/srep20359 9 References y g 2. Muegge, B. D. et al. Diet drives convergence in gut microbiome functions across mammalian phylogeny and within humans. Science 332, 970–974 (2011). 2. Muegge, B. D. et al. Diet drives convergence in gut microbiome functions across mammalian phylogeny and within humans. Science 332, 970–974 (2011). 3. LeBlanc, J. G. et al. Bacteria as vitamin suppliers to their host: a gut microbiota perspective. Curr. Opin. Biotechnol. 24, 160–168 (2013). 4. Carmody, R. N. & Turnbaugh, P. J. Host-microbial interactions in the metabolism of therapeutic and diet-derived xenobiotics J. Clin. Invest. 124, 4173–4181 (2014). 5. Bengmark, S. Gut microbiota, immune development and function. Pharmacol. Res. 69, 87–113 (2013).h 5. Bengmark, S. Gut microbiota, immune development and function. Pharmacol. Res. 69, 87–113 (2013). 6. Weng, M. & Walker, W. A. The role of gut microbiota in programming the immune phenotype. J. Dev. Orig Health. Dis. 4, 203–214 (2013).hl g p ( ) 6. Weng, M. & Walker, W. A. The role of gut microbiota in programming the immune phenotype. J. Dev. Orig Health. Dis. 4, 203–214 (2013).hl 7. West, C. E. et al. The gut microbiota and inflammatory noncommunicable diseases: associations and potentials for gut microbiota therapies. J. Allergy Clin. Immunol. 135, 3–13, quiz 14 (2015). 7. West, C. E. et al. The gut microbiota and inflammatory noncommunicable diseases: associations and potentials for gut microbiota therapies. J. Allergy Clin. Immunol. 135, 3–13, quiz 14 (2015). p gy q 8. Wopereis, H., Oozeer, R., Knipping, K., Belzer, C. & Knol, J. The first thousand days - intestinal microbiology of early life establishing a symbiosis. Pediatr. Allergy Immunol. 25, 428–438 (2014).l 8. Wopereis, H., Oozeer, R., Knipping, K., Belzer, C. & Knol, J. The first thousand days establishing a symbiosis. Pediatr. Allergy Immunol. 25, 428–438 (2014).l g y gy 9. Adlerberth, I. Factors influencing the establishment of the intestinal microbiota in infancy. Nestle Nutr. Workshop Ser. Pediatr Program. 62, 13–29; discussion 29–33 (2008). 9. Adlerberth, I. Factors influencing the establishment of the intestinal microbiota in infancy. Nestle Nutr. Workshop Ser. Pediatr. Program. 62, 13–29; discussion 29–33 (2008). Adl b h I & W ld A E E bl h f h b W f A P d ( ) g 10. Adlerberth, I. & Wold, A. E. Establishment of the gut microbiota in Western infants. Acta Paediatr. 98, 229–238 (2009). References 939, 35 45 (2013). 23. Milns, I., Beale, C. M. & Smith, V. A. Revealing ecological networks using Bayesian network inference algorithms. Ecology 91, 1892 1899 (2010) 3. Milns, I., Beale, C. M. & Smith, V. A. Revealing ecological networks using Bayesian network inference algorithms. Ecology 91 1892–1899 (2010). ( ) 24. van Gerven, M. A., Taal, B. G. & Lucas, P. J. Dynamic Bayesian networks as prognostic models for clinical patient management. J. Biomed. Inform. 41, 515–529 (2008). f 25. Zhu, J. et al. Characterizing dynamic changes in the human blo 25. Zhu, J. et al. Characterizing dynamic changes in the human blood transcriptional network. PLoS Comput. Biol. 6, e1000671 (2010). 26. Pe’er, D., Regev, A., Elidan, G. & Friedman, N. Inferring subnetworks from perturbed expression profiles. Bioinformatics 17 Suppl 1, S215–24 (2001). ( ) 7. McGeachie, M. J. CGBayesNets: Conditional Guassian Bayesian Networks package. (Accessed June 14, 2014, at www.cgbayesnets com) (2013). 28. McGeachie, M. J., Chang, H. H. & Weiss, S. T. CGBayesNets: conditional Gaussian Bayesian network learning and inference with mixed discrete and continuous data. PLoS Comput. Biol. 10, e1003676 (2014).t p 29. Kass, R. E. & Raftery, A. E. Bayes Factors. Journal of the American Statistical Association 90, 773–795 (1995). l f f b fl d b d l d d l d 30. Dogra, S. et al. Dynamics of infant gut microbiota are influenced by delivery mode and gestational duration and are associated subsequent adiposity. MBio 6, 10.1128/mBio.02419-14 (2015). l subsequent adiposity. MBio 6, 10.1128/mBio.02419-14 (2015). q p y ( ) 31. Koenig, J. E. et al. Succession of microbial consortia in the developing infant gut microbiome. Proc. Natl. Acad. Sci. USA. 108 Suppl 1, 4578–4585 (2011).t 2. Sharon, I. et al. Time series community genomics analysis reveals rapid shifts in bacterial species, strains, and phage during infan gut colonization. Genome Res. 23, 111–120 (2013). g 33. Trosvik, P., Stenseth, N. C. & Rudi, K. Convergent temporal dynamics of the human infant gut microbiota. ISME J. 4, 151–158 (2010). ( ) 34. Arumugam, M. et al. Enterotypes of the human gut microbiome. Nature 473, 174–180 (2011).hf 35. Dethlefsen, L., et al. The pervasive effects of an antibiotic on the human gut microbiota, as revealed by deep 16S rRNA sequencing. PLoS Biol. 6, 2383–2400 (2008). Acknowledgements g M.J.M. is supported by a grant from the Parker B. Francis Foundation (PI: McGeachie). J.E.S. is supported by NIH R00 HL109162. D.R.G. is supported by NIH grant R01 HL108818. We greatly appreciate the work of PS La Rosa, BB Warner, Y Zhou, G.M. Weinstock, E Sodergren, CM Hall-Moore, HJ Stevens, WE Bennett Jr, N Shaikh, LA Linneman, JA Hoffmann, A Hamvas, E Deych, BA Shands, WD Shannon, and PI Tarr, and their willingness to make their data of the infant gut microbiome publicly available and easily accessible. Author Contributions M.J.M. wrote the paper, conceived of the analysis, performed the computational analysis, and implemented the software. J.E.S. wrote the paper, conceived of the analysis, and performed the microbiome analysis. T.G. wrote the paper, and provided methodological support. G.M.W. aided the analysis and provided microbiome support. Y.Y.L. provided methodological and microbiome support, and helped fund the study. D.R.G. helped write the paper, provided microbiome support, and helped fund the study. S.T.W. helped write the paper, conceived of the analysis, and helped fund the study. A.L. conceived of the analysis, provided microbiome support, and helped fund the study. All authors read, approved and commented on the final manuscript. Scientific Reports \| 6:20359 \| DOI: 10.1038/srep20359 References d l b l d fi b l d h b h 11. Ardissone, A. N. et al. Meconium microbiome analysis identifies bacteria correlated with premature birth. PLoS One 9, e90784 (2014). 12 W T M & P i hi P I f d l i i il i ? L A l Mi bi l 59 572 579 (2014) 11. Ardissone, A. N. et al. Meconium microbiome analysis identifies bacteria correlated with premature birth. PLoS One 9 (2014). 12. Wassenaar, T. M. & Panigrahi, P. Is a foetus developing in a sterile environment? Lett. Appl. Microbiol. 59, 572–579 (2014). 13. La Rosa, P. S. et al. Patterned progression of bacterial populations in the premature infant gut. Proc. Natl. Acad. Sci. USA. 111, 12522–12527 (2014). ( ) 14. Faust, K. & Raes, J. Microbial interactions: from networks to models. Nat. Rev. Microbiol. 10, 538–550 (2012).i 14. Faust, K. & Raes, J. Microbial interactions: from networks to models. Nat. Rev. Microbiol. 10, 538 550 (2012). 15. Ljung, L. System identification: theory for the user (1999). 15. Ljung, L. System identification: theory for the user (1999). 15. Ljung, L. System identification: theory for the user (1999). j g yi y 6. Fisher, C. K. & Mehta, P. Identifying keystone species in the human gut microbiome from metagenomic timeseries using sparse linear regression. PLoS One 9, e102451 (2014). 17. Stein, R. R. et al. Ecological modeling from time-series inference: insight into dynamics and stability of intestinal microbiota. PLoS Comput. Biol. 9, e1003388 (2013).f p 18. Marino, S., Baxter, N. T., Huffnagle, G. B., Petrosino, J. F. & Schloss, P. D. Mathematical modeling of primary succession of murine intestinal microbiota. Proc. Natl. Acad. Sci. USA. 111, 439–444 (2014). 9. Bucci, V., Bradde, S., Biroli, G. & Xavier, J. B. Social interaction, noise and antibiotic-mediated switches in the intestinal microbiota PLoS Comput. Biol. 8, e1002497 (2012).fi p 20. Åström, K. J. & Eykhoff, P. System identification—a survey. Automatica 7, 123–162 (1971). p . Åström, K. J. & Eykhoff, P. System identification—a survey. Autom fi 1. Gerber, G. K., Onderdonk, A. B. & Bry, L. Inferring dynamic signatures of microbes in complex host ecosystems. PLoS Comput. Biol 8, e1002624 (2012). Scientific Reports \| 6:20359 \| DOI: 10.1038/srep20359 10 www.nature.com/scientificreports/ 2. Larjo, A., Shmulevich, I. & Lahdesmaki, H. Structure learning for Bayesian networks as models of biological networks. Methods Mol Biol. 939, 35–45 (2013). Biol. Additional Information Supplementary information accompanies this paper at http://www.nature.com/srepihi Competing financial interests: The authors declare no competing financial interests. How to cite this article: McGeachie, M. J. et al. Longitudinal Prediction of the Infant Gut Microbiome with Dynamic Bayesian Networks. Sci. Rep. 6, 20359; doi: 10.1038/srep20359 (2016). How to cite this article: McGeachie, M. J. et al. Longitudinal Prediction of the Infant Gut Microbiome with Dynamic Bayesian Networks. Sci. Rep. 6, 20359; doi: 10.1038/srep20359 (2016). This work is licensed under a Creative Commons Attribution 4.0 International License. The images or other third party material in this article are included in the article’s Creative Commons license, unless indicated otherwise in the credit line; if the material is not included under the Creative Commons license, users will need to obtain permission from the license holder to reproduce the material. To view a copy of this license, visit http://creativecommons.org/licenses/by/4.0/ Scientific Reports \| 6:20359 \| DOI: 10.1038/srep20359 11
https://openalex.org/W4381839115	https://www.researchsquare.com/article/rs-2117438/latest.pdf	English	null	“They want you to know who they really are inside of the old visage”—biographical storytelling as a methodological tool to explore emotional challenges in old age	BMC geriatrics	2,023	cc-by	10,840	“They want you to know who they really are inside of the old visage” - biographical storytelling as a methodological tool to explore emotional challenges in old age Chao Fang (  C.Fang@bath.ac.uk ) Research Article Keywords: biography, storytelling, older people, ageing, emotional challenges, qualitative methods Posted Date: October 6th, 2022 DOI: https://doi.org/10.21203/rs.3.rs-2117438/v1 License:   This work is licensed under a Creative Commons Attribution 4.0 International License. Read Full License DOI: https://doi.org/10.21203/rs.3.rs-2117438/v1 License:   This work is licensed under a Creative Commons Attribution 4.0 International License. Read Full License Additional Declarations: No competing interests reported. Additional Declarations: No competing interests reported. Version of Record: A version of this preprint was published at BMC Geriatrics on June 23rd, 2023. See the published version at https://doi.org/10.1186/s12877-023-04094-8. Page 1/23 Abstract Background: Older people are at risk of confronting emotional challenges, including loneliness, depression, anxiety and low-levels of wellbeing, as a result of varied ageing-related changes. These emotional challenges may further give rise to socio-health issues and even increased mortality, presenting a pressing public health issue in the context of population ageing. By employing a biographical lens, we aim to introduce storytelling as a methodological tool to more holistically explore the construction of their emotional challenges within the context of their life histories. Methods: Building upon theoretical understandings about the social construction of the life course, we draw on a qualitative study about older people’s loneliness to showcase the methodological value and feasibility of implementing biographical storytelling, to better understand the nuanced and sometimes painful emotional experiences facing people alongside their ageing. Results: Findings from the qualitative study we showcase highlight that not only older people’s emotional pains but also their (in)ability to deal with such are deeply rooted in their earlier lives. The biographical nature of these findings further illustrates the methodological merit of biographical storytelling, integrating meaningful narratives from across different life stages to more holistically understand and support people’s emotional challenges and wellbeing in old age. Conclusion: We argue that biographical storytelling is a particularly useful research methodology to explore the experiences and needs of older people, who have often accumulated rich life stories and who may be at increased risk of facing changes and challenges following bodily deterioration and other ageing-related losses. This methodological development can empower older people to reconstruct their ageing lives in the context of their biography as an ongoing construction. The development also lies in an insightful analytical lens to explore how older people’s earlier life experiences may be carried forward and confronted to shape their emotional stability in the present and future stages of their ageing lives. Introduction Ageing can have a profound impact on shaping people’s physical activeness, social relationships, family dynamics, and more deeply, their sense of meaning and being1. It is widely acknowledged that growing older is not necessarily connected to negative emotional experiences, but conversely, it is more likely to be associated with resilience, contentedness and inner growth2. Despite this, older people are still at risk of confronting emotional challenges, including loneliness, depression and anxiety, and low-levels of wellbeing, as a result of varied and even drastic ageing-related changes (e.g., bodily deterioration, diminished sociability and bereavement)3,4,5,6. These emotional challenges may further give rise to socio- health issues and even increased mortality, presenting a pressing public health issue in the context of population ageing7. The complexity of the disruptions to emotional lives in old age can be further compounded by broader socio-cultural environments, such as healthcare and social support systems, public attitudes to ageing and/or mental health and cultural values1,2. Page 2/23 Page 2/23 The emotional challenges facing older people are not fragmented experiences unique in the ageing process, but rather are often rooted in their lived lives. Ageing, in its broadest sense, begins with birth and continues throughout the life cycle - how people live and feel in old age is constantly building upon, shaped and even restricted by experiences accumulated in their earlier life histories8,9. Therefore, it is of a particular benefit to understand older people’s emotional challenges in the context of their biography. This biographical approach can afford an integrative and retrospective lens, interpreting how older people’s experiences of emotional challenges as rich and complex composites are interwoven by not only their present ageing but also individual life histories, personal meanings and broader socio-historical contexts10,11. Qualitative research methods have long been interested in the multifaceted aspects of the ageing process and older people’s lived experiences. The current methodological literature has largely focused on ‘involving’ and ‘empowering’ older people in research design, ranging from traditional interviews and focus groups to more participatory and creative methods, such as photovoice12,13,14. Storytelling has emerged as a scientifically insightful and emotionally meaningful tool that empowers older people to draw on socially accessible discourses (e.g., language, multimedia materials) to vividly reconstruct their lived lives with researchers, and to further make sense of their experiences and promote wellbeing15,16,17. Introduction More particularly, a study by Brotman, Ferrer and Koehn18 on ageing immigrants has illuminated the implications of storytelling on later life, highlighting that talking about life histories can “capture the complex and intersecting identities of older adults and expose the often-invisible institutional structures and relationships of power that mark older adults’ interactions with family, community, and the state in their everyday lives” (p.480). To build upon the above understandings, we argue that the methodological value of biographical storytelling can be further extended, from being an interventional tool to facilitate older people’s wellbeing and from solely researching 'marginalised older people. Drawing on a large qualitative project (as a showcase) that investigated the emotional lives of 80 older people in the UK and Australia, we aim to demonstrate this approach’s value as a methodological tool, which can help us begin to more holistically understand older people’s emotional challenges. This approach can also help to further to situate such experiences within their biographical contexts and through the lens/meanings that they ascribe to those experiences. This biographical storytelling method can uniquely allow the implications of these experiences on their ongoing later life to be ascertained by researchers. Background Understanding biographical storytelling as a methodology in the context of ageing Research has long argued for storytelling as an authentic methodological tool to capture enriched life encounters as experienced by individuals and intersected with wider socio-historical contexts,. The value of storytelling in research lies not only in generating rich, in-depth data in the form of stories to inform Page 3/23 Page 3/23 research outputs; but it is also closely connected to and reinforced by the storytelling process per se that enables both the storyteller and the researcher to co-construct, compare and contrast their varied interpretations of how experiences are created and given meaning. In a classic debate on the power of language, sociologist C. Wright Mills asserted that language provides speakers with a means of accessing available linguistic and cultural discourses to ‘tell’ (more accurately, interpret and justify) their experiences, and how they negotiate these within their unique situational contexts (e.g., with listeners, in socio-historical environments). The meaning-making processes of storytelling was also captured in the “Life Story Interview”, both as a research method and an intervention model, highlighting the benefits that revisiting lived experiences can have on constructing coherent narratives of self over time11,. The power of storytelling can be particularly evident when exploring older people’s lived experiences. Humans are natural storytellers, they can draw upon their past experiences and knowledge to present themselves to others and to retain their sense of identity. For older people, their long-lived lives can provide a rich source of material for storytelling. Gerontological research has captured the benefits of enabling older people to tell/share their stories to empower their voices and to restore their resilience in facing loss and other ageing-related transitions/challenges. Biographical storytelling in the context of ageing has thus far been largely approached from interventional and therapeutic perspectives, such as Westerhof’s “life-story work” model17 to encourage people to reflexively recollect, evaluate and ascribe meaning to memories as a means of promoting mental health and wellbeing (particularly in old age and during times of changes). Similarly, Bluck has argued the powerful functions of autobiographical memories for older people in creating a sense of continuity and purpose to enhance cognitive abilities and emotional status. Background These understandings can further enable researchers to capture both the ‘surface’ emotional reactions (e.g., sadness, anger and anxiety) and the deeper pain that is biographically/historically rooted in older people’s lived lives and identities. This in-depth lens afforded by storytelling as a research method is applicable to not only those marginalised and without a voice but also to older people more generally, as ageing is a universal experience that often gives rise to a dynamic process in which individuals lose and try to restore their long taken-for-granted lives and self4. Constructing theoretical framing for biographical storytelling Growing older may afford people unique opportunities to reflect on their lived lives. Old age is not a fragmented life stage that can be viewed in isolation, but instead is built upon, constructed and shaped by one’s prior life experiences. As aforementioned, older people may face a myriad of complex emotional, physical and social challenges; as they grow older, they may also increasingly lose their defence (to these emotional pains), in the form of social connections, physical strength, independence and confidence that they have accumulated across their life span4. Therefore, it is important to understand how and why older people experience emotional challenges, in the context of their life histories. This biographical lens can shed light on their taken-for-granted habits and values, and how these may be challenged, compromised and even lost in their ageing lives. Despite this, researchers are placed in a ‘disadvantaged’ position as they often have little knowledge about their participants (older people) prior to research activities, such as interviews. Unlike conventional interviews in which the focus of the conversation is largely on the participant’s present lives or experiences in recent time or in relation to specific research topics in question, biographical storytelling can enable the participant, within limited spatial-temporal frames, to explain and reflect on their experiences, feelings and their taken-for-granted meaning/being and how these may be deeply rooted in their biographies and wider socio-cultural-historical circumstances over time. From a social constructionist perspective, individuals’ lives are not a static entity but are both dynamic and ongoing. As shown in Fig. 1, lifespan consists of varied phases, starting from childhood all the way to old age and ultimately terminating at the end of life 8,9. Throughout their lifetime, humans experience and often learn from others about what is considered to be (or not to be) important, valuable and real at different stages. Background While the above functional approaches of biographical storytelling are valuable for improving older people’s wellbeing, it is also important to develop storytelling as methodological tool for research to explore the ‘hidden’ meaning of ageing experiences rather than making inferences (e.g., behaviour change). We argue that the experience of telling one’s lived life is not only an empowering interventional experience, but also an accessible tool more inherently embedded as part of everyday human life through which older people can naturally express the authenticity of self, reflecting on and re-telling experiences from across their life histories19,. As such, biographical storytelling has significant methodological value, providing convenient and engaging channels to gain access to the lived experiences of older people, whose voices and needs may be otherwise overlooked or unheard18 . Given growing older may be closely connected to varied emotional challenges as a result of facing increased risks of multifaceted loss (e.g., health, social resources, loved ones)1,, we believe biographical storytelling as a methodological tool can offer particularly deep insights into older people’s worries, fears and hopes alongside their ageing. The “Life Story Interview” method11,23 has provided a valuable direction to show how to use in-depth life review interviews to capture the ongoing construction of self- narratives and how the sense of self-continuity can inform future identity development. However, its generalised approach to interview participants from a wide range of age groups (varied from teenagers to Page 4/23 late mid-agers) and its emphasis on linking lived experiences to future goals may be less relevant to older people who face unique challenges and whose ongoing lives may be defined more by the past than the future. late mid-agers) and its emphasis on linking lived experiences to future goals may be less relevant to older people who face unique challenges and whose ongoing lives may be defined more by the past than the future. In order to better understand older people’s unique emotional challenges, it is of great importance to develop a tailored biographical storytelling method to centre their ageing lives within their life stories. As such, older people are afforded appropriate means of reconstructing the emotional challenges faced in their ongoing lives through stories about their own lived experience and wider socio-historical backgrounds. Background Meanwhile, they also tend to internalise these experiences and knowledge (despite being positive or negative, for example, personal growth or trauma) to construct a thread of narratives to nurture and maintain their sense of self as a unique being. These narratives are not always distinct but Page 5/23 are often closely connected with each other, contributing to a sense of self-identity which can be understand and made whole over time9. As such, people’s past experiences can be habitualised as a complex of meanings and resources providing references for them to make sense of who they are at present and further envisage their future self11. As life moves on, this spatial-temporal construction of identity is also ongoing across the life stages, in which individuals encounter varied experiences and socio-cultural scripts to continuously affirm, revise and justify their self-narratives31. To understand older people’s experiences, as will be showcased by our study, it is vital to link their ageing lives to and within their biography as a system of narratives imbued with individual life histories and intersected with broader socio-historical backgrounds. Such an approach can provide an insightful lens to capture the enriched and long process of older people’s interconnected life stages and their continual (re)construction of identity27. To research the biographical construction of older people’s lived experiences of emotional challenges, we argue an interpretivist epistemological position is required. This position is important when seeking to understand how older people recollect and make sense of their ‘cumulative fragments of a lived autobiography’ in order to reconstruct their life stories and meaning27,. Older people often draw upon language and other discourses to talk about and interpret their subjective experiences situated within the contexts of broader socio-historical circumstances and the particular situation where this reconstruction takes place. Biographical storytelling, thus, can provide older people with a unique social situation, in which they can enact upon their personal biography (life histories in different stages) to sequentially and consequentially reconstruct and represent their lived experiences to others. This can further allow for a phenomenological lens to provide insightful interpretations of what it means to be a person for older people in their unique life-world27. As such, we as researchers can access their lived experiences not necessarily as these actually happened but in a manner shaped by their meaning and identity that have been long taken for granted. Background Despite the subjective constraints of the retold life stories, biographical storytelling can provide us with an integrative tool to gain insight into older people’s experiences within their biography as a whole rather than as a stream of fragmented events. Methods A large qualitative study was conduct between 2019–2021, to explore the emotional challenges of 80 older people (25 male and 55 females, aged between 55–93 with an average age of 79 and a standard deviation of 7.6) living in retirement communities in the UK and Australia1. Biographical storytelling was employed and operationalised in the form of in-depth and semi-structured interviews with older people to retrospectively explore their emotional lives across the lifespan and how these could shape their emotional challenges in old age. We aim to use this study to showcase the methodological value of biographical storytelling in the context of ageing research. Interview design Interview design Page 6/23 Page 6/23 We designed our interviews to explore emotional challenges in old age through a lens of loneliness. The understanding of loneliness we adopted in this study was not just simply feelings associated with lacking meaningful relationships and interactions. Instead, we understood loneliness more broadly as a condition for eliciting emotional challenges, in which older people may lose resources and capability to retain the integrity of self in relation to others and the external world. As such, semi-structured interviews were used as a socially constructed vehicle encompassing and organising the dynamic aspects of how these older people engaged with scripts at different levels and phases of their lives to deal with and make sense of their emotional voids and losses. Adopting a biographical lens, these interviews enabled the older people to lead the conversation to tell their stories containing the following aspects (not necessarily in a strictly chronological order): (1) their lives up to this point (including childhood, adolescence, career, family, and anything else they wished to share and discuss), (2) their closest relationships, (3) their experiences of loss, (4) feelings of loneliness, isolation and other challenges, (5) their current living environments, and (6) their future concerns/plans. Such a structure is by no means limited to these topics and a biographical approach of this nature would encompass a wide breadth of research interests. The interview questions focused on each of the above areas, such broad outlooks enabled an enriched interview, where the intertwined nature of topics often became evident and sections often intersected with each other. A narrative approach was adopted, empowering the older people to freely tell their stories around these core subjects across their entire lifespan. Methods These questions were used as guides rather than verbatim, providing the interviewer with a thread of topics and directions to assist the interviewee (older people) to verbally reconstruct their ongoing emotional lives. We were not intended to capture their emotional lives in ongoing lives as a linear process. Instead, we deliberately designed the interview questions to contextualise and focus on interviewees’ lived experiences of their emotional lives and challenges in old age within the spatial-temporal dimensions of their biography (see Appendix 1). In so doing, we also aimed to minimise the distress of the storytelling experience and to more holistically understand their emotional challenges and needs in old age. Interview collection In light of the carefully designed interview schedule as outlined in Appendix 1, all the interviewers were professionally trained in sessions organised by the Principle Investigator of the study (the third author) prior to the interviews. Five researchers, including the three authors, conducted semi-structured interviews with 40 participants in the UK and the other four researchers interviewed the remaining 40 interviews in Australia. Approximately 8,000 minutes of rich audio data with the 80 older people were collected from both countries between 2019–2020. These Interviews ranged from 70 to 200 minutes and averaged around 100 minutes, capturing an in-depth picture about how older people’s emotional challenges and subsequent responses could be deeply rooted in their earlier life experiences. The interviews also highlighted the responsiveness of the older interviewees to such a biographical method, often being very receptive to the lines of questioning resulting in rich biographically informed data. Reflexivity Reflexivity Page 7/23 Page 7/23 We acknowledged that talking about earlier life experiences and current emotional challenges may be detrimental for some older people. Therefore, we were keen to be ‘a safe, interested stranger’ with whom the older people could feel at ease to share their sometimes painful and previously unspoken and overlooked life experiences30. While all our participants reported positive interview experiences, we as researchers and interviewers were conscious of the importance of our role in navigating and protecting these older people while walking them through their lived lives with a specific focus on emotional challenges. We felt it was particularly vital in the process of both designing and conducting the interviews, to ensure the participants’ needs and wellbeing were rigorously understood and addressed by the interviewers. Measures, such as organising interview questions in a certain order as outlined in Appendix 1, finishing the interview with uplifting conversations (e.g., psychological growth, hopes) and providing information on further support resources following the interview, were all in place to minimise the impact of the potentially sensitive topics addressed. This biographical storytelling method also allowed us to create a collective transformational space, to not only co-construct knowledge about emotional challenges in old age but also to deepen understandings about the value of reviewing lived experiences in life histories. It was only possible for us to collect such rich biographical interview data because of a trusting relationship between us and the participants. This trusting relationship was often developed in the conversations we had about their experiences of relationships, challenges and support at different life stages. Such conversations fostered a sense of empathy, by which we could more closely understand the older people’s feelings based on experiences of ourselves and our close others (e.g., our own (grand)parents). The experience of conducting the biographical storytelling interviews afforded us as researcher a transformational experience, granting us an insightful and vivid lens (not only for data analysis but also in our everyday lives more generally) to appreciate the value of both understanding the life stories of others and sharing our own. Despite the above benefits of being embedded in the biographical storytelling process with the older people, we were also aware of the potential risks of being overwhelmed by the relationships and the data we developed from the research. Reflexivity Therefore, regular meetings were held between the team members to debrief and resources on further/professional support were also provided throughout the project. Data analysis We adopted a reflexive thematic analysis method to interpret the rich data about these older people’s emotional challenges within their life stories. An inductive approach was embedded throughout to allow us (as researchers and also as ‘strangers’) to understand the older people’s lived experiences and nuanced feelings from their own perspectives, without being dominated by the assumptions and our pre-existing frameworks. We followed the six-phase thematic analysis process recommended by Braun and Clarke40, which involved familiarizing ourselves with the data, generating initial codes, searching for themes, reviewing and discussing potential themes, refining and naming themes, and writing themes up. The analysis was conducted by all authors, who read the interview transcripts and notes thoroughly before conducting independent coding of each interview transcript. A combination of NVivo 12, a qualitative Page 8/23 Page 8/23 analysis software package, and more traditional reading and coding was used to manage and analyse the data. The authors met frequently to discuss and compare findings and codes. If there was disagreement or divergence on codes, further reading and discussion were conducted until a consensus was met. analysis software package, and more traditional reading and coding was used to manage and analyse the data. The authors met frequently to discuss and compare findings and codes. If there was disagreement or divergence on codes, further reading and discussion were conducted until a consensus was met. To more systematically capture the complex experiences of emotional challenges in old age, we adopted an organising framework from Ettema et al., to contextualise our findings in three dimensions: (1) circumstances from which emotional challenges may arise, (2) experiences of multifaceted emotional challenges (3) responses to deal with these challenges alongside ageing. Figure 2 shows our coding process step: 1 captures an illustration of an early iteration of codes under the three dimensions above; step 2 was conducted for this article, aiming to re-categorise a wealth of qualitative findings to methodologically showcase how a biographical lens can be applied to more comprehensively analyse older people’s emotional challenges (for other qualitative findings, see 1,4). By employing our study as an example, we seek to illustrate why it is both important and useful to interpret older people’s life stories not as a series of segmented experiences but as rich and complex composites “embedded in a coherent, meaningful context, a biographical construct” (p.3). Results Paula (aged 72) felt she was nobody after losing her husband whom she had long lived with and cared for: Evident in these older people’s accounts of their ageing-related challenges is a myriad of barriers in retaining emotional strength in their ongoing lives. Our participants were often faced with challenging circumstances of losing people and things that have long helped them retain meanings and resources central to their identity. More than half of our participants had recently lost their spouse and other close families/friends. By analysing their reported experiences connected to their life histories, we found that these losses might not only prompt emotional voids of attachment, security and connectedness in old age, but could also shatter the long-term emotional resources that they had always employed to (re)affirm their taken-for-granted lives and identities. Paula (aged 72) felt she was nobody after losing her husband whom she had long lived with and cared for: Paula: “[W]hile he was alive and I was his full-time carer, companion, friend, we had a ball even though he was in a wheelchair, but when he was gone I didn’t know where I fitted anymore. I didn’t know who I was anymore because I wasn’t…” These emotional barriers could also arise from the loss of good health and physical capabilities that to date had determined their strength in fulfilling their roles and responsibilities in their lived lives. After suffering a stroke, David (aged 80) strongly conveyed his sadness and frustration about the changing caregiving roles between him and his wife and ultimately about the loss of his ‘old’ self as a healthy and capable husband. Only by employing such a biographical approach can one begin to appreciate the constructions of self that David’s loss of autonomy in old age confronted: David: “Well, my wife has always had a difficult hip and right from the days of ironing her blessed skirts I’ve been the one who can do things. I just got on with it… But now, it’s her who has to put my shirt on me. And at this present time, with my left arm sitting pretty well useless in my lap, I’m finding it so, so difficult.” The risks of losing one’s emotional strength in the past were not only attributed to ageing-related losses but might also be deeply rooted in their earlier life histories. Results We contextualised the older people’s reported life stories in their ongoing lives, in which the boundaries between past, present and even future were fluid and interconnected. As such, we sought to explore how both the explicit and hidden meanings of their experiences, not as a completely new composite but rooted in their biography, have been continuously constructed and negotiated alongside their life course. Four interconnected themes have been developed for this section: (1) biographical construction of emotional barriers, (2) biographical pain, (3) diminishing selfhood and (4) sharing life stories as a means of retaining biography. Theme 1 underlines the biographical circumstances that gave rise to emotional challenges in old age. Theme 2 and 3 capture the experiences of how life histories may persist, overshadowing older people’s emotional lives and conversely how losing these lived lives alongside ageing could also prompt deeply painful emotional distress. Theme 4 highlights the older people’s responses to various emotional challenges and how their emotional strength and growth were fostered or reaffirmed during the interview process. In so doing, we aimed to show the significance of retaining the integrity and consistency of biographies and how older people sought to do so by interacting with socio- cultural scripts and by constructing individual meanings. To protect the confidentiality of our participants, pseudonyms are used throughout and all identifying materials are removed. Biographical construction of emotional barriers In our study, many older people reported progressive declines in physical activity, social connections and emotional-cognitive strength due to bodily deterioration, immobility, bereavement and terminal illness. When conducting the analysis, we scrutinised these challenges not as isolated circumstances resulting from the process of ageing but as experiences deeply entwined within life histories41 . Page 9/23 Page 9/23 Evident in these older people’s accounts of their ageing-related challenges is a myriad of barriers in retaining emotional strength in their ongoing lives. Our participants were often faced with challenging circumstances of losing people and things that have long helped them retain meanings and resources central to their identity. More than half of our participants had recently lost their spouse and other close families/friends. By analysing their reported experiences connected to their life histories, we found that these losses might not only prompt emotional voids of attachment, security and connectedness in old age, but could also shatter the long-term emotional resources that they had always employed to (re)affirm their taken-for-granted lives and identities. Results By embedding their varied emotional barriers in the ongoing process of their lives, we discovered that many older people (especially men) embodied a ‘stiff upper lip’ mentality, which reflects unique cultural, generational, gendered and familial circumstances many of the older people shared in their earlier life. Like many others, Simon (aged 92) showed us how older people’s abilities to express emotions and seek support could be historically smothered and how this could undermine their strength in facing varied challenges in old age: Simon: “You had to keep a stiff upper lip as it was called and I suppose that’s stuck with me. I mean, people in my generations who came through the war, not that we were fighting, I was too young, a child, a boy, you had to get on with all sorts of things… [but] I feel sorry for myself at times when you are sat and you think oh god, I do wish she [my wife] was here and I do miss her.” From a methodological perspective, by situating our analysis within the framework of the participants’ biography and life history, we were afforded a useful analytic lens to more fully consider how the older people’s emotional barriers (circumstances) intersected with their prior lived experiences. Further, the Page 10/23 Page 10/23 analytic approach allowed us to more fully elucidate the significance of our findings not just in the now but across the life span of the older people, making succinct connections between events across their biography in relation to their current emotional challenges. In so doing, we were then provided a platform, from which to explore the older people’s experiences and responses to these varied emotional challenges in the ongoing construction of their biography moving forward (as shown below). analytic approach allowed us to more fully elucidate the significance of our findings not just in the now but across the life span of the older people, making succinct connections between events across their biography in relation to their current emotional challenges. In so doing, we were then provided a platform, from which to explore the older people’s experiences and responses to these varied emotional challenges in the ongoing construction of their biography moving forward (as shown below). Biographical pain By conducting a deep reading of these older people’s life stories, we were able to capture some profound emotional pains that they carried with them from across their life course. Whilst sharing the benefits and contentedness of growing older, these older people often also hinted at the (re)surfacing of painful and distressing narratives and memories that had previously been buried in the busyness of independent living and excitement of being ‘young’. Coined as ‘biographical pain’ by gerontologists, such deeply painful recollections may become increasingly difficult to ‘put right’ as older people lose close others and things key to their emotional strength alongside ageing30. When reflecting on their lived lives, the older people often reported traumas and regrets in the past deeply ‘engraved’ on their biography. Such deep pains might remain or even intensify in old age, as reported by Jane (aged 80) who spoke about her the insecurity and lack of confidence due to mistreatment during childhood that continued to haunt her emotional life in old age: Jane: “So from an early childhood I learnt that I was a bad person. My brother certainly told me that. I was stupid, ugly, etc… I mean, last year my daughter-in-law, the French one, started suddenly explaining to me what a worthless person I was and I burst out in tears and it took me months to think, well, poor you, if you are not able to like me, how horrible must that be for you, how horrible to hate.” As a result of losing resources to retain emotional strength when growing older, the regrets accumulated in earlier life could be revived, deeply questioning the worthiness of their lived lives and the integrity of self. Again, as we found from Jane’s interview, after having a busy life spending her time to the fullest, she felt guilt for not having provided enough support to her sons – she burst into tears and said: Jane: “I hope they forgive me for all the things I did not do and all different things they needed.” ane: “I hope they forgive me for all the things I did not do and all different th Such biographical pain was vividly illuminated by the storytelling approach outlined above. Key in our analysis was the focus on the coherence of human biography and how not only growth but also pain in earlier life may remain influential in old age. By contextualising confrontation of profoundly painful emotional challenges in old age within the life course, we were afforded a biographical lens to more holistically and historically understand the deeper roots of the older people’s emotional pains. Diminishing selfhood Prior to this point we have elucidated how our method can be used to extrapolate links to participants lived lives and to understand their emotional challenges in relation to their previous biographical Page 11/23 experiences. This theme will now illustrate how the biographical construction of their current ‘self’ in relation to their past is carried forward and confronted in the future as they age. Integral to their emotional challenges was a realisation that their lives, memories, narratives and identity were increasingly forgotten, unimportant and had been irretrievably lost alongside their ageing29. experiences. This theme will now illustrate how the biographical construction of their current ‘self’ in relation to their past is carried forward and confronted in the future as they age. Integral to their emotional challenges was a realisation that their lives, memories, narratives and identity were increasingly forgotten, unimportant and had been irretrievably lost alongside their ageing29. After living a long life full of experiences and memories, many older people talked about their fear of fading away in a sense that their lives would be forgotten as they grow older. As experienced by Iris (aged 90), living to be old could prompt deep and often growing pains of feeling disconnected and alienated from the external world due to varied ageing-related losses: Iris: “The only other thing is of course that most of my friends are dead. I’m 90 on Christmas day and when I go through my life, my school friends, most of them are gone, my college friends, most of them are gone. I’m the only in-law left, I’m the only great grandparent left. So that aspect of the extended family becomes less and less… Yes, then it becomes lonely as regard your future and your past life and the people that I was familiar with, people that I worked with, of course, and the people that I had social contact with and relatives. ane: “I hope they forgive me for all the things I did not do and all different th My own family, I’ve only got one brother left and my husband was Dutch, and the Dutch family, one sister-in-law. So I’m finding that I’m standing alone as regards my former life.” The older people could further confront pains of feeling that their lives no longer mattered and their biography was becoming increasingly unimportant. This finding was only unveiled by understanding what the participants perceived as important in their lived lives to date and then exploring what elements of their lived experiences may be confronted and undermined in the future. Both Emma (aged 77) and Harold (aged 91) highlighted their emotional struggles of longing for the authenticity of their own being: Emma: “Absolutely, and ‘who am I?’ This is why old people talk about their lives, their past, or whatever, because they want you to know who they really are inside of the old visage.” Harold: “My life? Nobody wants to know about my life!” Underneath these pains of fading away, we found that there was also a more existential awareness of the approaching mortality and the finite nature of their lives. As they age and continuously confront a multitude of losses and challenges, many older people felt death closer than it had ever been before. Robert (aged 72) painfully realised that death was no longer a distant reality after losing his friends: Robert: “I did sort of think, a couple of years back, ‘Oh, I’m now at the age that Jim died.’ But it’s one of the things that you come to learn, these things happen in life, I’m afraid. There was a young lady, younger than us, again died of cancer. That was one of our bridge group. We sort of knew it was going to happen, but it happened quicker than was expected, and again, that brings home your mortality.” The rich lives lived by the older people, as illustrated above, could come under threat as a result of ageing, losing others and ultimately confronting the inevitable fate of death. Our analysis ascribed a deeper meaning to older people’s emotional challenges of helplessness, worries and even fears. That is, we drew Page 12/23 Page 12/23 upon their biographical life stories to capture how their cherished memories and relationships (which have long defined who they are) could be left behind, thus causing emotional pain at a more existential level. upon their biographical life stories to capture how their cherished memories and relationships (which have long defined who they are) could be left behind, thus causing emotional pain at a more existential level. Sharing life stories as a means of retaining biography Employing our methodological approach of integrating older people’s narratives and experiences into their life histories afforded us a meaningful context to better understand why (circumstances) and how (experiences) these older people encountered varied emotional pains alongside ageing. Such an approach has also illuminated their dynamic responses to adopt, reject and revise existing resources to (re)negotiate meaning for varied emotional challenges. Research has long highlighted older people’s agency to reclaim their resilience and strength in facing challenges in varied life aspects. Our analysis has further enriched this understanding by shedding light on these older people’s emotional needs, and more precisely how their emotional needs in old age may be shaped, undermined and restored, in alignment with their life histories and lived experiences. Interviewer: “Yes, it’s just such an extraordinary story.” Helen: “Well, relating it to you makes me appreciate it more, because it’s something I just took for granted and never questioned. But the depth of it suddenly, talking to you, I thought oh wow, I must tell my brother and sister that.” By putting older people’s often under-recognised and undervalued stories into their broader social and biographical contexts, our analysis suggested that these older people faced a lack of accessible platforms to share their life stories in everyday life; this may have been due to losses of long-lasting resources for emotional support alongside ageing. This situation was further compounded by the deprivation of a story-telling literacy in wider society where both older people and others had little knowledge about the benefits of sharing and listening to life stories. This issue was illuminated by Jennifer (aged 75), who lost cherished resources to face and deal with her long-lasting emotional pains. After seeing her adult children become independent and moving from a long-lived neighbourhood to a retirement village, she found: Jennifer:” This is what I miss a lot, a private space to talk … All my life I’ve suffered … and some things I do find very hard, like this illness now. With everything that’s gone wrong, I would have liked to talk to somebody, no advice, I want to let off steam, I suppose. But it doesn’t happen…” It was only by adopting our biographical approach and enabling participants to tell their life stories so openly that the participants began to see the evident benefits/needs of storytelling and that we as researchers fully appreciated its methodological significance. This methodological value lies in better capturing older people’s emotional challenges and their dynamic responses to said challenges. The biographical lens could act as ‘cathartic’ tool for the older people, and also as an insightful analytical approach to understanding their emotional needs in old age more accurately and acutely within their broader biographical context. Harold: “My life? Nobody wants to know about my life!” The above analysis has shown that the older people’s emotional challenges could be experienced in a pervasive and often deeply painful manner as a result of multifaceted losses threatening their taken-for- granted lives and meaning. We found that the older people’s efforts to alleviate emotional challenges focused on maintaining/restoring key resources (e.g., family connection, social and community support) that have long help them stay strong and consistent. The unique biographical lens we employed further illuminated a deeper dimension of their responses, that was to preserve the meaningfulness of their lived lives and their being, potentially in the face of painful memories and fading existence in the world. For example, after moving closer to his daughter and her family, Craig (aged 72) and his wife Jacqueline (aged 71) confronted a profound need to carry on their memories and identities in the family beyond the finitude of their lives: Craig: “We are creating a memory box for each of them [grandsons] where we’ve written a kind of autobiography each of our life, the interesting bits anyway, the bits that we want them to know… We are finding it quite cathartic and quite nice and I would like to think that in 50 years’ time, our grandson will show this information to his grandchildren.” Some older people also demonstrated a concern with passing on their lived experiences as wisdom to others in wider society. Helen (aged 64) found sharing her views on positivity in life to be fulfilling and it helped her to reaffirm her meaning and being in old age: Helen: “So, my mantra is focus on strengths, which makes your weaknesses irrelevant … ‘we appreciate our strengths rather than constantly looking for problems?’ So that has become part of my thing here as well, when I’m talking to people. It’s very much reminding them of their life experiences and their strength and wisdom and things like that.” Page 13/23 Page 13/23 Despite this, we also noticed that not many older people had only realised the power of life-story-telling until the interview (much like Helen who felt the experience was eye-opening after talking about her fondness with her late parents), in which they were able to talk freely about their concerns and memories across their lifespan, helping them further make sense of their lived experienced and appreciate their lives: Discussion Page 14/23 Our study has highlighted the methodological value of biographical storytelling to collect and interpret older people’s emotional challenges. We believe this storytelling method is of particular relevance and importance to research older people’s experiences and needs as they tend to have accumulate more stories throughout their life histories that they may wish to share. By using our loneliness study as a showcase, we illustrated how a retrospective lens could be afforded in research design and interview collection to capture the biographical construction of older people’s emotional challenges. A biographical Page 14/23 analysis approach was also exampled to show the methodological values of interpreting older people’s emotional needs within the wider context of their biography (and associated socio-historical backgrounds). Our methodological procedures and the findings, as highlighted in the case study, have reaffirmed the value of life-story interviews for understanding older people’s present experiences within their life histories11,17. In addition, this article has further expanded the methodological scope of life-story interviewing methods by developing a more dedicated framework for researching (not seeking to change behaviour of) older people, who may focus more on past rather than future and may be more vulnerable to changes and challenges41. Our exploration of biographical storytelling has also highlighted the continuity and fluidity of identity construction across the human lifespan, providing an insightful and integrative methodological tool to illustrate how identity construction continues into old age. Much emphasis has been laid thus far on identity development in earlier stages of life such as childhood, adolescence, or the broader stage of adulthood. Conversely, old age is often seen as the last stage of life, and thus as reported by researchers and some older people themselves, ageing is closely connected a sense of tiredness and withdrawal from active human development of identity,. By conducting a deep listening and reading exercise of the older people’s enriched life stories, we illuminated the older people’s dynamic needs and actions to (re)construct their identity alongside ageing (and even beyond physical death) and how these processes were shaped by their earlier life experiences. As such, we methodologically showcased the importance of connecting biographical gaps between earlier life stages and old age. Discussion In addition to the above methodological benefits, our study has demonstrated the importance of developing appropriate tools and skills when implementing the biographical storytelling method with older people. First, a comprehensive and flexible interview schedule is key to allow for in-depth conversations about older people’s life stories while staying broadly on track about their emotional lives and challenges in old age. As shown in our study, the guide needs not to strictly follow a chronological trajectory but allow for the smooth progression of conversation to facilitate the development of a good rapport with the older people. Such rapport enabled researchers to gently walk older people through their lived lives to elicit recollections and reflections on their emotional ups and downs. The above caution about not too strictly prescribing an interview structure is also connected to the second requirement, that is to equip researchers (as interviewers) with transferable skills to learn how to be empathetic, supportive and focused throughout and across the often long-lasting interviews (e.g., over one hour and even longer). Given the sensitive and even painful nature of emotional challenges in old age, training prior to initial interviews and frequent follow-up self/group reflection sessions may be of use to help researchers develop their own ways to deal with unexpected and challenging disruptions confronted during or as a result of an interview (e.g., the participant suddenly bursting into tears or feeling distressed by certain conversations). Third, a biographical analytical lens is also needed for researchers (as interpreters) to examine older people’s emotional challenges within a wider context of their individual life histories and the associated social-historical circumstances. To gain such an integrative lens, researchers could benefit from developing in-depth theoretical understandings about the human life course and the construction of identity (see Fig. 1). Finally, the insightful responses by these older people have also highlighted the merits of extrapolating the biographical storytelling approach outside of this research paradigm and in a more societal setting. Extant research30 and our findings have both shown a lack of ‘safe and interested’ others and accessible platforms to facilitate life story telling between older people, their family, communities and wider society. Therefore, it would be of significant value to encourage older people and those in their support networks to further recognise the values of biographical storytelling and develop suitable knowledge about how to tell and/or listen to life stories17. Discussion That is, to provide a more longitudinal context to understand how the older people’s emotional lives have been continuously shaped by the ongoing construction of their meaning and identity throughout their lifespan, and more importantly how their emotional challenges in old age may arise from losing the consistency and integrity of their sense of self moving forward. By depicting an insightful picture of each participant’s life stories pertaining to their emotional lives, we further demonstrated the power of biographical storytelling to capture the nuanced experiences of emotional challenges in old age. Given emotions, particularly painful feelings, can be subtle and subjective, existing literature on qualitative methods in ageing studies has emphasised the importance of empowering older people to express their emotions, feelings and needs using their own voices and from their own perspectives12,14. Our study furthered such methodological approaches by encouraging older people to reconstruct varied stories across their life course, including not only big stories (major or life- changing events such as, moving home, illness and bereavement) but also small stories as lived and enriched within everyday settings (e.g., daily interactions and conversations with others). Although small stories may be easily forgotten or overlooked, the focus on such subtle but meaningful experiences is of particular methodological significance to capture how small emotional turbulences may accumulate and unknowingly contribute to the ongoing construction of older people’s meaning and identity15. We afforded the older people a unique opportunity to freely and reflexively tell both their big and small life stories, with which they were able to provide more enriched materials to revisit, reveal and even re- appreciate their memories and feelings of pains, worries and fears in varied aspects. From a data Page 15/23 Page 15/23 analysis perspective, the sensitivity to small stories could also avoid over-simplification and over- categorisation of only connecting the nuanced and complex experiences of older people’s emotional challenges to grand narratives of their life, but not through the lens of their own lived experiences and smaller contributing stories that can accumulatively significantly shape constructions of self. analysis perspective, the sensitivity to small stories could also avoid over-simplification and over- categorisation of only connecting the nuanced and complex experiences of older people’s emotional challenges to grand narratives of their life, but not through the lens of their own lived experiences and smaller contributing stories that can accumulatively significantly shape constructions of self. Ethical approval and consent to participate. This study was approved by the Social Sciences Research Ethics Committee (SSREC) at the University of Bath (SSREC reference S19-052). The study was conducted according to the guidelines of the Declaration of Helsinki. All participants provided their written informed consent prior to data collection and were informed that their anonymized data will be used in publications. Implications Further, biographical storytelling interviews can be conducted with cohort members to collate stories (especially small stories) that have not been picked up by previous qualitative data collections. Likewise, where longitudinal data is not yet available for a research topic, the biographical storytelling methodology outlined above may begin to allow researchers to draw links between lived life and current experience in a way which is not afforded by non-biographical approaches. Statement of funding This project was supported by a research grant (RE-ED1100) from Guild Living Ltd. The authors confirm that the funder played no role in the design, execution, analysis and interpretation of data, or writing of the study. Discussion We also believe that the practice of telling one’s life stories is not restricted solely to verbal forms (e.g., interviews or conversations) but has the potential to be developed through varied accessible and often creative forms depending older people’s preferences and capabilities. For example, the memory box approach as reported by Craig and Jacqueline in our study and other verbal and non-verbal mediums (e.g., music, painting, life-story writing, filming) may offer more diverse and participatory societal literacy to understand the circumstances, experiences and responses of older people’s emotional challenges and to further clarify their emotional needs as a biographical construct. Finally, the insightful responses by these older people have also highlighted the merits of extrapolating the biographical storytelling approach outside of this research paradigm and in a more societal setting. Page 16/23 Implications By illuminating how we explored older people’s emotional challenges in the context of their life histories, our study has promoted further questions about the applicability of this biographical storytelling method. The value of this method elucidated throughout this paper suggests that this method may be more broadly valuable in the paradigm of ageing studies. Despite our primary finding that older people’s emotional barriers, pains and responses may be rooted in their experiences during earlier life, we are aware of a risk in future qualitative inquires of mis- and/or over-interpreting the retrospective elements of biographical narratives and the impacts this may have on older people’s nuanced emotional experiences. This could be manifested in the application of a psycho-analytical lens that too heavily stresses causalities between certain emotional challenges in old age and traumatic experiences in early life. Instead of focusing on such a mechanical and prescriptive approach to linking specific and fragmented snapshots of experience between past and present, we feel it would be beneficial to embed more longitudinal elements into the biographical storytelling method to continuously follow up older people’s emotional experiences. For example, there is great potential to engage with cohort studies to use their longitudinal qualitative data often collected over years to explore the ongoing and dynamic construction of how cohort members deal with their emotional challenges as they grow older. Further, biographical storytelling interviews can be conducted with cohort members to collate stories (especially small stories) that have not been picked up by previous qualitative data collections. Likewise, where longitudinal data is not yet available for a research topic, the biographical storytelling methodology outlined above may begin to allow researchers to draw links between lived life and current experience in a way which is not afforded by non-biographical approaches. Instead of focusing on such a mechanical and prescriptive approach to linking specific and fragmented snapshots of experience between past and present, we feel it would be beneficial to embed more longitudinal elements into the biographical storytelling method to continuously follow up older people’s emotional experiences. For example, there is great potential to engage with cohort studies to use their longitudinal qualitative data often collected over years to explore the ongoing and dynamic construction of how cohort members deal with their emotional challenges as they grow older. Declaration of contribution of authors Page 17/23 Page 17/23 Page 17/23 All three authors contributed to the conception and design of this article. The first author developed and drafted the overall structure of this article. The second author contributed to the writing up of the theoretical frameworks, reflexive explanations of analysis methods in the finding section as well as the discussion and implications. The third author contributed to developing the theoretical foundation of biographical storytelling and also helped shaped the qualitative findings that have inspired this article. All authors approved the final version to be submitted for publication. All three authors contributed to the conception and design of this article. The first author developed and drafted the overall structure of this article. The second author contributed to the writing up of the theoretical frameworks, reflexive explanations of analysis methods in the finding section as well as the discussion and implications. The third author contributed to developing the theoretical foundation of biographical storytelling and also helped shaped the qualitative findings that have inspired this article. All authors approved the final version to be submitted for publication. Availability of Data The datasets used in this article are available from the corresponding author upon reasonable request. The datasets used in this article are available from the corresponding auth Acknowledgements We would like to acknowledge the following people and organisations for their kind support of this project: (1) Professor Malcolm Johnson, (2) Guild Living, (3) Dr Kate Barnett, (4) the research teams in the UK and Australia, (5) Life Care, Australia, (6) Inspired Villages Group, UK, and (7) he participants who kindly shared their life stories with us. Statement of conflict of interest The authors declare none. Consent for publication ll participants gave permission to use their anonymized data for publication References 1. Carr S, Fang C. A gradual separation from the world: A qualitative exploration of existential loneliness in old age. Ageing and Society. 2021;1-21. 1. Carr S, Fang C. A gradual separation from the world: A qualitative exploration of existential loneliness in old age. Ageing and Society. 2021;1-21. 2. Grossmann I, Karasawa M, Kan C, Kitayama S. A cultural perspective on emotional experiences across the life span. Emotion. 201;14(4), 679–92. 3. Charles S, Carstensen L. Social and emotional aging. Annual review of psychology. 2010;61:383– 409. 4. Fang C, Carr S. 'They're Going to Die at Some Point, but We're all Going to Die' - A Qualitative Exploration of Bereavement in Later Life. Omega. 2021;1-19. 4. Fang C, Carr S. 'They're Going to Die at Some Point, but We're all Going to Die' - A Qualitative Exploration of Bereavement in Later Life. Omega. 2021;1-19. 5. Rodda J, Walker Z, Carter J. Depression in older adults. BMJ. 2011;343(d5219). 5. Rodda J, Walker Z, Carter J. Depression in older adults. BMJ. 2011;343(d5219). Page 18/23 Page 18/23 6. Victor C, Bowling A. A Longitudinal Analysis of Loneliness Among Older People in Great Britain. The Journal of Psychology. 2012;146(3):313-31. 7. Gale C, Westbury L, Cooper C. Social isolation and loneliness as risk factors for the progression of frailty: the English Longitudinal Study of Ageing. Age and Ageing. 2018;47(3):392–7. doi: 10.1093/ageing/afx188 8. Erikson E, Erikson J. The life cycle completed (extended version). WW Norton & Company. 1998. 9. Carpentieri J, Elliott J. Understanding healthy aging using a qualitative approach: The value of narratives and individual biographies. In D. Kuh, et al. (Eds.), A life course approach to healthy ageing . Oxford: Oxford University Press. 2014;118–29 10. Plummer K. Documents of Life 2: An Invitation to a Critical Humanism. 2nd edition. London: Sage. 2001. 11. McAdams D. The redemptive self: Stories Americans live by (revised and expanded edition). New York: Oxford University Press. 2013. 12. Benjamin-Thomas T, Corrado A, McGrath C, Rudman D, Hand C. Working towards the promise of participatory action research: learning from ageing research exemplars. International Journal of Qualitative Methods. 2018;17:1-13. 13. Bryanton O, Weeks L, Townsend E, Montelpare W, Lees J, Moffatt L. The utilization and adaption of photovoice with rural women aged 85 and older. International Journal of Qualitative Methods: 2019;18:1-9. 14. Phillipson L, Hammond A. More than talking: a scoping review of innovative approaches to qualitative research involving people with dementia. References International Journal of Qualitative Methods. 2018;17:1-13. 15. Phoenix C, Sparkes A. Being Fred: big stories, small stories and the accomplishment of a positive ageing identity. Qualitative Research. 2009;9(2):219–36. 16. Scott K, DeBrew J. Helping older adults find meaning and purpose through storytelling. Journal of gerontological nursing. 2009;35(12):38–43. 16. Scott K, DeBrew J. Helping older adults find meaning and purpose through storytelling. Journal of gerontological nursing. 2009;35(12):38–43. 17. Westerhof G. Life Review and Life-Story Work. In S. Whitbourne, (Ed.). The Encyclopedia of Adulthood and Aging. 2015 18. Brotman S, Ferrer I, Koehn S. Situating the life story narratives of aging immigrants within a structural context: the intersectional life course perspective as research praxis. Qualitative Research. 2020;20(4):465–84. 19. Clough P, Goodley D, Lawthom R, Moore M. Researching life stories: Method, theory and analyses in a biographical age. Routledge.2004 20. Rice C, LaMarre A, Changfoot N, Douglas P. Making spaces: multimedia storytelling as reflexive, creative praxis, Qualitative Research in Psychology. 2020;17(2):222-39. 21. Koch T. Story telling: is it really research? Journal of Advanced Nursing. 1998;28:1182-90. Page 19/23 22. Mills C. Situated Actions and Vocabularies of Motive. American Sociological Review. 1940;5(6):904– 13. 23. McAdams D, Guo J. How Shall I Live? Constructing a Life Story in the College Years. New Directions for Higher Education. 2014;15-23. 24. Smith D, Schlaepfer P, Major K, et al. Cooperation and the evolution of hunter-gatherer storytelling. Nature Communications.2017;8:1853. 25. Mager B. Storytelling contributes to resilience in older adults, Activities, Adaptation & Aging. 2019;43(1):23-36. 26. Bluck S. Autobiographical memory: Exploring its functions in everyday life. Memory. 2003;11:113– 23. 27. Rodriguez A, Smith J. Phenomenology as a healthcare research method. Evidence-Based Nursing. 2018;21:96-8. 28. Scott S. Making sense of everyday life. Cambridge: Polity. 2009. 28. Scott S. Making sense of everyday life. Cambridge: Polity. 2009. 29. Sjöberg M, Beck I, Rasmussen B, Edberg, A. Being disconnected from life: meanings of existential loneliness as narrated by frail older people. Aging & Mental Health. 2018;22(10):1357–64. 30. Johnson M. Biography and generation: spirituality and biographical pain at the end of life in old age. In Silverstein, RGM (ed.), Kinship and Cohort in an Aging Society: From Generation to Generation. Baltimore, MD: Johns Hopkins University Press. 2013; 176–90. In Silverstein, RGM (ed.), Kinship and Cohort in an Aging Society: From Generation to Generation. Baltimore, MD: Johns Hopkins University Press. 2013; 176–90. 31. Berger P, T Luckmann. References The social construction of reality: a treatise in the sociology of knowledge. London: Allen Lane. 1967. 31. Berger P, T Luckmann. The social construction of reality: a treatise in the sociology of knowledge. London: Allen Lane. 1967. 32. Stryker S, Burke P. The Past, Present, and Future of an Identity Theory. Social Psychology Quarterly. 2000;63(4):284–97. 33. Davies B, Harré R. Positioning: The Discursive Production of Selves. Journal for the Theory of Social Behaviour. 1990;20:43-63. 34. Taylor S, Littleton K. Biographies in talk: A narrative-discursive research approach. Qualitative Sociology Review. 2006;2(1):22–38. 35. Roseneil S. Using biographical narrative and life story methods to research women's movements: FEMCIT, Women's Studies International Forum 2012;35(3):129-31. 36. van Tilburg T. Social, Emotional, and Existential Loneliness: A Test of the Multidimensional Concept. The Gerontologist. 2021;61(7):e335–44. 37. Thompson P. The voice of the past: Oral history. Oxford: Oxford University Press. 2000. 37. Thompson P. The voice of the past: Oral history. Oxford: Oxford Universit 38. Kvale S, Brinkmann S. Interviews: Learning the Craft of Qualitative Research Interviewing. Thousand Oaks, CA: Sage. 2009. 39. Rodriguez D. Storytelling in the field: Race, method, and the empowerment of Latina college students. Cultural Studies? Critical Methodologies. 2010;10(6):491-507. 40. Braun V, Clarke V. Thematic analysis. In H. Cooper, P. M. Camic, D. L. Long, A. T. Panter, D. Rindskopf, & K. J. Sher (Eds.), APA handbook of research methods in psychology, Vol. 2. Research designs: Page 20/23 Quantitative, qualitative, neuropsychological, and biological. American Psychological Association. 2012; 57-71. Quantitative, qualitative, neuropsychological, and biological. American Psychological Association. 2012; 57-71. 41. Ettema E, Derksen L, van Leeuwen E. Existential loneliness and end-of-life care: A systematic review. Theoretical Medicine and Bioethics. 2010;31(1):141–69. 41. Ettema E, Derksen L, van Leeuwen E. Existential loneliness and end-of-life care: A systematic review. Theoretical Medicine and Bioethics. 2010;31(1):141–69. 42. Rosenthal G. Reconstruction of life stories: principles of selection in generating stories for narrative biographical interviews. The Narrative Study of Lives. 1993;1(1):59-91. 42. Rosenthal G. Reconstruction of life stories: principles of selection in generating stories for narrative biographical interviews. The Narrative Study of Lives. 1993;1(1):59-91. 43. Rosenthal G. Biographical research. In C. Seale, G. Gobo, J. F. Gubrium, & D. Silverman (Eds.), Qualitative research practice. London: Sage. 2004;48-64. 43. Rosenthal G. Biographical research. In C. Seale, G. Gobo, J. F. Gubrium, & D. Silverman (Eds.), Qualitative research practice. London: Sage. 2004;48-64. 44. van Kessel G. References The ability of older people to overcome adversity: a review of the resilience concept. Geriatric nursing. 2013;34(2):122–7. 44. van Kessel G. The ability of older people to overcome adversity: a review of the resilience concept. Geriatric nursing. 2013;34(2):122–7. 45. McLean K, Syed M. The Oxford handbook of identity development. Oxford: Oxford University Press. 2015. 45. McLean K, Syed M. The Oxford handbook of identity development. Oxford: Oxford University Press. 2015. 46. van Wijngaarden E, Leget C, Goossensen A. Ready to give up on life: The lived experience of elderly people who feel life is completed and no longer worth living. Social Science & Medicine.2015;138:257–64. 46. van Wijngaarden E, Leget C, Goossensen A. Ready to give up on life: The lived experience of elderl people who feel life is completed and no longer worth living. Social Science & Medicine.2015;138:257–64. 47. Lloyd L, Calnan M, Cameron A, Seymour J, Mmith R. Identity in the fourth age: Perseverance, adaptation and maintaining dignity. Ageing and Society. 2014;34(1):1-19. 47. Lloyd L, Calnan M, Cameron A, Seymour J, Mmith R. Identity in the fourth age: Perseverance, adaptation and maintaining dignity. Ageing and Society. 2014;34(1):1-19. Figure 2 An illustration of the coding process Appendix1.docx Figures Page 21/23 Figure 1 theoretical construction of human biography Page 22/23 Figure 1 theoretical construction of human biography theoretical construction of human biography theoretical construction of human biography Page 22/23 Figure 2 An illustration of the coding process Supplementary Files This is a list of supplementary files associated with this preprint. Click to download. Appendix1.docx Supplementary Files This is a list of supplementary files associated with this preprint. Click to download. This is a list of supplementary files associated with this preprint. Click to d Appendix1.docx Appendix1.docx Page 23/23 Page 23/23 Page 23/23
https://openalex.org/W2050660554	https://www.scielo.br/j/pab/a/KCnZyY3jQH9tgbXBRqLT4fx/?lang=en&format=pdf	English	null	Genetic parameters and estimated genetic gains in young rubber tree progenies	Pesquisa Agropecuária Brasileira	2,013	cc-by	4,645	in young rubber tree progenies Cecília Khusala Verardi(1), Erivaldo José Scaloppi Junior(2), Guilherme Augusto Peres Silva(1), Lígia Regina Lima Gouvêa(1) and Paulo de Souza Gonçalves(1) (1)Instituto Agronômico, Caixa Postal 28, CEP 13012‑970 Campinas, SP, Brazil. E‑mail: ckverardi@yahoo.com.br, silva.gap@gmail.com, lgouvea@iac.sp.gov.br, paulog@iac.sp.gov (2)Agência Paulista de Tecnologia dos Agronegócios, Departamento de Descentralização do Desenvolvimento, Polo Regional Noroeste Paulista, Rodovia Péricles Belini, Km 121, Caixa Postal 61, CEP 15500‑970 Votuporanga, SP, Brazil. E-mail: scaloppi@apta.sp.gov.br Abstract – The objective of this work was to assess the genetic parameters and to estimate genetic gains in young rubber tree progenies. The experiments were carried out during three years, in a randomized block design, with six replicates and ten plants per plot, in three representative Hevea crop regions of the state of São Paulo, Brazil. Twenty‑two progenies were evaluated, from three to five years old, for rubber yield and annual girth growth. Genetic gain was estimated with the multi‑effect index (MEI). Selection by progenies means provided greater estimated genetic gain than selection based on individuals, since heritability values of progeny means were greater than the ones of individual heritability, for both evaluated variables, in all the assessment years. The selection of the three best progenies for rubber yield provided a selection gain of 1.28 g per plant. The genetic gains estimated with MEI using data from early assessments (from 3 to 5‑year‑old) were generally high for annual girth growth and rubber yield. The high genetic gains for annual girth growth in the first year of assessment indicate that progenies can be selected at the beginning of the breeding program. Population effective size was consistent with the three progenies selected, showing that they were not related and that the population genetic variability is ensured. Early selection with the genetic gains estimated by MEI can be made on rubber tree progenies. Index terms: Hevea brasiliensis, annual girth growth, early selection, effective population size, multi‑effect index, rubber yield. Parâmetros genéticos e ganhos de seleção estimados em progênies jovens de seringueira Resumo – O objetivo deste trabalho foi avaliar os parâmetros genéticos e estimar ganhos genéticos em progênies jovens de seringueira. Os experimentos foram realizados durante três anos, em delineamento de blocos ao acaso, com seis repetições e dez plantas por parcela, em três regiões representativas da heveicultura paulista. Vinte e duas progênies foram avaliadas, dos três aos cinco anos de idade, quanto à produção de borracha e incremento anual do caule. O ganho genético foi estimado com o índice multiefeitos (IME). A seleção pela média das progênies proporcionou maior ganho genético do que a seleção baseada em indivíduos, já que os valores de herdabilidade das médias de progênies foram maiores do que os da herdabilidade individual, para ambas as variáveis, em todos os anos de avaliação. A seleção das três melhores progênies quanto à produção de borracha proporcionou um ganho de seleção de 1,28 g por planta. Os ganhos genéticos estimados pelo IME com dados das primeiras avaliações (dos 3 aos 5 anos) foram altos em geral, para produção de borracha e incremento anual do caule. Os elevados ganhos genéticos obtidos para incremento anual no primeiro ano de avaliação indicam que as progênies podem ser selecionadas no início do programa de melhoramento. O tamanho efetivo da população foi condizente com as três progênies selecionadas e mostrou que elas não são aparentadas e que a variabilidade genética da população está garantida. A seleção precoce, com os ganhos genéticos estimados pelo IME, pode ser feita em progênies de seringueira. Termos para indexação: Hevea brasiliensis, crescimento anual do caule, seleção precoce, tamanho efetivo populacional, índice multiefeito, produção de borracha. Pesq. agropec. bras., Brasília, v.48, n.4, p.411-416, abr. 2013 DOI: 10.1590/S0100-204X2013000400009 Introduction have the main objective of uniting high yield and high vigor in a single individual. However, the breeding cycle of the species is very long, taking up to 30 years Breeding programs for rubber tree [Hevea brasiliensis (Willd. ex A. Juss.) Müell. Arg.] in Brazil Pesq. agropec. bras., Brasília, v.48, n.4, p.411-416, abr. 2013 DOI: 10.1590/S0100-204X2013000400009 412 C.K. Verardi et al. to achieve the final results. Therefore, research on young plants using early yield and plant vigor tests is very important. introduced in 1952 at the experimental center at the Instituto Agronômico, Campinas, SP, Brazil (Table 2). The experiments were set up in 2006, in a randomized block design, with 22 treatments (cross‑pollinated progenies), six replicates, and ten plants per plot, placed in a single row with 1.5x1.5 m spacing. introduced in 1952 at the experimental center at the Instituto Agronômico, Campinas, SP, Brazil (Table 2). The experiments were set up in 2006, in a randomized block design, with 22 treatments (cross‑pollinated progenies), six replicates, and ten plants per plot, placed in a single row with 1.5x1.5 m spacing. Knowledge of the genetic variability of a population is the basis for breeding and essential to estimate the genetic parameters, which describe the variation levels among progenies, the magnitude of genetic control of traits, and the expected gains with selection (Freitas et al., 2006). Researches in several parts of the world, such as Indonesia, India, Malaysia, and Brazil, have been trying to maximize genetic gains of rubber tree (Costa et al., 2008) using different selection techniques in order to accelerate the breeding programs. Rubber yield (g per tapping per plant) was obtained by the Hamaker Morris‑Mann test (HMM), modified for three‑year‑old seedlings (Dijkman, 1951), based on the average dry rubber yield of 30 tappings per plant. The tapping panel was opened 15 cm above the ground, using ½S d/3 tapping systems, with a total of 35 tappings, discarding the first five tapping samples which corresponded to the adaptation of the panel. According to Resende et al. (1999) the multi‑effect index (MEI) method is an excellent selection procedure, equivalent to the Blup (best linear unbiased prediction). MEI estimates uses the variances of the mixed linear model random effects when data are unbalanced or when survival is greater than 85%. Introduction The index is based on the multiplication of phenotypic values of individuals, plot means, family means, block means, and general means by the weighing coefficients of the indexes (heritability). These coefficients are calculated in order to maximize the correlation between heritability indexes and genetic values. This maximization is obtained from the regression between genetic values and phenotype values. The multiplication between these values and coefficients is made using a matrix system (Resende & Higa, 1994a). According to Resende & Higa (1994b), selection based on MEI maximizes the genetic progress of a trait, regardless of the experimental structure. Table 1. Characteristics of the experimental sites, in three municipalities of the state of São Paulo, Brazil. Characteristic Jaú Pindorama Votuporanga Latitude (S) 22°17' 21°13' 20°25' Longitude (W) 48°34' 48°56' 49°59' Altitude (m) 580 560 450 Mean annual temperature (ºC) 21.6 21.0 22.3 Mean annual precipitation (mm) 1,344 1,390 1,480 Soil type Ferralsol Acrisol Lixisol Table 1. Characteristics of the experimental sites, in three municipalities of the state of São Paulo, Brazil. Table 2. Name of the parental clones introduced from Southeast Asian to the Instituto Agronômico in 1952, which originated the 22 evaluated progenies. Table 2. Name of the parental clones introduced from Southeast Asian to the Instituto Agronômico in 1952, which originated the 22 evaluated progenies. Progeny Parental clone 1 Tjir 1 2 AVROS 49 3 C 290 4 GT 127 5 AVROS 255 6 C 318 7 PB 49 8 AVROS1328 9 AVROS 1126 10 AVROS 352 11 C 228 12 AVROS 363 13 RRIM 513 14 RRIM 600 15 C 256 16 PB 5/63 17 PR 107 18 C 297 19 GT 711 20 Tjir 16 21 PB 86 22 C 259 The objective of this work was to assess the genetic parameters and estimate genetic gains in young rubber tree progenies. Genetic parameters and estimated genetic gains 413 Selection gain (%) was calculated as Gs(%) = 100Gs/–x, in which: Gs is the genetic gain; and –x is the general mean The indication ½S stands for a half‑spiral cut, and d/3 indicates the interval between tappings, in this case, one tapping every three days. The multi‑effect index (MEI) was used to estimate the selection gains (Resende, 2002) by the following expression: Annual girth growth (cm per plant) was determined at 50 cm from the ground with a measuring tape. The analysis for individual variance was done according to the mathematical model that considers all the random effects, expressed by Yijk = m + pi + bj + eij + dijk, , in which: Yijk is the observation related to plant k in the j replicate, in the i progeny; m is the general mean; pi is the random effect of the ith progeny (i= 1, 2,..., p); bj is the effect of the j (j= 1, 2,..., b) replicate; eij is the effect of the random experimental error associated to the ij plot; dijk is the random deviation inherent to plant k within the ij plot. ˆI = ˆb1 Yijk + (ˆb2 - ˆb3) –Yi. + (ˆb3 - ˆb1) –Yij. - ˆb3 –Y.j. + (ˆb3 - ˆb2) –Y... in which: –Y... is the general mean of the experiment; Yijk is the individual value; –Yi. is the mean as a matrix in the experiment; –Yij. is the mean of the matrix in a determined block (plot mean); –Y.j. is the block mean; ˆb1 = ˆh2 d is the within plot heritability in the narrow sense; ˆh2 d = (3/4) ˆs2 A/ ˆs2 d; ˆb2 = ˆh2 m is the heritability in the narrow sense of the matrices ˆh2 m = [(3 + nb)/4nb]ˆs2 A/ˆs2 p + ˆs2 e/b + ˆs2 d/n; and ˆb3 = ˆh2 p is the plot heritability in the narrow sense, in which ˆh2 p = (3/4n) ˆs2 A/ˆs2 e + ˆs2 d /n. The variance estimates were obtained from the plot means in each location, following recommendations by Steel & Torrie (1980), analyzed by the genetic statistical program Selegen (Resende, 2002). Results and Discussion Individual heritability in narrow sense was estimated as ˆh2 r = ˆs2 a / (ˆs2 a + ˆs2 p + ˆs2 e), in which: ˆs2 a is the additive genetic variance; ˆs2 p is the genetic variance among progenies; ˆs2 e is the environmental variance among plots. Significant mean square for progenies show that they differed as to annual girth growth and rubber yield, and indicate that selection can be made with data already from the first year of assessment, except for rubber yield in Jaú, where the mean square for progenies was not significant in the first year (Table 3). These results agree with those of Gonçalves et al. (2004a, 2009) and Costa et al. (2006). Progeny mean heritability was calculated as ˆh2 mp = 0.25 ˆs2 a /(0.25 ˆs2 a + ˆs2 p /b + ˆs2 e /nb), in which: b is the number of blocks; n is the number of plots; ˆs2 a is the additive genetic variance; ˆs2 p is the genetic variance among progenies; ˆs2 e is the environmental variance among plots. The heritability values ranged from low to high for the evaluated variables (Table 4). Progeny mean heritability (ˆh2 mp) was higher than the individual heritability (ˆh2 r and ˆh2 ad); therefore, selection based on progeny mean would provide higher genetic gains than that based only on individuals. Gonçalves et al. (2005), in studies carried out in the same locations of the present one, with three‑year‑old progenies, and reported similar results, with greater (ˆh2 mp than ˆh2 r). These authors stated that high heritability estimates, together with a great variation within population, makes it more efficient to select for productive and vigorous progenies. Silva et al. (2012) also reported progeny mean‑based heritability greater than individual Within plot additive heritability was calculated as ˆh2 ad = 0.75 ˆs2 a /(0.75 ˆs2 a + ˆs2 e ), in which: ˆs2 p is the genetic variance among progenies; ˆs2 a is the additive genetic variance; ˆs2 e is the environmental variance among plots. Genetic parameters and estimated genetic gains p The effective population (Ne) was obtained according to Resende (2002) as Ne = (4Nf –kf)/[–kf + 3 + (ˆs2 kf /–kf)], in which: –kf is the mean number of selected individuals per matrix; ˆs2 kf is the variance of the number of selected individuals per matrix; Nf is the number of selected matrices. Within plot variances (s2 d) were calculated using individual plant data and averaging the individual plant square means. Genetic and phenotypic parameters were estimated using the genetic statistical software Selegen – Reml/ Blup (Model 93). Pesq. agropec. bras., Brasília, v.48, n.4, p.411-416, abr. 2013 DOI: 10.1590/S0100-204X2013000400009 Materials and Methods The study was done with a population of cross‑pollinated rubber tree progenies. It was set up in the regional research poles of the Agência Paulista de Agronegócios (Apta), in different rubber cropping regions of the state of São Paulo, Brazil, represented by the following municipalities: Jaú, Pindorama, and Votuporanga (Table 1). Cross‑pollinated progenies were obtained from parental clones from Southeast Asia germoplasm, Pesq. agropec. bras., Brasília, v.48, n.4, p.411-416, abr. 2013 DOI: 10.1590/S0100-204X2013000400009 Genetic parameters and estimated genetic gains Results and Discussion The annual girth growth had high selection gains, estimated with MEI, in all assessed years and locations Selecting 13.63% of the best progenies led to high selection gains, estimated with MEI, for rubber Table 4. Estimates of genetic parameters for annual girth growth and rubber yield of 22 rubber tree progenies assessed for three years, in three locations, in the state of São Paulo, Brazil. Table 3. Individual analysis for annual girth growth (cm per plant) and dry rubber yield (g per tapping per plant), of 22 rubber tree progenies assessed for three years, in Jaú, Pindorama, and Votuporanga, São Paulo state, Brazil, 2006. Table 3. Individual analysis for annual girth growth (cm per plant) and dry rubber yield (g per tapping per plant), of 22 rubber tree progenies assessed for three years, in Jaú, Pindorama, and Votuporanga, São Paulo state, Brazil, 2006. Variable Year Estimates ˆh2 r ˆh2 mp ˆh2 ad ˆrˆaa CVg CVr CVe Jaú Annual girth growth 1 0.73±0.15 0.82 0.86 0.91 12.13 0.88 13.82 2 0.13±0.06 0.55 0.11 0.74 10.35 0.45 22.74 3 0.31±0.09 0.77 0.27 0.88 10.50 0.74 14.05 Rubber yield 1 0.01±0.03 0.08 0.01 0.28 3.43 0.12 29.23 2 0.09±0.08 0.50 0.07 0.71 10.34 0.41 24.95 3 0.12±0.08 0.58 0.09 0.76 13.55 0.48 28.41 Pindorama Annual girth growth 1 0.60±0.12 0.84 0.60 0.91 8.91 0.92 9.67 2 0.14±0.06 0.47 0.13 0.68 11.21 0.38 29.33 3 0.39±0.10 0.84 0.33 0.92 12.16 0.93 13.13 Rubber yield 1 0.59±0.14 0.88 0.54 0.94 33.29 1.13 29.48 2 0.64±0.14 0.89 0.59 0.95 32.78 1.19 27.41 3 0.40±0.11 0.83 0.35 0.91 23.11 0.91 25.38 Votuporanga Annual girth growth 1 0.67±0.14 0.84 0.70 0.92 11.24 0.95 11.85 2 0.27±0.09 0.76 0.23 0.87 9.16 0.72 12.76 3 0.58±0.13 0.91 0.51 0.95 11.00 1.27 8.69 Rubber yield 1 0.34±0.11 0.84 0.27 0.91 24.06 0.93 25.87 2 0.29±0.10 0.82 0.24 0.91 23.60 0.87 27.11 3 0.19±0.08 0.75 0.15 0.87 19.91 0.71 28.21 ˆh2 r, individual heritability in the narrow sense; ˆh2 mp, progeny mean herita‑ bility; ˆh2 ad, within plot additive heritability; ˆrˆaa , accuracy; CVg (%), coef‑ ficient of genetic variation; CVr, index of relative variation (CVg/CVe); and CVe (%), coefficient of experimental variation. Pindorama, and Votuporanga, São Paulo state, Brazil, 2006. Results and Discussion Individual genetic coefficient of variation was calculated as CVg (%) = 100(ˆs2 a/–x)0.5, in which: ˆs2 a is the additive genetic variance; and –x is the general mean.i Within plot environmental coefficient of variation was calculated as CVe (%) = 100 [(c2 + ˆs2 parc /n)0.5/–x], in which: c2 is the plot coefficient of variation, ˆs2 parc is the plot variance; –x is the general mean. Pesq. agropec. bras., Brasília, v.48, n.4, p.411-416, abr. 2013 DOI: 10.1590/S0100-204X2013000400009 C.K. Verardi et al. 414 yield, using data from the three early tests (Table 5). However, low selection gains were observed in the first year of production in Jaú (3,92%), which was already expected since the observed heritabilities in this period (Table 4) were also low (ˆh2 mp = 0,08). Progeny selection in Jaú is recommended only from the third production year, when heritability value becomes high (59.52%). and within progeny heritabilities, for rubber yield and annual girth growth. Therefore, selecting by progeny mean seems to be a sound strategy to increase the genetic gain in rubber tree. The selective‑accuracy (ˆrˆaa) values were greater than 80%, within the indicated standard for the first and intermediate steps of a breeding program (Resende, 2007). High‑accuracy values are advantageous because they can provide high genetic gain values and increase selecting efficiency of progenies.i Genetic gains are based on the general mean of each year. In the present study, the improved mean was greater than the general mean, in all the locations and tests; thus, selecting the three best progenies would increase yield up to 1.28 g per plant. Because of the genetic gain increases over the assessment years, selection is recommended only at the end of the third test, in order to make the breeding programs more efficient. The coefficient of genetic variation (CVg%) and the experimental coefficient of variation (CVe%) (Table 4) were low, but within the expected range for rubber. Higher values were reported by Gonçalves et al. (2004b) and Gouvêa et al. (2012). The relative variation index (CVr) showed very high values, greater than 1.0 in some cases, indicating high genetic variability for the evaluated variables, in all the years and locations. These values also indicate a favorable situation for selection. Pesq. agropec. bras., Brasília, v.48, n.4, p.411-416, abr. 2013 DOI: 10.1590/S0100-204X2013000400009 (1)Progeny numbers related to identification in Table 2. Results and Discussion Selection gains, new means, and effective populational size (Ne) of selected individuals by the multi‑effect index, for annual girth growth (cm per plant) of 22 progenies assessed for three years, in Jaú, Pindorama, and Votuporanga, in the state of São Paulo, Brazil. Year Progeny(1) Gains New mean Ne Mean Jaú 1 6 0.03 0.54 1.00 0.51 19 0.02 0.53 2.00 22 0.02 0.53 3.00 2 19 0.33 1.30 1.00 0.98 15 0.30 1.28 2.00 19 0.29 1.27 2.48 3 17 0.84 2.10 1.00 1.26 19 0.76 2.04 2.00 19 0.75 2.01 2.48 Pindorama 1 20 1.37 2.06 1.00 0.69 20 1.35 2.04 1.60 20 1.30 2.01 2.00 2 4 1.91 2.76 2.76 0.85 15 1.77 2.62 2.61 20 1.71 2.56 2.56 3 19 1.70 2.93 1.00 1.23 5 1.62 2.84 2.00 15 1.54 2.76 3.00 Votuporanga 1 18 0.95 1.40 1.00 0.45 18 0.85 1.29 1.60 18 0.78 1.23 2.00 2 17 1.01 1.59 1.00 0.58 18 0.95 1.53 2.00 12 0.90 1.46 3.00 3 3 1.20 2.11 1.00 0.90 18 1.19 2.09 2.00 17 1.13 2.04 3.00 (1)Progeny numbers related to identification in Table 2 Year Progeny(1) Gains New mean Ne Mean Jaú 1 13 2.55 6.92 1.00 4.37 21 2.55 6.92 2.00 21 2.54 6.91 2.48 2 1 0.83 3.57 1.00 2.75 19 0.83 3.57 2.00 4 0.81 3.56 3.00 3 14 4.93 16.09 1.00 11.16 21 4.74 15.90 2.00 14 4.62 15.78 2.48 Pindorama 1 18 3.77 8.00 1.00 4.23 12 2.78 7.01 2.00 12 2.44 6.67 2.48 2 3 4.18 7.66 1.00 3.47 9 2.70 6.17 2.00 9 2.14 5.61 2.48 3 10 6.53 20.39 1.00 13.86 5 6.50 20.36 2.00 1 6.48 20.34 3.00 Votuporanga 1 11 3.15 7.17 1.00 4.01 17 3.04 7.05 2.00 1 2.79 6.81 3.00 2 14 1.89 7.22 1.00 5.32 3 1.86 7.18 2.00 21 1.83 7.15 3.00 3 5 9.52 26.45 1.00 16.93 5 9.40 26.33 1.60 18 9.26 26.19 2.48 (1)Progeny numbers related to identification in Table 2. Results and Discussion Variable Year Mean square Progeny Residual Within plot Jaú Annual girth growth 1 2.05 0.36 0.98 2 0.87 0.39 2.14 3 10.70 2.45 15.27 Rubber yield 1 0.02ns 0.02 0.12 2 0.12 0.06 0.44 3 0.30 0.12 0.92 Pindorama Annual girth growth 1 1.02 0.17 0.71 2 1.94* 1.04 3.49 3 20.37 3.31 25.82 Rubber yield 1 0.36 0.04 0.29 2 0.52 0.05 0.39 3 0.58 0.09 0.69 Votuporanga Annual girth growth 1 1.45 0.23 0.87 2 1.88 0.46 3.15 3 22.99 2.17 20.22 Rubber yield 1 0.08 0.01 0.13 2 0.14 0.02 0.24 3 0.26 0.06 0.63 nsNonsignificant. andSignificant by the F test, at 5 and 1% probability. Genetic parameters and estimated genetic gains 415 (Table 6). Due to the high genetic gain obtained with data already from the first year, the most vigorous progenies can be selected at the beginning of the breeding program. Early selection can reduce rubber tree breeding cycle, which is very long, taking from 25 to 30 years to reach the final results. The positive selection gains obtained in all the assessment years and locations allows for a more reliable early selection. The values reported here are compatible with other studies, such as those by Gonçalves et al. (2004b), which used a different methodology. other variables that contribute to yield increase, especially the annual girth growth gain, which is related to plant vigor. The effective population size was in agreement with the number of selected progenies (three), indicating that they were not related , thus ensuring the genetic variability of the population. The mean inbreeding coefficient was high with F=[1/(2Ne)100]= 17%, but this value can be reduced by increasing the number of selected individuals. Arantes et al. (2010) reported endogamy coefficient values lower (F=2.88%) than those observed in the present study, selecting the 20 best rubber tree progenies for rubber yield. A more effective selection must take yield into consideration and account for the performance of Table 5. Selection gains, new means, and effective population size (Ne) of selected individuals by the multi‑effect index method, for rubber yield (g per tapping per plant) of 22 progenies assessed for three years, in Jaú, Pindorama, and Votuporanga, in the state of São Paulo, Brazil. Table 6. (1)Progeny numbers related to identification in Table 2. Pesq. agropec. bras., Brasília, v.48, n.4, p.411-416, abr. 2013 , , , , p , DOI: 10.1590/S0100-204X2013000400009 Pesq. agropec. bras., Brasília, v.48, n.4, p.411-416, abr. 2013 DOI: 10.1590/S0100-204X2013000400009 Conclusions 1. Genetic variability for the studied variables in rubber tree progenies can be detected in young plants, which indicates favorable conditions for early selection. GONÇALVES, P. de S.; MARTINS, A.L.M.; BORTOLETTO, N.; SAES, L.A. Selection and genetic gains for juvenile trains in progenies of Hevea in São Paulo state, Brazil. Genetics and Molecular Biology, v.27, p.207‑214, 2004b. DOI: 10.1590/ S1415‑47572004000200014. 2. Selection is more efficient when based on the progeny mean than when based only on individuals. 2. Selection is more efficient when based on the progeny mean than when based only on individuals. 3. Early selection can be made on rubber tree progenies using the genetic gains estimated with multi‑effects index. GONÇALVES, P. de S.; MORAES, M.L.T. de; SILVA, M. de A.; GOUVÊA, L.R.L.; AGUIAR, A.T. da E.; COSTA, R.B. Prediction of Hevea progeny performance in the presence of genotype‑environment interaction. Brazilian Archives of Biology and Technology, v.52, p.25‑33, 2009. DOI: 10.1590/ S1516‑89132009000100004. References RESENDE, M.D.V. de. Genética biométrica e estatística no melhoramento de plantas perenes. Brasília: Embrapa Informação Tecnológica; Colombo: Embrapa Florestas, 2002. 975p. ARANTES, F.C.; GONÇALVES, P. de S.; SCALOPPI JUNIOR, E.J.; MORAES, M.L.T. de; RESENDE, M.D.V. de. Ganho genético com base no tamanho efetivo populacional de progênies de seringueira. Pesquisa Agropecuária Brasileira, v.45, p.1419‑1424, 2010. DOI: 10.1590/S0100‑204X2010001200012. RESENDE, M.D.V. de. Matemática e estatística na análise de experimentos e no melhoramento genético. Colombo: Embrapa Florestas, 2007. 561p. COSTA, R.B. da; RESENDE, M.D.V. de; GONÇALVES, P. de S.; CHICHORRO, J.F.; ROA, R.A.R. Variabilidade genética e seleção para caracteres de crescimento da seringueira. Bragantia, v.67, p.299‑305, 2008. DOI: 10.1590/ S0006‑87052008000200005. RESENDE, M.D.V. de; FERNANDES, J.S.C.; SIMEÃO, R.M. Blup individual multivariado em presença de interação genótipo x ambiente para delineamentos experimentais repetidos em vários ambientes. Revista de Matemática e Estatística, v.17, p.209‑228, 1999. COSTA, R.B. da; RESENDE, M.D.V. de; GONÇALVES, P. de S.; REGO, F.L.H.; CONTINI, A.Z.; ROA, R.A.R. Genetic evaluation of Hevea brasiliensis [(Willd ex Adr. de Juss.) Müell. Arg.] for juvenile vigour using the Reml/Blup method. Floresta e Ambiente, v.12, p.8‑14, 2006. RESENDE, M.D.V. de; HIGA, A.R. Estimação de valores genéticos no melhoramento de Eucalyptus: seleção em um caráter com base em informações do indivíduo e de seus parentes. Boletim de Pesquisa Florestal, v.28/29, p.11‑35, 1994a. RESENDE, M.D.V. de; HIGA, A.R. Maximização da eficiência da seleção em testes de progênies de Eucalyptus através da utilização de todos os efeitos do modelo matemático. Boletim de Pesquisa Florestal, v.28/29, p.37‑55, 1994b. DIJKMAN, M. J. Hevea: thirty years of research in Far East Florida. Miami: University of Miami, 1951. 329p. FREITAS, M.L.M.; SEBBENN, A.M.; MORAIS, E.; ZANATTO, A.C.S.; VERARDI, C.K.; PINHEIRO, A.N. Parâmetros genéticos em progênies de polinização aberta de Cordia trichotoma (Vell.) ex Steud. Revista do Instituto Florestal, v.18, p.95‑102, 2006. SILVA, G.A.P.; GOUVÊA, L.R.L.; VERARDI, C.K.; RESENDE, M.D.V. de; SCALOPPI JUNIOR, E.J.; GONÇALVES, P. de S. Genetic parameters and correlation in early measurement cycles in rubber trees. Euphytica, v.189, p.343‑350, 2012. DOI: 10.1007/ s10681‑012‑0751‑8. GONÇALVES, P. de S.; BORTOLETTO, N.; COSTA, R.B. da; SILVA, M. de A.; CARDINAL, Á.B.B.; GOUVÊA, L.R.L. Estimates of annual genetic parameters and expected gains in the second cycle of Hevea genotype selection. Crop Breeding and Applied Biotechnology, v.4, p.416‑421, 2004a. STEEL, R.G.D.; TORRIE, J.H. Principles and procedures of statistics: a biometrical approach. 2nd ed. New York: McGraw‑Hill, 1980. 633p. Acknowledgments To Fundação de Amparo à Pesquisa do Estado de São Paulo (Fapesp), for financial support and scholarship. GOUVÊA, L.R.L.; SILVA, G.A.P.; VERARDI, C.K.; SILVA, J.Q.; SCALOPPI‑JUNIOR, E.J.; GONÇALVES, P. de S. Temporal stability of vigor in rubber tree genotypes in the pre‑ and post‑tapping phases using different methods. Euphytica, v.186, p.625‑634, 2012. DOI: 10.1007/s10681‑012‑0688‑y. Pesq. agropec. bras., Brasília, v.48, n.4, p.411-416, abr. 2013 DOI: 10.1590/S0100-204X2013000400009 Results and Discussion Year Progeny(1) Gains New mean Ne Mean Jaú 1 6 0.03 0.54 1.00 0.51 19 0.02 0.53 2.00 22 0.02 0.53 3.00 2 19 0.33 1.30 1.00 0.98 15 0.30 1.28 2.00 19 0.29 1.27 2.48 3 17 0.84 2.10 1.00 1.26 19 0.76 2.04 2.00 19 0.75 2.01 2.48 Pindorama 1 20 1.37 2.06 1.00 0.69 20 1.35 2.04 1.60 20 1.30 2.01 2.00 2 4 1.91 2.76 2.76 0.85 15 1.77 2.62 2.61 20 1.71 2.56 2.56 3 19 1.70 2.93 1.00 1.23 5 1.62 2.84 2.00 15 1.54 2.76 3.00 Votuporanga 1 18 0.95 1.40 1.00 0.45 18 0.85 1.29 1.60 18 0.78 1.23 2.00 2 17 1.01 1.59 1.00 0.58 18 0.95 1.53 2.00 12 0.90 1.46 3.00 3 3 1.20 2.11 1.00 0.90 18 1.19 2.09 2.00 17 1.13 2.04 3.00 (1)Progeny numbers related to identification in Table 2. Year Progeny(1) Gains New mean Ne Mean Jaú 1 13 2.55 6.92 1.00 4.37 21 2.55 6.92 2.00 21 2.54 6.91 2.48 2 1 0.83 3.57 1.00 2.75 19 0.83 3.57 2.00 4 0.81 3.56 3.00 3 14 4.93 16.09 1.00 11.16 21 4.74 15.90 2.00 14 4.62 15.78 2.48 Pindorama 1 18 3.77 8.00 1.00 4.23 12 2.78 7.01 2.00 12 2.44 6.67 2.48 2 3 4.18 7.66 1.00 3.47 9 2.70 6.17 2.00 9 2.14 5.61 2.48 3 10 6.53 20.39 1.00 13.86 5 6.50 20.36 2.00 1 6.48 20.34 3.00 Votuporanga 1 11 3.15 7.17 1.00 4.01 17 3.04 7.05 2.00 1 2.79 6.81 3.00 2 14 1.89 7.22 1.00 5.32 3 1.86 7.18 2.00 21 1.83 7.15 3.00 3 5 9.52 26.45 1.00 16.93 5 9.40 26.33 1.60 18 9.26 26.19 2.48 (1)Progeny numbers related to identification in Table 2. ropec. bras., Brasília, v.48, n.4, p.411-416, abr. 2013 DOI: 10.1590/S0100-204X2013000400009 C.K. Verardi et al. 416 GONÇALVES, P. de S.; CARDINAL, A.B.B.; COSTA, R.B. da; BORTOLETTO, N.; GOUVÊA, L.R.L. Genetic variability and selection for laticiferous system characters in Hevea brasiliensis. Genetics and Molecular Biology, v.28, p.414‑422, 2005. DOI: 10.1590/S1415‑47572005000300014. Conclusions References Received on July 16, 2012 and accepted on March 28, 2013
https://openalex.org/W4313487973	https://nottingham-repository.worktribe.com/file/20003741/1/Self-management%20interventions%20for%20children%20and%20young%20people%20with%20sickle%20cell%20disease%3A%20A%20systematic%20review	English	null	Self‐management interventions for children and young people with sickle cell disease: A systematic review	Health expectations	2,023	cc-by	22,232	R E V I E W A R T I C L E Received: 5 September 2022 \| Revised: 3 December 2022 \| Accepted: 11 December 2022 Received: 5 September 2022 \| Revised: 3 December 2022 \| Accepted: 11 December 2022 DOI: 10.1111/hex.13692 This is an open access article under the terms of the Creative Commons Attribution License, which permits use, distribution and reproduction in any medium, provided the original work is properly cited. © 2023 The Authors. Health Expectations published by John Wiley & Sons Ltd. Health Expectations. 2023;26:579–612. Brenda A. Poku PhD, Research Associate1 \| Karl Michael Atkin PhD, Professor of Sociology2 \| Sue Kirk PhD, Professor of Family and Child Health1 1Division of Nursing, Midwifery and Social Work, School of Health Sciences, University of Manchester, Manchester, UK 2Department of Sociology, University of York, York, UK 1Division of Nursing, Midwifery and Social Work, School of Health Sciences, University of Manchester, Manchester, UK 1Division of Nursing, Midwifery and Social Work, School of Health Sciences, University of Manchester, Manchester, UK 2Department of Sociology, University of York, York, UK Abstract Background: Increasing numbers of interventions are being developed to support self‐management for children and young people (CYP) with sickle cell disease (SCD), but no systematic review has systematically synthesized this evidence regarding their characteristics, effectiveness, acceptability and feasibility for all published intervention types. Abstract 2Department of Sociology, University of York, York, UK Correspondence Brenda A. Poku, PhD, Research Associate, Division of Nursing, Midwifery and Social Work, School of Health Sciences, University of Manchester, Jean McFarlane Bldg, Oxford Rd, Manchester M13 9PL, UK. E il b d k @ h t k Methods: The Joanna Briggs Institute guidelines for mixed‐method reviews were followed. A systematic search of eight databases and key journals was conducted from their inception to November 2021. Primary research of self‐management interventions targeting CYP with SCD aged 8–24 years and reporting any health/ social outcome and acceptability data were included. Design‐specific standardized critical appraisal instruments were used. Two independent reviewers screened and appraised the articles. A third reviewer resolved disagreements. Funding information National Institute for Health and Care Research, Grant/Award Number: NIHR202169 K E Y W O R D S K E Y W O R D S children, interventions, self‐care, self‐management, sickle cell disease, strategies, young people 2 \| METHODS iley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License The Joanna Briggs Institute (JBI) guidelines for conducting mixed‐ method reviews22 informed the conduct of this review. The review protocol was registered (PROSPERO: CRD42021286422). The review is reported according to the expanded Preferred Reporting Items for Systematic Reviews and Meta‐Analyses (PRISMA) reporting checklist.23 1 \| INTRODUCTION Previous reviews have focused on specific psychological,18 medication adherence,19 transition programmes20 and e‐health interventions,21 and assessed their effects on particular out- comes. No systematic review has been identified that focused on all types of self‐management interventions for CYP with SCD and included the full range of health and social outcomes that have been evaluated. This review aimed to: (1) summarize the range and characteristics of self‐management interventions for CYP with SCD; and (2) critically evaluate the effectiveness, acceptability and feasibility of different self‐management inter- ventions, identifying factors influencing their acceptability and implementation. 1 Sickle cell disease (SCD) is globally the most common paediatric genetic blood disorder. Approximately 300,000 babies are born with the condition annually, with a projected 30% increase in incidence by 2050.1 Most births occur in Africa, but annually an estimated 2000 and 300 babies are born with SCD in the United States and the United Kingdom, respectively.2 Alterations in the shape of red blood cells characterize SCD, leading to blood vessel occlusion and inflammation, infarction, organ damage, pain and profound chronic anaemia.3 Consequently, children born with SCD experience comorbidities, have high healthcare needs and reduced life expectancy.4,5 With medical advances, mainly in high‐income countries, children born with SCD transition into adulthood. However, SCD is associated with significant medical and psychosocial challenges that often worsen when children and young people (CYP) reach an age to become responsible for managing their condition. As CYP with SCD mature, they have increased decision‐making autonomy and self‐care independence. Self‐management is the purposeful performance of specific learned tasks, activities and behaviours to manage the medical, psychosocial and life impact of a chronic illness.6‐8 SCD self‐ management strategies include symptom monitoring, following treatment plans, and ensuring health maintenance practices. Self‐ management is crucial as CYP with SCD begin caring for themselves and their illness. Funding information Results: Of 1654 articles identified, 38 studies were included. Methodological quality was moderate. Most studies evaluated SCD education, psycho‐behavioural, psychosocial and skills training and/or social support interventions. They appear to demonstrate short‐term improvements in knowledge, social functioning and medical adherence outcomes. Interventions that were multifaceted in content, combined technological platforms and in‐person group‐based formats and involved peers, family and care providers were more acceptable and effective. The long‐term impact of interventions was limited, including CYP's involvement in the intervention development and implementation. Conclusions: There is inconclusive evidence for any self‐management programme. Nonetheless, support from family, peers and care providers appears to be important for self‐management interventions' effectiveness and acceptability. Future research needs to prioritize CYP involvement in both intervention design and delivery, their wider social context and include CYP with SCD from non‐Black backgrounds. wileyonlinelibrary.com/journal/hex Health Expectations. 2023;26:579–612. 579 580 \| 580 POKU ET AL. Patient and Public Contribution: Three young people with SCD recruited acted as the review advisors. They were formally trained in the review process and involved in every aspect of the review: the design, conduct and interpretation of the findings. CYP involvement in the interventions' development and implementation was analysed as part of the review. This systematic review was conducted as part of a wider research project titled: Understanding fatigue experiences of CYP with SCD to guide the co‐development of a fatigue self‐management intervention. Two of the young advisors involved in the review were also involved in the development of the project funding application. 2.1 \| Search strategy First, eight databases: Ovid MEDLINE, EMBASE, PsycINFO and EBM Reviews, ASSIA, CINAHL, Web of Science and Engineering Village were searched from their inception to November 2021. A comprehensive search strategy was developed and modified appropriately across the databases to reflect the various MeSH terms. The following main concepts: ‘child’, ‘young people’, ‘sickle cell disease’, ‘self‐management’, ‘self‐care’ and ‘intervention’ guided the development of the search strategy, with support from a medical reference librarian. See Supporting Information: File 1 for full search strategies applied across the databases. Second, searches SCD‐related management strategies can negatively impact CYP's perceived quality of life, disrupt school and work attend- ance and participation and influence their social interactions and relationships.9‐15 This may reduce commitment to treatment plans and health maintenance practices.16 Effective self‐ management support can help mitigate treatment challenges, engage CYP, improve health and social outcomes and reduce illness burden and health costs.16,17 Consequently, recent efforts have focused on developing and implementing interventions to support and improve self‐management. POKU ET AL. 581 2.3 \| Quality assessment Included papers were appraised for methodological quality indepen- dently by B. A. P. and S. K. using design‐specific standardized critical appraisal instruments from the JBI System for the Unified Manage- ment, Assessment and Review of Information (SUMARI).24 For each of the appraisal criteria, the individual appraisers assessed whether TABLE 1 Eligibility criteria PICOS framework Inclusion criteria Population CYP aged 8–25 years with SCD. The upper age limit is consistent with definitions of young people by the UN, WHO and UK NHS. The lower age limit is informed by the evidence that children with long‐term health conditions may begin to engage (informally) in self‐management around this age.8 Papers including samples described as CYP without further specificity around age. Papers involving samples primarily encompassing CYP with a reported mean age between 8 and 25 years. Papers including children and/or young people with long‐term conditions that reported findings for CYP with SCD. Intervention We defined self‐management as purposeful performance of specific learned tasks, activities and behaviours to manage the medical, psychosocial and life impact of a chronic illness.6‐8 And we defined a self‐management intervention as any intervention, programme or approach designed to develop the ability of and/or support CYP with SCD to manage their long‐term health condition through education, training and support to develop their knowledge, skills or psychosocial resources. All self‐management interventions that target CYP with SCD. All types/forms of self‐management intervention—health, social care and educational interventions—designed to support or facilitate CYP with SCD to take control of and manage their condition, promote their capacity for self‐care and/or maintain and enhance their physical and/or mental health. Comparator Usual care Another self‐management intervention(s) No comparator Outcome No restrictions Study design Primary research, including qualitative, mixed‐methods and quantitative studies of all designs. Evaluations and discursive articles about SCD self‐management interventions. Conference papers that present detailed information about the intervention, research methods and outcome(s). Abbreviations: CYP children and young people; SCD sickle cell disease Exclusion criteria S. h Only included participants over 25 years old. Mean age > 25 years. 2.2 \| Study selection of online abstract archives of key paediatric and haematology journals (Blood, Journal of Pediatric Psychology, Journal of Pediatric Hematology Oncology, Pediatric Blood Cancer, Journal of Adolescent Health), trial registries (ISRCTN, ClinicalTrials.gov), review registries (PROSPERO, Cochrane Database of Systematic Reviews, JBI Systematic review Register) internet‐based resources (Google, Google Scholar) and OpenGrey from their inception to November 2021 supplemen- ted database searches. Third, reference lists of included studies and systematic review18‐21 also provided sources for studies not identified in the database searches. The 3‐step search strategy was implemented by one reviewer (B. A. P.). Authors of recent and ongoing studies were contacted for available publications and/or unpublished data. The review included evidence that met the eligibility criteria detailed in Table 1. The search was unrestricted by country or setting, language or publication year. All citation records from the database and manual searches were exported into EndNote 20® and deduplicated. Records were independently screened in Covidence® by two reviewers (B. A. P. and S. K.). Any disagreements were adjudicated by the third author (K. M. A.). Reasons for excluding papers were noted. Abbreviations: CYP, children and young people; SCD, sickle cell disease. 3.2 \| Methodological quality of included studies Overall, the body of evidence was of moderate methodological quality (Tables 3–7). Several of the randomised controlled trial (RCTs) (n = 7) were assessed as being of poor quality. Four were feasibility RCTs with poor reporting quality. Only four RCTs were rated as high quality, three of which were feasibility/pilot RCTs. The RCT studies lacked details of randomization methods, group allocations, blinding of assessors, group comparativeness and statistical analysis methods. The limitations of the quasi‐experimental studies were their short follow‐up length, the multiplicity of outcome measurements and the reliability of the outcome measurement tools used. The cohort and cross‐sectional studies failed to report on how confounding factors were managed in the analysis. Only two quantitative study article46,64 sreported sample size calculation or provided reasons for the chosen sample sizes. While all articles reported the number of participants Overall, the body of evidence was of moderate methodological quality (Tables 3–7). Several of the randomised controlled trial (RCTs) (n = 7) were assessed as being of poor quality. Four were feasibility RCTs with poor reporting quality. Only four RCTs were rated as high quality, three of which were feasibility/pilot RCTs. The RCT studies lacked details of randomization methods, group allocations, blinding of assessors, group comparativeness and statistical analysis methods. 2.5 \| Data synthesis Results from the included papers were synthesized narratively in line with the JBI approach. As only one qualitative research paper was included, it was impossible to follow the JBI convergent integrated approach for mixed‐method review. The characteristics of the different interventions were examined and summarized in relation to their effectiveness, acceptability and feasibility for implementation in ‘real life’ settings. 3.1 \| Study and sample characteristics A typology from previous studies on self‐management interventions guided data extraction on intervention characteristics.25,26 Data were extracted on the following dimensions: study aims and objectives, study location, publication year, research design, participant demo- graphics, intervention characteristics and all outcomes measured for effectiveness and acceptability, as well as reported feasibility and implementation issues. For quantitative studies, extracted outcomes comprised descriptive and/or inferential statistical results and themes from the analysis of open‐ended survey questions. For the qualitative studies, data extracted included themes with corresponding quotes. One reviewer (B. A. P.) undertook data extraction, whilst a second reviewer (S. K.) independently extracted data from 10% of all the included papers as a measure to ensure quality control and consistency. Additional information was located and extracted from published protocols/formative papers linked to the included studies. See Supporting Information: File 2 for the detailed data extraction domains. Table 2 presents a summary of the included studies. The 38 articles published between 1987 and 2021 were predominantly from the United States (n = 34, 89.5%). More than half (n = 26, 68.4%) were published within the last 5 years. There was a preponderance of relatively small‐scale (e.g., single‐site, small‐sample) feasibility/pilot studies (n = 21)28,34‐36,40,44,46‐51,55,57,58,60‐64 whose primary objective was to report effectiveness data, acceptability data and/or assess the feasibility and trial design of interventions. The majority (n = 31) of the included studies used a randomized30,31,37,42‐44,46‐50,54,55,58,64 or non‐randomized comparative study design27,28,32,38,39,41,45,51,52,56,57,59‐63 in which self‐ management interventions were either compared with standard care37,42,45,47‐49 or specific control interventions.30,36,44,46,54,55,58 Most studies (n = 24) evaluated self‐management interventions targeting CYP, ten targeted CYP‐parent dyads27,33,44,48‐51,58,60,62 and four targeted family units.30,37,47,55 Most studies (n = 23) included self‐selected convenience samples, with sizes ranging from 534 to 11029 participants recruited mainly from single sites. Nine studies recruited participants from multiple settings.27,37,41,44,48,50,54,57,58 CYP participants were aged between 6 and 24 years, predominantly African American and had homozygous SCD (i.e., HbSS genotype). Exclusion criteria Only included participants over 25 years old. Mean age > 25 years. Interventions in which children and/or young people are not actively engaged and/or remain passive recipients of knowledge or instructions. Self‐management interventions that target professionals, parents/carers/siblings or families as a whole, without a distinct component(s) for CYP. Papers only describing the development of an intervention. Papers where the effects of the self‐management intervention cannot be distinguished from broader interventions for SCD. Dissertations/theses Secondary research (secondary data analysis and literature reviews) Case studies Posters/conference abstracts and proceedings 582 POKU ET AL. the article met each quality criterion (‘yes’), if it failed to meet the criterion (‘no’), if insufficient information was presented to assess the specific criterion adequately (‘unclear’) or if the criterion was not applicable due to the study type, design and/or intervention (‘N/A’) (Tables 3–7). The total number of appraisal criteria met determined the overall quality of a paper. Papers that met three‐quarters or more of the appraisal criteria were considered to be of high quality, between half and three‐quarters as moderate quality and less than half as poor quality (Tables 3–7). The quality appraisal did not impact study inclusion or exclusion but was considered in the synthesis. recommendations after reviewing the extracted and synthe- sized data. rom https://onlinelibrary.wiley.com/doi/10.1111/hex.13692 by Test, Wiley Online Library on [02/05/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons Licens 3 \| RESULTS A total of 1654 records were imported into EndNote®. After deduplicating the records, 1024 were screened in Covidence®; 38 (representing 32 distinct studies) met all the inclusion criteria and were included in the final review (Figure 1). Of these, 37 were quantitative studies27‐63 and 1 was a qualitative study.64 FIGURE 1 PRISMA flow diagram 2.6 \| Patient and Public Involvement (PPI) The review team worked with a PPI group of three young people with SCD recruited from a patient support group. They were trained and involved at every review stage to ensure that patient perspectives (and concerns) were reflected in the review. They commented on the search strategy and the data extraction dimensions developed by the reviewers, helping to identify additional search terms and quality criteria. In addition, they highlighted issues for discussion and The limitations of the quasi‐experimental studies were their short follow‐up length, the multiplicity of outcome measurements and the reliability of the outcome measurement tools used. The cohort and cross‐sectional studies failed to report on how confounding factors were managed in the analysis. Only two quantitative study article46,64 sreported sample size calculation or provided reasons for the chosen sample sizes. While all articles reported the number of participants 13697625, 2023, 2, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/hex.13692 by Test, Wiley Online Library on [02/05/20 583 POKU ET AL. 2023, 2, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/hex.13692 by Test, Wiley Online Library on [02/05/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles FIGURE 1 PRISMA flow diagram 584 \| POKU ET AL. POKU ET AL. 584 \| udy and participant characteristics Country Study design Target population CYP participant characteristics (number, mean age, % ethnicity, % gender, % SCD diagnosis) Intervention name and type Control No. of study sites Attrition rate (follow‐up) (%) Nigeria Noncontrolled, pre‐ post experimentala CYP N = 20; 19.8 years; 100% HbSS; gender and ethnicity not specified Behavioural N/A 1 0 Egypt Noncontrolled, pre‐ post experimental (full trial) CYP‐parent dyads N = 100; 12 ± 3.12 years; 55% males; 100% HbSS; ethnicity not specified Educational, skill training N/A 2 0 Canada Cohort— retrospective CYP N = 110; 18.1 years. Gender, diagnosis and ethnicity not specified Educational, behavioural N/A 1 14 USA RCT CYP & family N = 53 (27 IG, 26 CG); 14.24 years IG, 14.10 years CG; 52.9% males IG, 30% males CG; 100% African American IG, 95% African American CG PAIN (educational, behavioural) Attention control: DIS- EASE ED (educa- tional) 1 40 USA RCT CYP N = 67 (35 IG children, 32 IG adolescents); 9.2 years IG children; 15.3 years IG adolescents; 51% males IG children, 65% males IG adolescents; 100% African American. 2.6 \| Patient and Public Involvement (PPI) No data provided on CG Educational, behavioural, social support Attention control (picnic, trip to a museum) Not spec- ified 48 USA Cross‐sectionalb CYP & parents N = 43; 12.8 ± 3.98 years; 39.5% males; 79% HbSS; 100% African American Take‐Charge Programme (skill training) N/A 1 O USA Cross‐sectionala CYP N = 5; 13–24 years. Gender, diagnosis and ethnicity not specified iManage (skill training, social support N/A 1 0 USA Cross‐sectionala CYP N = 22; 18.8 years; 55% males; 100% African American. Diagnosis not specified Chronic Disease Self‐ Management Programme (CDSMP) (educational, skill training) N/A 1 36 13697625, 2023, 2, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/hex.13692 by Test, Wiley Online Library on [02/05/2 585 POKU ET AL. 585 TABLE 2 (Continued) References Country Study design Target population CYP participant characteristics (number, mean age, % ethnicity, % gender, % SCD diagnosis) Intervention name and type Control No. of study sites Attrition rate (follow‐up) (%) Crosby et al.36 USA RCTa CYP N = 58 (27 IG, 31 CG); 16.7 years IG, 16.3 years CG; 52% females IG, 54% females CG; 46% HbSS IG, 63% HbSS CG; 100% African American SC Thrive (educational, skill training) Attention control: SCHeal- thED (educa- tional) 1 9 Crosby et al.64 USA Qualitative CYP N = 19; 17.1 years; 58% females; 53% HbSS; 100% African American SC Thrive (educational, skill training) N/A 1 N/A Cozzi et al.32 USA Noncontrolled, pre‐ post experimental CYP N = 8; 10–20 years; 63% females; 88% HbSS; ethnicity not specified Educational, behavioural N/A 1 0 Daniel et al.37 USA RCT CYP & family N = 83; 8.29 years; 50% males; 59% HbSS; ethnicity not specified Family Taking Control (educational, skill training) Delayed inter- vention control 2 27 Dobson and Byrne38 USA Noncontrolled, pre‐ post experimental CYP N = 20; 8.4 years; 60% males; 95% HbSS; 75% African American Behavioural N/A 1 0 Dobson39 USA Noncontrolled, pre‐ post experimental CYP N = 20; 8.4 years; 60% males; 95% HbSS; 75% African American Behavioural N/A 1 0 Estepp et al.40 USA Cohort— retrospectivea CYP N = 83; 13.9 years; 53% females; 91% HbSS; ethnicity not specified Scheduled Instant Messaging Over the Network (SIMON) (behavioural) N/A 1 32 Fouda et al.41 Egypt Noncontrolled, pre‐ post experimental CYP N = 50; 14.4 ± 1.15 years; 54% males. 2.6 \| Patient and Public Involvement (PPI) Gender and ethnicity not specified Educational, skill training N/A 2 0 Gil et al.42 USA RCT CYP N = 49 (25 IG, 24 CG); 11.9 ± 3 years; 63% males; 67% HbSS; 100% African American Behavioural Standard care 1 0 Gil et al.43 USA RCT CYP N = 49 (25 IG, 24 CG); 11.9 ± 3 years; 63% males; 67% Behavioural Standard care 1 0 KU ET AL. \| 5 y.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License 586 \| POKU ET AL. POKU ET AL. 586 \| TABLE 2 (Continued) References Country Study design Target population CYP participant characteristics (number, mean age, % ethnicity, % gender, % SCD diagnosis) Intervention name and type Control No. of study sites Attrition rate (follow‐up) (%) Green et al.44 USA RCTa CYP & parents N = 28; 14.3 years; 43% females; 50% Hispanic. IG and CG are not specified. HABIT (educational, behavioural, skill training) Two educa- tional bro- chures 2 11 Hazzard et al.45 USA Noncontrolled, pre‐ post experimental CYP N = 47. Other characteristics not specified STARBRIGHT World (educational, social support) Usual educa- tion pro- gramme 1 0 Hood et al.46 USA RCTa CYP N = 58 (27 IG, 31 CG); 16.7 years IG, 16.3 years CG; 52% females IG, 54% females CG; 46% HbSS IG, 63% HbSS CG; 100% African American iManage (behavioural, skill training, social support) Attention control SCHeal- thED (educa- tional) 1 0 Kaslow et al.47 USA RCTb CYP & family N = 39 (IG = 20, CG = 19); 10.3 years; 62% females; 69% HbSS; 100% African American Educational, behavioural Treatment as usual 1 34 Ketchen et al.48 USA RCTb CYP & parents N = 56; 12.7 years IG, 12.18 years CG; 50% males IG, 53% males CG; 100% African American. Diagnosis not specified STARBRIGHT World (educational, behavioural, social support) Wait‐list 3 34 McClellan et al.49 USA RCTa CYP & parents N = 19 (9 IG, 10 CG); 13.4 ± 2.9 years; 68% females; 100% African American. 2.6 \| Patient and Public Involvement (PPI) Diagnosis not specified Behavioural Wait‐list 1 0 Palermo et al.50 USA RCTa CYP & parents N = 25 (15 IG, 10 CG); 14.8 years; 64% females; 64% HbSS; 64% African American Web‐based Management of Adolescent Pain (Web‐ MAP) (behavioural, skill training) Internet‐ deliv- ered pain edu- cation 9 20 \| 13697625, 2023, 2, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/hex.13692 by Test, Wiley Online Library on [02/05/20 587 POKU ET AL. ( ) References Country Study design Target population CYP participant characteristics (number, mean age, % ethnicity, % gender, % SCD diagnosis) Intervention name and type Control No. of study sites Attrition rate (follow‐up) (%) Phillips et al.51 USA Noncontrolled, pre‐ post experimentala CYP & parents N = 60; 7.8 years; 53% males. Diagnosis and ethnicity not specified Voice Crisis Alert V2 (educational, behavioural, skill training) N/A 1 20 Rodgers‐ Melnick et al.52 USA Noncontrolled, pre‐ post experimental CYP N = 30; 21.04 years; 50% males; 63% HbSS; 100% black ethnicity Build, Educate, Advance, Transition in Sickle cell disease (BEATS) (educational, behavioural, skill training) N/A 1 18% attended 0/4 sessions Saulsberry et al.53 USA Cohort— longitudinalb CYP N = 53; 14 years; 62% males; 66% HbSS; 100% African American Sickle Cell Transition E‐Learning Programme (STEP) (educational, behavioural) N/A 1 26 Schatz et al.54 USA RCT CYP N = 48 (23 IG, 25 CG); 12.3 ± 2.2 years CG, 13.8 ± 2.7 years IG; 43% males CG, 39% males IG; 78% HbSS CG, 74% HbSS IG; 96% African American CG, 91% African American IG Behavioural Wait‐list standard care 2 4 Schwartz et al.55 USA RCTa CYP & family N = 49 (25 IG, 24 CG); 14.32 years IG; 56% females IG; 88% HbSS IG; 100% African American Behavioural Disease educa- tion atten- tion control 1 0 (Continues) 588 \| POKU ET AL. POKU ET AL. 588 ABLE 2 (Continued) References Country Study design Target population CYP participant characteristics (number, mean age, % ethnicity, % gender, % SCD diagnosis) Intervention name and type Control No. 2.6 \| Patient and Public Involvement (PPI) of study sites Attrition rate (follow‐up) (%) Wihak et al.62 USA Noncontrolled, pre‐ post experimentala CYP & parents N = 8; 10–17 years; 100% black. Diagnosis and gender not specified Comfort Ability Programme for Sickle Cell Pain (CAP for SCP) (educational, behavioural) N/A 1 0 Yoon and Godwin63 USA Noncontrolled, pre‐ post experimentalb CYP N = 22; 10.86 ± 2.58 years; gender, ethnicity and diagnosis not specified Sickle Cell Slime‐ O‐Rama Game (educational) N/A 1 0 Abbreviations: CG, control group; CYP, children and young people; HbSS, haemoglobin SS genotype; IG, intervention group; N, number; N/A, not applicable; RCT, randomised controlled trial; SCD, sickle ce disease While studies reported differences in the characteristics of those who completed the study and those who dropped out, they failed to report how these had been factored into their analysis. Some RCTs only reported data on the intervention groups. In some studies, not all outcomes were measured for the comparison groups or intervention delivery and the timing of outcome assessments was variable between and within the groups. Several studies assessed outcomes using nonvalidated tools, or there were missing outcome data. Eleven papers reported high follow‐up attrition rates (Table 2). Many studies experienced recruitment challenges, achieving less than 50% of their target sample size. Commonly reported reasons for recruitment difficulties included people's lack of interest in the research, time constraints, inconvenience of intervention format and delivery approaches and perceived lack of intervention benefits. 3.3 \| Description of self‐management interventions Thirty‐two self‐management interventions were reported in the 38 included articles (Table 8). The majority (n = 22) combined different types of self‐management support,28‐32,34‐39,41,44,45,51,52,57‐60,62,64 but all had an educational component on general information about SCD or combined this with more specific information about pain, pharmaco- logical treatments, self‐care, clinic appointments, school and academic support, medical insurance and social care supports. Twenty‐eight interventions included skills training programmes (self‐assessment and planning, goal‐setting, problem‐solving, coping and interpersonal com- munication)27‐39,41‐44,47‐49,51‐61,64 and 15 were cognitive‐behavioural therapy programmes.28,30‐32,38,39,42,43,49,50,54‐57,62 Several interventions all had an educational component on general information about SCD or combined this with more specific information about pain, pharmaco- logical treatments, self‐care, clinic appointments, school and academic support, medical insurance and social care supports. Twenty‐eight interventions included skills training programmes (self‐assessment and planning, goal‐setting, problem‐solving, coping and interpersonal com- munication)27‐39,41‐44,47‐49,51‐61,64 and 15 were cognitive‐behavioural therapy programmes.28,30‐32,38,39,42,43,49,50,54‐57,62 Several interventions included elements to facilitate peer and social interactions and support.34,37,45,46,48,52,57,60,61 Others29,33,40,44,58‐61 focused on promot- ing medical adherence (i.e., medication adherence and clinic attendance). Several interventions were referred to by names and/or acronyms that reflected their intent or objective(s) (see Table 2). ley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License 2.6 \| Patient and Public Involvement (PPI) of study sites Attrition rate (follow‐up) (%) Sil et al.57 USA Noncontrolled, pre‐ post experimentala CYP N = 57; 13.33 years; 61% females; 73.7% HbSS; 100 non‐Hispanic black Comfort Ability Programme for Sickle Cell Pain (CAP for SCP) (educational, behavioural) N/A 3 53 Sil et al.56 USA Noncontrolled, pre‐ post experimental CYP N = 57; 13.4 years; 56.4% females; 68‐1% HbSS; 98% African American Behavioural, skill training N/A 1 38 Smaldone et al.58 USA RCTa CYP & parents N = 28; 14.3 years; 43% females; 50% Hispanic. IG and CG are not specified. HABIT (educational, behavioural, skill training) Two educa- tional bro- chures 2 11 Smith et al.59 USA Noncontrolled, pre‐ post experimental CYP N = 32. Other characteristics not specified Virginia Common- wealth University Transition Intervention Programme (educational, skill training) N/A 1 0 Treadwell and Weissman60 USA Noncontrolled, pre‐ post experimentalb CYP & parents N = 11; 11.6 years; 100% African American. Gender and diagnosis not specified Desferal Day Camp (educational, behavioural, social support) N/A 1 0 Viola et al.61 USA Noncontrolled, pre‐ post experimentala CYP N = 24; 20.8 ± 2–3 years; 45.8% females; 75% HbSS; 87.5% black or African American Students Helping Individuals Facilitate Transition (SHIFT) (educational, social support) N/A 1 0 POKU ET AL. 589 lost to follow‐up, none of the full trials reported conducting an intention‐to‐treat analysis. References Country Study design Target population CYP participant characteristics (number, mean age, % ethnicity, % gender, % SCD diagnosis) Intervention name and type Control No. of study sites Attrition rate (follow‐up) (%) Wihak et al.62 USA Noncontrolled, pre‐ post experimentala CYP & parents N = 8; 10–17 years; 100% black. Diagnosis and gender not specified Comfort Ability Programme for Sickle Cell Pain (CAP for SCP) (educational, behavioural) N/A 1 0 Yoon and Godwin63 USA Noncontrolled, pre‐ post experimentalb CYP N = 22; 10.86 ± 2.58 years; gender, ethnicity and diagnosis not specified Sickle Cell Slime‐ O‐Rama Game (educational) N/A 1 0 Abbreviations: CG, control group; CYP, children and young people; HbSS, haemoglobin SS genotype; IG, intervention group; N, number; N/A, not applicable; RCT, randomised controlled trial; SCD, sickle ce disease References Country Study design Target population CYP participant characteristics (number, mean age, % ethnicity, % gender, % SCD diagnosis) Intervention name and type Control No. 3.3.1 \| Format and delivery methods Twelve interventions were delivered both in‐person and remotely,29,30,32,37‐39,42‐44,47‐49,54,58,62 combining lectures, group activities, printed materials, prerecorded audio‐visuals, telephone calls and online features.Combining different delivery methods aimed at promoting flexibility, convenience and financial accessibility, as well as including methods to engage young people. Twelve interventions were technology‐based (e.g., mobile apps, virtual chatrooms, text messaging, e‐modules, electronic or online games, e‐diaries).33,34,36,44‐46,48‐51,53,54,58,61,63,64 Some of these interven- tions included symptom monitoring and tracking programmes aimed to change and reinforce behaviours believed to promote medical adherence and pain management. 13697625, 2023, 2, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/hex.13692 by Test, Wiley Online Library on [02/05/2 590 \| POKU ET AL. POKU ET AL. 590 TABLE 3 Methodological quality of included RCTs Quality criterion Green et al.44 Barakat et al.30 Broome et al.31 Crosby et al.36 Daniel et al.37 Gil et al.42 Gil et al.43 Hood et al.46 Kaslow et al.47 Ketchen et al.48 McClellan et al.49 Palermo et al.50 Schatz et al.54 Schwartz et al.55 Smaldone et al.58 Was randomization used for assignment of participants to treatment groups? Yes Unclear Unclear Yes Yes Unclear Unclear Unclear Unclear Yes Unclear Yes Yes Unclear Unclear Was allocation to treatment group concealed? Yes Unclear Unclear Unclear Yes Unclear Unclear Unclear Unclear Yes Unclear Yes Yes Unclear Yes Were treatment groups similar at the baseline? Unclear Yes Unclear Yes Yes Yes Unclear Unclear Yes Yes Unclear Unclear Unclear Unclear No Were participants blind to treatment assignment? N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A Were those delivering treatment blind to treatment assignment? N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A Were outcomes assessors blind to treatment assignment? Unclear Unclear Unclear Unclear Unclear Unclear Unclear Unclear Unclear Unclear Unclear Unclear Unclear Unclear Unclear Were treatment groups treated identically other than the treatment of interest? Yes Unclear Yes Yes Yes Yes Yes Unclear Yes N Unclear Unclear Unclear Unclear Unclear Was follow‐up completed and if not, were diff Unclear N Yes N Yes Unclear Unclear Unclear Yes Yes Yes N Unclear N Unclear iley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License TABLE 3 Methodological quality of included RCTs 13697625, 2023, 2, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/hex.13692 by Test, Wiley Online Library on [02/05/2 \| 591 591 POKU ET AL. 3.3.1 \| Format and delivery methods ( ) Quality criterion Green et al.44 Barakat et al.30 Broome et al.31 Crosby et al.36 Daniel et al.37 Gil et al.42 Gil et al.43 Hood et al.46 Kaslow et al.47 Ketchen et al.48 McClellan et al.49 Palermo et al.50 Schatz et al.54 Schwartz et al.55 Smaldone et al.58 Were participants analysed in the groups to which they were randomized? Yes Yes Yes Yes Yes Yes Yes Unclear Yes Yes Unclear Unclear Yes Unclear Yes Were outcomes measured in the same way for treatment groups? Yes Yes Yes Yes Yes Yes Yes Unclear Yes Yes Unclear Yes Yes Unclear Yes Were outcomes measured in a reliable way? Unclear N Yes Unclear Unclear Unclear Unclear Yes Yes Unclear Unclear Yes Unclear Unclear Yes Was appropriate statistical analysis used? Yes Yes Yes Yes Yes Unclear Unclear Yes Yes Unclear Unclear Yes Unclear Unclear Yes Was the trial design appropriate, and any deviations N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A 592 \| POKU ET AL. 592 POKU ET AL. 592 no confusion about which variable comes first) Were the participants included in any comparison similar? No N/A N/A N/A N/A N/A N/A N/A N/A Unclear Yes N/A N/A Yes N/A N/A N/A Were the participants included in any comparison receiving similar treatment/ care, other than the exposure of intervention of interest? Unclear N/A N/A N/A N/A N/A N/A N/A N/A Unclear Yes N/A N/A Unclear N/A N/A N/A Was there a control group? 3.3.1 \| Format and delivery methods Yes No No No No No No No No No No No No No No No No Were there multiple measure- No No No Unclear Yes No No No No No No No No No No No No Methodological quality of included quasi‐experimental studies n Hazzard et al.54 Abd Elaziz and Abd Elghany27 Adegbolagun et al.28 Cozzi et al.32 Crosby et al.34 Dobson andByrne38 Dobso- n39 Fouda et al.41 Phillips et al.51 Rodgers‐ Melnick et al.52 Sil et al.56 Sil et al.57 Smith et al.59 Treadwell and Weissman60 Viola et al.61 Wihak et al.62 Yoon and Godwin63 is s n h Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Unclear Yes Yes Yes Yes No N/A N/A N/A N/A N/A N/A N/A N/A Unclear Yes N/A N/A Yes N/A N/A N/A Unclear N/A N/A N/A N/A N/A N/A N/A N/A Unclear Yes N/A N/A Unclear N/A N/A N/A Yes No No No No No No No No No No No No No No No No e No No No Unclear Yes No No No No No No No No No No No No TABLE 4 Methodological quality of included quasi‐experimental studies Quality criterion Hazzard et al.54 Abd Elaziz and Abd Elghany27 Adegbolagun et al.28 Cozzi et al.32 Crosby et al.34 Dobson andByrne38 Dobso- n39 Fouda et al.41 Phillips et al.51 Rodgers‐ Melnick et al.52 Sil et al.56 Sil et al.57 Smith et al.59 Treadwell and Weissman60 Viola et al.61 Wihak et al.62 Yoon and Godwin63 Is it clear in the study what is the ‘cause’ and what is the ‘effect’? (i.e., there is no confusion about which variable comes first) Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Unclear Yes Yes Yes Yes Were the participants included in any comparison similar? No N/A N/A N/A N/A N/A N/A N/A N/A Unclear Yes N/A N/A Yes N/A N/A N/A Were the participants included in any Unclear N/A N/A N/A N/A N/A N/A N/A N/A Unclear Yes N/A N/A Unclear N/A N/A N/A TABLE 4 Methodological quality of included quasi‐experimental studies 13697625, 2023, 2, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/hex.13692 by Test, Wiley Online Library on [02/05/2 \| 593 \| 593 593 POKU ET AL. TABLE 5 Methodological quality of included cohort studies Quality criterion Allemang et al.29 Estepp et al.40 Saulsberry et al.53 Were the two groups similar and recruited from the same population? Yes N/A N/A Were the exposures measured similarly to assign people to both exposed and unexposed groups? N/A N/A N/A Was the exposure measured in a valid and reliable way? N/A No Yes Were confounding factors identified? Unclear No No Were strategies to deal with confounding factors stated? Unclear No No Were the groups/participants free of the outcome at the start of the study (or at the moment of exposure)? N/A Yes Yes Were the outcomes measured in a valid and reliable way? Yes Unclear No Was the follow‐up time reported and sufficient to be long enough for outcomes to occur? Yes Yes Unclear Was follow‐up completed, and if not, were the reasons to loss to follow‐up described and explored? Yes Yes Yes Were strategies to address incomplete follow‐up utilized? N/A N/A N/A Was appropriate statistical analysis used? Yes Yes Yes Overall quality Moderate Moderate Moderate TABLE 6 Methodological quality of included cross‐sectional studies Quality criterion Crosby et al.33 Crosby et al.34 Crosby et al.35 Were the criteria for inclusion in the sample clearly defined? Yes Yes Yes Were the study subjects and the setting described in detail? Yes Yes Yes Was the exposure measured in a valid and reliable way? Unclear N/A Yes Were objective standard criteria used for measurement of the condition? Yes N/A Yes Were confounding factors identified? No N/A No Were strategies to deal with confounding factors stated? No N/A No Were the outcomes measured in a valid and reliable way? Unclear Yes Unclear Was appropriate statistical analysis used? Yes Yes Yes Overall quality Moderate High Moderate TABLE 6 Methodological quality of included cross‐sectional studies ley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License Overall quality 3.3.1 \| Format and delivery methods ( ) criterion Hazzard et al.54 Abd Elaziz and Abd Elghany27 Adegbolagun et al.28 Cozzi et al.32 Crosby et al.34 Dobson andByrne38 Dobso- n39 Fouda et al.41 Phillips et al.51 Rodgers‐ Melnick et al.52 Sil et al.56 Sil et al.57 Smith et al.59 Treadwell and Weissman60 Viola et al.61 Wihak et al.62 Yoon and Godwin63 llow‐up mpleted if not, re erences ween ups in ms of their ow‐up quately cribed and lysed? Unclear Yes Unclear Unclear Unclear Yes Unclear Yes No No No No Unclear Unclear Yes Unclear Yes he comes of ticipants uded in mparison asured in me way? Yes N/A N/A N/A N/A N/A N/A Unclear N/A Unclear No N/A N/A Yes N/A N/A N/A outcomes asured in a able way? Unclear Unclear Yes Yes Yes Unclear Unclear Unclear Yes Unclear Unclear Yes No Unclear Yes unclear No ppropriate tistical lysis d? Yes Yes Yes Yes Yes Yes Unclear Yes Yes Yes Yes Yes Unclear Yes Yes Unclear Yes quality Moderate Moderate Moderate Moderate High Moderate Poor Moderate Moderate Poor Moderate Moderate Poor Moderate High Poor Moderate any comparison measured in the same way? Were outcomes measured in a reliable way? Unclear Unclear Yes Yes Yes Unclear Unclear Unclear Yes Unclear Unclear Yes No Unclear Yes unclear No Was appropriate statistical analysis used? Yes Yes Yes Yes Yes Yes Unclear Yes Yes Yes Yes Yes Unclear Yes Yes Unclear Yes Overall quality Moderate Moderate Moderate Moderate High Moderate Poor Moderate Moderate Poor Moderate Moderate Poor Moderate High Poor Moderate Quality criterion Hazzard et al.54 Abd Elaziz and Abd Elghany27 Adegbolagun et al.28 Cozzi et al.32 Crosby et al.34 Dobson andByrne38 Dobso- n39 Fouda et al.41 Phillips et al.51 Rodgers‐ Melnick et al.52 Sil et al.56 Sil et al.57 Smith et al.59 Treadwell and Weissman60 Viola et al.61 Wihak et al.62 Yoon and Godwin63 Was follow‐up completed and if not, were differences between groups in terms of their follow‐up adequately described and analysed? Unclear Yes Unclear Unclear Unclear Yes Unclear Yes No No No No Unclear Unclear Yes Unclear Yes Were the outcomes of Yes N/A N/A N/A N/A N/A N/A Unclear N/A Unclear No N/A N/A Yes N/A N/A N/A 13697625, 2023, 2, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/hex.13692 by Test, Wiley Online Library on [02/05/20 POKU ET AL. POKU ET AL. 594 TABLE 7 Methodological quality of included qualitative study Quality criterion No studies provided data on cost‐effectiveness or interven- tion development and/or implementation costs. In all the studies where race and ethnicity were identified, the SCD population were of African American background. There appeared to be an assumption that participants would be of low socioeconomic status and have strong family systems due to their racial and ethnic identity. These assumptions influenced the intervention format, delivery and target population, with attempts to make interventions financially accessible and/or involve family relatives. Some studies also highlighted the black culture of the SCD population and included black music, sports celebrities, mentors, navigators, interventionists and content from black TV shows to make them ‘culturally sensitive’.37,47,49,55,61,62 The agency of CYP, however, appeared to be overlooked, with most interventions developed with minimal or no CYP involvement (see Table 8) to tailor them to the values, expectations and hopes of the CYP. All the interventions were individual‐ or family‐focused, aimed at knowledge, skill acquisition, reinforcement and/or behaviour change, underpinned pre- dominately by psychological models (see Table 8). There was little or no consideration for CYP's wider social contexts like school and/or workplace. Outcomes were measured at different time points between the final intervention session and 12 months postintervention implementation (Table 9). Only four studies assessed outcomes over a longer period.30,31,52,60 Three studies measured several CYP‐focused outcomes via proxy reports.27,51,58 These studies also employed measurement tools not validated for CYP. Further, several studies involving self‐guided interventions did not assess engagement levels, and only two48,53 accounted for the connec- tion between engagement and outcomes. No studies reported if CYP with SCD participated in deciding how self‐management interventions' effectiveness, acceptability and feasibility should be evaluated. ey.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License Different validated and unvalidated generic and SCD‐specific outcome measurement tools were used (Table 4). Most tools had not been formally evaluated for use with CYP with SCD, and different tools were used to measure the same outcomes across the papers (Table 9). Only three papers44,58,60 reported the measurement of physiological parameters (Table 9). Most of the studies reporting intervention effectiveness failed to report on the effect size and/or 95% confidence intervals of the results. The evidence below must, therefore, be considered in the context of these limitations. Table 9 summarizes the effective- ness data. TABLE 7 Methodological quality of included qualitative study Quality criterion TABLE 7 Methodological quality of included qualitative study Quality criterion Crosby et al.64 Is there congruity between the stated philosophical perspective and the research methodology? N/A Is there congruity between the research methodology and the research questions or objectives? N/A Is there congruity between research methodology and the methods used to collect data? N/A Is there congruity between the research methodology and the representation and data analysis? No Is there congruity between the research methodology and the interpretation of results? No Is there a statement locating the researcher culturally or theoretically? No Is the influence of the researcher on the research, and vice‐versa addressed? No Are participants and their voices adequately represented? Unclear Is the research ethical according to current criteria or for recent studies, and is there evidence of ethical approval by an appropriate body? No Do the conclusions drawn in the research report flow from the analysis and interpretation of the data? Unclear Overall quality Poor nloaded from https://onlinelibrary.wiley.com/doi/10.1111/hex.13692 by Test, Wiley Online Library on [02/05/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Common Some studies reported involving different lay and professional stakeholders in intervention development or adaptation. However, only five (18.5%) studies involved CYP with SCD in developing their self‐ management programmes.33,34,36,37,46,51,61,64 Of these, one intervention described a co‐creation process where CYP were engaged throughout the development stages.34,36,46,64 In the remaining four studies, CYP with SCD appear to have been involved at the initial needs assessment stage61 or the review stage following the intervention development.33,37,51 CYP involvement appeared limited to intervention content. were feasibility/pilot studies that assessed and reported intervention effectiveness as primary outcomes.28,35,36,40,44,46‐48,53,57,58,60,61,63,64 were feasibility/pilot studies that assessed and reported intervention effectiveness as primary outcomes.28,35,36,40,44,46‐48,53,57,58,60,61,63,64 The most commonly assessed outcomes were disease knowledge, coping, self‐efficacy, quality of life, family and social functioning and support, self‐management confidence and behaviours, healthcare utilization, treatment adherence, parent–child relationship and tran- sition readiness and success. Twenty‐four papers reported on intervention acceptability and feasibility and/or implementation issues. No studies provided data on cost‐effectiveness or interven- tion development and/or implementation costs. The most commonly assessed outcomes were disease knowledge, coping, self‐efficacy, quality of life, family and social functioning and support, self‐management confidence and behaviours, healthcare utilization, treatment adherence, parent–child relationship and tran- sition readiness and success. Twenty‐four papers reported on intervention acceptability and feasibility and/or implementation issues. 3.3.2 \| Intervention development Most interventions were delivered on an individual basis, with only eight interventions (21.1%) being group‐based.27,28,35,37,52,57,59,60 Intervention duration ranged from 45 min38,39 to 36 months.59 Interven- tions were delivered as one‐off sessions, multiple sessions or were self‐ guided. Most were either delivered at a hospital (n = 16)27,28,31,38,39,41‐ 43,45,47,52,53,55,56,59,63 or at multiple locations (n = 12) (internet, hospital, telephone or community/home29,32,36,37,44,48,49,54,57,59,62,64). One inter- vention was delivered at a camp.51 Several articles (n = 13) explicitly stated the frameworks underpinning the interventions' development or implementation. These were predominantly psychosocial and behavioural change theories (Table 8). Only one intervention was underpinned by a self‐ management model48 and none by the self‐care management model for SCD.65 Information on intervention sources or development methods was provided for 27 (84.4%) of the 32 interventions. Thirteen were new interventions developed for the study,27,29,33,34,36,37,41,46,47,51‐53,59,64 six were adapted from existing generic (noncondition specific) self‐management pro- grammes,33,40,41,48,49,56 four were self‐management existing pro- grammes for SCD,42,43,54,63 and three were generic self‐management interventions implemented without adaptation for SCD.45,48,50 Several articles (n = 13) explicitly stated the frameworks underpinning the interventions' development or implementation. These were predominantly psychosocial and behavioural change theories (Table 8). Only one intervention was underpinned by a self‐ management model48 and none by the self‐care management model for SCD.65 Information on intervention sources or development methods was provided for 27 (84.4%) of the 32 interventions. Thirteen were new interventions developed for the 27 29 33 34 36 37 41 46 47 51 53 59 64 Most interventions (n = 29, 76.3%) were delivered by the researchers themselves or health and social care staff or students, who were predominantly from a psychology background. Only two interventions were delivered by lay counsellors.35,60 Some studies explicitly detailed the training and supervision of interventionists and measures to maintain and monitor intervention consistency and fidelity.30,37,44,47,54,55,58,61 13697625, 2023, 2, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/hex.13692 by Test, Wiley Online Library on [02/05/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library POKU ET AL. 595 TABLE 7 Methodological quality of included qualitative study Quality criterion 3.4 Thirty‐one studies assessed the effectiveness of self‐management interventions using multiple outcome measures (Table 9). Fifteen 596 \| POKU ET AL. POKU ET AL. 596 References Intervention content Theoretical framework Format/delivery Duration Location Facilitators Intervention development (PPI) Abd Elaziz and Abd Elghany27 SCD education and self‐care skill training Not specified In‐person, group‐based, Face‐to‐face teaching (group discussions, brainstorming, role play, demonstration & redemonstration), printed materials 3 daily sessions, 20–30 min/session Outpatient Research staff (nursing background) Developed by researchers with no PPI or CYP involvement. Adegbolagun et al.28 SCD education, pain management, skill training (activity scheduling, relaxation techniques, attention diversion, effective communication), general self‐management, Not specified In‐person, group‐based, Face‐to‐face teaching 5 weekly sessions. Duration of each session not indicated Outpatient Research staff (psychology background) Adapted from a previous study. Adaptation done by researchers with no PPI. Allemang et al.29 SCD education, self‐ management counselling (medication appointment adherence), health and social system navigation support Model for improvement (a quality improvement model) Face‐to‐face teaching, telephone and e‐mail support, printed resources 120–180 min/in‐ person sessions. The overall duration was 12 months Outpatient, remote (phone call, e‐mail) Transition navigator (social worker) Developed based on survey on the facilitators and barrier to successful transition from young people with SCD. Programme content and format developed with input from clinicians. No PPI or CYP involvement References Intervention content Theoretical framework Format/delivery Duration Location Facilitators Intervention development (PPI) Abd Elaziz and Abd Elghany27 SCD education and self‐care skill training Not specified In‐person, group‐based, Face‐to‐face teaching (group discussions, brainstorming, role play, demonstration & redemonstration), printed materials 3 daily sessions, 20–30 min/session Outpatient Research staff (nursing background) Developed by researchers with no PPI or CYP involvement. Adegbolagun et al.28 SCD education, pain management, skill training (activity scheduling, relaxation techniques, attention diversion, effective communication), general self‐management, Not specified In‐person, group‐based, Face‐to‐face teaching 5 weekly sessions. Duration of each session not indicated Outpatient Research staff (psychology background) Adapted from a previous study. Adaptation done by researchers with no PPI. Allemang et al.29 SCD education, self‐ management counselling (medication appointment adherence), health and social system navigation support Model for improvement (a quality improvement model) Face‐to‐face teaching, telephone and e‐mail support, printed resources 120–180 min/in‐ person sessions. The overall duration was 12 months Outpatient, remote (phone call, e‐mail) Transition navigator (social worker) Developed based on survey on the facilitators and barrier to successful transition from young people with SCD. 3.4 CYP involved in all the development phases. C b l 64 SCD d i kill i i N ifi d I b d f 6 kl i (fi O i P h l d 3 h d d l competition) development phases Crosby et al.35 SCD education, action planning and problem‐ solving skill training Not specified In‐person, group‐based, face‐ to‐face teaching 6 weekly sessions, 150 min/session Outpatient Lay leaders Pre‐existing programme (Stanford Chronic Disease Self‐ Management Programme) without adaptation for CYP with SCD. Crosby et al.36 SCD education, skills training (communication, problem‐ solving, pain, mood, mind and health management) Not specified In‐person, group‐based and remote (Zoom & mobile app) face‐to‐face teaching (brainstorming, video vignettes, modeling, role‐ play, rehearsal, group exercises, group discussions), e‐handouts 6 weekly sessions (first 3 in‐person and last 3 via Zoom) plus optional booster session after 2 weeks, 90 min/ session Outpatient, web‐based Psychology student, psychologist 3‐phased development process (survey, design and usability testing). CYP involved in all the development phases. Crosby et al.64 SCD education, skills training (communication, problem‐ solving, pain, mood, mind and health management) Not specified In‐person, group‐based, face‐ to‐face teaching (brainstorming, video vignettes, modelling, role‐ play, rehearsal, group exercises, group discussions), e‐handouts, web‐based (individual‐ based) 6 weekly sessions (first 3 in‐person & last 3 via Zoom) plus optional booster session, 90 min/ session Outpatient, web‐based Psychology student, psychologist 3‐phased development process. (survey, design and usability testing). CYP involved in all the development phases. Cozzi et al.32 behavioural skill training for pain management Not specified In‐person, individual‐based, telephone‐based, Face‐ 13 weekly sessions, 20–30 min/session Outpatient, home Clinical psychologist and prerecorded Pre‐existing intervention adapted for CYP with with SCD. Crosby et al.36 SCD education, skills training (communication, problem‐ solving, pain, mood, mind and health management) Not specified In‐person, group‐based and remote (Zoom & mobile app) face‐to‐face teaching (brainstorming, video vignettes, modeling, role‐ play, rehearsal, group exercises, group discussions), e‐handouts 6 weekly sessions (first 3 in‐person and last 3 via Zoom) plus optional booster session after 2 weeks, 90 min/ session Outpatient, web‐based Psychology student, psychologist 3‐phased development process (survey, design and usability testing). CYP involved in all the development phases. 3.4 Crosby et al.64 SCD education, skills training (communication, problem‐ solving, pain, mood, mind and health management) Not specified In‐person, group‐based, face‐ to‐face teaching (brainstorming, video vignettes, modelling, role‐ play, rehearsal, group exercises, group discussions), e‐handouts, web‐based (individual‐ based) 6 weekly sessions (first 3 in‐person & last 3 via Zoom) plus optional booster session, 90 min/ session Outpatient, web‐based Psychology student, psychologist 3‐phased development process. (survey, design and usability testing). CYP involved in all the development phases. Cozzi et al.32 behavioural skill training for pain management (breathing techniques, electromyographic and thermal biofeedback) Not specified In‐person, individual‐based, telephone‐based, Face‐ to‐face teaching, audiotapes and phone calls 13 weekly sessions, 20–30 min/session Outpatient, home Clinical psychologist and prerecorded audiotapes Pre‐existing intervention adapted for CYP with SCD but no PPI or CYP involvement in the adaptation process. (Continues) 25, 2023, 2, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/hex.13692 by Test, Wiley Online Library on [02/05/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Com ences Intervention content framework Format/delivery Duration Location Facilitators development (PPI) y et al.33 Self‐assessment of barriers to treatment adherence (medication and appointments), action planning, problem‐solving skills Not specified Voice automated interactive web‐based assessment tool Not specified Remote (web‐ based) Self‐guided CYP with SCD and their parents involved in the second phase (identification of content domains) of development. y et al.34 Pain monitoring/tracking, skill training (goal setting, goal evaluation), social support (group formation and competition) Not specified Digital, Mobile app, individual‐based 6 weeks Web‐based (digital) Self‐directed 3‐phased development process (survey, design and usability testing). CYP involved in all the development phases y et al.35 SCD education, action planning and problem‐ solving skill training Not specified In‐person, group‐based, face‐ to‐face teaching 6 weekly sessions, 150 min/session Outpatient Lay leaders Pre‐existing programme (Stanford Chronic Disease Self‐ Management Programme) without adaptation for CYP with SCD. y et al.36 SCD education, skills training (communication, problem‐ solving, pain, mood, mind and health management) Not specified In‐person, group‐based and remote (Zoom & mobile app) face‐to‐face teaching (brainstorming, video vignettes, modeling, role‐ play, rehearsal, group exercises, group discussions), e‐handouts 6 weekly sessions (first 3 in‐person and last 3 via Zoom) plus optional booster session after 2 weeks, 90 min/ session Outpatient, web‐based Psychology student, psychologist 3‐phased development process (survey, design and usability testing). CYP involved in all the development phases. 3.4 y et al.64 SCD education, skills training (communication, problem‐ solving, pain, mood, mind and health management) Not specified In‐person, group‐based, face‐ to‐face teaching (brainstorming, video vignettes, modelling, role‐ play, rehearsal, group exercises, group discussions), e‐handouts, web‐based (individual‐ based) 6 weekly sessions (first 3 in‐person & last 3 via Zoom) plus optional booster session, 90 min/ session Outpatient, web‐based Psychology student, psychologist 3‐phased development process. (survey, design and usability testing). CYP involved in all the development phases. et al.32 behavioural skill training for pain management (breathing techniques, electromyographic and thermal biofeedback) Not specified In‐person, individual‐based, telephone‐based, Face‐ to‐face teaching, audiotapes and phone calls 13 weekly sessions, 20–30 min/session Outpatient, home Clinical psychologist and prerecorded audiotapes Pre‐existing intervention adapted for CYP with SCD but no PPI or CYP involvement in the adaptation process. (Continues) 5, 2023, 2, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/hex.13692 by Test, Wiley Online Library on [02/05/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable C ( ) nces Intervention content Theoretical framework Format/delivery Duration Location Facilitators Intervention development (PPI) y et al.33 Self‐assessment of barriers to treatment adherence (medication and appointments), action planning, problem‐solving skills Not specified Voice automated interactive web‐based assessment tool Not specified Remote (web‐ based) Self‐guided CYP with SCD and their parents involved in the second phase (identification of content domains) of development. y et al.34 Pain monitoring/tracking, skill training (goal setting, goal evaluation), social support (group formation and competition) Not specified Digital, Mobile app, individual‐based 6 weeks Web‐based (digital) Self‐directed 3‐phased development process (survey, design and usability testing). CYP involved in all the development phases y et al.35 SCD education, action planning and problem‐ solving skill training Not specified In‐person, group‐based, face‐ to‐face teaching 6 weekly sessions, 150 min/session Outpatient Lay leaders Pre‐existing programme (Stanford Chronic Disease Self‐ Management Programme) without adaptation for CYP with SCD. y et al.36 SCD education, skills training (communication, problem‐ solving, pain, mood, mind and health management) Not specified In‐person, group‐based and remote (Zoom & mobile app) face‐to‐face teaching (brainstorming, video vignettes, modeling, role‐ play, rehearsal, group exercises, group discussions), e‐handouts 6 weekly sessions (first 3 in‐person and last 3 via Zoom) plus optional booster session after 2 weeks, 90 min/ session Outpatient, web‐based Psychology student, psychologist 3‐phased development process (survey, design and usability testing). CYP involved in all the development phases. 3.4 Programme content and format developed with input from clinicians. No PPI or CYP involvement. Barakat et al.30 Knowledge and skill training for pain management (relaxation, guided imagery, positive coping self‐statements) Not specified In‐person, family‐based, Face‐to‐face teaching, digital resources (guided imagery audiotapes), phone calls 4 sessions, the first 3 sessions were 2 weeks apart with a booster session 1 month apart. 90 min/sessions Home, phone calls Psychology students & psychologists Not specified Broome et al.31 SCD education, cognitive‐ behavioural skill training (relaxation, distraction breathing, guided imagery), art therapy (drawing, murals, art, interaction) to Not specified In‐person, group‐based, Face‐to‐face teaching, printed materials, digital resources (videotapes, audiotapes) 4 sessions over 6 weeks. Duration of each session not indicated. Outpatient Clinical nurse specialists, art therapist No PPI or CYP involvement. Barakat et al.30 Knowledge and skill training for pain management (relaxation, guided imagery, positive coping self‐statements) Not specified In‐person, family‐based, Face‐to‐face teaching, digital resources (guided imagery audiotapes), phone calls 4 sessions, the first 3 sessions were 2 weeks apart with a booster session 1 month apart. 90 min/sessions Home, phone calls Psychology students & psychologists Not specified Broome et al.31 SCD education, cognitive‐ behavioural skill training (relaxation, distraction breathing, guided imagery), art therapy (drawing, murals, art, interaction) to express feelings about pain and develop social skills Not specified In‐person, group‐based, Face‐to‐face teaching, printed materials, digital resources (videotapes, audiotapes) 4 sessions over 6 weeks. Duration of each session not indicated. Outpatient Clinical nurse specialists, art therapist No PPI or CYP involvement. 13697625, 2023, 2, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/hex.13692 by Test, Wiley Online Library on [02/05 597 POKU ET AL. 597 evaluation), social support (group formation and competition) and usability testing). CYP involved in all the development phases Crosby et al.35 SCD education, action planning and problem‐ solving skill training Not specified In‐person, group‐based, face‐ to‐face teaching 6 weekly sessions, 150 min/session Outpatient Lay leaders Pre‐existing programme (Stanford Chronic Disease Self‐ Management Programme) without adaptation for CYP with SCD. Crosby et al.36 SCD education, skills training (communication, problem‐ solving, pain, mood, mind and health management) Not specified In‐person, group‐based and remote (Zoom & mobile app) face‐to‐face teaching (brainstorming, video vignettes, modeling, role‐ play, rehearsal, group exercises, group discussions), e‐handouts 6 weekly sessions (first 3 in‐person and last 3 via Zoom) plus optional booster session after 2 weeks, 90 min/ session Outpatient, web‐based Psychology student, psychologist 3‐phased development process (survey, design and usability testing). 3.4 y et al.64 SCD education, skills training (communication, problem‐ Not specified In‐person, group‐based, face‐ to‐face teaching 6 weekly sessions (first 3 in‐person & last 3 Outpatient, web‐based Psychology student, psychologist 3‐phased development process. (survey, design References Intervention content Theoretical framework Format/delivery Duration Location Facilitators Intervention development (PPI) Crosby et al.33 Self‐assessment of barriers to treatment adherence (medication and appointments), action planning, problem‐solving skills Not specified Voice automated interactive web‐based assessment tool Not specified Remote (web‐ based) Self‐guided CYP with SCD and their parents involved in the second phase (identification of content domains) of development. Crosby et al.34 Pain monitoring/tracking, skill training (goal setting, goal evaluation), social support (group formation and competition) Not specified Digital, Mobile app, individual‐based 6 weeks Web‐based (digital) Self‐directed 3‐phased development process (survey, design and usability testing). CYP involved in all the development phases Crosby et al.35 SCD education, action planning and problem‐ l i kill t i i Not specified In‐person, group‐based, face‐ to‐face teaching 6 weekly sessions, 150 min/session Outpatient Lay leaders Pre‐existing programme (Stanford Chronic Di S lf 598 \| POKU ET AL. POKU ET AL. 598 Fouda et al.41 SCD education, self‐care practices/skill training Not specified Face‐to‐face teaching (group discussion, role‐play), printed materials 6 sessions over 3 days, 30–60 min/session Outpatient Research staff (nursing background) Developed by researchers based on the assessed needs of CYP with SCD. No PPI or CYP involvement. Gil et al.42 Cognitive behavioural skill training for coping with pain (breathing, relaxation, guided imagery, calming self‐statements) Not specified Face‐to‐face, individual‐ based, in‐person teaching, prerecorded audiotapes for self‐directed practice, homework assignments, phone calls 2 sessions. First session 45 min, second session a week later to review skills taught/learned Outpatient, remote (telephone) Research staff (psychology background) plus self‐ directed practice for 1 week Modelled after one used in an SCD adult population. PPI or CYP involvement not specified. References Intervention content Theoretical framework Format/delivery Duration Location Facilitators Intervention development (PPI) Daniel et al.37 SCD information, school support information, problem‐identification and problem‐solving skills Problem‐solving Face‐to‐face (role‐play, problem‐solving) telephone teaching 4 weekly sessions plus 3 booster phone call sessions. 75–90 min/in‐ person session & 30 min/booster session Outpatient, telephone Psychology students Intervention materials were developed based on focus group feedback and reviewed by a focus group of parents and CYP with SCD. 3.4 Dobson and Byrne38 guided imagery education and skill training Cognitive‐ behavioural theory & Bandura self‐ efficacy theory In‐person, individual‐based, Face‐to‐face teaching, audiotapes One session, 15–45 min Outpatient Child life specialist Not specified. Dobson39 guided imagery education and skill training Cognitive‐ behavioural theory & Bandura self‐ efficacy theory In‐person, individual‐based, Face‐to‐face teaching, audiotapes One session, 15–45 min Outpatient Child life specialist Not specified. Estepp et al.40 Daily text reminders about taking hydroxyurea. Participants can reply to the system Not specified Telephone‐based, individual‐ based 12 months Remote (tele- phone‐ based) Automated Text message content customized for each CYP participant. No PPI or CYP involvement in the intervention development. Intervention developed by clinical informatics staff. Fouda et al.41 SCD education, self‐care practices/skill training Not specified Face‐to‐face teaching (group discussion, role‐play), printed materials 6 sessions over 3 days, 30–60 min/session Outpatient Research staff (nursing background) Developed by researchers based on the assessed needs of CYP with SCD. No PPI or CYP involvement. Gil et al.42 Cognitive behavioural skill training for coping with pain (breathing, relaxation, guided imagery, calming self‐statements) Not specified Face‐to‐face, individual‐ based, in‐person teaching, prerecorded audiotapes for self‐directed practice, homework assignments, phone calls 2 sessions. First session 45 min, second session a week later to review skills taught/learned Outpatient, remote (telephone) Research staff (psychology background) plus self‐ directed practice for 1 week Modelled after one used in an SCD adult population. PPI or CYP involvement not specified. development. Intervention developed by clinical informatics staff. et al.41 SCD education, self‐care practices/skill training Not specified Face‐to‐face teaching (group discussion, role‐play), printed materials 6 sessions over 3 days, 30–60 min/session Outpatient Research staff (nursing background) Developed by researchers based on the assessed needs of CYP with SCD. No PPI or CYP involvement. .42 Cognitive behavioural skill training for coping with pain (breathing, relaxation, guided imagery, calming self‐statements) Not specified Face‐to‐face, individual‐ based, in‐person teaching, prerecorded audiotapes for self‐directed practice, homework assignments, phone calls 2 sessions. First session 45 min, second session a week later to review skills taught/learned Outpatient, remote (telephone) Research staff (psychology background) plus self‐ directed practice for 1 week Modelled after one used in an SCD adult population. PPI or CYP involvement not specified. 3.4 References Intervention content Theoretical framework Format/delivery Duration Location Facilitators Intervention development (PPI) Daniel et al.37 SCD information, school support information, problem‐identification and problem‐solving skills Problem‐solving Face‐to‐face (role‐play, problem‐solving) telephone teaching 4 weekly sessions plus 3 booster phone call sessions. 75–90 min/in‐ person session & 30 min/booster session Outpatient, telephone Psychology students Intervention materials were developed based on focus group feedback and reviewed by a focus group of parents and CYP with SCD. Dobson and Byrne38 guided imagery education and skill training Cognitive‐ behavioural theory & Bandura self‐ efficacy theory In‐person, individual‐based, Face‐to‐face teaching, audiotapes One session, 15–45 min Outpatient Child life specialist Not specified. Dobson39 guided imagery education and skill training Cognitive‐ behavioural theory & Bandura self‐ efficacy theory In‐person, individual‐based, Face‐to‐face teaching, audiotapes One session, 15–45 min Outpatient Child life specialist Not specified. Estepp et al.40 Daily text reminders about taking hydroxyurea. Participants can reply to the system Not specified Telephone‐based, individual‐ based 12 months Remote (tele- phone‐ based) Automated Text message content customized for each CYP participant. No PPI or CYP involvement in the intervention development. Intervention developed 13697625, 2023, 2, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/hex.13692 by Test, Wiley Online Library on [02/05/2 POKU ET AL. 599 medication text reminders Hazzard et al.45 SCD education, recreational activities (games, arts, crafts) to enhance coping and social support Face‐to‐face teaching, web‐ based (video conferencing, chatrooms, e‐games, arts, crafts Up to the time of hospital discharge In‐patient, web‐based self‐guided. Healthcare professionals present educational activity (15–45 min). Pre‐existing programme with the educational component adapted for CYP with SCD. Adapted component reviewed by a team of clinicians. No PPI or CYP involvement in the adaptation process. Hood et al.46 Pain monitoring/tracking, skill training (goal setting, goal evaluation), social support (group formation and competition) Not specified Digital, Mobile app, individual‐based 6 weeks Web‐based (digital) Self‐directed 3‐phased development process (survey, design and usability testing). CYP involved in all the development phases. Kaslow et al.47 SCD education, pain management techniques, problem‐based coping practices Stress‐coping‐ adjustment framework In‐person, family‐focused, Face‐to‐face teaching (role‐play, games), homework, printed materials 6 weekly sessions, 60 min/session Outpatient Psychology students Input from clinicians and a SCD family member. No CYP involvement. Ketchen et al.48 SCD education, skill training (interpersonal communication, self‐ expression, distraction‐ effective elevation) Not specified Web‐based (video conferencing, chatrooms, e‐recreational activities), a CD‐ROM game, homework 6 weeks intervention In‐patient, home, web‐based Self‐guided. 3.4 Weekly reminder phone calls from the research team Pre‐existing programme used without adaption for CYP with SCD. Programme development not identified. (Continues) References Intervention content Theoretical framework Format/delivery Duration Location Facilitators Intervention development (PPI) Gil et al.43 Cognitive behavioural skill training for coping with pain (breathing, relaxation, guided imagery, calming self‐statements) Not specified Face‐to‐face, individual‐ based, in‐person teaching, prerecorded audiotapes for self‐directed practice, homework assignments, phone calls In‐person session— 45 min. Weekly telephone contacts for 4 weeks Outpatient, remote (telephone) Research staff (psychology background) plus self‐ directed practice for 1 month Modelled after one used in an SCD adult population. PPI or CYP involvement not specified. Green et al.44 SCD and hydroxyurea education, adherence barriers identification, action planning, problem‐ solving skills, automated medication text reminders Not specified Face‐to‐face teaching, text messages, phone calls, printed brochures 5 sessions, 60–90 min/ session Outpatient, remote Community health workers Not specified. Hazzard et al.45 SCD education, recreational activities (games, arts, crafts) to enhance coping and social support Face‐to‐face teaching, web‐ based (video conferencing, chatrooms, e‐games, arts, crafts Up to the time of hospital discharge In‐patient, web‐based self‐guided. Healthcare professionals present educational activity (15–45 min). Pre‐existing programme with the educational component adapted for CYP with SCD. Adapted component reviewed by a team of clinicians. No PPI CYP i l t References Intervention content Theoretical framework Format/delivery Duration Location Facilitators Intervention development (PPI) Gil et al.43 Cognitive behavioural skill training for coping with pain (breathing, relaxation, guided imagery, calming self‐statements) Not specified Face‐to‐face, individual‐ based, in‐person teaching, prerecorded audiotapes for self‐directed practice, homework assignments, phone calls In‐person session— 45 min. Weekly telephone contacts for 4 weeks Outpatient, remote (telephone) Research staff (psychology background) plus self‐ directed practice for 1 month Modelled after one used in an SCD adult population. PPI or CYP involvement not specified. Green et al.44 SCD and hydroxyurea education, adherence barriers identification, Not specified Face‐to‐face teaching, text messages, phone calls, printed brochures 5 sessions, 60–90 min/ session Outpatient, remote Community health workers Not specified. ley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License 600 \| POKU ET AL. 3.4 600 References Intervention content Theoretical framework Format/delivery Duration Location Facilitators Intervention development (PPI) McClellan et al.49 Cognitive‐behavioural coping skills training for pain management (deep breathing, guided imagery, relaxation) Not specified In‐person, family‐based, face‐ to‐face teaching (participatory activities and demonstrations), printed materials, e‐ modules, telephone calls, e‐mails One 120 min in‐person session plus digital engagement over 2 months and weekly telephone calls within the month Outpatient, web‐based Psychologists, self‐ directed Not identified Palermo et al.50 Cognitive and behavioural skill training for pain management (relaxation techniques, healthy lifestyles, operant strategies for adaptive behaviours), goal setting Cognitive behavioural theory Web‐based, family‐based, e‐ modules (presented through videos, vignettes, activities, illustrations), e‐ mail contact 8 weekly e‐modules Web‐based Self‐directed A generic pre‐existing programme used without adaption for the study population. Intervention development not detailed. Phillips et al.51 SCD education, pain tracking/ monitoring, pain management behaviour modification Paediatric self‐ management model Digital, mobile app Intervention duration— 12 weeks Web‐based Self‐guided CYP with SCD and their parents or caregivers reviewed the intervention and it was revised based on the feedback. Rodgers‐ Melnick et al.52 SCD education, pain management, medication management, skill training (relationship‐building), peer support Not stated In‐person, group‐based, group‐drumming, songwriting, music production/creation, printed action plans, 4 sessions every 3 months for 12 months Outpatient Multidisciplinary team (haematologists, nurse practitioner, social worker, paly patient navigator, music therapist) Incorporated resources (song stories) developed with CYP with SCD. Saulsberry et al.53 SCD information and transition skills Not specified Web‐based, individual‐based, e‐modules, videos, quizzes 6 modules, completed within 1 year Outpatient Healthcare professionals (nurse‐led) Pre‐existing intervention adapted for CYP with SCD. Adaption involved a team of clinicians and hospital administrative staff. No PPI or CYP involvement. Schatz et al.54 Cognitive and behavioural skill training for pain management (deep breathing, relaxation, guided imagery) Not specified In‐person, individual‐based, face‐to‐face teaching plus home‐based practices and skill acquisition using a mobile app, written materials, telephone calls One in‐person session 45–60 min. 8 weeks of engagement with the mobile app Outpatient, home, telephone Psychology students. Self‐ directed Modelled from a pre‐ existing intervention developed for SCD adult population. No PPI or CYP involvement in the adaption. 3.4 Relaxation training, activity pacing, cognitive restructuring techniques, healthy lifestyle habits) Cognitive‐ behavioural theory In‐person, individual‐based, face‐to‐face teaching and skill acquisition 8 bi‐weekly sessions, 45–60 min/session. Outpatient Psychologists Intervention personalized for each CYP‐parent dyad but no PPI or CYP involvement in the development. one al.58 SCD and hydroxyurea education, adherence barriers identification, action planning, problem‐ solving skills, automated medication text reminders Not specified Face‐to‐face teaching, text messages, phone calls, printed brochures 5 sessions, 60–90 min/ session Outpatient, remote Community health workers Not specified. et al.59 SCD education, information about transition policy, culture of adult medicine, school and work support and self‐advocacy skill training Not specified In‐person, group‐based, individual‐based, face‐to‐ face teaching 3 years Outpatient Multidisciplinary team (medical providers, educational coordinator, clinical psychologist, social worker, transition coordinator) Not specified. (Continues) Continued) Intervention content Theoretical framework Format/delivery Duration Location Facilitators Intervention development (PPI) Cognitive and behavioural skill training for pain management (deep breathing, relaxation, guided imagery, positive self‐statements) Not specified In‐person, individual‐based, face‐to‐face teaching plus home‐based practices using prerecorded audiotapes 4 sessions, sessions 1–3 occur 2 weeks intervals, session 4 a month after 3. Each session is 90 min Home and telephone‐ based Psychology students and psychologists Guided imagery content chosen by CYP participants but no PPI or CYP involvement in the development. Psychological and behavioural pain management skill training Dynamic adaption process model In‐person, group‐based, face‐ to‐face teaching (group discussion), videotapes, workbook exercises 3 sessions, 60 min/ session over duration of hospitalization In‐patient Psychologists Pre‐existing programme with the educational component adapted for CYP with SCD. Adapted component reviewed by a team of clinicians. No PPI or CYP involvement in the adaption process. Cognitive and behavioural education and skill training for pain and illness management (education about pain, behavioural strategies for parents. Relaxation training, activity pacing, cognitive restructuring techniques, healthy lifestyle habits) Cognitive‐ behavioural theory In‐person, individual‐based, face‐to‐face teaching and skill acquisition 8 bi‐weekly sessions, 45–60 min/session. Outpatient Psychologists Intervention personalized for each CYP‐parent dyad but no PPI or CYP involvement in the development. SCD and hydroxyurea education, adherence barriers identification, action planning, problem‐ solving skills, automated medication text reminders Not specified Face‐to‐face teaching, text messages, phone calls, printed brochures 5 sessions, 60–90 min/ session Outpatient, remote Community health workers Not specified. 3.4 SCD education, information about transition policy, culture of adult medicine, school and work support and self‐advocacy skill training Not specified In‐person, group‐based, individual‐based, face‐to‐ face teaching 3 years Outpatient Multidisciplinary team (medical providers, educational coordinator, clinical psychologist, social worker, transition coordinator) Not specified. (Continues) 2023, 2, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/hex.13692 by Test, Wiley Online Library on [02/05/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the a References Intervention content Theoretical framework Format/delivery Duration Location Facilitators Intervention development (PPI) Schwartz et al.55 Cognitive and behavioural skill training for pain management (deep breathing, relaxation, guided imagery, positive self‐statements) Not specified In‐person, individual‐based, face‐to‐face teaching plus home‐based practices using prerecorded audiotapes 4 sessions, sessions 1–3 occur 2 weeks intervals, session 4 a month after 3. Each session is 90 min Home and telephone‐ based Psychology students and psychologists Guided imagery content chosen by CYP participants but no PPI or CYP involvement in the development. Sil et al.57 Psychological and behavioural pain management skill training Dynamic adaption process model In‐person, group‐based, face‐ to‐face teaching (group discussion), videotapes, workbook exercises 3 sessions, 60 min/ session over duration of hospitalization In‐patient Psychologists Pre‐existing programme with the educational component adapted for CYP with SCD. Adapted component reviewed by a team of clinicians. No PPI or CYP involvement in the adaption process. Sil et al.56 Cognitive and behavioural education and skill training for pain and illness Cognitive‐ behavioural theory In‐person, individual‐based, face‐to‐face teaching and skill acquisition 8 bi‐weekly sessions, 45–60 min/session. Outpatient Psychologists Intervention personalized for each CYP‐parent dyad but no PPI or CYP p g g restructuring techniques, healthy lifestyle habits) Smaldone et al.58 SCD and hydroxyurea education, adherence barriers identification, action planning, problem‐ solving skills, automated medication text reminders Not specified Face‐to‐face teaching, text messages, phone calls, printed brochures 5 sessions, 60–90 min/ session Outpatient, remote Community health workers Not specified. Smith et al.59 SCD education, information about transition policy, culture of adult medicine, school and work support and self‐advocacy skill training Not specified In‐person, group‐based, individual‐based, face‐to‐ face teaching 3 years Outpatient Multidisciplinary team (medical providers, educational coordinator, clinical psychologist, social worker, transition coordinator) Not specified. management (education about pain, behavioural strategies for parents. Relaxation training, activity pacing, cognitive restructuring techniques, healthy lifestyle habits) involvement in the development. 3.4 References Intervention content Theoretical framework Format/delivery Duration Location Facilitators Intervention development (PPI) McClellan et al.49 Cognitive‐behavioural coping skills training for pain management (deep breathing, guided imagery, relaxation) Not specified In‐person, family‐based, face‐ to‐face teaching (participatory activities and demonstrations), printed materials, e‐ modules, telephone calls, e‐mails One 120 min in‐person session plus digital engagement over 2 months and weekly telephone calls within the month Outpatient, web‐based Psychologists, self‐ directed Not identified Palermo et al.50 Cognitive and behavioural skill training for pain management (relaxation techniques, healthy lifestyles, operant strategies for adaptive behaviours), goal setting Cognitive behavioural theory Web‐based, family‐based, e‐ modules (presented through videos, vignettes, activities, illustrations), e‐ mail contact 8 weekly e‐modules Web‐based Self‐directed A generic pre‐existing programme used without adaption for the study population. Intervention development not detailed. Phillips et al.51 SCD education, pain tracking/ monitoring, pain t b h i Paediatric self‐ management d l Digital, mobile app Intervention duration— 12 weeks Web‐based Self‐guided CYP with SCD and their parents or caregivers i d th 13697625, 2023, 2, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/hex.13692 by Test, Wiley Online Library on [02/05/20 POKU ET AL. 601 p p Sil et al.56 Cognitive and behavioural education and skill training for pain and illness management (education about pain, behavioural strategies for parents. Relaxation training, activity pacing, cognitive restructuring techniques, healthy lifestyle habits) Cognitive‐ behavioural theory In‐person, individual‐based, face‐to‐face teaching and skill acquisition 8 bi‐weekly sessions, 45–60 min/session. Outpatient Psychologists Intervention personalized for each CYP‐parent dyad but no PPI or CYP involvement in the development. Smaldone et al.58 SCD and hydroxyurea education, adherence barriers identification, action planning, problem‐ solving skills, automated medication text reminders Not specified Face‐to‐face teaching, text messages, phone calls, printed brochures 5 sessions, 60–90 min/ session Outpatient, remote Community health workers Not specified. Smith et al.59 SCD education, information about transition policy, culture of adult medicine, Not specified In‐person, group‐based, individual‐based, face‐to‐ face teaching 3 years Outpatient Multidisciplinary team (medical providers, educational Not specified. p g training p g (g p discussion), videotapes, workbook exercises duration of hospitalization component adapted for CYP with SCD. Adapted component reviewed by a team of clinicians. No PPI or CYP involvement in the adaption process. al.56 Cognitive and behavioural education and skill training for pain and illness management (education about pain, behavioural strategies for parents. 3.4 Smaldone et al.58 SCD and hydroxyurea education, adherence barriers identification, action planning, problem‐ solving skills, automated medication text reminders Not specified Face‐to‐face teaching, text messages, phone calls, printed brochures 5 sessions, 60–90 min/ session Outpatient, remote Community health workers Not specified. Smith et al.59 SCD education, information about transition policy, culture of adult medicine, school and work support and self‐advocacy skill training Not specified In‐person, group‐based, individual‐based, face‐to‐ face teaching 3 years Outpatient Multidisciplinary team (medical providers, educational coordinator, clinical psychologist, social worker, transition coordinator) Not specified. (Continues) Continued) Intervention content Theoretical framework Format/delivery Duration Location Facilitators Intervention development (PPI) Cognitive and behavioural skill training for pain management (deep breathing, relaxation, guided imagery, positive self‐statements) Not specified In‐person, individual‐based, face‐to‐face teaching plus home‐based practices using prerecorded audiotapes 4 sessions, sessions 1–3 occur 2 weeks intervals, session 4 a month after 3. Each session is 90 min Home and telephone‐ based Psychology students and psychologists Guided imagery content chosen by CYP participants but no PPI or CYP involvement in the development. Psychological and behavioural pain management skill training Dynamic adaption process model In‐person, group‐based, face‐ to‐face teaching (group discussion), videotapes, workbook exercises 3 sessions, 60 min/ session over duration of hospitalization In‐patient Psychologists Pre‐existing programme with the educational component adapted for CYP with SCD. Adapted component reviewed by a team of clinicians. No PPI or CYP involvement in the adaption process. Cognitive and behavioural education and skill training for pain and illness management (education about pain, behavioural strategies for parents. Relaxation training, activity pacing, cognitive restructuring techniques, healthy lifestyle habits) Cognitive‐ behavioural theory In‐person, individual‐based, face‐to‐face teaching and skill acquisition 8 bi‐weekly sessions, 45–60 min/session. Outpatient Psychologists Intervention personalized for each CYP‐parent dyad but no PPI or CYP involvement in the development. iley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License 13697625, 2023, 2, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/hex.13692 by Test, Wiley Online Library on [02/05/20 602 \| POKU ET AL. 3.4 Last session, 12 m (adolescents): increase In the frequency of use of coping strategies (p = .018). decrease in ER and clinic visits (p = .05) and admissions (p = .03) Moderate Crosby et al.33 Clinic Attendance, Hydroxyurea Adherence Barriers to Care Questionnaire, Web‐based Assessment Tool (Take Charge Programme), Medical Record Review No real data on the impact of the intervention reported Moderate Crosby et al.35 General and Disease‐Specific Self‐Efficacy, Self‐ Management Behaviours, Health Status, QOL, ER Visits, Patient Activation SCSES, Self‐Efficacy for Managing Chronic Disease Scale, Transition Readiness Assessment Questionnaire, National Health Interview Survey, Patient Activation Measure, Pediatric QoL Inventory 6 w, 3 m, 6 m: significant improvement in general self‐ efficacy (p = .015) but not in disease‐specific self‐ efficacy (p = .293), self‐management behaviours (p = .39)), health status (p = .30), quality of life (p = .17) and ER visits (p = .59). Medium effect size was noted for patient activation (p = .70). Moderate (Continues) y References Outcome Measure Result or finding (d, day; w, week; m, month) Study quality rating Abd Elaziz and Abd Elghany27 Knowledge, Pain level, Pain Coping Strategies, Self‐ Care Practices Verbally administered Numerical Rating Scale for pain assessment. Other outcomes were measured using checklists designed for the study. 3 d: Improvement in outcomes (p < .001) Moderate Adegbolagun et al.28 Anxiety, Depression, Coping Strategies, Quality of Life (QOL), Knowledge Hospital Anxiety and Depression Scale, Coping Skills Questionnaire for Sickle Cell Disease‐Adult Version, SF‐36, Knowledge Questionnaire (designed for the study) 1 w: Reduction in anxiety (p < .001) and depression (p < .02). Improvement in active and passive coping skills (p < .004), affective coping skills (p = .58), self‐reported knowledge, the practice of coping skills and confidence (p < .0001) Moderate Allemang et al.29 Lost to Follow‐up, Appointment Attendance Rate, Medication Adherence, Hospitalization Medical Records and self‐reported checklists 12 m: Reduction in the proportion of transition patients lost to follow‐up (p = .00034). Improvement in appointment (p = .096), medication adherence (p = .047) and hospitalization rates. 3.4 602 References Intervention content Theoretical framework Format/delivery Duration Location Facilitators Intervention development (PPI) Treadwell and Weissman60 Medication (deferoxamine) information and management, general information (genetics, nutrition), social support Not specified In‐person, group‐based, face‐ to‐face teaching, games, visit to a theme park 4 day‐camp Community Lay leaders Not specified Viola et al.61 Educational, skill training and social support mentorship programme focused on transition Social‐ecological model of adolescent and young adult readiness to transition (SMART) Web‐based, individual‐based communication via video conferencing and text messaging Monthly video calls and weekly text messaging for 6 months. Video call last 25–60 min. Web‐based Medical students Content informed by formative need assessment of CYP population. Wihak et al.62 Cognitive and behavioural pain management training (breathing, relaxation, guided imagery), skill training (building confidence and positive thoughts) Not specified In‐person, individual‐based, face‐to‐face teaching, audio‐visual resources, workbook One‐off, 150 min Outpatient, in‐ patient Multidisciplinary team (haematology nurse, child life provider, social worker) PPI in the 4‐phased adaptation process of a pre‐existing programme. PPI included clinicians, patient advocate and community consultant. CYP involvement not specified. Yoon and Godwin63 SCD information, self‐care and pain management Lieberman's theory on interactive health games and health outcomes in children CD‐ROM, individual‐based, e‐game 3 rounds, one‐off. Intervention duration not indicated Outpatient Self‐directed Pre‐existing programme with the educational component adapted for CYP with SCD. Adapted component reviewed by a team of clinicians and no PPI/CYP involvement. Abbreviations: CYP, children and young people; SCD, sickle cell disease. 13697625, 2023, 2, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/hex.13692 by Test, Wiley Online Library on [02/05/ POKU ET AL. 603 Barakat et al.30 Pain (presence, healthcare use, medication), School Attendance, Health‐Related Hindrance, Coping Strategies, Knowledge, Teen and Family Physical and Psychosocial Well‐being, Family Cohesion Pain Diary (designed for the study), Health‐related Hindrance Inventory (designed for the study), Coping Strategies Questionnaire, SCD Knowledge Questionnaire, SCD Transition Knowledge Questionnaire (designed for the study), Disease Self‐Efficacy Scale (adapted for this study from a cancer‐specific scale), Child Health Questionnaire 10 w, 12 m: No difference in outcomes between groups Poor Broome et al.31 Coping School‐agers' Coping Strategies Inventory, Adolescent Coping Orientation for Problem Experiences 6 w, 12 m (children): No difference in the frequency of use of coping strategies among the groups. Increase in the number of strategies used and rated effective over time in the intervention groups (p < .0003). Nonsignificant reduction in clinic and ER visits and admissions over time. 3.4 Moderate Barakat et al.30 Pain (presence, healthcare use, medication), School Attendance, Health‐Related Hindrance, Coping Strategies, Knowledge, Teen and Family Physical and Psychosocial Well‐being, Family Cohesion Pain Diary (designed for the study), Health‐related Hindrance Inventory (designed for the study), Coping Strategies Questionnaire, SCD Knowledge Questionnaire, SCD Transition Knowledge Questionnaire (designed for the study), Disease Self‐Efficacy Scale (adapted for this study from a cancer‐specific scale), Child Health Questionnaire 10 w, 12 m: No difference in outcomes between groups Poor Broome et al.31 Coping School‐agers' Coping Strategies Inventory, Adolescent Coping Orientation for Problem Experiences 6 w, 12 m (children): No difference in the frequency of use of coping strategies among the groups. Increase in the number of strategies used and rated effective over time in the intervention Moderate References Outcome Measure Result or finding (d, day; w, week; m, month) Study quality rating Abd Elaziz and Abd Elghany27 Knowledge, Pain level, Pain Coping Strategies, Self‐ Care Practices Verbally administered Numerical Rating Scale for pain assessment. Other outcomes were measured using checklists designed for the study. 3 d: Improvement in outcomes (p < .001) Moderate Adegbolagun et al.28 Anxiety, Depression, Coping Strategies, Quality of Life (QOL), Knowledge Hospital Anxiety and Depression Scale, Coping Skills Questionnaire for Sickle Cell Disease‐Adult Version, SF‐36, Knowledge Questionnaire (designed for the study) 1 w: Reduction in anxiety (p < .001) and depression (p < .02). Improvement in active and passive coping skills (p < .004), affective coping skills (p = .58), self‐reported knowledge, the practice of coping skills and confidence (p < .0001) Moderate Allemang et al.29 Lost to Follow‐up, Appointment Attendance Rate, Medication Adherence, Hospitalization Medical Records and self‐reported checklists 12 m: Reduction in the proportion of transition patients lost to follow‐up (p = .00034). Improvement in appointment (p = .096), medication adherence (p = .047) and hospitalization rates. 3.4 Moderate Barakat et al.30 Pain (presence, healthcare use, medication), School Attendance, Health‐Related Hindrance, Coping Strategies, Knowledge, Teen and Family Physical and Psychosocial Well‐being, Family Cohesion Pain Diary (designed for the study), Health‐related Hindrance Inventory (designed for the study), Coping Strategies Questionnaire, SCD Knowledge Questionnaire, SCD Transition Knowledge Questionnaire (designed for the study), Disease Self‐Efficacy Scale (adapted for this study from a cancer‐specific scale), Child Health Questionnaire 10 w, 12 m: No difference in outcomes between groups Poor Broome et al.31 Coping School‐agers' Coping Strategies Inventory, Adolescent Coping Orientation for Problem Experiences 6 w, 12 m (children): No difference in the frequency of use of coping strategies among the groups. Increase in the number of strategies used and rated effective over time in the intervention groups (p < .0003). Nonsignificant reduction in clinic and ER visits and admissions over time. Last session, 12 m (adolescents): increase In the frequency of use of coping strategies (p = .018). decrease in ER and clinic visits (p = .05) and admissions (p = .03) Moderate Crosby et al.33 Clinic Attendance, Hydroxyurea Adherence Barriers to Care Questionnaire, Web‐based Assessment Tool (Take Charge Programme), Medical Record Review No real data on the impact of the intervention reported Moderate Crosby et al.35 General and Disease‐Specific Self‐Efficacy, Self‐ Management Behaviours, Health Status, QOL, ER Visits, Patient Activation SCSES, Self‐Efficacy for Managing Chronic Disease Scale, Transition Readiness Assessment Questionnaire, National Health Interview Survey, 6 w, 3 m, 6 m: significant improvement in general self‐ efficacy (p = .015) but not in disease‐specific self‐ efficacy (p = .293), self‐management behaviours Moderate TABLE 9 Summary of the effectiveness of interventions on outcomes 604 \| POKU ET AL. 604 References Outcome Measure Result or finding (d, day; w, week; m, month) Study quality rating Crosby et al.64 Health‐Related Quality of Life (HRQOL), Self‐ Efficacy, Self‐Management Skills, Knowledge, Health Motivation Patient Activation Measure, Transition Readiness Questionnaire, Disease‐Specific Knowledge Questionnaire, Treatment Self‐Regulation Questionnaire, Paediatric QoL Inventory SCD Module 2 w: Higher scores in quality of life (p = .01). Medium effect size in self‐efficacy (p = .90). No difference in health motivation, self‐management skills and knowledge (p > .05). 3.4 Moderate Cozzi et al.32 Self‐Concept, Pain, Anxiety, In‐patient Hospital Visits, Emergency Room (ER) Visits State‐Trait Anxiety Inventory for adults or children, Tennessee Self‐Concept Scale for adults, Piers‐ Harris Children's Self‐Concept Scale, Human Drawings Test 13 w: Nonsignificant change in in‐patient and emergency room visits and self‐concept. Improvement in anxiety for children (p < .01) and adolescents (p < .05); reported headaches as crises symptoms (p < .01); reported pain medication days (p < .001); reported pain intensity (p < .001) and self‐treated crises (p < .05). Moderate Daniel et al.37 HRQOL, School Attendance, Access to School Resources Haematology/Oncology Psycho‐Educational Needs Assessment, Paediatric QOL Inventory, Woodcock‐Johnson III, Wechsler Abbreviated Scale of Intelligence, Social Problem‐Solving Inventory‐Revised Short Form 4 w, 6 m: No impact on all the outcomes. High Dobson and Byrne38 Self‐Efficacy, School Attendance, Analgesic Use, ER Visits, Pain Intensity Pain Assessment Tool (investigator‐adapted), Sickle Cell Self‐Efficacy Scale (SCSES), Pain Diary (designed for the study) 1 m: improvement in self‐efficacy score (p = .000) and pain intensity (p = .003), nonsignificant reduction in missed school days, use of analgesics and acute care visits. Moderate Dobson39 Disease‐specific Self‐Efficacy SCSES 1 m: Greater disease‐specific self‐efficacy (p < .000) Poor Estepp et al.40 Hydroxyurea Adherence, Hospitalization Rates, Clinic Visits Medication Possession Ratio (days medication in family's possession/days prescribed medication), Haematologic Laboratory Parameters (HbF levels, mean corpuscular volumes, bilirubin levels, absolute reticulocyte counts), Medical Record Review 3–12 m: Significant improvement in haematological parameters (p < .03) and ER attendance (p = .0013). Nonsignificant improvement in hospitalization rates (p = .79) and medication possession ratio (p = .99). A change in participant's medication possession ratio was the only predictor of improved HbF levels after intervention initiation Moderate Fouda et al.41 Knowledge, QOL, Self‐Care Practices Paediatric QOL Inventory. Other outcomes were measured using checklists designed for the study 3 d, 3 m: improvement in all outcomes (p ≤.05). Positive correlation between knowledge and QOL and reported self‐care practices and QOL (p ≤.001) Moderate Gil et al.42 Pain Sensitivity, Coping Strategies Forgione‐Barber Focal Pressure Stimulator, Coping Strategies Questionnaire for SCD 1 w: Lower negative thinking scores (p < .05) but nonsignificant interaction effects for coping attempts and illness‐focused strategies. 3.4 Less pain Poor References Outcome Measure Result or finding (d, day; w, week; m, month) Study quality rating Crosby et al.64 Health‐Related Quality of Life (HRQOL), Self‐ Efficacy, Self‐Management Skills, Knowledge, Health Motivation Patient Activation Measure, Transition Readiness Questionnaire, Disease‐Specific Knowledge Questionnaire, Treatment Self‐Regulation Questionnaire, Paediatric QoL Inventory SCD Module 2 w: Higher scores in quality of life (p = .01). Medium effect size in self‐efficacy (p = .90). No difference in health motivation, self‐management skills and knowledge (p > .05). Moderate Cozzi et al.32 Self‐Concept, Pain, Anxiety, In‐patient Hospital Visits, Emergency Room (ER) Visits State‐Trait Anxiety Inventory for adults or children, Tennessee Self‐Concept Scale for adults, Piers‐ Harris Children's Self‐Concept Scale, Human Drawings Test 13 w: Nonsignificant change in in‐patient and emergency room visits and self‐concept. Improvement in anxiety for children (p < .01) and adolescents (p < .05); reported headaches as crises symptoms (p < .01); reported pain medication days (p < .001); reported pain intensity (p < .001) and self‐treated crises (p < .05). Moderate Daniel et al.37 HRQOL, School Attendance, Access to School Resources Haematology/Oncology Psycho‐Educational Needs Assessment, Paediatric QOL Inventory, Woodcock‐Johnson III Wechsler Abbreviated 4 w, 6 m: No impact on all the outcomes. High Woodcock‐Johnson III, Wechsler Abbreviated Scale of Intelligence, Social Problem‐Solving Inventory‐Revised Short Form Dobson and Byrne38 Self‐Efficacy, School Attendance, Analgesic Use, ER Visits, Pain Intensity Pain Assessment Tool (investigator‐adapted), Sickle Cell Self‐Efficacy Scale (SCSES), Pain Diary (designed for the study) 1 m: improvement in self‐efficacy score (p = .000) and pain intensity (p = .003), nonsignificant reduction in missed school days, use of analgesics and acute care visits. Moderate Dobson39 Disease‐specific Self‐Efficacy SCSES 1 m: Greater disease‐specific self‐efficacy (p < .000) Poor Estepp et al.40 Hydroxyurea Adherence, Hospitalization Rates, Clinic Visits Medication Possession Ratio (days medication in family's possession/days prescribed medication), Haematologic Laboratory Parameters (HbF levels, mean corpuscular volumes, bilirubin levels, absolute reticulocyte counts), Medical Record Review 3–12 m: Significant improvement in haematological parameters (p < .03) and ER attendance (p = .0013). Nonsignificant improvement in hospitalization rates (p = .79) and medication possession ratio (p = .99). A change in participant's medication possession ratio was the only predictor of improved HbF levels after intervention initiation Moderate Fouda et al.41 Knowledge, QOL, Self‐Care Practices Paediatric QOL Inventory. Other outcomes were measured using checklists designed for the study 3 d, 3 m: improvement in all outcomes (p ≤.05). 3.4 Positive correlation between knowledge and QOL and reported self‐care practices and QOL (p ≤.001) Moderate Gil et al.42 Pain Sensitivity, Coping Strategies Forgione‐Barber Focal Pressure Stimulator, Coping Strategies Questionnaire for SCD 1 w: Lower negative thinking scores (p < .05) but nonsignificant interaction effects for coping attempts and illness‐focused strategies. Less pain Poor iley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License 13697625, 2023, 2, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/hex.13692 by Test, Wiley Online Library on [02/05/2 POKU ET AL. 605 References Outcome Measure Result or finding (d, day; w, week; m, month) Study quality rating during low levels of laboratory pain stimulation (p < .02). positive and significant correlation between improvement in coping strategies (negative thinking) and improvement in pain sensitivity (p < .01) Gil et al.47 Depression, Anxiety, Pain (sensitivity, sleep quality, functionality, healthcare contacts) Coping Strategies Children's Depression Inventory, Revised Children's Manifest Anxiety Scale, Pain Diary, Coping Strategies Questionnaire, Structured Pain Interview 2–3 w, 1 m: nonsignificant improvement in all outcomes (p > .05) Poor Green et al.44 Hydroxyurea Adherence Prescription Refill Adherence, Morisky Self‐Report Scale, Foetal Haemoglobin (HbF) level 1–6 m: Nonsignificant improvement in pharmacy refill (p = .33) and self‐reported medication adherence. Less decrease in Personal best HbF in the intervention group compared to controls (p = .009). Controlling for intervention month and group assignment, the intervention group, progressed to Personal best by 2.3% per month during Months 0–4 (p = .30). Moderate Hazzard et al.45 Knowledge, Perceived Social Support, Coping Strategies How Much Do I Know About SCD Scale, Perceived Social Support‐Friends Scale, Kidcope At hospital discharge: Significant effect on all outcomes (p < .001). Moderate Hood et al.46 Self‐Efficacy, Self‐Management (goal and confidence) Transition Readiness Assessment Questionnaire, Patient Activation Measure 6 w: Nonsignificant improvement in self‐efficacy (p > .05). improvement in self‐management confidence, goal‐setting and progress tracking but the significance of improvement is not indicated Poor Kaslow et al.47 Knowledge, Internalizing and Externalizing Behaviours, Family/Social Functioning and Support Children's Depression Inventory, Family Adaptability and Cohesion Evaluation Scale, Child Behaviour Checklist, Sickle Cell Disease Knowledge Test (designed for the study) 6 w: Improvement in outcomes after the last session (p < .01). 6 m: Improvement in CYP's knowledge sustained. 3.4 Improvement in parental knowledge and internalizing behaviours reduced (p < .05) High Ketchen et al.48 Knowledge, Health promotion Activities, QOL, Depression, Parent‐Child Relationship Quality How Much Do I Know About SCD Scale, Health Promotion Activities Checklist (designed for the study), Paediatric QOL Inventory, Children's Depression Inventory, Interaction Behaviour Questionnaire At hospital discharge, 2 m: No significant effect on knowledge, health promotion and depressive symptoms. Significant improvement in QOL (p = .05) and parent‐child relationship (p = .02)) High Rodgers‐Melnick et al.52 Self‐Efficacy, Trust, Knowledge, ER Visits, Hospitalization SCSES, Seidman Sickle Cell Knowledge Quiz (designed for the study), Wake Forest Trust in Medical Profession Scale 3–12 m: Significant improvement in knowledge (p = .0002). No effect on self‐efficacy (p = .37), trust (p = .08), ER visits (p = .33) and hospitalization (p = .2) Poor (Continues) during Months 0–4 (p = .30). Hazzard et al.45 Knowledge, Perceived Social Support, Coping Strategies How Much Do I Know About SCD Scale, Perceived Social Support‐Friends Scale, Kidcope At hospital discharge: Significant effect on all outcomes (p < .001). Moderate Hood et al.46 Self‐Efficacy, Self‐Management (goal and confidence) Transition Readiness Assessment Questionnaire, Patient Activation Measure 6 w: Nonsignificant improvement in self‐efficacy (p > .05). improvement in self‐management confidence, goal‐setting and progress tracking but the significance of improvement is not indicated Poor Kaslow et al.47 Knowledge, Internalizing and Externalizing Behaviours, Family/Social Functioning and Support Children's Depression Inventory, Family Adaptability and Cohesion Evaluation Scale, Child Behaviour Checklist, Sickle Cell Disease Knowledge Test (designed for the study) 6 w: Improvement in outcomes after the last session (p < .01). 6 m: Improvement in CYP's knowledge sustained. Improvement in parental knowledge and internalizing behaviours reduced (p < .05) High Ketchen et al.48 Knowledge, Health promotion Activities, QOL, Depression, Parent‐Child Relationship Quality How Much Do I Know About SCD Scale, Health Promotion Activities Checklist (designed for the study), Paediatric QOL Inventory, Children's Depression Inventory, Interaction Behaviour Questionnaire At hospital discharge, 2 m: No significant effect on knowledge, health promotion and depressive symptoms. Significant improvement in QOL (p = .05) and parent‐child relationship (p = .02)) High Rodgers‐Melnick et al.52 Self‐Efficacy, Trust, Knowledge, ER Visits, Hospitalization SCSES, Seidman Sickle Cell Knowledge Quiz (designed for the study), Wake Forest Trust in Medical Profession Scale 3–12 m: Significant improvement in knowledge (p = .0002). 3.4 No effect on self‐efficacy (p = .37), trust (p = .08), ER visits (p = .33) and hospitalization (p = .2) Poor (Continues) References Outcome Measure Result or finding (d, day; w, week; m, month) Study quality rating during low levels of laboratory pain stimulation (p < .02). positive and significant correlation between improvement in coping strategies (negative thinking) and improvement in pain sensitivity (p < .01) Gil et al.47 Depression, Anxiety, Pain (sensitivity, sleep quality, functionality, healthcare contacts) Coping Strategies Children's Depression Inventory, Revised Children's Manifest Anxiety Scale, Pain Diary, Coping Strategies Questionnaire, Structured Pain Interview 2–3 w, 1 m: nonsignificant improvement in all outcomes (p > .05) Poor Green et al.44 Hydroxyurea Adherence Prescription Refill Adherence, Morisky Self‐Report Scale, Foetal Haemoglobin (HbF) level 1–6 m: Nonsignificant improvement in pharmacy refill (p = .33) and self‐reported medication adherence. Less decrease in Personal best HbF in the intervention group compared to controls (p = .009). Controlling for intervention month and group assignment, the intervention group, progressed to Personal best by 2.3% per month during Months 0–4 (p = .30). Moderate Hazzard et al.45 Knowledge, Perceived Social Support, Coping Strategies How Much Do I Know About SCD Scale, Perceived Social Support‐Friends Scale, Kidcope At hospital discharge: Significant effect on all outcomes (p < .001). Moderate Hood et al.46 Self‐Efficacy, Self‐Management (goal and confidence) Transition Readiness Assessment Questionnaire, Patient Activation Measure 6 w: Nonsignificant improvement in self‐efficacy (p > .05). improvement in self‐management confidence, goal‐setting and progress tracking but Poor ley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License 606 \| POKU ET AL. POKU ET AL. 606 References Outcome Measure Result or finding (d, day; w, week; m, month) Study quality rating Saulsberry et al.53 Knowledge, Self‐Management Confidence Disease Knowledge Tool, Self‐Management Skills Checklist (designed for the study) 12 m: Median self‐management confidence score was 8 (range 5–10), and the median knowledge assessment score was 79 (range 37–100). A positive correlation between the number of modules completed and the disease knowledge score (p = .003). No correlation was found between the number of modules completed and the self‐management confidence ratings (p = .945). 3.5 \| Effectiveness References Outcome Measure Result or finding (d, day; w, week; m, month) Study quality rating Viola et al.61 Transition Readiness, Disease‐Related QOL, Disease‐Related Self‐Efficacy, Health Literacy, Knowledge Sickle Cell Transition Intervention Programme‐ Readiness for Transition, Adult Sickle Cell QOL Measurement System, SCSES, Morisky Medication Adherence Scale 6 m: Significant improvement in transition readiness (p < .01) and self‐efficacy (p = .002). Nonsignificant improvement in QOL, health literacy and knowledge High Yoon and Godwin63 Knowledge, Self‐Management Confidence Knowledge checklist (designed for the study), 100 mm horizontal Visual Analogy Scale 0 d: improvement in knowledge (p = .010) and self‐ management confidence (p = .001) Moderate bbreviations: CYP, children and young people; HRQOL, health‐related quality of life; QOL, quality of life; SCD, sickle cell disease. Seventeen studies (20 articles) evaluated the impact of self‐management programmes on psycho‐behavioural outcomes (self‐efficacy, anxiety, depression and coping).28,30‐32,34,36,38,39,42,43,45‐48,52,54,57‐59,61 Findings from five studies associated self‐management interventions with statistically significant positive impacts on self‐efficacy.28,38,39,61,64 In one study, improved self‐efficacy was correlated with improved self‐ management confidence and behaviours.46,64 Similarly, five studies reported significant improvements in coping and the use of positive coping strategies.27,28,45,54,56 Coping outcomes were reported primarily for pain. Mixed findings were reported regarding self‐management interventions' impact on pain frequency and intensity.30,37,42,43,49,54‐56 Only one study reported significant improvement in anxiety and depression postintervention.28 The effective interventions for psycho‐ behavioural outcomes were predominantly in‐person, hospital‐based and multisessional. The overall strength of the evidence base for the psycho‐ behavioural outcomes limits confidence in the results, particularly since 16 of the 20 articles were of moderate to poor methodological quality. Only one high‐quality study associated the self‐management interven- tion (e‐mentorship by medical students) with significant improvements in self‐efficacy.61 Furthermore, effectiveness appeared to reduce overtime for self‐efficacy and coping.30,31,47,52 Eight studies of mixed quality examined the impact of self‐ management interventions on healthcare utilization which included emergency room visits, in‐patient admission duration and/or hospi- talization rate.29,30,32,35,38,40,52,56 Only a 12‐month CYP‐personalized transition programme delivered by a designated social worker reported a statistically significant reduction in hospitalization rate at 12 months follow‐up.29 This study involved the largest number of participants (n = 110); however, outcome measurement was based on a retrospective record review, which would not have included hospitalizations outside the study centre. 3.4 Moderate Schatz et al.54 Coping Strategies, Pain (frequency, intensity, functionality) Coping Strategies Questionnaire for SCD, Daily Pain and Activity Diary (electronic) 2 m: Increase in self‐reported use of cognitive‐ behavioural skills (p = .03) and beliefs in pain controllability (p = .02). Moderate Sil et al.57 Knowledge, Pain Coping Efficacy Knowledge of Psychological Interventions for Pain, Child Pain Coping Self‐Efficacy Scale End of each session: Only significant improvement postintervention for session 1 (<.01). No significant difference in coping scores and self‐ efficacy remained relatively similar across all sessions Moderate Sil et al.56 Healthcare Utilization, Treatment Adherence, Pain Intensity, Functional Disability, Pain Coping Efficacy Medical Record Review, Numeric Rating Scale (designed for the study), Functional Disability Inventory, Pain Coping Questionnaire Every 2–3 sessions, 8 w: No significant improvement in healthcare utilization (p > .17). 31.6% of participants in the treatment group reported a statistically significant decrease in pain intensity, functional disability and improved coping efficacy. Poor Smaldone et al.58 Hydroxyurea Adherence, HRQOL, Youth‐Parent Self‐Management Responsibility Concordance Paediatric QOL Inventory, Paediatric QOL SCD Module, Sickle Cell Family Responsibility Scale (adapted from the Diabetes Family Responsibility Questionnaire) 3 m: Improved dyad self‐management responsibility concordance (3.5 points, 95% CI: −0.2, 7.1). 6 m: Improved generic HRQOL (9.8 points, 95% CI: 0.4, 19.2), No difference in QOL Moderate Smith et al.59 Transition Readiness, Disease‐Related Self‐Efficacy, Disease‐Related Stress, Feelings about Transition, Transfer Success, Patient Retention Transition Intervention Programme‐Readiness For Transition Assessment Tool (designed for the study), SCSES, Feelings‐Sickle Cell Transfer Questionnaire, Sickle Cell Stress Scale‐Adolescent (designed for the study) 36 m: Improvements in all outcomes but statistical/ clinical significance not indicated Poor Treadwell and Weissman60 Treatment Compliance, Knowledge, Perceived Social Support, Parent‐Child Responsibility for Medication Management 24‐h recall interview, Multi‐Choice Questionnaires (designed for the study), Perceived Competence Scale for Children 2 w: Nonsignificant improvement in outcomes Moderate iley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License 607 POKU ET AL. 607 Negative aspects included the burden of home‐based self‐directed activities,47 the lack of mobile phone‐based technology interventions,50,64 the frequency of reminder notifications and the surveillance features of some interventions.44,50,58,61 Interestingly, participants recommended incorporating in‐person sessions for interventions involving virtual platforms for social interaction and strengthening peer support.61,64 Some studies described how the professionals delivering the interven- tions31,61 reported improved communication and relationships with CYP and improved understanding of their challenges. unvalidated measurement tools and had small sample sizes and short‐ term follow‐ups. Six studies measured medical adherence as medication adher- ence, CYP responsibility for treatment, clinic attendance or transition success.29,40,44,58,60,61 Four studies reported significant improve- ments in these outcomes.29,44,58 The effective interventions were behavioural and skills training programmes delivered by a designated social care provider29,44,58 and an educational programme delivered by lay counsellors.60 None of the technological interventions using treatment reminders improved treatment adherence.40,61 TABLE 10 Recommendations for future self‐management interventions TABLE 10 Recommendations for future self‐management interventions Self‐management programmes/interventions need to: Self‐management programmes/interventions need to: 3.5 \| Effectiveness y.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License Health‐related quality of life was measured in eight studies.28,35‐ 37,41,48,58,61 Only three studies of moderate quality reported statistically significant improvement in quality of life.28,41,64 How- ever, in two, improvement was only observed in the social domain.28,64 The high‐quality studies identified no significant improvement in quality of life.37,61 Overall, our greatest confidence is in the evidence of improvements in SCD knowledge and social outcomes, where studies of predominantly moderate methodological quality reported positive outcomes (Table 9). Sixteen studies of strong to moderate quality reported a statistically significant impact of self‐management interventions on these out- comes.27,29,36,38,40,41,44,45,47,48,52,57‐61,63 Eight studies found that educa- tional programmes on SCD and pain improved knowledge and under- standing.27,40,47,52,57,60,63,64 Six studies found a positive association between improved knowledge and understanding and self‐management confidence, behaviours and practices.27,41,46,53,63,64 Similarly, six studies also found improvements in social functioning and perceived social support.38,45,47,48,59,60 Although evidence on knowledge and social outcomes is based upon high‐quality studies, many used different POKU ET AL. 608 Abbreviations: CYP, children and young people; SCD, sickle cell disease. 3.6 \| Acceptability and engagement ne Library for rules of use; OA articles are governed by the applicable Creative Commons License TABLE 10 Recommendations for future self‐management interventions Self‐management programmes/interventions need to: m‐Health self‐management interventions need to: Be personalized to CYP's goals and values.34,45,46 Be family‐oriented.30 Use multiple formats and delivery approaches to enhance flexibility, convenience, acceptability, usefulness and engagement.37,50 Include components to facilitate communication with health and social care professionals.50 Develop the broad range of knowledge and skills for managing SCD.50 Support older CYP's capacity to navigate psychological, relational, educational, vocational, economic and health and social care system challenges.29,30,43,47,64 Be integrated within the wider healthcare system.46 Involve schools.47 Designate a lead for self‐management in the multidisciplinary team to support CYP.30,59,61,62 Employ user‐centred design.51 Mitigate risks to privacy.46 Consider time, resources, monitoring and technology maintenance.46 Include features to allow CYP to self‐assess their progress and intervention impact.51 Help CYP communicate health, illness and self‐management information with health care professionals by enabling them to capture visual feedback in real‐time to share.51 Abbreviations: CYP, children and young people; SCD, sickle cell disease. ne Library for rules of use; OA articles are governed by the applicable Creative Commons License 3.6 \| Acceptability and engagement Ten RCTs,30,37,43,44,46,47,49,50,55,58 eight quasi‐experimental stud- ies,28,32,45,51,52,57,61,62 three cross‐sectional studies,33,35 two cohort studies40,53 and one qualitative study64 reported on participants' satisfaction and engagement with the interventions. Levels of satisfaction and engagement were measured by survey and were assessed only for participants who completed the interventions. Studies reported high levels of satisfaction and acceptance, with those associated with the highest satisfaction levels being interventions that were delivered in person or remotely;28,30,32‐37,44,47,51,55,61,62,64 involved peers with SCD or parents32,34,45‐47,52 and supported pain coping skills and strate- gies,30,32,34,46,47 SCD knowledge and understanding,30,34,46,47,52 advocacy skills46 and direct interaction with care providers.40,61 Participants valued interventions that were engaging, interactive, motivating and enjoy- able.33‐37,46,52,64 The main challenges reported for intervention implementation related to study recruitment and retention and provider (interven- tionist) availability. Eleven studies reported high follow‐up attrition rates, which ranged between 32%40 and 53%57 (Table 2). Participants' responses for not completing the intervention and/ or outcome assessments included the burden of completing outcome measures,44 loss of interest,37,52,54 dislike of intervention format,50,54 acute illness and hospitalization,35,48,57 and transport and time constraints.48,56 Some eligible participants were excluded or lost to follow‐up due to the intervention times being limited to the clinic's working hours. Barriers to participation and engage- ment in the m‐Health interventions included device and technical malfunctions.46,49,54 Implementation facilitators included patient navigation systems, study schedule flexibility, parental, family and clinical staff involvement, third‐sector organization partnership and accessible m‐Health programmes.31,34,37,46,48,51,52 The only healthcare system implementation issue reported was provider availability57,59 when competing for clinical demands and emer- gency consultations often interfered with intervention delivery. Training each multidisciplinary team member to deliver self‐ management interventions was reportedly facilitative. Several studies made recommendations to facilitate future intervention acceptability and engagement (Table 10). However, participants' engagement with digital interventions was reported to be variable, decreasing over time. 3.6 \| Acceptability and engagement Engagement with interventions was reported to be higher amongst parent participants than CYP for the interventions that involved CYP‐parent dyads.44,51,58 In addition, engagement levels were reported to be higher among younger CYP participants than in the older CYP cohort in the same studies.34,43,46,47 Some studies reported on CYP participants' feedback regarding the negative aspects of the interventions and their preferences.44,47,50,58,61,64 TABLE 10 Recommendations for future self‐management interventions Self‐management programmes/interventions need to: m‐Health self‐management interventions need to: Be personalized to CYP's goals and values.34,45,46 Be family‐oriented.30 Use multiple formats and delivery approaches to enhance flexibility, convenience, acceptability, usefulness and engagement.37,50 Include components to facilitate communication with health and social care professionals.50 Develop the broad range of knowledge and skills for managing SCD.50 Support older CYP's capacity to navigate psychological, relational, educational, vocational, economic and health and social care system challenges.29,30,43,47,64 Be integrated within the wider healthcare system.46 Involve schools.47 Designate a lead for self‐management in the multidisciplinary team to support CYP.30,59,61,62 Employ user‐centred design.51 Mitigate risks to privacy.46 Consider time, resources, monitoring and technology maintenance.46 Include features to allow CYP to self‐assess their progress and intervention impact.51 Help CYP communicate health, illness and self‐management information with health care professionals by enabling them to capture visual feedback in real‐time to share.51 Abbreviations: CYP, children and young people; SCD, sickle cell disease. nline Library for rules of use; OA articles are governed by the applicable Creative Commons License TABLE 10 Recommendations for future self‐management interventions Self‐management programmes/interventions need to: m‐Health self‐management interventions need to: Be personalized to CYP's goals and values.34,45,46 Be family‐oriented.30 Use multiple formats and delivery approaches to enhance flexibility, convenience, acceptability, usefulness and engagement.37,50 Include components to facilitate communication with health and social care professionals.50 Develop the broad range of knowledge and skills for managing SCD.50 Support older CYP's capacity to navigate psychological, relational, educational, vocational, economic and health and social care system challenges.29,30,43,47,64 Be integrated within the wider healthcare system.46 Involve schools.47 Designate a lead for self‐management in the multidisciplinary team to support CYP.30,59,61,62 Employ user‐centred design.51 Mitigate risks to privacy.46 Consider time, resources, monitoring and technology maintenance.46 Include features to allow CYP to self‐assess their progress and intervention impact.51 Help CYP communicate health, illness and self‐management information with health care professionals by enabling them to capture visual feedback in real‐time to share.51 Abbreviations: CYP, children and young people; SCD, sickle cell disease. 4 \| DISCUSSION A recent review suggests that involving CYP in designing and delivering self‐management interventions improves effectiveness.69 Ignoring the agency of CYP, particularly from minority backgrounds like CYP with SCD, also has the potential to create and deepen health disparities and inequalities. This supports embedded co‐ production to ensure self‐management intervention design and implementation are consistent with CYP's needs, voices, experiences and expectations rather than prioritizing medical adherence. Only a third of the technological interventions (4/12) showed effectiveness. These findings are consistent with the findings of other reviews in which limited data was found to support the effectiveness of mobile or web‐based applications for self‐management of adolescents with a chronic illness68 and for treatment adherence among CYP and adults with SCD.19,21 There was limited evidence for interventions improving healthcare utilization or quality of life. Conflicting and inconsistent evidence for the effectiveness of self‐ management interventions on quality of life in CYP with chronic illnesses has been found and reported by others.66 In addition to previous research on the effectiveness of self‐management for CYP with chronic illnesses,66 this review revealed that Interventions that combined technological platforms and in‐person group‐based ap- proaches and involved family and care providers appeared to be more acceptable to CYP. Further, interventions focused on role management (i.e., social domains and social participation) were limited. Despite the impor- tance of the social context on self‐management,70,71 the included studies overlooked home, school and/or work environments. This is despite several research studies reporting the difficulties of engaging in SCD self‐management practices in these settings.9,11,13‐15 The lack of accommodation for SCD self‐management practices in the home, school and workplace has been highlighted, including how school policies and practices contradict self‐management.9,11,13‐15 At the same time, self‐management practices may be (dis)regarded as interference and perceived as disruptive and stigmatizing by others and CYP themselves.9‐11,13‐15,72 Future interventions should, there- fore, consider how families and school and/or work contexts can support CYP's ability to self‐management. iley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License However, confidence in these results is limited by the methodo- logical quality of the included papers. Most interventions are in the early stages of evaluation, and most papers report findings from retrospective observational studies, single‐arm pre‐ and poststudies or small feasibility/pilot trials. Several studies were characterized by high follow‐up attrition rates and demonstrated significant potential risk for bias. 4 \| DISCUSSION Consequently, more robust definitive RCTs conducted across multiple and different settings are needed. Future trials should be methodologically robust in terms of randomization, larger sample size and power calculations, blinding and outcome measures (i.e., use of age‐appropriate validated tools, effect size calculations, clinical significance indications and longer follow‐up periods). Qualitative studies should be nested within future trials to help understand intervention effectiveness and acceptability. This is particularly important as patient‐reported outcomes measures (e.g., quality of life, psychosocial well‐being) are highly variable and qualitative methods will provide additional explanatory information that can facilitate the interpretation of trial outcomes. More economic evaluations of interventions are also needed. In addition, more research is needed to develop and validate health‐related outcome measurement tools for CYP with SCD to support intervention and programme evaluations. The development and evaluation of self‐management interventions to support CYP with SCD is a growing area of research, particularly in the United States, with 50% of the included papers being published in the last 5 years. This paper reports the first comprehensive review of the evidence on their effectiveness and acceptability. In terms of their characteristics, the interventions developed are wide‐ranging in relation to their purpose, content, format and delivery methods. However, only a third of these (10/32) demonstrated statistically significant improvements in the outcomes measured.27‐29,38,39,41,45,47,54,63 These interventions were delivered in‐person, involved educational components, skills training for pain management and medical adherence and/or promoted family and/or peer support; and reported improvements in knowledge, social functioning and/or medical adherence. However, it is unclear if these improvements are maintained over time. Similar to the findings of previous reviews on CYP with chronic conditions,66,67 the majority of the included interventions were focused on medical management (i.e., disease‐specific and healthcare‐related) and/or emotional management (i.e., feelings and intrinsic characteristics). This notwith- standing, no clear effects were found in favour of any interventions aimed at symptom reduction. The lack of, or where sought, the minimal involvement of CYP in the intervention development and implementation and study design and conduct is a limitation of the included studies. This may explain why several interventions included features that CYP found burden- some and unhelpful, as well as problems with recruitment, retention and engagement. Self‐management programmes/interventions need to: Be personalized to CYP's goals and values.34,45,46 Be family‐oriented.30 Use multiple formats and delivery approaches to enhance flexibility, convenience, acceptability, usefulness and engagement.37,50 Be integrated within the wider healthcare system.46 Involve schools.47 Designate a lead for self‐management in the multidisciplinary team to support CYP.30,59,61,62 POKU ET AL. 609 ACKNOWLEDGEMENTS We thank the following people for their insights into the conduct of the review: Dr. John Grainger (MD, Manchester Royal Children's Hospital), Ms. Iyamide Thomas (Sickle Cell Society UK) and our PPI advisors: Miss. Rachael Oshinbolu, Miss. Edith Kyewalyanga and Mr. Abubakar Mohammed. This project is funded by the National Institute for Health and Care Research (NIHR) under its Research for Patient Benefit (RfPB) Programme (Grant Reference Number NIHR202169). The views expressed are those of the authors and not necessarily those of the NIHR or the Department of Health and Social Care. REFERENCES 1. Piel FB, Hay SI, Gupta S, Weatherall DJ, Williams TN. Global burden of sickle cell anaemia in children under five, 2010‐2050: modelling based on demographics, excess mortality, and interventions. PLoS Med. 2013;10:e1001484. 2. Preiss DJ. The young child with sickle cell disease. Adv Nurse Pract. 1998;6(6):32‐39. 2. Preiss DJ. The young child with sickle cell disease. Adv Nurse Pract. 1998;6(6):32‐39. 3. Rees DC, Williams TN, Gladwin MT. Sickle‐cell disease. The Lancet. 2010;376(9757):2018‐2031. 3. Rees DC, Williams TN, Gladwin MT. Sickle‐cell disease. The Lancet. 2010;376(9757):2018‐2031. 4. Platt OS, Brambilla DJ, Rosse WF, et al. Mortality in sickle cell disease – life expectancy and risk factors for early death. N Engl J Med. 1994;330(23):1639‐1644. The authors declare no conflict of interest. The authors declare no conflict of interest. The authors declare no conflict of interest. This is the first review of self‐management interventions for CYP with SCD that synthesizes the evidence from the range of programmes developed, the range of outcomes assessed and the diversity of evaluative study designs. Including young people with SCD in the review's conduct has enhanced the credibility of the process. However, there are some limitations. Only peer‐reviewed articles published in English were included, posing a potential publication bias. It was impossible to synthesize study results using meta‐analysis due to the heterogeneity of the interventions and outcome measures. As previously described, the robustness of the included studies limits the confidence that can be placed on the results. In addition, the cost‐effectiveness or cost‐benefits of interventions has not been evaluated, and there is a lack of qualitative research to inform understanding of intervention effec- tiveness and acceptability. Future studies need to be robust in design, conducted across multiple and different settings and collect long‐term outcome data to assess the effectiveness, acceptability, cost‐benefits and generalizability of interventions across populations and settings. In addition, it is important that more validated health‐related outcome tools are developed for CYP with SCD; CYP are involved in intervention development, study design and implementation; CYP participants from non‐Black backgrounds are included and that attention is paid to the wider context of CYP's lives when designing and evaluating self‐management interventions. Overall, the highlighted limitations mean that there is limited evidence for any single self‐management programme or programme components in relation to promoting health and social outcomes for CYP with SCD. However, we found effective and acceptable interventions appear to include those that provide support from peers and trusted adults (e.g., parents, teachers, relatives and care providers). While the ultimate goal for self‐management is to increase CYP with SCD autonomy and independence in managing their illness, they depend on others for encouragement, communication and treatment support. Adults (e.g., parents, relatives, care providers, teachers and adults with SCD) can provide information on how CYP can independently handle health and social care needs, while peers can learn from and support one another by sharing personal experiences. A recent review highlights the pivotal role peers and adults play in supporting the adoption and promotion of self‐ management strategies in CYP with chronic conditions.69 Future interventions should, therefore, include components that facilitate peer and significant adult support. The involvement of peers, families, care providers and teachers is particularly important because CYP with SCD have been found to experience significant stigma in relation to their illness management, even from families, care providers and teachers due to their illness and race/ethnicity.9,11,72,73 Brenda A. Poku https://orcid.org/0000-0002-2265-7041 Brenda A. Poku https://orcid.org/0000-0002-2265-7041 DATA AVAILABILITY STATEMENT Data sharing does not apply to this article as no data sets were generated or analysed during the current study. 610 \| 610 POKU ET AL. 4 \| DISCUSSION There were inconsistencies and wide variations in the outcome measurements used across the studies, with some unvalidated tools. This reflects the current state of limited availability of validated health‐ related outcome tools for CYP with SCD specifically and the SCD population in general. Further, none of the studies compared the different types of intervention components or examined if specific self‐ management intervention components are more or less effective and for which patient and health and social setting characteristics, which would allow for creating targeted interventions. Including CYP with SCD from diverse backgrounds or developing interventions inclusive of diverse cultures is important. However, no study included CYP with SCD from Eastern Mediterranean, Middle Eastern and Southeast Asian backgrounds. Indeed, SCD appeared to be ‘problematised’ as a ‘black illness’ in the included studies. Future research should, therefore, strive to involve CYP with SCD from non‐Black backgrounds. This could be achieved through cross‐country, cross‐ institutional and cross‐centre research collaborations; development of interventions that are diverse (in terms of language and content); targeted recruitment of non‐Black populations and more non‐US‐based studies. 13697625, 2023, 2, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/hex.13692 by Test, Wiley Online Library on [02/05/2023]. See the Terms and C 2017;64(1):139‐145. 16. Modi AC, Pai AL, Hommel KA, et al. Pediatric self‐management: a framework for research, practice and policy. Pediatrics. 2012;129(2): e473‐e485. 35. Crosby LE, Joffe NE, Peugh J, Ware RE, Britto MT. Pilot of the chronic disease self‐management program for adolescents and young adults with sickle cell disease. J Adolesc Health. 2017;60(1): 120‐123. 17. Paterson GA, Nayda RJ, Paterson JA. Chronic condition self‐ management: working in partnership toward appropriate models for age and culturally diverse clients. Contemp Nurse. 2012;40(2): 169‐178. 36. Crosby LE, Hood A, Kidwell K, et al. Improving self‐management in adolescents with sickle cell disease. Pediatr Blood Cancer. 2020;67(10):e.28492. doi:10.1002/pbc.28492 18. Edwards LY, Edwards CL. Psychosocial treatments in pain manage- ment of sickle cell disease. J Natl Med Assoc. 2010;102(11): 1084‐1094. 37. Daniel LC, Li Y, Smith K, et al. Lessons learned from a randomized controlled trial of a family‐based intervention to promote school functioning for school‐age children with sickle cell disease. J Pediatr Psychol. 2015;40(10):1085‐1094. 19. Shih S, Cohen LL. A systematic review of medication adherence interventions in pediatric sickle cell disease. J Pediatr Psychol. 2020;45(6):593‐606. 20. Viola A, Porter J, Shipman J, Brooks E, Valrie C. A scoping review of transition interventions for young adults with sickle cell disease. Pediatr Blood Cancer. 2021;68:e29135. 38. Dobson CE, Byrne MW. Original research: using guided imagery to manage pain in young children with sickle cell disease. Am J Nurs. 2014;114(4):26‐36. 21. Badawy SM, Cronin RM, Hankins J, et al. Patient‐centered eHealth interventions for children, adolescents and adults with sickle cell disease: systematic review. J Med Internet Res. 2018;20(7):e10940. 39. Dobson C. Outcome results of self‐efficacy in children with sickle disease pain who were trained to use guided imagery. Appl Nurs Res. 2015;28:384‐390. 40. Estepp JH, Winter B, Johnson M, Smeltzer MP, Howard SC, Hankins JS. Improved hydroxyurea effect with the use of text messaging in children with sickle cell anemia. Pediatr Blood Cancer. 2014;61:2031‐2036. 22. Stern C, Lizarondo L, Carrier J, et al. Methodological guidance for the conduct of mixed methods systematic reviews. JBI Evidence Synthesis. 2020;18(10):2108‐2118. 23. Page MJ, McKenzie JE, Bossuyt PM, et al. The PRISMA 2020 statement: an updated guideline for reporting systematic reviews. Syst Rev. 2021;10(89):89. doi:10.1186/s13643-021-01626-4 41. Mahdy Fouda N, Mohamed Adly R, Shafik Mahmoud F, Abd el‐Ghany mohamed R. Effect of self‐learning guidelines on quality of life and self‐care reported practice of adolescents with sickle cell anemia. J Nurs Sci Benha Univ. 13697625, 2023, 2, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/hex.13692 by Test, Wiley Online Library on [02/05/2023]. See the Terms and C POKU ET AL. 611 psychological wellbeing, quality of life and coping of students with sickle cell disease in Nigeria. Int J Disabil Dev Educ. 2020;69: 1095‐1104. doi:10.1080/1034912X.2020.1755422 9. Poku BA, Caress A‐L, Kirk S. “Body as a machine”: how adolescents with sickle cell disease construct their fatigue experiences. Qual Health Res. 2020;30:1431‐1444. 29. 10. Poku BA, Caress A‐L, Kirk S. Adolescents' experiences of living with sickle cell disease: An integrative narrative review of the literature. Int J Nurs Stud. 2018;80:20‐28. 29. Allemang B, Allan K, Johnson C, et al. Impact of a transition program with navigator on loss to follow‐up, medication adherence, and appointment attendance in hemoglobinopathies. Pediatr Blood Cancer. 2019;66:e27781. 11. Poku BA, Pilnick A. Biographical accounts of the impact of fatigue in young people with sickle cell disease. Sociol Health Illn. 2022;44(6): 1027‐1046. 30. Barakat LP, Schwartz LA, Salamon KS, Radcliffe J. A family‐based randomized controlled trial of pain intervention for adolescents with sickle cell disease. J Pediatr Hematol Oncol. 2010;32(7):540‐547. 12. Dyson SM, Abuateya H, Atkin K, et al. Reported school experiences of young people living with sickle cell disorder in England. Br Educ Res J. 2010;36(1):125‐142. 31. Broome ME, Maikler V, Kelber S, Bailey P, Lea G. An intervention to increase coping and reduce health care utilization for school‐age children and adolescents with sickle cell disease. J Natl Black Nurses Assoc. 2001;12(2):6‐14. 13. Dyson SM, Atkin K, Culley LA, Dyson SE, Evans H. Sickle cell, habitual dys‐positions and fragile dispositions: young people with sickle cell at school. Sociol Health Illn. 2011;33(3):465‐483. 32. Cozzi L, Tryon WW, Sedlacek K. The effectiveness of biofeedback‐ assisted relaxation in modifying sickle cell crises. Biofeedback Self‐ Regul. 1987;12(1):51‐61. 14. Dyson SM, Atkin KM, Berghs MJ, Greene AM. On the possibility of a disabled life in capitalist ruins: black workers with sickle cell disorder in England. Soc Sci Med. 2021;272:113713. 33. Crosby LE, Barach I, McGrady ME, Kalinyak KA, Eastin AR, Mitchell MJ. Integrating interactive web‐based technology to assess adherence and clinical outcomes in pediatric sickle cell disease. Anemia. 2012;2012:1‐8. doi:10.1155/2012/492428 15. Berghs M, Dyson S, Gabba A, et al. I want to become someone!” gender, reproduction and the moral career of motherhood for women with sickle cell disorders. Cult Health Sex. 2022:1‐15. doi:10. 1080/13691058.2022.2083239 34. Crosby LE, Ware RE, Goldstein A, et al. Development and evaluation of iManage: a self‐management app co‐designed by adolescents with sickle cell disease. Pediatr Blood Cancer. 6 \| CONCLUSION 5. Powars DR, Chan LS, Hiti A, Ramicone E, Johnson C. Outcome of sickle cell anemia: a 4‐decade observational study of 1056 patients. Medicine. 2005;84(6):363‐376. 5. Powars DR, Chan LS, Hiti A, Ramicone E, Johnson C. Outcome of sickle cell anemia: a 4‐decade observational study of 1056 patients. Medicine. 2005;84(6):363‐376. Findings generated from this systematic review provide a critical summary of the characteristics, effectiveness, acceptability and feasibility of self‐management interventions for CYP with SCD. This review found that interventions that promoted peer, parent/family and care provider support and combined technology and in‐person group methods appear to be associated with the effectiveness and acceptability to CYP. However, there is limited evidence for any single self‐management programme or programme components in relation to promoting health/social outcomes for CYP with SCD. 6. Lorig KR, Holman HR. Self‐management education: history, defini- tion, outcomes and mechanisms. Ann Behav Med. 2003;26(1):1‐7. 6. Lorig KR, Holman HR. Self‐management education: history, defini- tion, outcomes and mechanisms. Ann Behav Med. 2003;26(1):1‐7. 7. Schulman‐Green D, Jaser S, Martin F, et al. Processes of self‐ management in chronic illness. J Nurs Scholarsh. 2012;44(2): 136‐144. 7. Schulman‐Green D, Jaser S, Martin F, et al. Processes of self‐ management in chronic illness. J Nurs Scholarsh. 2012;44(2): 136‐144. 8. Nightingale R, McHugh G, Kirk S, Swallow V. Supporting children and young people to assume responsibility from their parents for the self‐management of their long‐term condition: An integrative review. Child Care Health Dev. 2019;45(2):175‐188. 612 \| 612 \| 612 POKU ET AL. for sickle cell pain: a patient‐informed, video‐based pain manage- ment intervention for adolescents with sickle cell disease. Clin Pract Pediatr Psychol. 2020;8(2):150‐163. 47. Kaslow NJ, Collins MH, Rashid FL, et al. The efficacy of a pilot family psychoeducational intervention for pediatric sickle cell disease (SCD). Fam Syst Health. 2000;18(4):381‐404. 63. 48. Ketchen B, Hazzard A, Lassiter S, et al. STARBRIGHT world: a pilot study of a home‐based sickle cell psychoeducational intervention. Chil Health Care. 2006;35(4):321‐338. 63. Yoon SL, Godwin A. Enhancing self‐management in children with sickle cell disease through playing a CD‐ROM educational game: a pilot study. Pediatr Nurs. 2007;33(1):60‐63. 49. McClellan CB, Schatz JC, Puffer E, Sanchez CE, Stancil MT, Roberts CW. Use of handheld wireless technology for a home‐ based sickle cell pain management protocol. J Pediatr Psychol. 2009;34(5):564‐573. 64. Crosby LE, Joffe NE, Kidwell KM, et al. Perceptions of a self‐ management intervention for adolescents with sickle cell disease. Clin Pract Pediatr Psychol. 2022;10(1):79‐90. 65. Jenerette CM, Murdaugh C. Testing the theory of self‐care management for sickle cell disease. Res Nurs Health. 2008;31(4): 355‐369. 50. Palermo TM, Dudeney J, Santanelli JP, Carletti A, Zempsky WT. Feasibility and acceptability of internet‐delivered cognitive behavioral therapy for chronic pain in adolescents with sickle cell disease and their parents. J Pediatr Hematol Oncol. 2018;40(2):122‐127. 66. Bal MI, Sattoe JNT, Roelofs PDDM, Bal R, van Staa A, Miedema HS. Exploring effectiveness and effective components of self‐manage- ment interventions for young people with chronic physical condi- tions: A systematic review. Patient Educ Couns. 2016;99(8): 1293‐1309. 51. Levin ME, Petersen JM, Durward C, et al. A randomized controlled trial of online acceptance and commitment therapy to improve diet and physical activity among adults who are overweight/obese. Transl Behav Med. 2021;11:1216‐1225. doi:10.1093/tbm/ibaa123 67. Sattoe JNT, Bal MI, Roelofs PDDM, Bal R, Miedema HS, van Staa A. Self‐management interventions for young people with chronic conditions: a systematic overview. Patient Educ Couns. 2015;98(6): 704‐715. 52. Rodgers‐Melnick SN, Pell TJG, Lane D, et al. The effects of music therapy on transition outcomes in adolescents and young adults with sickle cell disease. Int J Adolesc Med Health. 2019;31: e20170004. doi:10.1515/ijamh-2017-0004 68. Virella Pérez YI, Medlow S, Ho J, Steinbeck K. Mobile and web‐ based apps that support self‐management and transition in young people with chronic illness: systematic review. J Med Internet Res. 2019;21(11):e13579. doi:10.2196/13579 53. 612 \| Saulsberry AC, Hodges JR, Cole A, Porter JS, Hankins J. Web‐ based technology to improve disease knowledge among adoles- cents with sickle cell disease: pilot study. JMIR Pediatr Parent. 2020;3(1):e15093. doi:10.2196/15093 69. Camp‐Spivey LJ, Logan A, Nichols M. Theoretical and contextual considerations for self‐management strategies of children and adolescents with chronic diseases: an integrative review. J Child Health Care. 2021;26(0):242‐261. doi:10.1177/136749352 11013697 54. Schatz J, Schlenz AM, McClellan CB, et al. Changes in coping, pain, and activity after cognitive‐behavioral training: a randomized clinical trial for pediatric sickle cell disease using smartphones. Clin J Pain. 2015;31(6):536‐547. 70. Ong BN, Rogers A, Kennedy A, et al. Behaviour change and social blinkers? The role of sociology in trials of self‐management behaviour in chronic conditions. Sociol Health Illn. 2014;36(2): 226‐238. 55. Schwartz LA, Radcliffe J, Barakat LP. The development of a culturally sensitive pediatric pain management intervention for African American adolescents with sickle cell disease. Chil Health Care. 2007;36(3):267‐283. 71. Renedo A, Miles S, Marston C. Transitions to adulthood: self‐ governance and disciplining in the making of patient citizens. Sociol Health Illn. 2020;42(3):481‐495. 56. Sil S, Lai K, Lee JL, et al. Preliminary evaluation of the clinical implementation of cognitive‐behavioral therapy for chronic pain management in pediatric sickle cell disease. Complement Ther Med. 2020;49:102348. doi:10.1016/j.ctim.2020.102348 72. Poku BA, Pilnick A, Kirk S. How a child's gender mediates maternal care and expectations in the fatigue experiences of adolescents with sickle cell disease. J Fam Stud. 2022:1‐22. doi:10.1080/13229400. 2022.2060851 57. Sil S, Lee JL, Klosky J, et al. The comfort ability program for adolescents with sickle cell pain: Evaluating feasibility and accept- ability of an inpatient group‐based clinical implementation. Pediatr Blood Cancer. 2021;68(6):e29013. doi:10.1002/pbc.29013 73. Poku BA, Pilnick A. Research knowledge transfer to improve care and support for adolescents with sickle cell disease in Ghana. Health Expect. 2022. doi:10.1111/hex.13573 58. Smaldone A, Findley S, Manwani D, Jia H, Green NS. HABIT, a randomized feasibility trial to increase hydroxyurea adherence, suggests improved health‐related quality of life in youths with sickle cell disease. J Pediatr. 2018;197:177‐185. 13697625, 2023, 2, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/hex.13692 by Test, Wiley Online Library on [02/05/2023]. See the Terms and C 2021;2:400‐415. 24. Joanne Briggs Institute System for the Unified Management, Assessment and Review of Information (JBI SUMARI). Accessed November 24, 2021. https://sumari.jbi.global/ 42. Gil KM, Wilson JJ, Edens JL, et al. Cognitive coping skills training in children with sickle cell disease pain. Int J Behav Med. 1997;4(4): 364‐377. 25. Kirk S, Beatty S, Callery P, Milnes L, Pryjmachuk S. Evaluating self‐ care support for children and young people with long‐term conditions. National Institute for Health and Social Care. 2010. Accessed May 30, 2022. http://www.netscc.ac.uk/hsdr/projdetails. php?ref=08-1715-162 43. Gil KM. Daily coping practice predicts treatment effects in children with sickle cell disease. J Pediatr Psychol. 2001;26(3):163‐173. 44. Green NS, Manwani D, Matos S, et al. Randomized feasibility trial to improve hydroxyurea adherence in youth ages 10‐18 years through community health workers: the HABIT study. Pediatr Blood Cancer. 2017;64(12):e26689. doi:10.1002/pbc.26689 26. Pryjmachuk S, Elvey R, Kirk S, Kendal S, Bower P, Catchpole R. Developing a model of mental health self‐care support for children and young people through an integrated evaluation of available types of provision involving systematic review, meta‐analysis and case study. Health Serv Del Res. 2014;2(18):1‐212. 45. Hazzard A, Celano M, Collins M, Markov Y. Effects of STARBRIGHT world on knowledge, social support, and coping in hospitalized children with sickle cell disease and asthma. Child Health Care. 2002;31(1):69‐86. 27. Abd Elaziz SM, Abd, Elghany RM. Effect of self care management program on pain and fatigue in sickle cell children. Int J Novel Res Healthc Nurs. 2019;6(3):747‐761. 46. Hood AM, Nwankwo C, Walton A, et al. Mobile health use predicts self‐efficacy and self‐management in adolescents with sickle cell disease. Transl Behav Med. 2021;11:1823‐1831. 28. Adegbolagun A, Ani C, Adejumo O, James B, Omigbodun O. Effect of a group‐based cognitive behavioural therapy program on the 612 \| SUPPORTING INFORMATION Additional supporting information can be found online in the Supporting Information section at the end of this article Additional supporting information can be found online in the Additional supporting information can be found online in the Supporting Information section at the end of this article. 59. Smith WR, Sisler IY, Johnson S, et al. Lessons learned from building a pediatric‐to‐adult sickle cell transition program. South Med J. 2019;112(3):190‐197. Supporting Information section at the end of this article. 60. Treadwell MJ, Weissman L. Improving adherence with deferoxamine regimens for patients receiving chronic transfusion therapy. Sem Hematol. 2001;38(suppl 1):77‐84. How to cite this article: Poku BA, Atkin KM, Kirk S. Self‐ management interventions for children and young people with sickle cell disease: a systematic review. Health Expect. 2023;26:579‐612. doi:10.1111/hex.13692 How to cite this article: Poku BA, Atkin KM, Kirk S. Self‐ management interventions for children and young people with sickle cell disease: a systematic review. Health Expect. 2023;26:579‐612. doi:10.1111/hex.13692 61. Viola AS, Drachtman R, Kaveney A, et al. Feasibility of medical student mentors to improve transition in sickle cell disease. J Pediatr Psychol. 2021;46(6):650‐661. 62. Wihak T, Burns M, Miranda J, Windmueller G, Oakley C, Coakley R. Development and feasibility testing of the comfort ability program
https://openalex.org/W2807716735	https://europepmc.org/articles/pmc6006174?pdf=render	English	null	PreDicta chip-based high resolution diagnosis of rhinovirus-induced wheeze	Nature communications	2,018	cc-by	12,462	1 Division of Immunopathology, Department of Pathophysiology and Allergy Research, Center for Pathophysiology, Infectiology and Immunology, Medical University of Vienna, A-1090 Vienna, Austria. 2 Astrid Lindgren Children’s Hospital, Karolinska University Hospital, SE-171 76 Stockholm, Sweden. 3 Department of Women’s and Children’s Health, Karolinska Institutet, SE-171 77 Stockholm, Sweden. 4 Division of Infection, Immunity & Respiratory Medicine, University of Manchester, Manchester M13 9NT, UK. 5 Phadia Austria GmbH, Part of Thermo Fisher Scientiﬁc ImmunoDiagnostics, A-1220 Vienna, Austria. 6 Immunology and Allergy Unit, Department of Medicine Solna, Karolinska Institutet and University Hospital, SE-171 77 Stockholm, Sweden. 7 Allergy Department, 2nd Pediatric Clinic, University of Athens, 106 79 Athens, Greece. These authors contributed equally: Katarina Stenberg-Hammar, Spyridon Megremis. Correspondence and requests for materials should be addressed to N.G.P. (email: ngpallergy@gmail.com) or to R.V. (email: rudolf.valenta@meduniwien.ac.at) ARTICLE ARTICLE NATURE COMMUNICATIONS\| (2018) 9:2382 \| DOI: 10.1038/s41467-018-04591-0\| www.nature.com/naturecommunications ARTICLE ARTICLE NATURE COMMUNICATIONS \| DOI: 10.1038/s41467-018-04591-0 R e t a R espiratory viral infections are among the most common triggers of acute exacerbations of pre-school wheeze, asthma, and chronic obstructive pulmonary disease (COPD)1–3. Asthma and COPD are severe and disabling diseases of the respiratory tract and hence represent a serious global health problem affecting different age groups. Acute pre-school wheeze and community-acquired pneumonia (CAP) are other common causes of emergency visits with possible viral etiology. There is an increasing prevalence of these airways diseases, rising treatment costs, and therefore virus-induced respiratory illnesses are a heavy burden for patients and the community4,5. Respiratory viral infections, mainly due to rhinovirus (RV), are responsible for approximately 80% of wheeze and asthma exacerbations in children6,7. Moreover, infants with rhinovirus-induced wheeze have a signiﬁcantly increased risk for subsequent development of recurrent wheeze and childhood asthma8. Since exposure to RV does not lead to wheezing illness in all children, additional factors such as the host genotype, defects of the respiratory epithelial barrier, and/or atopic predisposition have been suggested to play important roles in asthma9–11. found in a high proportion of asymptomatic infants and chil- dren30–32. Furthermore, little is known about levels and epitope- speciﬁcities of natural antibody responses capable of neutralizing rhinoviruses, and thus protecting individuals against RV infec- tions. Such information would be helpful for the development of new immunological strategies for the treatment and prevention of RV-induced exacerbations of respiratory diseases. Therefore, there is a huge and so far unmet need for high-resolution ser- ological detection of rhinovirus infections. We have previously identiﬁed the capsid protein VP1 and an N-terminal VP1 peptide as major target for the natural antibody response of RV-infected subjects33. Then we have demonstrated that in vivo inoculation of subjects with RV16 indeed induced increases of VP1-speciﬁc antibody responses which were best detected with the VP1 pro- tein of the corresponding species33. Similar increases of RV- speciﬁc antibodies were found in pre-school children after asthma attacks using complete recombinant RV capsid proteins34. In a recent study performed in pre-school children with acute asthma attacks, we observed that increases of RV-speciﬁc antibody responses reﬂected the severity of respiratory symptoms35. In the present study funded by the European Union project “PreDicta” (https://cordis.europa.eu/project/rcn/96868_en.html), we investigated if it is possible to generate a microarray-based serological test which can discriminate RV-A, RV-B, and RV-C as culprit species involved in childhood asthma attacks. Results l Development of a high-resolution PreDicta microarray. Fig- ure 1 shows the arrangement of RV-derived proteins, peptides as well as control proteins on the PreDicta chip and the selection procedure for the N-terminal VP1 peptides. Recombinant capsid proteins (VP1-VP4) and fragments thereof representing RV-A, B, C species as well as non-structural proteins from RV89 were included (Supplementary Tables 1–3). Based on our previous ﬁnding that antibodies from RV-infected patients react pre- ferentially with the N-terminus of VP1 (ref. 33), we included synthetic N-terminal VP1 peptides from 30 RV strains which were selected in a rational, multistep process to represent distinct RV strains (Fig. 1a, Supplementary Table 2). Starting with 107 VP1 sequences retrieved from the NCBI database (RV-A: 76; RV- B: 25; RV-C: 6), multiple sequence alignments were performed to identify clusters of peptides with high degrees of sequence iden- tities (Fig. 1a, Supplementary Table 2: Clusters A1–A17; B1–B9; C1–C3). In a next step, peptides were re-clustered into groups taking the chemical properties of amino acids into consideration (Supplementary Table 2: AI–AXVIII; BI–BIX; CI–CIII). From these groups 30 peptides with the most distantly related sequences were selected (Fig. 1b, c; Supplementary Tables 2 and 3). For a further reﬁnement of VP1, VP2, and VP3 antibody responses, VP2 and VP3 fragments as well as peptides spanning the complete VP1 sequence from RV89 were included (Fig. 1d, Supplementary Tables 4 and 5). For control and calibration purposes we added recombinant allergens and control proteins (Fig. 1d, Supplementary Table 6) to the PreDicta chip. p p The identiﬁcation of the culprit rhinovirus species responsible for severe exacerbations of respiratory disease is an extremely important topic as certain RV species (e.g., RV-C) are suspected to be associated with more severe wheezing illnesses and acute asthma exacerbations in infants and children compared to others18,19. In fact, there are also several preventive and ther- apeutic strategies for RV infections under development which require a precise knowledge of the clinically relevant RV species to be targeted. For example, several approaches for developing vaccines based on polyvalent inactivated RV, synthetic RV- derived peptides and recombinant RV proteins have been reported20–25. The formulation of a broadly protective vaccine obviously requires the inclusion of the clinically most relevant and common RV species. Furthermore, it has been shown that blocking of the viral receptor on respiratory epithelial cells (e.g., ICAM-1) can prevent RV infection26. ARTICLE RV is genetically a highly diverse virus group with more than 160 distinct RV types which have been divided into three distinct RV species, RV-A, RV-B, and RV-C12,13. Rhinoviruses can also be classiﬁed according to which cellular receptor on human respiratory epithelial cells they use for entry14. RV-B and most RV-A variants bind to the intercellular adhesion molecule-1 (ICAM-1) (i.e., major RV group), while a subset of RV-A species binds to the low-density lipoprotein receptor (i.e., minor RV group)15,16. More recently, a cadherin-related family member 3 protein (CDHR3) has been reported as a one of the probable receptors for the RV-C species17. \| DOI: 10.1038/s41467-018-04591-0\| www.nature.com/naturecommunications PreDicta chip-based high resolution diagnosis of rhinovirus-induced wheeze Katarzyna Niespodziana1, Katarina Stenberg-Hammar2,3, Spyridon Megremis4, Clarissa R. Cabauatan1, Kamila Napora-Wijata1, Phyllis C. Vacal 1, Daniela Gallerano1, Christian Lupinek1, Daniel Ebner5, Thomas Schlederer5, Christian Harwanegg5, Cilla Söderhäll3, Marianne van Hage6, Gunilla Hedlin2,3, Nikolaos G. Papadopoulos4,7 & Rudolf Valenta1 Rhinovirus (RV) infections are major triggers of acute exacerbations of severe respiratory diseases such as pre-school wheeze, asthma and chronic obstructive pulmonary disease (COPD). The occurrence of numerous RV types is a major challenge for the identiﬁcation of the culprit virus types and for the improvement of virus type-speciﬁc treatment strategies. Here, we develop a chip containing 130 different micro-arrayed RV proteins and peptides and demonstrate in a cohort of 120 pre-school children, most of whom had been hospitalized due to acute wheeze, that it is possible to determine the culprit RV species with a minute blood sample by serology. Importantly, we identify RV-A and RV-C species as giving rise to most severe respiratory symptoms. Thus, we have generated a chip for the serological identiﬁ- cation of RV-induced respiratory illness which should be useful for the rational development of preventive and therapeutic strategies targeting the most important RV types. 1 Division of Immunopathology, Department of Pathophysiology and Allergy Research, Center for Pathophysiology, Infectiology and Immunology, Medical University of Vienna, A-1090 Vienna, Austria. 2 Astrid Lindgren Children’s Hospital, Karolinska University Hospital, SE-171 76 Stockholm, Sweden. 3 Department of Women’s and Children’s Health, Karolinska Institutet, SE-171 77 Stockholm, Sweden. 4 Division of Infection, Immunity & Respiratory Medicine, University of Manchester, Manchester M13 9NT, UK. 5 Phadia Austria GmbH, Part of Thermo Fisher Scientiﬁc ImmunoDiagnostics, A-1220 Vienna, Austria. 6 Immunology and Allergy Unit, Department of Medicine Solna, Karolinska Institutet and University Hospital, SE-171 77 Stockholm, Sweden. 7 Allergy Department, 2nd Pediatric Clinic, University of Athens, 106 79 Athens, Greece. These authors contributed equally: Katarina Stenberg-Hammar, Spyridon Megremis. Correspondence and requests for materials should be addressed to N.G.P. (email: ngpallergy@gmail.com) or to R.V. (email: rudolf.valenta@meduniwien.ac.at) \| DOI: 10.1038/s41467-018-04591-0\| www.nature.com/naturecommunications 1 NATURE COMMUNICATIONS\| (2018) 9:2382 Results l Sequence identities (%) with peptide 1 from RV 1 (01p1; top line) are shown f id h i h i Ph l i f h VP1 id i b d RV i l d i i h i i f bi a Retrieval of VP1 amino acid sequences from the NCBI database RV-A: n=76; RV-B: n=25; RV-C: n=6; EV=1 Multiple sequence alignment of VP1 N-terminal peptides (VP1-p1: aa 1–40) using Clustal W2 Clustering of RV strains according to sequence identities of VP1-p1 peptides Difference among RV peptides in 1 cluster <12.5% (i.e., 5 aa) RV-A=17 clusters; RV-B=9 clusters; RV-C=3 clusters Re-clustering of VP1-p1 peptides according to chemical properties Selection of one representative peptide sequence from each re-cluster b [%] 01p1 : NPVENYIDEVLNEVLVVPNIKESHHTTSNSAPLLDAAETG : 100.0 02p1 : ....................NS.NP.......A....... : 87.5 18p1 : ...................VN...AI......A....... : 87.5 29p1 : ......V.............R...PS......I....... : 87.5 16p1 : ....R.V.............N...P....A..V....... : 85.0 25p1 : ..I...V.Q...............PS......I....... : 85.0 11p1 : ....D.V.GI.............QA.......A....... : 82.5 59p1 : ......VND...........Q...P....A..A....... : 82.5 43p1 : ......V..I..Q......TV... S....A..A.......: 80.0 78p1 : ....E.V.Q..............KPQS.....V....... : 80.0 19p1 : ....K.V.TI..........N...PS...A..A....... : 77.5 12p1 : ....R.V.............NK.NGQL..A..A....... : 75.0 89p1 : ........S...........QP.TSVS.HA..A....... : 75.0 68p1 : ....K.TEA...........PA.NTQ...A..A....... : 72.5 20p1 : ....R.TEAI..........TS.NSQ...A..A....... : 70.0 28p1 : ....K.TEAL..........NP.NAQ.T.A..A....... : 67.5 08p1 : ..I.QFTEA..........TQA.NGSIA....A....... : 62.5 45p1 : ....QFAEA..DQ......TRP.DGLIA....A....... : 60.0 YPp1 : ....D...K.VDT..Q...TQP.GPQH.IQPSA.G.M.I. : 47.5 c025p1 : ....QFV.N..E.......TQP.GPIHTTKPTA.S.M.I. : 47.5 QPMp1 : ....EFVEHT.K......DTQA.GPVHTTKPQA.G.V.I. : 40.0 84p1 : GLEDVLEEVIVDKAKQTIASIN.NSKYTQQV.T.S.S... : 35.0 69p1 : GLG.ELEEV.IDKMKQ.TASVQ.GSKHTQKV.A.S.S... : 25.0 17p1 : GFEDELEEV.IDKMKQ.TASSQ.GPKHTQKV.A.S.N... : 22.5 70p1 : GFEGELEEV.IDKMKQ.TASSQ.GPKYTQKV.A.S.N... : 22.5 27p1 : GLG.ELEEVIVDKAKQTIASVS.NSKHTQKV.T.S.S... : 20.0 05p1 : GLEDDLVEVIVDKAQQTLASIK.DSKHTQKV.S.T.N... : 17.5 04p1 : GLG.DLMEVIVDKTHQTLASVK.DSKHTQKV.A.T.N... : 20.0 14p1 : GLGDELEEVIVEKTKQTVASIS.GPKHTQKV.I.T.N... : 15.0 86p1 : GLGDELEEVIVEKTKQTLASVATGSKYTQKV.S.S.N... : 15.0 68p1 : LDHLHGAEAAYQVESIIKTATDTVKSEIDAELGVVPSLNA : 12.5 1 10 20 30 40 a p VP1 MBP RV89 RV16 RV02 RV-C RV14 2A 2C 3A 3C 3D 3B VP1-PI VP1-PII VP1-PIII RV89 VP2 VP3 VP4 VP1-PI VP1-PII VP1-PIII VP1 VP2 VP3 VP4 VP1-PI VP1-PII VP1-PIII VP1 VP2 VP3 VP4 VP1-PI VP1-PII VP1-PIII VP1 VP2 VP3 VP4 VP1-PI VP1-PII VP1-PIII VP1 VP2 VP3 VP4 RV proteins/fragments d RV-25 RV-29 RV-78 RV-11 RV-02 RV-1B RV-18 RV-43 RV-59 RV-16 RV-19 RV-12 RV-20 RV-28 RV-68 RV-89 RV-YP RV-QPM RV-c25 RV-45 RV-08 RV-17 RV-70 RV-69 RV-04 RV-27 RV-84 RV-05 RV-14 RV-86 EV-68 RV-A RV-B RV-C EV c d d c Fig. 1 Composition of the PreDicta chip. a Selection process of VP1 peptides representing different RV strains as summarized in Supplementary Table 2. Results l Again therapeutic approaches targeting the viral receptors require knowledge which RV species are the most frequent and relevant ones. Finally, it is important to investigate the role of the different RV species for exacerbations of severe bronchial obstruction in different popu- lations and for different age groups and manifestations of respiratory illness (e.g., pre-school wheeze, asthma, COPD, asthma-COPD overlap: ACO, CAP). While RV is well established as an important trigger factor for childhood wheeze and asthma, less is known regarding the role of RV infections in exacerbations of COPD and in respiratory disease exacerbations of older sub- jects3. Furthermore, the causal relationship of RV with CAP is also unknown27. Proteins and peptides were spotted in triplicates on a pre- activated glass slide containing six microarrays surrounded by a Teﬂon frame so that one chip allows testing of six serum samples (Fig. 1d). Identities and purities of recombinant proteins and peptides were tested by sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE) followed by Coomassie Brilliant Blue staining, western-blotting, and by mass spectrometry, respectively (Supplementary Tables 3–5, Supplementary Fig. 1). Supplementary Table 7 shows that the PreDicta microarray allows reproducible measurement of IgG levels to the antigens according to intra- and inter-assay variations. Currently, the detection of RV in the course of respiratory infections is mainly based on reverse transcription of viral RNA and DNA ampliﬁcation by polymerase chain reaction (PCR)28. Such tests can demonstrate the presence of virus-derived nucleic acid but they do not necessarily indicate that the particular virus had caused an infection and is indeed responsible for clinical symptoms in the patient29. In fact, rhinovirus RNA has been NATURE COMMUNICATIONS\| (2018) 9:2382 \| DOI: 10.1038/s41467-018-04591-0\| www.nature.com/naturecommunications 2 ARTICLE a Retrieval of VP1 amino acid sequences from the NCBI database RV-A: n=76; RV-B: n=25; RV-C: n=6; EV=1 Multiple sequence alignment of VP1 N-terminal peptides (VP1-p1: aa 1–40) using Clustal W2 b RV-25 RV-29 RV-78 RV-11 RV-02 RV-1B RV-18 RV-43 RV-59 RV-16 RV-19 RV-12 RV-20 RV-28 RV-68 RV-89 RV-YP RV-QPM RV-c25 RV-45 RV-08 RV-17 RV-70 RV-69 RV-04 RV-27 RV-84 RV-05 RV-14 RV-86 EV-68 RV-A RV-B RV-C EV [%] 01p1 : NPVENYIDEVLNEVLVVPNIKESHHTTSNSAPLLDAAETG : 100.0 02p1 : ....................NS.NP.......A....... : 87.5 18p1 : ...................VN...AI......A....... : 87.5 29p1 : ......V.............R...PS......I....... : 87.5 16p1 : ....R.V.............N...P....A..V....... : 85.0 25p1 : ..I...V.Q...............PS......I....... : 85.0 11p1 : ....D.V.GI.............QA.......A....... : 82.5 59p1 : ......VND...........Q...P....A..A....... : 82.5 43p1 : ......V..I..Q......TV... S....A..A.......: 80.0 78p1 : ....E.V.Q..............KPQS.....V....... : 80.0 19p1 : ....K.V.TI..........N...PS...A..A....... Results l : 77.5 12p1 : ....R.V.............NK.NGQL..A..A....... : 75.0 89p1 : ........S...........QP.TSVS.HA..A....... : 75.0 68p1 : ....K.TEA...........PA.NTQ...A..A....... : 72.5 20p1 : ....R.TEAI..........TS.NSQ...A..A....... : 70.0 28p1 : ....K.TEAL..........NP.NAQ.T.A..A....... : 67.5 08p1 : ..I.QFTEA..........TQA.NGSIA....A....... : 62.5 45p1 : ....QFAEA..DQ......TRP.DGLIA....A....... : 60.0 YPp1 : ....D...K.VDT..Q...TQP.GPQH.IQPSA.G.M.I. : 47.5 c025p1 : ....QFV.N..E.......TQP.GPIHTTKPTA.S.M.I. : 47.5 QPMp1 : ....EFVEHT.K......DTQA.GPVHTTKPQA.G.V.I. : 40.0 84p1 : GLEDVLEEVIVDKAKQTIASIN.NSKYTQQV.T.S.S... : 35.0 69p1 : GLG.ELEEV.IDKMKQ.TASVQ.GSKHTQKV.A.S.S... : 25.0 17p1 : GFEDELEEV.IDKMKQ.TASSQ.GPKHTQKV.A.S.N... : 22.5 70p1 : GFEGELEEV.IDKMKQ.TASSQ.GPKYTQKV.A.S.N... : 22.5 27p1 : GLG.ELEEVIVDKAKQTIASVS.NSKHTQKV.T.S.S... : 20.0 05p1 : GLEDDLVEVIVDKAQQTLASIK.DSKHTQKV.S.T.N... : 17.5 04p1 : GLG.DLMEVIVDKTHQTLASVK.DSKHTQKV.A.T.N... : 20.0 14p1 : GLGDELEEVIVEKTKQTVASIS.GPKHTQKV.I.T.N... : 15.0 86p1 : GLGDELEEVIVEKTKQTLASVATGSKYTQKV.S.S.N... : 15.0 68p1 : LDHLHGAEAAYQVESIIKTATDTVKSEIDAELGVVPSLNA : 12.5 c 1 10 20 30 40 p10S p11S p12S p13S p14S p15S p16S p17S p18S p19S p1aS p1bS p2S p3S p4S p5S p6S p7S p8S p9S p1L p2L p3L p4L p5L p3aL p5aL p6L p7L p8L p9L p7aL p1A p1B p1C VP1 MBP Ara h 1 Bet v 1 BSA Can f 1 Clone 110 GG1 m43 Ole e 9 Cyn d 1 Clone 79 Clone 85 Can f 3 01p1 Der p 1 Gal d 1 Der p 10 HSA RV89 RV16 RV02 RV-C RV14 2A 2C 3A 3C 3D 3B VP2-PI VP2-PII VP2-PIII VP3-PI VP3-PII VP3-PIII VP1-PI VP1-PII VP1-PIII 02p1 18p1 29p1 16p1 25p1 11p1 59p1 43p1 78p1 19p1 12p1 89p1 68p1 20p1 28p1 08p1 45p1 c025p1 YPp1 QPMp1 04p1 14p1 86p1 68p1 RV89 VP2 VP3 VP4 VP1-PI VP1-PII VP1-PIII VP1 VP2 VP3 VP4 VP1-PI VP1-PII VP1-PIII VP1 VP2 VP3 VP4 VP1-PI VP1-PII VP1-PIII VP1 VP2 VP3 VP4 VP1-PI VP1-PII VP1-PIII VP1 VP2 VP3 VP4 VP1-p1 (aa 1–40) RV proteins/fragments RV89 fragments d 84p1 69p1 17p1 70p1 27p1 05p1 0.5 Clustering of RV strains according to sequence identities of VP1-p1 peptides Difference among RV peptides in 1 cluster <12.5% (i.e., 5 aa) RV-A=17 clusters; RV-B=9 clusters; RV-C=3 clusters Re-clustering of VP1-p1 peptides according to chemical properties Selection of one representative peptide sequence from each re-cluster Fig. 1 Composition of the PreDicta chip. a Selection process of VP1 peptides representing different RV strains as summarized in Supplementary Ta Multiple sequence alignment of selected VP1 peptides. Left margin: Strain numbers (red circle: RV-A; blue triangle: RV-C, green square: RV-B, black Enterovirus 68). Dots in the alignment represent identical amino acids. \| DOI: 10.1038/s41467-018-04591-0\| www.nature.com/naturecommunications \| DOI: 10.1038/s41467-018-04591-0\| www.nature.com/naturecommunications 3 Results l b Multiple sequence alignment of selected VP1 peptides. Left margin: Strain numbers (red circle: RV-A; blue triangle: RV-C, green square: RV-B, black circle: Enterovirus 68). Dots in the alignment represent identical amino acids. Sequence identities (%) with peptide 1 from RV 1 (01p1; top line) are shown for each peptide on the right margin. c Phylogenetic tree of the VP1 peptide sequences in b. d RV microarray layout depicting the positions of recombinant RV proteins/protein fragments, synthetic peptides (VP1-p1, 89VP1-derived peptides), MBP (maltose binding protein), and other control proteins which were spotted in triplicates. Proteins and peptides are designated as described in Supplementary Tables 3–6 RV strain recognition is broader in older wheezing children. The PreDicta chip was tested with sera from 120 pre-school children who were admitted to the hospital due to an acute wheezing episode35. Table 1 summarizes demographic and clin- ical data of the children investigated in this study. To investigate if the spectrum of recognized RV peptides varies by age, children were grouped according to age (Group I: <1 year, n = 35; Group II: 1–2 years, n = 53; Group III: >2 years, n = 32). Figure 2a, b show the frequency and intensity of IgG and IgA antibody reactivity of the children to the 30 N-terminal VP1 peptides representing RV-A, RV-B, and RV-C species. Results l Results obtained thus conﬁrmed our earlier observations showing that the majority of RV-speciﬁc antibody responses are directed against the N-terminus of VP1 as represented by the 30 peptides from the N-terminus of VP1 proteins from the different RV species33. Some children showed an IgG response against VP2 and in particular to a fragment representing the middle portion of VP2 whereas VP3 and VP4 showed no relevant IgG reactivity (Supplementary Fig. 2a, b). Among the non-structural proteins, 2C, 3A, and 3C showed some IgG reactivity (Supplementary Fig. 2). Identiﬁcation of RV species-speciﬁc antibody increases. Based on our previous observations that antibody increases speciﬁc for the N-terminal portion of VP1 can be detected in serum samples obtained from subjects after RV infection36, the PreDicta chip was equipped with a VP1 peptide set which should allow detecting species-speciﬁc immune responses at high resolution (Fig. 1). Supplementary Figure 4 shows images of RV-speciﬁc antibody responses measured in sera from six representative children obtained at the time of the acute episode of wheeze and in sera at a follow-up visit 2–3 months later. In sera from the follow-up visit increased antibody responses to RV-A (sera #33, #108), RV-C (sera #119, #84), and RV-B (sera #66, #89) peptides could be detected (Supplementary Fig. 4). The analysis of IgG reactivity according to the age of children at the acute visit showed that children <1 year of age had much lower RV-speciﬁc IgG levels compared to children who were older than 1 year. There was almost no change of the RV strain- speciﬁcity pattern among the three age groups (Fig. 2c). However, we found a positive correlation between the number of recognized VP1 peptides and the age of the children (Fig. 2d). Children between 6 and 12 months of age recognized a signiﬁcantly lower number of N-terminal VP1 peptides than older (13–42 months) children. Children older than 2 years reacted with signiﬁcantly more peptides than both groups of younger children (Fig. 2e). pp y g We then compared the peptide-speciﬁc IgG antibody levels in sera obtained from the 120 children at the time of the acute wheeze and at the follow-up 2–3 months later (i.e., median 11 weeks later) (Table 1). Supplementary Figure 5 shows a color-coded map of the peptide-speciﬁc antibody responses for the acute phase and follow-up of each of the children demonstrating species-speciﬁc IgG increases. Results l The most frequent and highest IgG responses were directed to RV-C-derived \| DOI: 10.1038/s41467-018-04591-0\| www.nature.com/naturecommunications NATURE COMMUNICATIONS\| (2018) 9:2382 3 ARTICLE NATURE COMMUNICATIONS \| DOI: 10.1038/s41467-018-04591-0 Table 1 Demographic and clinical characterization of wheezing children Characteristic Group I(N=35) Group II(N=53) Group III(N=32) Total(N=120) Age (months) Median 10 17 32 18 Range (Min–Max) 6–12 13–24 25–42 6–42 Gender Male:female ratio 26:9 31:22 19:12a 76:43 Male (%) 74 58 61 63 Ever wheeze before, n (%) 23 (66) 44 (83) 25 (83)a 92 (77) Allergic sensitizationb Food allergens, n (%) 5 (14) 7 (13) 10 (31) 22 (18) Respiratory allergens, n (%) 1 (3) 4 (7) 4 (12.5) 9 (7.5) Hospitalized at the acute visit, n (%) 26 (74) 44 (83) 26 (87)a 96 (80) Weeks until follow-up visit Median 11a 11.5a 12.5 11 Range (Min–Max) 7–27 7–24 9–30 7–30 Presence:absence of cold symptoms 35:0 50:3 30:2 115:5 Days with respiratory symptoms (%) Median 13 11a 6.5 11 Range (Min–Max) 1–85 0–100 1–63 0–100 Days with β2-agonists (%) Median 20 22a 29.5 22 Range (Min–Max) 0–100 0–100 0–100 0–100 a1 or 2 values are missing for this variable bAllergen-speciﬁc IgE to food and/or respiratory allergens determined by MEDALL allergen chip for the acute and follow-up visits (≥0.3 ISU-E was considered positive) Table 1 Demographic and clinical characterization of wheezing children responses. Finally, the results of these analyses were super- imposed. It turned out that there was a very strong correlation between the two groupings, one based on antibody reactivity and the other based on sequence identities among peptides made through the unsupervised analysis. Thus antibody response patterns reﬂected very closely the peptide sequence similarities (Supplementary Fig. 3). peptides followed by RV-A peptides whereas IgG reactivity to RV-B peptides was less frequent and intense (Fig. 2a). Similar results were obtained for IgA responses which in general were lower than the IgG responses (Fig. 2b). The analysis of IgG reactivity to structural and non-structural proteins and to recombinant fragments and synthetic peptides spanning VP1, VP2, and VP3 from RV89 is shown in Supplementary Fig. 2a for all 120 children and in Supplementary Fig. 2b for those children (n = 41) who had shown increases of RV89-speciﬁc antibody responses in follow-up serum samples taken after recovery. \| DOI: 10.1038/s41467-018-04591-0\| www.nature.com/naturecommunications Results l Next, we com- pared the peptide-speciﬁc increases in relative numbers for each child (Fig. 3a). According to these peptide-speciﬁc IgG increases, children could be identiﬁed who responded prefer- entially to RV-A-derived peptides (n = 41), RV-C-derived peptides (n = 33), RV-B-derived peptides (n = 23), and some with a mixed response pattern (n = 7) (Fig. 3a). For 16 children no increases of peptide-speciﬁc IgG responses were found (Fig. 3a, bottom). The same analysis was performed for increases of antibody responses against complete recombinant VP1 proteins from the three RV-A strains (i.e., RV89, 16, 2), The patterns of antibody response against different RV peptides were also analyzed using an independent bioinformatics approach (Supplementary Fig. 3). A phylogenetic clustering of all peptides was prepared and used to generate peptide groups according to sequence homology which represented to a large extent the RV subgroups A, B, and C. Then an unsupervised computer algorithm was used to cluster the patterns of antibody NATURE COMMUNICATIONS\| (2018) 9:2382 \| DOI: 10.1038/s41467-018-04591-0\| www.nature.com/naturecommunications 4 he RV-C strain YP, and the RV-B strain 14 (Fig. 3a) but sensitivity of serology vs. PCR in these children. Moreove a 86 14 04 05 27 70 17 69 84 QP c025 YP 45 08 28 20 68 89 12 19 78 43 59 11 25 16 29 18 02 01 RV-A RV-C RV-B Number of IgG reactive sera (n) 120 100 80 60 40 20 0 25–75 >75 5–25 A B C ISU-G level c 0 20 40 60 86 14 04 05 27 70 17 69 84 QP c025 YP 45 08 28 20 68 89 12 19 78 43 59 11 25 16 29 18 02 01 RV-A RV-C RV-B 6–12 13–24 25–42 IgG levels (ISU-G) d Age [months] Number of IgG-reactive peptides (n) = 0.5433 p<0.0001 0 10 20 30 40 50 0 10 20 30 40 e 0 10 20 30 40 Age [months] 6–12 13–24 25–42 Number of IgG-reactive peptides (n) 86 14 04 05 27 70 17 69 84 QP c025 YP 45 08 28 20 68 89 12 19 78 43 59 11 25 16 29 18 02 01 RV-A RV-C RV-B Number of IgA reactive sera (n) 120 100 80 60 40 20 0 ISU-A level 25–75 >75 5–25 A B C b ig. 2 RV-speciﬁc antibody responses in sera from children with acute wheeze. Results l Horizontal lin \| a 86 14 04 05 27 70 17 69 84 QP c025 YP 45 08 28 20 68 89 12 19 78 43 59 11 25 16 29 18 02 01 RV-A RV-C RV-B Number of IgG reactive sera (n) 120 100 80 60 40 20 0 25–75 >75 5–25 A B C ISU-G level 86 14 04 05 27 70 17 69 84 QP c025 YP 45 08 28 20 68 89 12 19 78 43 59 11 25 16 29 18 02 01 RV-A RV-C RV-B Number of IgA reactive sera (n) 120 100 80 60 40 20 0 ISU-A level 25–75 >75 5–25 A B C b b a 86 14 04 05 27 70 17 69 84 QP c025 YP 45 08 28 20 68 89 12 19 78 43 59 11 25 16 29 18 02 01 RV-A RV-C RV-B 0 c 0 20 40 60 86 14 04 05 27 70 17 69 84 QP c025 YP 45 08 28 20 68 89 12 19 78 43 59 11 25 16 29 18 02 01 RV-A RV-C RV-B 6–12 13–24 25–42 IgG levels (ISU-G) d Age [months] Number of IgG-reactive peptides (n) = 0.5433 p<0.0001 0 10 20 30 40 50 0 10 20 30 40 e 40 ** 86 14 04 05 27 70 17 69 84 QP c025 YP 45 08 28 20 68 89 12 19 78 43 59 11 25 16 29 18 02 01 RV-A RV-C RV-B 0 c 0 20 40 60 86 14 04 05 27 70 17 69 84 QP c025 YP 45 08 28 20 68 89 12 19 78 43 59 11 25 16 29 18 02 01 RV-A RV-C RV-B 6–12 13–24 25–42 IgG levels (ISU-G) Number of IgG-reactive peptides (n) e 0 10 20 30 40 ** Age [months] 6–12 13–24 25–42 Number of IgG-reactive peptides (n) e Age [months] 13–24 Fig. 2 RV-speciﬁc antibody responses in sera from children with acute wheeze. Frequencies and levels of a IgG and b IgA responses (y-axes: n, number of reactive sera) to the N-terminal VP1 peptides from 30 RV strains (Supplementary Tables 2 and 4) (x-axes: red: RV-A species; green: RV-B species; blue: RV-C species). Antibody levels are color-coded and expressed as ISAC standardized units, ISU-G and ISU-A, respectively. \| DOI: 10.1038/s41467-018-04591-0\| www.nature.com/naturecommunications Results l Frequencies and levels of a IgG and b IgA responses (y-axes: n, number eactive sera) to the N-terminal VP1 peptides from 30 RV strains (Supplementary Tables 2 and 4) (x-axes: red: RV-A species; green: RV-B species; bl RV-C species). Antibody levels are color-coded and expressed as ISAC standardized units, ISU-G and ISU-A, respectively. c Median IgG levels (y-axis: IS G) to VP1 peptides (x-axis) in children grouped according to age (6–12 months: squares; 13–24 months: triangles; 25–42: circles). d Spearman’s rank orrelation between the number of IgG-reactive peptides (n, y-axis; median IgG >15 ISU) and age (x-axis: months). Correlation coefﬁcient (ρ) and p-va re shown. e Comparison of the number of IgG-reactive VP1 peptides (n, y-axis; median IgG >15 ISU) in children according to age (x-axis). Results l Frequencies and levels of a IgG and b IgA responses (y-axes: n, number eactive sera) to the N-terminal VP1 peptides from 30 RV strains (Supplementary Tables 2 and 4) (x-axes: red: RV-A species; green: RV-B species; bl V-C species). Antibody levels are color-coded and expressed as ISAC standardized units, ISU-G and ISU-A, respectively. c Median IgG levels (y-axis: IS G) to VP1 peptides (x-axis) in children grouped according to age (6–12 months: squares; 13–24 months: triangles; 25–42: circles). d Spearman’s rank orrelation between the number of IgG-reactive peptides (n, y-axis; median IgG >15 ISU) and age (x-axis: months). Correlation coefﬁcient (ρ) and p-va re shown. e Comparison of the number of IgG-reactive VP1 peptides (n, y-axis; median IgG >15 ISU) in children according to age (x-axis). Horizontal lin ndicate medians. Statistically signiﬁcant differences between groups are indicated (p < 0.01, **p < 0.0001) (Mann–Whitney U-test) NATURE COMMUNICATIONS \| DOI: 10.1038/s41467-018-04591-0 ARTICL ARTICLE NATURE COMMUNICATIONS \| DOI: 10.1038/s41467-018-04591-0 a 86 14 04 05 27 70 17 69 84 QP c025 YP 45 08 28 20 68 89 12 19 78 43 59 11 25 16 29 18 02 01 RV-A RV-C RV-B Number of IgG reactive sera (n) 120 100 80 60 40 20 0 25–75 >75 5–25 A B C ISU-G level c 0 20 40 60 86 14 04 05 27 70 17 69 84 QP c025 YP 45 08 28 20 68 89 12 19 78 43 59 11 25 16 29 18 02 01 RV-A RV-C RV-B 6–12 13–24 25–42 IgG levels (ISU-G) d Age [months] Number of IgG-reactive peptides (n) = 0.5433 p<0.0001 0 10 20 30 40 50 0 10 20 30 40 e 0 10 20 30 40 ** Age [months] 6–12 13–24 25–42 Number of IgG-reactive peptides (n) 86 14 04 05 27 70 17 69 84 QP c025 YP 45 08 28 20 68 89 12 19 78 43 59 11 25 16 29 18 02 01 RV-A RV-C RV-B Number of IgA reactive sera (n) 120 100 80 60 40 20 0 ISU-A level 25–75 >75 5–25 A B C b Fig. 2 RV-speciﬁc antibody responses in sera from children with acute wheeze. Results l c Median IgG levels (y-axis: ISU- G) to VP1 peptides (x-axis) in children grouped according to age (6–12 months: squares; 13–24 months: triangles; 25–42: circles). d Spearman’s rank correlation between the number of IgG-reactive peptides (n, y-axis; median IgG >15 ISU) and age (x-axis: months). Correlation coefﬁcient (ρ) and p-value are shown. e Comparison of the number of IgG-reactive VP1 peptides (n, y-axis; median IgG >15 ISU) in children according to age (x-axis). Horizontal lines indicate medians. Statistically signiﬁcant differences between groups are indicated (p < 0.01, **p < 0.0001) (Mann–Whitney U-test) the RV-C strain YP, and the RV-B strain 14 (Fig. 3a) but the results were less clear and negative for several children because only few strains were covered with the recombinant proteins. We have also included in Fig. 3a (right column) results from the PCR testing performed using VP4–VP2- speciﬁc primers in 108 of the 120 children35, which showed that the nucleic acid-based detection of virus strains was negative for approximately 25% of children with increases of RV peptide-speciﬁc IgG levels which may indicate a higher sensitivity of serology vs. PCR in these children. Moreover, PCR results did not correspond well with the speciﬁcities of the antibody responses. There were also 14 children without increases of RV peptide-speciﬁc antibody responses who had positive PCR results (Fig. 3a, bottom). Figure 3b provides a summary of the absolute increases of peptide-speciﬁc IgG responses showing that RV-A and RV-C- speciﬁc antibody responses dominated over RV-B-speciﬁc increases. Results l \| DOI: 10.1038/s41467-018-04591-0\| www.nature.com/naturecommunications 5 NATURE COMMUNICATIONS\| (2018) 9:2382 ARTICLE NATURE COMMUNICATIONS \| DOI: 10.1038/s41467-018-04591-0 Microarray PCR VP1 P1 VP4/VP2 RV-A RV-C RV-B Mixed Negative a #2 #4 #6 #7 #32 #40 #50 #53 #55 #56 #57 #58 #60 #63 #70 #104 1 2 18 29 16 25 11 59 43 78 19 12 89 68 20 28 8 45 YP c025 QPM 84 69 17 70 27 5 4 14 86 68 89 16 2 14 YP #8 #12 #18 #20 #27 #28 #33 #34 #36 #39 #43 #48 #52 #59 #65 #68 #71 #76 #77 #78 #80 #82 #83 #85 #88 #90 #93 #95 #97 #99 #100 #103 #108 #109 #110 #113 #114 #115 #116 #118 #87 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 A B C EV #9 #10 #11 #15 #21 #22 #24 #25 #26 #29 #31 #38 #42 #44 #46 #47 #49 #54 #61 #66 #67 #69 #79 #84 #91 #92 #94 #96 #106 #107 #112 #117 #119 #13 #45 #62 #101 #102 #120 #74 #1 #3 #5 #14 #16 #17 #19 #23 #30 #35 #37 #41 #51 #64 #72 #73 #75 #81 #86 #89 #105 #111 #98 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 b 86 14 04 05 27 70 17 69 84 QP c025 YP 45 08 28 20 68 89 12 19 78 43 59 11 25 16 29 18 02 01 RV-A RV-C RV-B 80 60 40 20 0 Number of sera with increases of IgG ISU-G increase 25–75 >75 1–25 A B C ncreases of IgG antibodies to N-terminal VP1 peptides detect culprit RV species. a A map showing IgG antibody increases (gradient colo from minimal to maximal value for each of the subject: white to dark red) to recombinant VP1 proteins and N-terminal VP1 peptides from t (top lines RV-A: red; RV-C: blue; RV-B: green) measured by the microarray and PCR data for the wheezing children (white squares: PCR n uares: PCR positive; no squares: not tested due to lack of samples). Results l Shown are children for which RV-A (n = 41), RV-C (n = 33), and RV-B ( entiﬁed as culprit species according to chip analysis, as well as children with mixed (n = 7) and no responses (n = 15). b Absolute increase y levels between the acute and follow-up visit for the wheezing children (y-axes: n, number of reactive sera) to the VP1 peptides red: RV-A species; green: RV-B species; blue: RV-C species) are shown. Antibody increases are color-coded and expressed as ISAC stand SU-G) ICLE NATURE COMMUNICATIONS \| DOI: 10.1038/s41467 018 0 b 86 14 04 05 27 70 17 69 84 QP c025 YP 45 08 28 20 68 89 12 19 78 43 59 11 25 16 29 18 02 01 RV-A RV-C RV-B 80 60 40 20 0 Number of sera with increases of IgG ISU-G increase 25–75 >75 1–25 A B C a Fig. 3 Increases of IgG antibodies to N-terminal VP1 peptides detect culprit RV species. a A map showing IgG antibody increases (gradient color scale ranging from minimal to maximal value for each of the subject: white to dark red) to recombinant VP1 proteins and N-terminal VP1 peptides from three RV species (top lines RV-A: red; RV-C: blue; RV-B: green) measured by the microarray and PCR data for the wheezing children (white squares: PCR negative; black squares: PCR positive; no squares: not tested due to lack of samples). Shown are children for which RV-A (n = 41), RV-C (n = 33), and RV-B (n = 23) were identiﬁed as culprit species according to chip analysis, as well as children with mixed (n = 7) and no responses (n = 15). b Absolute increases of IgG antibody levels between the acute and follow-up visit for the wheezing children (y-axes: n, number of reactive sera) to the VP1 peptides (x-axes: red: RV-A species; green: RV-B species; blue: RV-C species) are shown. Antibody increases are color-coded and expressed as ISAC standardized units (ISU-G) Fig. 3 Increases of IgG antibodies to N-terminal VP1 peptides detect culprit RV species. Discussion Wi hi h Within the European Union-funded project PreDicta (https:// cordis.europa.eu/project/rcn/96868_en.html) we developed the PreDicta chip which is based on 130 micro-arrayed RV-derived proteins and peptides selected to represent the three RV species (RV-A, RV-B, and RV-C). Using the PreDicta chip we could demonstrate in a cohort of 120 pre-school children with acute wheeze and a follow-up visit that the RV-speciﬁc antibody response (IgG > IgA) is directed against an N-terminal peptide of the major capsid protein VP1 which conﬁrms earlier results obtained by the mapping of RV89-speciﬁc antibody respon- ses33. Since the PreDicta chip was designed to contain a panel of 30 synthetic peptides which represented the most diverse RV strains of the three genetic RV species in terms of sequence identity and physicochemical properties we were able to per- form a high-resolution mapping of RV species-speciﬁc antibody responses by serology. In a cohort of clinically well-described Swedish pre-school children from whom sera were available from an episode of acute wheeze requiring an emergency room visit and from a follow-up visit after convalescence, peptides from RV-A and RV-C species were most frequently recognized whereas RV-B species were much less commonly recognized. Interestingly, we found that older children (i.e., children older than 2 years of age) recognized peptides from more RV strains than younger children. This result would indicate that children encounter in their life different RV strains and thus may broaden their IgG reactivity proﬁles later in life but this needs to be conﬁrmed in longitudinal studies with samples taken from the same children at different ages, as has recently been done in the analysis of the evolution of IgE reactivity proﬁles in allergic children in birth cohorts37–40. One of the important ﬁndings of our study was that we could demonstrate that IgG reactivity to peptides from certain RV strains increased in the children which may allow identifying the culprit RV species responsible for the acute wheeze by serology. Furthermore, it turned out that increases of IgG responses to RV-A and RV-C species were signiﬁcantly associated with more severe illness as compared to IgG increases to RV-B. The PreDicta chip thus seems to be not only suitable for identifying the culprit RV species responsible for an exacerbation of respiratory illness by simple serology but also allows to determine those RV species giving rise to severe symptoms. NATURE COMMUNICATIONS \| DOI: 10.1038/s41467-018-04591-0 Interestingly, we found that older children (i.e., children older than 2 years of age) recognized peptides from more RV strains than younger children. This result would indicate that children encounter in their life different RV strains and thus may broaden their IgG reactivity proﬁles later in life but this needs to be conﬁrmed in longitudinal studies with samples taken from the same children at different ages, as has recently been done in the analysis of the evolution of IgE reactivity proﬁles in allergic children in birth cohorts37–40. One of the important ﬁndings of our study was that we could demonstrate that IgG reactivity to peptides from certain RV strains increased in the children which may allow identifying the culprit RV species responsible for the acute wheeze by serology. Furthermore, it turned out that increases of IgG responses to RV-A and RV-C species were signiﬁcantly associated with more severe illness as compared to IgG increases to RV-B. The PreDicta chip thus seems to be not only suitable for identifying the culprit RV species responsible for an exacerbation of respiratory illness by simple serology but also allows to determine those RV species giving rise to severe symptoms. Since we also had results from PCR-based testing of nasal swab samples from the same children we could compare the detection of strain-speciﬁc nucleic acid with antibody results. In fact, we found that all children without species- speciﬁc antibody increases were also negative in the PCR test. However, there was poor correlation between the PCR results from the nasal samples and the chip-based serological results. Furthermore, with the exception of four children (#3, #61, #65, #72) for which a positive PCR result has been obtained at the follow-up visit, all PCR results were negative at both the acute and follow-up visit in approximately 25% of the children for whom species-speciﬁc antibody increases could be clearly demonstrated. Several possibilities for the discrepancies may be considered. For example, it may be possible that the time interval after acute infection chosen for serology was not optimal. However, we have previously investigated the time interval required for the appearance of VP1-speciﬁc antibody increases in a controlled infection study36 and it is therefore Results l a A map showing IgG antibody increases (gradient color scale ranging from minimal to maximal value for each of the subject: white to dark red) to recombinant VP1 proteins and N-terminal VP1 peptides from three RV species (top lines RV-A: red; RV-C: blue; RV-B: green) measured by the microarray and PCR data for the wheezing children (white squares: PCR negative; black squares: PCR positive; no squares: not tested due to lack of samples). Shown are children for which RV-A (n = 41), RV-C (n = 33), and RV-B (n = 23) were identiﬁed as culprit species according to chip analysis, as well as children with mixed (n = 7) and no responses (n = 15). b Absolute increases of IgG antibody levels between the acute and follow-up visit for the wheezing children (y-axes: n, number of reactive sera) to the VP1 peptides (x-axes: red: RV-A species; green: RV-B species; blue: RV-C species) are shown. Antibody increases are color-coded and expressed as ISAC standardized units (ISU-G) \| DOI: 10.1038/s41467-018-04591-0\| www.nature.com/naturecommunications \| DOI: 10.1038/s41467-018-04591-0\| www.nature.com/naturecommunications NATURE COMMUNICATIONS\| (2018) 9:2382 6 NATURE COMMUNICATIONS \| DOI: 10.1038/s41467-018-04591-0 \| DOI: 10.1038/s41467-018-04591-0\| www.nature.com/naturecommunications NATURE COMMUNICATIONS \| DOI: 10.1038/s41467-018-04591-0 ARTICLE Antibody signatures associated with severity of wheeze. Next, we investigated whether antibody responses to certain RV species 0 50 100 150 * * * RV-A RV-C RV-B Mixed Negative Days with respiratory symptoms 0 50 100 150 200 RV-A RV-C RV-B Mixed Negative Days with medication use 0 50 100 150 200 250 300 * * RV-A RV-C RV-B Mixed Negative Days with respiratory symptoms and medication use Fig. 4 Days with respiratory symptoms and medication use in children without and with RV-A, RV-B, RV-C, or mixed peptide-speciﬁc IgG increases. Days with respiratory symptoms (top panel, y-axis), with medication use (middle panel, y-axis), and with symptoms and medication use (bottom panel, y-axis) are shown for children without (negative), mixed, RV-A, RV-C, or RV-B increases of IgG (x-axes). Horizontal lines show median values. Statistically signiﬁcant differences between groups are indicated (*p < 0.001; p < 0.01, p < 0.05) (Mann–Whitney U-test) associated with the highest number of days with symptoms and medication (Fig. 4, bottom panel). Children with RV-A- or RV- C-speciﬁc antibody increases had signiﬁcantly more days with symptoms and medication than children with RV-B-speciﬁc antibody increases (Fig. 4, bottom panel). Discussion Within the European Union-funded project PreDicta (https:// cordis.europa.eu/project/rcn/96868_en.html) we developed the PreDicta chip which is based on 130 micro-arrayed RV-derived proteins and peptides selected to represent the three RV species (RV-A, RV-B, and RV-C). Using the PreDicta chip we could demonstrate in a cohort of 120 pre-school children with acute wheeze and a follow-up visit that the RV-speciﬁc antibody response (IgG > IgA) is directed against an N-terminal peptide of the major capsid protein VP1 which conﬁrms earlier results obtained by the mapping of RV89-speciﬁc antibody respon- ses33. Since the PreDicta chip was designed to contain a panel of 30 synthetic peptides which represented the most diverse RV strains of the three genetic RV species in terms of sequence identity and physicochemical properties we were able to per- form a high-resolution mapping of RV species-speciﬁc antibody responses by serology. In a cohort of clinically well-described Swedish pre-school children from whom sera were available from an episode of acute wheeze requiring an emergency room visit and from a follow-up visit after convalescence, peptides from RV-A and RV-C species were most frequently recognized whereas RV-B species were much less commonly recognized. Methods S l ti Selection and production of N-terminal VP1 peptides. VP1 amino acid sequences of RV strains representing the three RV species (RV-A: n = 76; RV-B: n = 25; RV-C: n = 6) were retrieved from the NCBI database (https://www.ncbi. nlm.nih.gov/) (Supplementary Table 1). Multiple sequence alignments of the VP1 N-terminal peptides (aa 1–40) were performed using ClustalW2 software available at the EMBL-EBI website (http://www.ebi.ac.uk/tools/clustalw2) to determine amino acid sequence identities among the peptides. Peptide sequences showing sequence identities greater than 87.5% (i.e. differences of ≤5 aa) were grouped together into clusters (Fig. 1). Sequences among the clusters (Supplementary Tables 1 and 2, A1–A17, B1–B9, C1–C3) were re-aligned using GeneDoc software (http://iubio.bio.indiana.edu/soft/molbio/ibmpc/genedoc-readme.html) and each amino acid mismatch was analyzed regarding physicochemical properties of the amino acids. This procedure led to re-clustering of the peptides (Supplementary Table 2, AI–XVIII, BI–BIX; CI–CIII). From each re-clustered group one repre- sentative RV strain peptide was selected for printing onto the chip (Fig. 1b, Sup- plementary Tables 2 and 3: RV-A: n = 18; RV-B: n = 9; RV-C: n = 3). An enterovirus-derived peptide was also included (Fig. 1). For the set of peptides to be printed a multiple sequence alignment was performed by ClustalW2 and a phy- logenetic tree was constructed by the Neighbor-Joining (N-J) method using MEGA 6 software (www.megasoftware.net) (Fig. 1c). The evolutionary distances between sequences were computed using the Kimura 2-parameter model with bootstrap values calculated from 1000 replicates. Additional 36 peptides spanning the RV89VP1 protein (Fig. 1, Supplementary Table 4) were selected to detect anti- bodies towards VP1 epitopes other than the N-terminal portion. The peptides as well as the non-structural 3B protein from strain 89 (VPg: GPYSGEPKPKSRAPERRVVTQ) were produced by solid-phase synthesis with the 9-ﬂuorenyl-methoxy carbonyl (Fmoc)-method (CEM-Liberty, Matthews, NC, USA and Applied Biosystems, Carlsbad, CA, USA) on PEG-PS preloaded resins (Applied Biosystems). After synthesis, peptides were washed with dichloromethane, cleaved from the resins using 19 ml triﬂuoroacetic acid (TFA), 0.5 ml silane, and 0.5 ml H2O and precipitated into pre-chilled tert- butylmethylether. Peptides were puriﬁed by reversed-phase high-performance liquid chromatography in a 10–70% acetonitrile gradient using a Jupiter 4 μm Proteo 90 Å, LC column (Phenomenex, Torrance, CA, USA) and an UltiMate 3000 Pump (Dionex, Sunnyvale, CA, USA) to a purity >90%. Their identities and molecular weights were veriﬁed by mass spectrometry (Microﬂex MALDI-TOF, Bruker, Billerica, MA, USA)49. Discussion Wi hi h Since we also had results from PCR-based testing of nasal swab samples from the same children we could compare the detection of strain-speciﬁc nucleic acid with antibody results. In fact, we found that all children without species- speciﬁc antibody increases were also negative in the PCR test. However, there was poor correlation between the PCR results from the nasal samples and the chip-based serological results. Furthermore, with the exception of four children (#3, #61, #65, #72) for which a positive PCR result has been obtained at the follow-up visit, all PCR results were negative at both the acute and follow-up visit in approximately 25% of the children for whom species-speciﬁc antibody increases could be clearly demonstrated. Several possibilities for the discrepancies may be considered. For example, it may be possible that the time interval after acute infection chosen for serology was not optimal. However, we have previously investigated the time interval required for the appearance of VP1-speciﬁc antibody increases in a controlled infection study36 and it is therefore Fig. 4 Days with respiratory symptoms and medication use in children without and with RV-A, RV-B, RV-C, or mixed peptide-speciﬁc IgG increases. Days with respiratory symptoms (top panel, y-axis), with medication use (middle panel, y-axis), and with symptoms and medication use (bottom panel, y-axis) are shown for children without (negative), mixed, RV-A, RV-C, or RV-B increases of IgG (x-axes). Horizontal lines show median values. Statistically signiﬁcant differences between groups are indicated (p < 0.001; p < 0.01, *p < 0.05) (Mann–Whitney U-test) Antibody signatures associated with severity of wheeze. Next, we investigated whether antibody responses to certain RV species were associated with the severity of RV-induced wheeze. For this purpose, we determined the number of days with respiratory symptoms and the number of days when medication was required in the period between the acute and follow-up visit. The number of days with respiratory symptoms but not with medication was signiﬁcantly higher in subjects with an increase in RV-A > Mixed > RV-C > RV-B speciﬁc signal (Fig. 4, top and middle panels). We then analyzed the sum of days with symptoms and medication and found that RV-A and RV-C antibody increases were 7 \| DOI: 10.1038/s41467-018-04591-0\| www.nature.com/naturecommunications 7 NATURE COMMUNICATIONS\| (2018) 9:2382 ARTICLE NATURE COMMUNICATIONS \| DOI: 10.1038/s41467-018-04591-0 very unlikely that the time interval used in our study was too long and responsible for discrepancies between PCR testing and serology. Discussion Wi hi h We also do not think that cross-reactivity among strains or the original antigenic sin41 is responsible for the observed differences because we have taken care to include on the chip sequences from several different RV strains and the results in Fig. 3a clearly show that the serological results obtained with the chip allow a bona ﬁde discrimination of RV- A, RV-B, and RV-C infections because we have used a large panel of RV peptides on the chip. Finally, we found that older children recognized more RV peptides from different strains than younger children (Fig. 2c–e), which indicates that the children develop antibodies against new viruses and thus the concept of the original antigenic sin does not seem to apply here. One limitation of our study is, however, that we cannot exclude that infections with additional RV strains have occur- red in the time window between the ﬁrst and second blood sampling and thus are responsible for the discrepancy between PCR results and serology. Nevertheless we think that the dis- crepancy between PCR results and serology is rather due to the fact that not every virus detected by PCR causes an infection with a consecutive immune response. In fact, studies performed in young children report that up to 35% of asymptomatic subjects have positive PCR results30,42. One more likely possi- bility for the poor correlation of PCR results and antibody results could be that we used a PCR strategy based on primers speciﬁc for VP4- and VP2-encoding regions of the viral gen- ome, which may be less speciﬁc than PCR strategies based on the ampliﬁcation and sequencing of the VP1-encoding region or of the complete RV genome12,28,43,44. In general, nucleic acid-based strategies for virus detection only demonstrate the presence of virus-speciﬁc nucleic acid but provide no evidence that an infection has taken place which gave rise to a speciﬁc immune response. We therefore think that the PreDicta chip and future versions of it containing an even larger repertoire of N-terminal VP1 peptides from more RV strains will be a complementary tool in addition to PCR testing and eventually turn out to be even superior. The antibody test is actually fast and economical: It takes only few hours and requires only microliter amounts of serum. Furthermore, serological analysis is robust and can be easily performed without need for PCR cyclers and subsequent sequencing. Discussion Wi hi h However, more prospective studies will be needed to investigate the diagnostic sensitivity and speciﬁcity of the RV chip. The PreDicta chip may be extremely useful to determine RV species-speciﬁc antibody responses in serum samples from existing cohorts world-wide to deﬁne the most common and relevant RV species involved in respiratory illness. Chip-based measurements will allow exploring in prospective studies the role of RV infections in a variety of respiratory disease exacerbations. For example, it will be possible to discriminate whether asthma exacerbations have been triggered by an RV infection or by allergen exposure because it has been shown that both factors (i.e., RV infections as well as allergen exposure) induce increases of speciﬁc anti- body responses when serum samples collected at the acute visit and during a follow-up visit are compared35,36,45,46. The identiﬁcation of the culprit factors triggering asthma attacks becomes increasingly important in respiratory medicine due to availability of selective treatments of allergic asthma such as anti-IgE antibodies and a variety of other biologics such as anti- cytokine antibodies targeting different forms of asthma47,48 respiratory viruses by PCR and the immune reaction by serol- ogy in close intervals and for extended periods after exacerbation. The reliable determination of the most common RV species involved in triggering severe respiratory illness will ultimately provide a rational basis for the development of RV vaccines and RV species-targeting therapeutic approaches20–26. In conclusion, we developed and evaluated a high-resolution antibody assay based on micro-arrayed peptides and recombinant antigens from the most common RV strains to identify antibody signatures discriminating RV infections at the levels of different RV species and allowed to point towards the culprit species responsible for the triggering of acute pre-school wheezing. The PreDicta chip has the potential to be useful for a serological global mapping of RV infections, the identiﬁcation of RV species involved in triggering different forms of severe respiratory illness, and for paving the road for RV-speciﬁc therapeutic and pro- phylactic treatment strategies, such as vaccines. \| DOI: 10.1038/s41467-018-04591-0\| www.nature.com/naturecommunications Methods S l ti Protein concentrations were determined using BCA Protein Assay Kit (Thermo Fisher Scientiﬁc, Rockford, IL, USA). The secondary structure of the proteins was measured by circular dichroism spectroscopy on a Jasko J-810 spectropolarimeter (Japan Spectroscopic, Tokyo, Japan) at a protein concentration of 0.1 mg/ml in 10 mM NaH2PO4. People’s Republic of China) and the image analysis was evaluated by Microarray Image Analyzer v3.1.2 software (Phadia-Thermo Fisher)50. For calibration and determination of background signals, a calibrator serum (i.e., a pool of allergic patients sera, diluted 1:10) and sample diluent were included in each analysis run. Calibration and variability of the microarray. The Phadia Microarray Image Analysis software was used to process images, to calculate the mean ﬂuorescence intensities (FI) of triplicate analyses and to calibrate the results. A calibration curve was generated by relating ﬂuorescence intensities obtained by scanning the Pre- Dicta microarray with allergen-speciﬁc antibody levels measured by ImmunoCAP. Results were reported in ISAC standardized units Germany). Refolding of recombinant proteins was achieved by a stepwise dialysis against 10 mM NaH2PO4 for structural and non-structural proteins and 20 mM Tris-HCl, 200 mM NaCl, 1 mM EDTA for MBP fusion proteins, respectively. The purity of recombinant proteins was veriﬁed by SDS-PAGE followed by Coomassie Brilliant Blue staining and the identity by immunoblotting using a monoclonal mouse anti-His-tag antibody 1:1000 diluted (Cat: DIA-900, Dianova, Hamburg, Germany). Bound antibodies were detected with 1:1000 diluted alkaline p (ISU)50,52. Background reactivities of ﬂuorescence-labeled α-huIgG antibodies towards all microarray components were determined by testing six replicates of a sample diluent alone. For the characterization of the assay variability, one cali- brator serum and three normal sera were proﬁled at 1:100, 1:200, 1:500, and 1:1000 in order to ﬁnd a dilution at which the broadest spectrum of reactivity levels were covered. To assess intra-assay (i.e. within experiment) variability, six replicates of four serum samples were used to measure IgG reactivities towards all microarray components on the same day. To assess inter-assay (i.e. between experiments) variability, four serum samples were evaluated in experiments conducted on ﬁve consecutive days. The evaluation of four different samples allowed determining whether the inter-assay variability was sample-dependent. For each microarray component, the mean ISU-G, standard deviation (SD), coefﬁcient of variation (CV % = SD/mean ISU-G) across the six replicates and ﬁve different experiments, respectively, were calculated for each sample. Methods S l ti This cohort and the genotyping of RV strains in the nasopharyngeal swab samples of 108 of the 120 children by nested PCR and sequencing have been previously 35 g ISU-G, 25% and 12.5% were determined as CV% for intra-assay variability and subtracted from the data, respectively. RV-speciﬁc IgM antibodies were not mea- sured because we found in an earlier study that no increases of RV-speciﬁc IgM antibody levels were detectable on days 4, 7, and 42 after experimental RV inoculation36. g ISU-G, 25% and 12.5% were determined as CV% for intra-assay variability and subtracted from the data, respectively. RV-speciﬁc IgM antibodies were not mea- sured because we found in an earlier study that no increases of RV-speciﬁc IgM antibody levels were detectable on days 4, 7, and 42 after experimental RV inoculation36. described . A molecular diagnostic platform for the rapid detection of 15 respiratory strains was used in 118 of the 120 children. The following respiratory viruses were found: Adenovirus: 7 children; Bocavirus: 8 children; Coronavirus: 6 children; Inﬂuenza A/B: 1 child; Metapneumovirus: 3 children; Parainﬂuenzavirus: 4 children: RSV: 22 children52. For 108 of the 120 children nasopharyngeal swabs had been available for the RV PCR targeting VP4/2 sequences. Written informed consent was obtained from the parents or by the legal guardians and the study was approved by the Regional Ethics Committee of Karolinska Institutet, Stockholm, Sweden. Peripheral blood samples had been obtained within 24 h of presentation in the emergency unit and sera were stored at −80 °C. In addition to blood samples, nasopharyngeal swab samples were obtained at the acute visit and again at the follow-up visit by the research nurse and stored in the biobank at the Department of Clinical Microbiology, Karolinska University Hospital35. Follow-up samples were obtained between 6 and 30 weeks after the initial recruitment (median 11 weeks) at a scheduled visit after recovery. Although this study was not planned as a prospective study for the assessment of increases of RV-speciﬁc antibodies, the time interval of 6–30 weeks was suited for this purpose because we found in an earlier study that increase of RV-speciﬁc IgG responses emerge 42 days after experimental inoculation36. Methods S l ti Microarray components were clas- siﬁed according to the ISU-G level (>50, 25–50, 1–25) and averages across all components within these groups on the array were also calculated for each of these quantities. Detection antibodies and printing of microarrays. Anti-huIgG (Cat: 309-005- 008; Jackson ImmunoResearch Laboratories, West Grove, PA, USA) and anti- huIgA (Cat.: 555885; Becton Dickinson, Franklin Lakes, NJ) were labeled with DyLight 650 (Pierce, Thermo Fisher Scientiﬁc, Rockford, IL, USA). Customized printing of RV microarrays was done by Phadia-ThermoFisher using ImmunoCAP ISAC (Immuno Solid-phase Allergen Chip) technology50,51. Spotting was per- formed by slow pin mode printing using the Aushon 2410 Printer (Aushon, Bill- erica, MA, USA). Stock solutions of peptides (5 mg/ml) were diluted 1:4 in a phosphate buffer, pH 8.4 and then used for spotting. Antigens were spotted in triplicates on a glass surface coated with an amino-reactive organic polymer, each spot containing 50–200 fg of microarray component, corresponding to 1–5 atto- mol. Allergens used for the calibration and other control proteins spotted on the microarray are listed in Supplementary Table 6. Data analysis. Initial data processing was performed with Microsoft Excel. Fre- quencies (i.e., the number of reactive sera) and intensities (i.e., ISU-G levels) of peptide-speciﬁc IgG and IgA antibody levels were calculated using IBM SPSS Statistics (version 24; IBM Corp., Armonk, NY, USA). Median values of peptide- speciﬁc IgG levels (ISU-G) were calculated using GraphPad Prism 6 (La Jolla, CA, USA). The cut-off for a positive IgG reactivity was set at 1 ISU-G. The numbers of IgG-reactive peptides in Fig. 2d, e were calculated for speciﬁc IgG levels >15 ISU-G. Increases of RV-speciﬁc antibody responses were calculated as differences between ISU-G values measured at the follow-up (F) and the acute (A) visits (ΔISU-G = ISU-GF −ISU-GA) followed by subtraction of the double coefﬁcient of variation (2 × CV%) calculated from the baseline values (ISU-GA) for each of the antigens. For IgG levels of 1–5 ISU-G and >5 Cohort of pre-school children with acute wheeze. Serum samples examined in this study were from a cohort of 120 pre-school children who had been admitted to the Paediatric Emergency Ward as a result of acute wheeze, at Astrid Lindgren Children’s Hospital, Stockholm, Sweden (Table 1). Methods S l ti For the unsupervised analysis of antibody responses to RV peptides and proteins, unsupervised K-means clustering (K = 4) was used to deﬁne clusters of peptides with similar antibody response measurements. The K number of clusters was pre-determined in order to match the number of the peptide homology groups. The peptides’ amino acid sequences were aligned with a Gap open cost of 10.0 and a Gap extension cost of 1.0. Based on the alignment, a homology distance cladogram was built using the Neighbor-Joining algorithm and 1000 bootstrap replicates. The peptides were then color-coded based on the antibody response cluster that they belonged to. Data were processed using the CLC Genomics Workbench (CLC, CLCbio, Qiagen, Hilden, Germany). The heat map representing RV-speciﬁc antibody responses was generated by Qlucore Omics Explorer (Qlucore, Lund, Sweden). Expression and puriﬁcation of recombinant RV proteins. Recombinant his- tagged structural (VP1–4) proteins from ﬁve representative RV strains (RV-A2, -A16, -A89, -B14, and -CYP) and MBP fusion proteins containing fragments thereof (VP1–3) as well as non-structural (2A, 2C, 3A, 3C, and 3D) proteins from RV strain 89 were expressed in Escherichia coli as previously described33,36. DNA sequences coding for the complete genes or fragments thereof (accession numbers NATURE COMMUNICATIONS\| (2018) 9:2382 8 \| DOI: 10.1038/s41467-018-04591-0\| www.nature.com/naturecommunications \| DOI: 10.1038/s41467-018-04591-0\| www.nature.com/naturecommunications ARTICLE NATURE COMMUNICATIONS \| DOI: 10.1038/s41467-018-04591-0 are shown in Supplementary Table 5) were codon optimized for bacterial expression, synthesized with the addition of the 3′ sequence coding for a C- terminal hexa-histidine tag and cloned into the NdeI and EcoRI sites of plasmid pET27b (Genscript, Piscataway, NJ, USA). Transformed E. coli BL21 (DE3) cells (Agilent Technologies, Santa Clara, CA, USA) were induced with 1 mM isopropyl- β-thiogalactopyranoside (IPTG) and cells were harvested at time-points of max- imal expression. Recombinant proteins were puriﬁed by Nickel-afﬁnity chroma- tography under denaturing conditions as previously described (Qiagen, Hilden, Germany). Refolding of recombinant proteins was achieved by a stepwise dialysis against 10 mM NaH2PO4 for structural and non-structural proteins and 20 mM Tris-HCl, 200 mM NaCl, 1 mM EDTA for MBP fusion proteins, respectively. The purity of recombinant proteins was veriﬁed by SDS-PAGE followed by Coomassie Brilliant Blue staining and the identity by immunoblotting using a monoclonal mouse anti-His-tag antibody 1:1000 diluted (Cat: DIA-900, Dianova, Hamburg, Germany). Bound antibodies were detected with 1:1000 diluted alkaline phosphatase-coupled rat anti-mouse IgG antibodies (Cat: 557272; BD Biosciences, Erembodegem, Belgium). Methods S l ti At the follow-up visit, the guardians also ﬁlled out a standardized questionnaire concerning the number of days the child had suffered from respiratory symptoms at home (i.e., ‘cough and/or wheeze’), use of medica- tion (i.e., β2-agonists, inhaled corticosteroids, leukotriene receptor antagonist), as well as about any emergency visits between the acute and follow-up visits. The chip analysis of the anonymized sera was performed with the approval of the ethics committee of Medical University of Vienna (EK1721/2014)37,50. Statistical analysis. GraphPad Prism 6 (La Jolla, CA, USA) was used to evaluate all statistical parameters. Correlation between the number of reactive peptides and age was evaluated by calculating the Spearman’s rank correlation coefﬁcient (ρ). Differences between groups in number of reactive peptides and in number of days the children spent with respiratory symptoms, medication use, or with both were assessed by Mann–Whitney U-test (two-tailed). Values of p < 0.05 were considered statistically signiﬁcant. Data availability. Primary data that support the ﬁndings of this study are available from the corresponding author on request. Received: 3 January 2018 Accepted: 7 May 2018 \| DOI: 10.1038/s41467-018-04591-0\| www.nature.com/naturecommunications References Screening respiratory samples for detection of human rhinoviruses (HRVs) and enteroviruses: comprehensive VP4-VP2 typing reveals high incidence and genetic diversity of HRV species C. J. Clin. Microbiol. 47, 3958–3967 (2009). 15. Greve, J. M. et al. The major human rhinovirus receptor is ICAM-1. Cell 56, 839–847 (1989). 16. Hofer, F. et al. Members of the low density lipoprotein receptor family mediate cell entry of a minor-group common cold virus. Proc. Natl. Acad. Sci. USA 91, 1839–1842 (1994). 44. Ledford, R. M. et al. VP1 sequencing of all human rhinovirus serotypes: insights into genus phylogeny and susceptibility to antiviral capsid-binding compounds. J. Virol. 78, 3663–3674 (2004). 17. Bochkov, Y. A. et al. Cadherin-related family member 3, a childhood asthma susceptibility gene product, mediates rhinovirus C binding and replication. Proc. Natl. Acad. Sci. USA 112, 5485–5490 (2015). p 45. Niederberger, V. et al. Antigens drive memory IgE responses in human allergy via the nasal mucosa. Int. Arch. Allergy Immunol. 142, 133–144 (2007). 46. Egger, C., Horak, F., Vrtala, S., Valenta, R. & Niederberger, V. Nasal application of rBet v 1 or non-IgE-reactive T-cell epitope-containing rBet v 1 fragments has different effects on systemic allergen-speciﬁc antibody responses. J. Allergy Clin. Immunol. 126, 1312–1315 (2010). 18. Bizzintino, J. et al. Association between human rhinovirus C and severity of acute asthma in children. Eur. Respir. J. 37, 1037–1042 (2011). p 19. Miller, E. K. et al. Human rhinovirus C associated with wheezing in hospitalised children in the Middle East. J. Clin. Virol. 46, 85–89 (2009). responses. J. Allergy Clin. Immunol. 126, 1312–1315 (2010). 47. Heck, S., Nguyen, J., Le, D. D., Bals, R. & Dinh, Q. T. Pharmacological therapy of bronchial asthma: the role of biologicals. Int. Arch. Allergy Immunol. 168, 241–252 (2015). 20. Lee, S. et al. A polyvalent inactivated rhinovirus vaccine is broadly immunogenic in rhesus macaques. Nat. Commun. 7, 12838 (2016). immunogenic in rhesus macaques. Nat. Commun. 7, 12838 (2016). 21. McCray, J. & Werner, G. Different rhinovirus serotypes neutralized by antipeptide antibodies. Nature 329, 736–738 (1987). 21. McCray, J. & Werner, G. Different rhinovirus serotypes neutralized by antipeptide antibodies. Nature 329, 736–738 (1987). 48. Gauthier, M., Ray, A. & Wenzel, S. E. Evolving concepts of asthma. Am. J. Respir. Crit. Care Med. 192, 660–668 (2015). 22. Edlmayr, J. et al. References 28, 513–520 (2017). 37. Westman, M. et al. Mechanisms for the development of allergies consortium. Early childhood IgE reactivity to pathogenesis-related class 10 proteins predicts allergic rhinitis in adolescence. J. Allergy Clin. Immunol. 135, 1199–1206 (2015). 8. Jackson, D. J. et al. Wheezing rhinovirus illnesses in early life predict asthma development in high-risk children. Am. J. Respir. Crit. Care Med. 178, 667–672 (2008). 38. Posa, D. et al. Evolution and predictive value of IgE responses toward a comprehensive panel of house dust mite allergens during the ﬁrst 2 decades of life. J. Allergy Clin. Immunol. 139, 541–549 (2017). 9. Caliskan, M. et al. Rhinovirus wheezing illness and genetic risk of childhood- onset asthma. N. Engl. J. Med. 368, 1398–1407 (2013). g 10. Holgate, S. T. The sentinel role of the airway epithelium in asthma pathogenesis. Immunol. Rev. 242, 205–219 (2011). 39. Westman, M., Asarnoj, A., Hamsten, C., Wickman, M. & van Hage, M. Windows of opportunity for tolerance induction for allergy by studying the evolution of allergic sensitization in birth cohorts. Semin. Immunol. 30, 61–66 (2017). 11. Heymann, P. W. & Kennedy, J. L. Rhinovirus-induced asthma exacerbations during childhood: the importance of understanding the atopic status of the host. J. Allergy Clin. Immunol. 130, 1315–1316 (2012). 40. Wickman, M. et al. Detection of IgE reactivity to a handful of allergen molecules in early childhood predicts respiratory allergy in adolescence. EBioMedicine 26, 91–99 (2017). gy 12. Palmenberg, A. C. et al. Sequencing and analyses of all known human rhinovirus genomes reveal structure and evolution. Science 324, 55–59 (2009). 41. Henry, C., Palm, A. E., Krammer, F. & Wilson, P. C. From original antigenic sin to the universal inﬂuenza virus vaccine. Trends Immunol. 39, 70–79 (2018). 13. McIntyre, C. L., Knowles, N. J. & Simmonds, P. Proposals for the classiﬁcation of human rhinovirus species A, B and C into genotypically assigned types. J. Gen. Virol. 94, 1791–1806 (2013). 42. Jartti, T. et al. Serial viral infections in infants with recurrent respiratory illnesses. Eur. Respir. J. 32, 314–320 (2008). 14. Uncapher, C. R., DeWitt, C. M. & Colonno, R. J. The major and minor group receptor families contain all but one human rhinovirus serotype. Virology 180, 814–817 (1991). 43. Wisdom, A., Leitch, E. C., Gaunt, E., Harvala, H. & Simmonds, P. References 1. Holgate, S. T., Roberts, G., Arshad, H. S., Howarth, P. H. & Davies, D. E. The role of the airway epithelium and its interaction with environmental factors in asthma pathogenesis. Proc. Am. Thorac. Soc. 6, 655–659 (2009). Microarray-based determination of antibody proﬁles. Microarrays were washed in a washing buffer (Phadia-Thermo Fisher) for 5 min by stirring. After drying by centrifugation (1 min, 1000 g, RT), 35 µl of serum samples were applied on each microarray and the slides were incubated for 2 h at gentle rocking (RT). For the detection of RV-speciﬁc IgG and IgA antibodies, serum samples were diluted 1:300 and 1:20 in a sample dilution buffer (Phadia-Thermo Fisher), respectively. Microarrays were then rinsed with washing buffer and washed for 5 min as described above. After centrifugation, 30 µl of ﬂuorescence-labeled antibodies (1 µg/ml) was added and the slides were incubated 30 min at gentle rocking (in dark, RT). After further rinsing, washing and drying, microarrays were scanned using a confocal laser scanner (LuxScan-10K microarray scanner, Capital-Bio, Beijing, p g 2. Edwards, M. R., Bartlett, N. W., Hussell, T., Openshaw, P. & Johnston, S. L. The microbiology of asthma. Nat. Rev. Microbiol. 10, 459–471 (2012). 3. Hansel, T. T., Johnston, S. L. & Openshaw, P. J. Microbes and mucosal immune responses in asthma. Lancet 381, 861–873 (2013). 4. Global Initiative for Asthma (GINA). Global strategy for asthma management and prevention (updated 2015) http://ginasthma.org/wp-content/uploads/ 2016/01/GINA_Report_2015_Aug11-1.pdf (2015). p g p 5. Global Initiative for Chronic Obstructive Lung Disease (GOLD). Global strategy for the diagnosis, management, and prevention of chronic obstructive 9 \| DOI: 10.1038/s41467-018-04591-0\| www.nature.com/naturecommunications \| DOI: 10.1038/s41467-018-04591-0\| www.nature.com/naturecommunications NATURE COMMUNICATIONS\| (2018) 9:2382 ARTICLE NATURE COMMUNICATIONS \| DOI: 10.1038/s41467-018-04591-0 pulmonary disease (updated 2015). http://www.goldcopd.org/uploads/users/ ﬁles/GOLD_Report_2015_Apr2.pdf (2015). pulmonary disease (updated 2015). http://www.goldcopd.org/uploads/users/ ﬁles/GOLD_Report_2015_Apr2.pdf (2015). 35. Stenberg-Hammar, K. et al. Rhinovirus-speciﬁc antibody responses in preschool children with acute wheeze reﬂect severity of respiratory symptoms. Allergy 71, 1728–1735 (2016). p p p 6. Heymann, P. W., Platts-Mills, T. A. & Johnston, S. L. Role of viral infections, atopy and antiviral immunity in the etiology of wheezing exacerbations among children and young adults. Pediatr. Infect. Dis. J. 24, 217–222 (2005). gy 36. Niespodziana, K. et al. Rhinovirus-induced VP1-speciﬁc antibodies are group- speciﬁc and associated with severity of respiratory symptoms. EBioMedicine 2, 64–70 (2014). g y g f 7. Stenberg-Hammar, K., Hedlin, G. & Soderhall, C. Rhinovirus and preschool wheeze. Pediatr. Allergy Immunol. References A combination vaccine for allergy and rhinovirus infections based on rhinovirus-derived surface protein VP1 and a nonallergenic peptide of the major timothy grass pollen allergen Phl p 1. J. Immunol. 182, 6298–6306 (2009). 49. Focke-Tejkl, M. et al. Development and characterization of a recombinant, hypoallergenic, peptide-based vaccine for grass pollen allergy. J. Allergy Clin. Immunol. 2, 1207–1207 (2015). 50. Lupinek, C. et al. Advances in allergen-microarray technology for diagnosis and monitoring of allergy: the MeDALL allergen-chip. Methods 66, 106–119 (2014). 23. Edlmayr, J. et al. Antibodies induced with recombinant VP1 from human rhinovirus exhibit cross-neutralisation. Eur. Respir. J. 37, 44–52 (2011). 51. Gallerano, D. et al. HIV microarray for the mapping and characterization of HIV-speciﬁc antibody responses. Lab Chip 15, 1574–1589 (2015). 24. McLean, G. R. et al. Rhinovirus infections and immunisation induce cross- serotype reactive antibodies to VP1. Antivir. Res. 95, 193–201 (2012). 52. Stenberg-Hammar, K. et al. Subnormal levels of vitamin D are associated with acute wheeze in young children. Acta Paediatr. 103, 856–861 (2014). 25. Glanville, N. et al. Cross-serotype immunity induced by immunization with a conserved rhinovirus capsid protein. PLoS Pathog. 9, e1003669 (2013). 26. Traub, S. et al. An anti-human ICAM-1 antibody inhibits rhinovirus-induced exacerbations of lung inﬂammation. PLoS Pathog. 9, e1003520 (2013). Acknowledgements This study was funded by PreDicta, a FP7-funded EU project (No. 260895), by the FWF- funded projects F4605 and P29398 of the Austrian Science Fund, by research grants from Biomay AG and Viravaxx, Vienna, Austria, by the Swedish Research Council, the Swedish Heart-Lung Foundation, Stockholm County Council (ALF project), the Swedish Asthma and Allergy Association´s Research Foundation, the King Gustaf V´s 80-year Foundation, the Centre for Allergy Research at Karolinska Institutet and the Swedish Cancer and Allergy Foundation. g g 27. Jain, S. et al. Community-acquired pneumonia requiring hospitalization amoug U.S children. N. Engl. J. Med. 372, 835–845 (2015). 28. Bochkov, Y. A., Grindle, K., Vang, F., Evans, M. D. & Gern, J. E. Improved molecular typing assay for rhinovirus species A, B, and C. J. Clin. Microbiol. 52, 2461–2471 (2014). 29. Pavia, A. T. Viral infections of the lower respiratory tract: old viruses, new viruses, and the role of diagnosis. Clin. Infect. Dis. 52, 284–289 (2011). ( ) 30. Advani, S., Sengupta, A., Forman, M., Valsamakis, A. & Milstone, A. M. Detecting respiratory viruses in asymptomatic children. Pediatr. Infect. Dis. J. 31, 1221–1226 (2012). Author contributions Conception and design: K.N., S.M., N.G.P., and R.V.; analysis and interpretation: K.N., K.S.-H., S.M., C.R.C., K.N.-W., P.C.V., D.G., C.L., D.E., T.S., .C.H., C.S., M.v.H., G.H., N.G.P., and R.V. Drafting and or reading the manuscript for important intellectual content: K.N., K.S.-H., S.M., C.R.C., K.N.W., P.C.V., D.G., C.L., D.E., T.S., .C.H., C.S., M.v.H., G.H., N.G.P., and R.V. 31. van der Zalm, M. M. et al. Respiratory pathogens in children with and without respiratory symptoms. J. Pediatr. 154, 396–400 (2009). 32. Winther, B., Hayden, F. G. & Hendley, J. O. Picornavirus infections in children diagnosed by RT-PCR during longitudinal surveillance with weekly sampling: association with symptomatic illness and effect of season. J. Med. Virol. 78, 644–650 (2006). NATURE COMMUNICATIONS \| DOI: 10.1038/s41467-018-04591-0 NATURE COMMUNICATIONS \| DOI: 10.1038/s41467-018-04591-0 Novartis, Faes Farma, Biomay AG, Vienna, Austria, HAL, Nutricia Research, Menarini, MEDA, Abbvie, MSD, Omega Pharma, Danone, grants from Menarini, outside the submitted work. R.V. reports grants from European Union, grants and personal fees from Biomay AG, Vienna, Austria, grants and personal fees from Viravaxx, Vienna, Austria, during the conduct of the study; grants from Austrian Science Fund (FWF), grants and personal fees from Biomay AG, Vienna, Austria, grants and personal fees from Viravaxx AG, Vienna, Austria, outside the submitted work. In addition, R.V. and K.N. are co-inventors in a patent application (PCT/AT2010/000416) regarding the rhinovirus diagnosis reported in this paper. The remaining authors declare no competing interests. Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this license, visit http://creativecommons.org/ licenses/by/4.0/. Additional information 33. Niespodziana, K. et al. Misdirected antibody responses against an N-terminal epitope on human rhinovirus VP1 as explanation for recurrent RV infections. FASEB J. 26, 1001–1008 (2012). Supplementary Information accompanies this paper at https://doi.org/10.1038/s41467- 018-04591-0. Supplementary Information accompanies this paper at https://doi.org/10.1038/s41467- 018-04591-0. Competing interests: C.L. and M.v.H. report personal fees from Thermo Fisher Scientiﬁc, outside the submitted work. M.v.H. serves as a consultant for Biomay AG, Vienna, Austria and Hycor Biomedical LLC, CA, US. N.G.P. reports personal fees from 34. Iwasaki, J. et al. Comparison of rhinovirus antibody titers in children with asthma exacerbations and species-speciﬁc rhinovirus infection. J. Allergy Clin. Immunol. 134, 25–32 (2014). NATURE COMMUNICATIONS\| (2018) 9:2382 10 NATURE COMMUNICATIONS\| (20 \| DOI: 10.1038/s41467-018-04591-0\| www.nature.com/naturecommunications \| DOI: 10.1038/s41467-018-04591-0\| www.nature.com/naturecommunications © The Author(s) 2018 Reprints and permission information is available online at http://npg.nature.com/ reprintsandpermissions/ Reprints and permission information is available online at http://npg.nature.com/ reprintsandpermissions/ © The Author(s) 2018 Publisher's note: Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional afﬁliations. NATURE COMMUNICATIONS\| (2018) 9:2382 11 \| DOI: 10.1038/s41467-018-04591-0\| www.nature.com/naturecommunications
https://openalex.org/W2994957387	https://europepmc.org/articles/pmc6956016?pdf=render	English	null	Self-Filtering Monochromatic Infrared Detectors Based on Bi2Se3 (Sb2Te3)/Silicon Heterojunctions	Nanomaterials	2,019	cc-by	6,234	Self-Filtering Monochromatic Infrared Detectors Based on Bi2Se3 (Sb2Te3)/Silicon Heterojunctions Xujie Pan, Jing He, Lei Gao * and Handong Li * School of Materials and Energy, University of Electronic Science and Technology of China, Chengdu 61005 China; xujiepan@163.com (X.P.); hjing@std.uestc.edu.cn (J.H.) * Correspondence: gaolei@std.uestc.edu.cn (L.G.); hdli@uestc.edu.cn (H.L.) Xujie Pan, Jing He, Lei Gao * and Handong Li * School of Materials and Energy, University of Electronic Science and Technology of China, Chengdu 610054, China; xujiepan@163.com (X.P.); hjing@std.uestc.edu.cn (J.H.) * Correspondence: gaolei@std.uestc.edu.cn (L.G.); hdli@uestc.edu.cn (H.L.) Received: 8 November 2019; Accepted: 10 December 2019; Published: 12 December 2019 Received: 8 November 2019; Accepted: 10 December 2019; Published: 12 December 2019 Abstract: This paper focuses on the photoelectric properties of heterostructures formed by surface-modiﬁed Si (111) and hexagonal, quintuple-layered selenides (Bi2Se3 and Sb2Te3). It was shown that H-passivated Si (111) can form robust Schottky junctions with either Bi2Se3 or Sb2Te3. When back illuminated (i.e., light incident towards the Si side of the junction), both the Bi2Se3/Si and Sb2Te3/Si junctions exhibited signiﬁcant photovoltaic response at 1030 nm, which is right within the near-infrared (NIR) light wavelength range. A maximum external quantum eﬃciency of 14.7% with a detection response time of 2 ms for Bi2Se3/Si junction, and of 15.5% with a 0.8 ms response time for the Sb2Te3/Si junction, were achieved. Therefore, utilizing Si constituents as high-pass ﬁlters, the Bi2Se3 (Sb2Te3)/Si heterojunctions can serve as monochromatic NIR photodetectors. Keywords: Bi2Se3; Sb2Te3; Si photovoltaic detector; external quantum eﬃciency; response time Keywords: Bi2Se3; Sb2Te3; Si photovoltaic detector; external quantum eﬃciency; response time nanomaterials nanomaterials nanomaterials www.mdpi.com/journal/nanomaterials Nanomaterials 2019, 9, 1771; doi:10.3390/nano9121771 1. Introduction Topological insulators (TIs) possess helical surface states locked by time-reversal symmetry (TRS), and have been predicted to be promising candidates for electronic and optoelectronic device applications [1–4]. Moreover, the growth of these TIs on Si substrates has been optimized, and such heterostructures were shown to exhibit a Schottky junction which has already been widely studied [5–7]. However, most research has focused on the growth and optoelectronic properties of TIs. In fact, combining TI with other materials is also worthy of attention. Silicon is well known for its application in near-infrared optical detection due to its low dark current fast response and easy integration. Near-infrared light detectors are typically based on diode or Schottky structures; the key unit of a diode detector is its P–N or PIN junction, which requires a basic bias voltage. Although the diode detector has a large dark current and low response speed, its quantum eﬃciency is high. Compared to those simple diode-based detectors, Schottky junction-based detectors require metal to couple with silicon; compared with PIN detection, the latter has a faster response speed, but also has the problem of a low potential barrier, which causes a huge leakage of current. In addition, the problem of reactions between metal and silicon at the interface cannot be ignored, as it may cause poor device performance. Based on these issues, we innovatively propose the use of Bi2Se3 and Sb2Te3 to construct a heterostructure with Si. The layered nature of these two materials can oﬀer a good interfacial characteristic; the performance of the Schottky junction is also likely to be further improved. On the other hand, the band gap of Si lies in the near infrared wavelength region (~1.12 eV), which implies that photons in the infrared region can be absorbed by Si in principle. However, for practical applications, monochromatic optical detectors are more desirable. So, some more work is still to be done. The traditional method is to add a ﬁlter structure to ﬁlter out the light of high frequencies, but this increases the diﬃculty of implementing the device. Meanwhile, the ﬁlter or membrane structure is costly, and therefore, not conducive to commercialization. As an indirect bandgap semiconductor, www.mdpi.com/journal/nanomaterials 2 of 8 Nanomaterials 2019, 9, 1771 the absorption coeﬃcient of Si for diﬀerent frequencies photons varies greatly. 2. Materials and Methods The high quality, single-crystalline Bi2Se3 thin ﬁlm on a Si (111) substrate was grown by molecular beam epitaxy (MBE), whereas the Sb2Te3 ﬁlm was grown using the physical vapor deposition (PVD) technique [8,9]. As a large lattice mismatch exists between Bi2Se3 and Si (111), surface modiﬁcation engineering was employed for Si (111). H-passivated Si (111) was obtained by rinsing Si (111) in hydroﬂuoric acid (HF) for 10 min after a standard cleaning procedure. This H-passivated Si (111) was then loaded into a MBE chamber and degassed for 24 h to get rid of the impurities and obtain a clean surface. The surface structure of Si (111) was modiﬁed by ﬂashing at high temperature (over 1000 ◦C with Si exhibiting a bright orange-yellow color) for about 30 s. Bi was grown ﬁrstly to further minimize the lattice mismatch, followed by the growth of the Bi2Se3 ﬁlm. Sb2Te3 ﬁlm was prepared on a H-passivated Si (111) substrate via PVD. The Bi2Se3 ﬁlm was prepared on Bi-passivated Si (111). After the growth, the crystalline quality was studied by High resolution X-ray diﬀraction (XRD, Bede 1, from Jordan Valley Semiconductors). The photodetector device from the Bi2Se3 on H-passivated Si (111) (BS/H–Si) was fabricated to analyze its photoelectric properties. A self-ﬁltered silicon Schottky monochrome detector was fabricated according to the following steps. The device has four layers. The ﬁrst layer is a window electrode. The window electrode comprises a Ga-In alloy to make ohmic contact with the Si substrate. The second layer is the Si substrate with a low-pass ﬁlter function. The Si substrate was single crystalline along (111) orientation, with both sides polished, a thickness of 300 µm, and a resistance of 1~10 Ω/cm. The third layer consisted of a metal ﬁlm (typically Bi2Se3 or Sb2Te3) with high infrared reﬂectivity. Furthermore, the metal ﬁlm here is Bi2Se3 or Sb2Te3 ﬁlm grown along a (001) crystalline direction, with a thickness ranging from 50 nm to 200 nm. Indium was used as the bottom electrode to form ohmic contact with the Bi2Se3 or Sb2Te3 ﬁlm. An external bias was applied on the sample device using source meter (Keithley 2400, from Tektronix, Beaverton, Oregon, USA). Then, solar light from a class 3A sun simulator was used to shine light vertically incident on the sample from the Si substrate side. 2. Materials and Methods Bi2Se3 on the Bi-passivated Si (111) (BS/Bi–Si), Sb2Te3 H-passivated Si (111) device was fabricated and tested in the same way. QEX 10 equipment was employed to study the photoelectric eﬀect. The surface morphology of the samples grown by MEB was observed by in situ scanning tunneling microscopy (STM). 1. Introduction For example, light with a 600 nm wavelength cannot penetrate a Si substrate with a thickness of 300 microns, while a wavelength of 1000 nm can still have 25% strength after passing through the Si substrate of the same thickness. Therefore, selective high-pass ﬁltering can be realized by adjusting the thickness of the Si substrate. In this report, the photoelectric eﬀect of the heterostructure formed by Bi2Se3 and Sb2Te3 with P–Si (111) is analyzed. The high wavelength light was incident from the Si substrate side, which is transparent to wavelengths. The substrate of Si as a high-pass ﬁlter formed a near infrared monochromatic light detector with the Bi2Se3 and Sb2Te3 ﬁlm. The external quantum eﬃciency and response speed of the photodetector are also analyzed. These results demonstrated the great potential of this TI/Si heterostructure for high-performance optoelectronic device applications. 3. Results The curve is also in good accordance with Richardson-Dushman thermionic emission theory [10,11]. It is worth noting that the light was incident from the Si substrate side, in order to study the Bi2Se3/H–Si heterostructure. This is because the strong reflection and absorption of Si in the visible range can be used as a natural low-pass filter for the device [13]. Part of the light passed through the Si substrate and reached the heterojunction, and the different type of junctions caused the I-V curve to change. Furthermore, an evident photoelectric effect was observed in the Bi2Se3/H– Si heterostructure, from which we can note that the open circuit voltage (Voc) was 0.16V, the short circuit current (Isc) was 0 05 mA and the photoelectric conversion efficiency was 0 083% Figure 2a presents a schematic illustration of the Bi2Se3/H–Si device under illumination. Figure 2b shows the current-voltage (I-V) characteristics of the device under sunlit and dark conditions at room temperature while and external bias voltage is applied. According to the typical diode characteristics, it can be concluded that the ﬁlm formed a robust Schottky contact with the P–Silicon substrate. The curve is also in good accordance with Richardson-Dushman thermionic emission theory [10–12]. It is worth noting that the light was incident from the Si substrate side, in order to study the Bi2Se3/H–Si heterostructure. This is because the strong reﬂection and absorption of Si in the visible range can be used as a natural low-pass ﬁlter for the device [13]. Part of the light passed through the Si substrate and reached the heterojunction, and the diﬀerent type of junctions caused the I-V curve to change. Furthermore, an evident photoelectric eﬀect was observed in the Bi2Se3/H–Si heterostructure, from which we can note that the open circuit voltage (Voc) was 0.16 V, the short circuit current (Isc) was 0.05 mA, and the photoelectric conversion eﬃciency was 0.083%. circuit current (Isc) was 0.05 mA, and the photoelectric conversion efficiency was 0.083%. (a) (b) -1.0 -0.5 0.0 0.5 1.0 -0.01 0.00 0.01 0.02 0.03 0.04 0.05 0.06 Currnet (mA) Voltage(V) dark sun Figure 2c shows the typical current-voltage curve, from which we can determine that a Schottky contact exists between BS and Bi–Si. However, when light was incident from the substrate, the current changed slightly. For the heterostructure formed by the Sb2Te3 ﬁlm with H-passivation Si (ST/H–Si), the same test was done. 3. Results Bi2Se3 and Sb2Te3 were grown on H-passivated Si (111) via MBE and PVD respectively. During the growth of Bi2Se3 via MBE, an additional Bi layer was grown on H-passivated Si (111) as a buﬀer layer to minimize the lattice mismatch. The crystal structures of the as-grown ﬁlms were studied by XRD. The XRD patterns are displayed in Figure 1a; they demonstrate diﬀraction peaks corresponding to the (00l) planes of the Bi2Se3 and Sb2Te3. The XRD results indicate that the Bi2Se3 and Sb2Te3 ﬁlms are crystalline, and of high quality, as demonstrated by the full width at half maxima (FWHMs) of the (006) plane of the Bi2Se3 and Sb2Te3 ﬁlm. 3 of 8 3 of 8 Nanomaterials 2019, 9, 1771 Na o ate ial 2020 10 FOR Figure 1. XRD spectrum of Bi2Se3 and Sb2Te3 thin film on a Si (111) substrate. (a) 2Theta scan, and (b) rocking curve of Bi2Se3 (006) and Sb2Te3 (006). The blue line represents Sb2Te3 film on H–Silicon, the red line represents Bi2Se3 film on Bi–Silicon; the black line represents the Bi2Se3 film on H–Silicon. Figure 1. XRD spectrum of Bi2Se3 and Sb2Te3 thin ﬁlm on a Si (111) substrate. (a) 2Theta scan, and (b) rocking curve of Bi2Se3 (006) and Sb2Te3 (006). The blue line represents Sb2Te3 ﬁlm on H–Silicon, the red line represents Bi2Se3 ﬁlm on Bi–Silicon; the black line represents the Bi2Se3 ﬁlm on H–Silicon. Figure 1. XRD spectrum of Bi2Se3 and Sb2Te3 thin film on a Si (111) substrate. (a) 2Theta scan, and (b) rocking curve of Bi2Se3 (006) and Sb2Te3 (006). The blue line represents Sb2Te3 film on H–Silicon, the red line represents Bi2Se3 film on Bi–Silicon; the black line represents the Bi2Se3 film on H–Silicon. Figure 1. XRD spectrum of Bi2Se3 and Sb2Te3 thin ﬁlm on a Si (111) substrate. (a) 2Theta scan, and (b) rocking curve of Bi2Se3 (006) and Sb2Te3 (006). The blue line represents Sb2Te3 ﬁlm on H–Silicon, the red line represents Bi2Se3 ﬁlm on Bi–Silicon; the black line represents the Bi2Se3 ﬁlm on H–Silicon. Figure 2a presents a schematic illustration of the Bi2Se3/H–Si device under illumination. Figure 2b shows the current-voltage (I-V) characteristics of the device under sunlit and dark conditions at room temperature while and external bias voltage is applied. According to the typical diode characteristics, it can be concluded that the film formed a robust Schottky contact with the P–Silicon substrate. 3. Results Figure 2d indicates that ST/H–Si formed a Schottky contact and had a photoelectric conversion eﬃciency when the solar was incident from the substrate side. Figure 3a shows that the quantum eﬃciency of the BS/H–Si heterojunction changed signiﬁcantly at wavelengths between 800 nm and 1200 nm. However, at other wavelengths, the external quantum eﬃciency was not even detected. This indicates that when the wavelength of light was larger than 1200 nm, electrons could not be excited, but at less than 800 nm, photons could not reach the heterojunction due to being blocked by the silicon. The eﬀect of diﬀerent reverse bias voltages applied on the BS/H–Si heterojunction were also studied. When the reverse voltage was 0.1 V, the external quantum eﬃciency was low. But when it increased to 0.5 V, it increased rapidly, indicating that the space charge layer had enlarged. However, as the voltage continued to increase, the external quantum eﬃciency increased slowly. Furthermore, the sample broke down when the reverse voltage was more than 2 V. This indicated that the space charge layer was nearly saturated at 2 V. Meanwhile, the maximum value reached 14.7% when the wavelength was 1030 nm and the reverse voltage was 2 V. The ST/H–Si heterojunction also had a similar external quantum eﬃciency as that shown in Figure 3b. Obvious quantum eﬃciency occurred between 800 nm and 1200 nm, when the reverse voltage was applied. The space charge layer was approaching saturation at 2 V. Meanwhile, the quantum conversion eﬃciency was the highest at 1030 nm, reaching 15.5%, under the same reverse voltage. 4 of 8 he shor Nanomaterials 2019, 9, 1771 Si heterostructure, fr circuit current (Isc) w (a) (b) -1.0 -0.5 0.0 0.5 1.0 -0.01 0.00 0.01 0.02 0.03 0.04 0.05 0.06 Currnet (mA) Voltage(V) dark sun nomaterials 2020, 10, x FOR PEER REVIEW 4 of (c) (d) Figure 2. (a) Schematic of BS/Si heterostructure photodetector (b) I-V curve of BS/H–Si heterojunction (c) I-V curve of BS/Bi–Si heterojunction (d) I-V curve of ST/H–Si heterojunction in the under illumination and dark. -2.0 -1.0 0.0 1.0 2.0 -4 -2 0 2 4 6 8 10 12 14 Current(mA) Voltage(V) dark sun -1.0 -0.5 0.0 0.5 1.0 -1.00 -0.50 0.00 0.50 1.00 1.50 2.00 2.50 dark sun Current(mA) Voltage(V) Figure 2. 3. Results (a) Schematic of BS/Si heterostructure photodetector (b) I-V curve of BS/H–Si heterojunction (c) I-V curve of BS/Bi–Si heterojunction (d) I-V curve of ST/H–Si heterojunction in the under illumination and dark. (a) Nanomaterials 2020, 10, x FOR PEER REVIEW -1.0 -0.5 0.0 0.5 1.0 -0.01 0.00 0.01 0.02 0.03 0.04 0.05 0.06 Currnet (mA) Voltage(V) dark sun (b) (a) (d) -1.0 -0.5 0.0 0.5 1.0 -1.00 -0.50 0.00 0.50 1.00 1.50 2.00 2.50 dark sun Current(mA) Voltage(V) (c) -2.0 -1.0 0.0 1.0 2.0 -4 -2 0 2 4 6 8 10 12 14 Current(mA) Voltage(V) dark sun (d) (c) Figure 2. (a) Schematic of BS/Si heterostructure photodetector (b) I-V curve of BS/H–Si heterojunction (c) I-V curve of BS/Bi–Si heterojunction (d) I-V curve of ST/H–Si heterojunction in the under illumination and dark Figure 2. (a) Schematic of BS/Si heterostructure photodetector (b) I-V curve of BS/H–Si heterojunction (c) I-V curve of BS/Bi–Si heterojunction (d) I-V curve of ST/H–Si heterojunction in the under illumination and dark. Figure 2c shows the typical current-voltage curve, from which we can determine that a Schottky contact exists between BS and Bi–Si. However, when light was incident from the substrate, the current changed slightly. For the heterostructure formed by the Sb2Te3 film with H-passivation Si (ST/H–Si), the same test was done. Figure 2d indicates that ST/H–Si formed a Schottky contact and had a photoelectric conversion efficiency when the solar was incident from the substrate side. Figure 3a shows that the quantum efficiency of the BS/H–Si heterojunction changed significantly at wavelengths between 800 nm and 1200 nm. However, at other wavelengths, the external quantum efficiency was not even detected. This indicates that when the wavelength of light was larger than 1200 nm, electrons could not be excited, but at less than 800 nm, photons could not reach the heterojunction due to being blocked by the silicon. The effect of different reverse bias voltages applied on the BS/H–Si heterojunction were also studied. When the reverse voltage was 0.1V, the external quantum efficiency was low. But when it increased to 0.5V, it increased rapidly, indicating that the space charge layer had enlarged. However, as the voltage continued to increase, the external quantum efficiency increased slowly. Furthermore, the sample broke down when the reverse voltage was more than 2V. This indicated that the space charge layer was nearly saturated at 2V. 3. Results Meanwhile, the maximum value reached 14 7% when the wavelength was 1030 nm and the reverse voltage was STM can provide us with surface morphology information. From the Figure 3c, the surface of H-passivated Si (111) is atomic rough, which indicates that monohydride, dihydride, and trihydride coexist on the surface of H-passivated Si [14]. The H–Si bonding causes the ﬁrst silicon layer to be slightly relaxed inward [15]. This indicates that the lattice constant of Si becomes larger when H bonds with Si, resulting in the narrowing of the band gap. The band gap of silicon at room temperature was 1.12 eV, and light longer than 1100 nm cannot be absorbed [16]. However, due to the compression of the band gap, light with a wavelength of 1100 nm to 1200 nm can be absorbed. The H-passivated surface of Si near the ﬁlm was protected, but the other side which was exposed to the air was oxidized [17]; this led to the lattice constant having no change. Silicon with diﬀerent surface lattice constants is formed, which results in external quantum eﬃciency at 1100 nm to 1200 nm. Light from 800 nm to 1000 nm has low transmittance and high absorptance when passing through the Si substrate, while light from 1000 nm to 1200 nm has high transmittance and low absorptance. Considering these two eﬀects, it shows that an external quantum eﬃciency peak appears at 1030 nm. The Bi2Se3 ﬁlm has a good reﬂectivity in the near-infrared waveband, and the light passing through the substrate is reﬂected back to the substrate by the ﬁlm, causing secondary absorption, which increases the external quantum eﬃciency value, but causes the external quantum eﬃciency peak to be incompletely symmetric. 5 of 8 5 of 8 Nanomaterials 2019, 9, 1771 Nanomaterials 2020, 10, x FOR (a) (b) (c) (d) Figure 3. (a)The EQE of BS/H–Si heterostructure (b)The EQE of ST/H–Si heterostructure (c) STM of H–Si (111) (d) STM of Bi–Si (111). Figure 3. (a) The EQE of BS/H–Si heterostructure (b) The EQE of ST/H–Si heterostructure (c) STM of H–Si (111) (d) STM of Bi–Si (111). Nanomaterials 2020, 10, x FOR PEER REVIEW 6 (a) (b) Figure 4. (a) Response speed of BS/H–Si under 250 Hz light pulse, and (b) response speed of ST/H–Si under 600 Hz light pulse. 3. Results 0 10 20 30 40 50 0 5 10 15 20 25 Test signal(mV) reference square wave(mV) voltage(mV) Time (ms) 0 5 10 15 20 -2 0 2 4 6 8 10 12 14 Test signal(mV) reference square wave(mV) voltage(mV) Time (ms) Nanomaterials 2020, 10, x FOR PEER REVIEW (b) (d) (b) 0 5 10 15 20 -2 0 2 4 6 8 10 12 14 Test signal(mV) reference square wave(mV) voltage(mV) Time (ms) (b) (a) (c) (a) 0 10 20 30 40 50 0 5 10 15 20 25 Test signal(mV) reference square wave(mV) voltage(mV) Time (ms) (a) (d) (b (c) (a Figure 3. (a)The EQE of BS/H–Si heterostructure (b)The EQE of ST/H–Si heterostructure (c) STM of H–Si (111) (d) STM of Bi–Si (111) Figure 3. (a) The EQE of BS/H–Si heterostructure (b) The EQE of ST/H–Si heterostructure (c) STM of H–Si (111) (d) STM of Bi–Si (111). Figure 4. (a) Response speed of BS/H–Si under 250 Hz light pulse, and (b) response speed of ST/H–Si under 600 Hz light pulse STM can provide us with surface morphology information. From the Figure 3c, the surface of H- passivated Si (111) is atomic rough, which indicates that monohydride, dihydride, and trihydride coexist on the surface of H-passivated Si [14] The H–Si bonding causes the first silicon layer to be slightly relaxed inward [15] This indicates that the lattice constant of Si becomes larger when H From Figure 4a, the BS/H–Si device exhibited a fast response at 250 Hz pulse. So, the response speed in one cycle was studied under a 250 Hz light pulse; the rising time and decay time were 2 ms. The response speed of the ST/H–Si device was tested. As shown in Figure 4b, the response speed in one cycle was studied under a 600 Hz light pulse; the rising time and decay time was 0.8 ms. 4. Discussion nds with Si, resulting in the narrowing of the band gap. The band gap of silicon at room mperature was 1.12 eV, and light longer than 1100 nm cannot be absorbed [16]. However, due to e compression of the band gap, light with a wavelength of 1100 nm to 1200 nm can be absorbed he H-passivated surface of Si near the film was protected, but the other side which was exposed to e air was oxidized [17]; this led to the lattice constant having no change. 3. Results Silicon with different rface lattice constants is formed, which results in external quantum efficiency at 1100 nm to 1200 m. Light from 800 nm to 1000 nm has low transmittance and high absorptance when passing rough the Si substrate, while light from 1000 nm to 1200 nm has high transmittance and low sorptance. Considering these two effects, it shows that an external quantum efficiency peak pears at 1030 nm. The Bi2Se3 film has a good reflectivity in the near-infrared waveband, and the ht passing through the substrate is reflected back to the substrate by the film, causing secondary sorption, which increases the external quantum efficiency value, but causes the external quantum iciency peak to be incompletely symmetric. From Figure 4a, the BS/H–Si device exhibited a fast response at 250 Hz pulse. So, the response eed in one cycle was studied under a 250 Hz light pulse; the rising time and decay time were 2 ms he response speed of the ST/H–Si device was tested. As shown in Figure 4b, the response speed in e cycle was studied under a 600 Hz light pulse; the rising time and decay time was 0.8 ms. To further explain the photocurrent generation process, the Bi2Se3/H–Si heterostructure w selected as an example The presence of a large number of dangling bonds on the Si (111) surfa Figure 4. (a) Response speed of BS/H–Si under 250 Hz light pulse, and (b) response speed of ST/H–Si under 600 Hz light pulse. onds with Si, resulting in the narrowing mperature was 1.12 eV, and light longer th he compression of the band gap, light with he H-passivated surface of Si near the film w he air was oxidized [17]; this led to the latt urface lattice constants is formed, which res m. Light from 800 nm to 1000 nm has low hrough the Si substrate, while light from 1 bsorptance. Considering these two effects, ppears at 1030 nm. The Bi2Se3 film has a go ght passing through the substrate is reflecte bsorption, which increases the external quan ficiency peak to be incompletely symmetric From Figure 4a, the BS/H–Si device exhi peed in one cycle was studied under a 250 H d gap. The band gap of silicon at room m cannot be absorbed [16]. 4. Discussion To further explain the photocurrent generation process, the Bi2Se3/H–Si heterostructure was selected as an example. The presence of a large number of dangling bonds on the Si (111) surface leads to a high density of surface states [18,19]; however, H-passivation can reduce the density of states on the Si surface eﬀectively [14,15]. Because of the diﬀerent work function of Si and Bi2Se3 [20,21], the electrons will ﬂow from the ﬁlm to the Si, and a space charge layer will be formed at their interface. The incident light stimulates the transition of electrons from the bulk valence band to the empty bulk conduction band or to unoccupied surface-states, which generates electron-hole pairs [22,23]. The electrons and holes are separated by the built-in ﬁeld at the interface. Electrons were injected into the Bi2Se3 ﬁlm and the holes were drifted to the electrodes. Thereby, a photocurrent is generated in the external circuit. The existence of the built-in electric ﬁeld at the interface reduces the recombination of hole carriers, which extends the lifetime of the carriers. The increasing reverse voltage causes the space charge layer to expand, which also promotes external quantum eﬃciency. The conclusion is consistent with our experimental results. From Figure 3d, the honeycomb structure shows that a Bi trimer was formed on the Si surface which could not reduce the surface state [24]. Because of the high surface state density of Si [18,19], the Fermi level was pinned when Si came into contact with Bi [25], which limited the space charge layer enlargement. The rectiﬁcation was relatively low, and thus, didn’t show any photoelectric response. This also veriﬁes the phenomena observed in our experiments. To further explain the photocurrent generation process, the Bi2Se3/H–Si heterostructure was selected as an example. The presence of a large number of dangling bonds on the Si (111) surface leads to a high density of surface states [18,19]; however, H-passivation can reduce the density of states on the Si surface eﬀectively [14,15]. Because of the diﬀerent work function of Si and Bi2Se3 [20,21], the electrons will ﬂow from the ﬁlm to the Si, and a space charge layer will be formed at their interface. The incident light stimulates the transition of electrons from the bulk valence band to the empty bulk conduction band or to unoccupied surface-states, which generates electron-hole pairs [22,23]. The electrons and holes are separated by the built-in ﬁeld at the interface. 4. Discussion Electrons were injected into the Bi2Se3 ﬁlm and the holes were drifted to the electrodes. Thereby, a photocurrent is generated in the external circuit. The existence of the built-in electric ﬁeld at the interface reduces the recombination of hole carriers, which extends the lifetime of the carriers. The increasing reverse voltage causes the space charge layer to expand, which also promotes external quantum eﬃciency. The conclusion is consistent with our experimental results. From Figure 3d, the honeycomb structure shows that a Bi trimer was formed on the Si surface which could not reduce the surface state [24]. Because of the high surface state density of Si [18,19], the Fermi level was pinned when Si came into contact with Bi [25], which limited the space charge layer enlargement. The rectiﬁcation was relatively low, and thus, didn’t show any photoelectric response. This also veriﬁes the phenomena observed in our experiments. 3. Results However, due t th of 1100 nm to 1200 nm can be absorbed ed, but the other side which was exposed t nt having no change. Silicon with differen rnal quantum efficiency at 1100 nm to 120 ance and high absorptance when passin 1200 nm has high transmittance and low that an external quantum efficiency pea ity in the near-infrared waveband, and th he substrate by the film, causing secondary ncy value, but causes the external quantum response at 250 Hz pulse. So, the respons pulse. S the BS/ d ed in one cycle was studied under a 250 Hz light pulse; the rising time and decay time were 2 m response speed of the ST/H–Si device was tested. As shown in Figure 4b, the response speed cycle was studied under a 600 Hz light pulse; the rising time and decay time was 0.8 ms. To further explain the photocurrent generation process, the Bi2Se3/H–Si heterostructure l t d l Th f l b f d li b d th Si (111) Figure 4. (a) Response speed of BS/H–Si under 250 Hz light pulse, and (b) response speed of ST/H–Si under 600 Hz light pulse. 6 of 8 Nanomaterials 2019, 9, 1771 Funding: This paper is supported by the National Key R&D Program of China (Grant Nos. 2018YFA0306703 and 2018YFA0306102), the National Natural Science Foundation of China (Grant No. 61474014), and the Sichuan Science and Technology Program under Grant No. 2019YJ0202. Acknowledgments: We thank Ali for helping polishing the manuscript. References In Proceedings of the APS March Meeting, Dallas, TX, USA, 21–25 March 2011. 7. Zhang, H.; Zhang, X.; Liu, C.; Lee, S.-T.; Jie, J. High-Responsivity, High-Detectivity, Ultrafast Topological Insulator Bi2Se3/Silicon Heterostructure Broadband Photodetectors. ACS Nano 2016, 10, 5113. [CrossRef] 7. Zhang, H.; Zhang, X.; Liu, C.; Lee, S.-T.; Jie, J. High-Responsivity, High-Detectivity, Ultrafast Topological Insulator Bi2Se3/Silicon Heterostructure Broadband Photodetectors. ACS Nano 2016, 10, 5113. [CrossRef] 8. Li, H.D.; Gao, L.; Li, H.; Wang, G.; Wu, J.; Zhou, Z.; Wang, Z. Growth and band alignment of Bi2Se3 topological insulator on H-terminated Si(111) van der Waals surface Appl Phys Lett 2013 102 074106 7. Zhang, H.; Zhang, X.; Liu, C.; Lee, S. T.; Jie, J. High Responsivity, High Detectivity, Ultrafast Topological Insulator Bi2Se3/Silicon Heterostructure Broadband Photodetectors. ACS Nano 2016, 10, 5113. [CrossRef] 8. Li, H.D.; Gao, L.; Li, H.; Wang, G.; Wu, J.; Zhou, Z.; Wang, Z. Growth and band alignment of Bi2Se3 topological insulator on H-terminated Si(111) van der Waals surface. Appl. Phys. Lett. 2013, 102, 074106. [CrossRef] 8. Li, H.D.; Gao, L.; Li, H.; Wang, G.; Wu, J.; Zhou, Z.; Wang, Z. Growth and band alignment of Bi2Se3 topological insulator on H-terminated Si(111) van der Waals surface. Appl. Phys. Lett. 2013, 102, 074106. [CrossRef] 9. Li, H.; Yu, S.; Li, Y.; Channa, A.I.; Ji, H.; Wu, J.; Niu, X.; Wang, Z. The coherent heteroepitaxy of high-index Bi2Se3 thin ﬁlm on nanofaceted Si(211) substrate. Appl. Phys. Lett. 2019, 115, 041602. [CrossRef] 10. TascOglu, I.L.; Aydemir, U.; AltNdal, S.; Kınacı, B.; Özçelik, S. Analysis of the forward and reverse bias I-V characteristics on Au/PVA:Zn/n-Si Schottky barrier diodes in the wide temperature range. J. Appl. Phys. 2011, 109, 054502. [CrossRef] 11. Toyama, N. Variation in the eﬀective Richardson constant of a metal-silicon contact due to metal-ﬁlm thickness. J. Appl. Phys. 1988, 63, 2720. [CrossRef] 12. Karata¸s, ¸S.; Altındal, ¸S.; Türüt, A.; Özmen, A. Temperature dependence of characteristic parameters of the H-terminated Sn/p-Si(1 0 0) Schottky contacts. Appl. Surf. Sci. 2003, 217, 250–260. [CrossRef] 3. Rajkanan, K.; Singh, R.; Shewchun, J. Absorption coeﬃcient of silicon for solar cell calculations. Solid S Electron. 1979, 22, 793–795. [CrossRef] 14. Trucks, G.W.; Raghavachari, K.; Higashi, G.S.; Chabal, Y.J. Mechanism of HF etching of silicon surfaces: A theoretical understanding of hydrogen passivation. Phys. Rev. Lett. 1990, 65, 504. [CrossRef] [PubMed] 15. References 1. Peng, H.; Dang, W.; Cao, J.; Chen, Y.; Wu, D.; Zheng, W.; Li, H.; Shen, Z.-X.; Liu, Z. Topological insulator nanostructures for near-infrared transparent ﬂexible electrodes. Nat. Chem. 2012, 4, 281–286. [CrossRef] [PubMed] 1. Peng, H.; Dang, W.; Cao, J.; Chen, Y.; Wu, D.; Zheng, W.; Li, H.; Shen, Z.-X.; Liu, Z. Topological insulator nanostructures for near-infrared transparent ﬂexible electrodes. Nat. Chem. 2012, 4, 281–286. [CrossRef] [PubMed] 2. Zhang, H.B.; Yu, H.L.; Bao, D.H.; Li, S.W.; Wang, C.X.; Yang, G.W. Magnetoresistance switch eﬀect of a Sn-doped Bi2Te3 topological insulator. Adv. Mater. 2012, 24, 132–136. [CrossRef] [PubMed] 2. Zhang, H.B.; Yu, H.L.; Bao, D.H.; Li, S.W.; Wang, C.X.; Yang, G.W. Magnetoresistance switch eﬀect of a Sn-doped Bi2Te3 topological insulator. Adv. Mater. 2012, 24, 132–136. [CrossRef] [PubMed] 3. Xiu, F.; He, L.; Wang, Y.; Cheng, L.; Chang, L.; Lang, M.; Huang, G.; Kou, X.; Zhou, Y.; Jiang, X.; et al. Manipulating surface states in topological insulator nanoribbons. Nat. Nanotechnol. 2011, 6, 216–221. [CrossRef] [PubMed] 3. Xiu, F.; He, L.; Wang, Y.; Cheng, L.; Chang, L.; Lang, M.; Huang, G.; Kou, X.; Zhou, Y.; Jiang, X.; et al. Manipulating surface states in topological insulator nanoribbons. Nat. Nanotechnol. 2011, 6, 216–221. [CrossRef] [PubMed] 4. Liu, L.; Nakano, S.; Kakimoto, K. Carbon concentration and particle precipitation during directional solidiﬁcation of multicrystalline silicon for solar cells. J. Cryst. Growth 2008, 310, 2192–2197. [CrossRef] 4. Liu, L.; Nakano, S.; Kakimoto, K. Carbon concentration and particle precipitation during directional solidiﬁcation of multicrystalline silicon for solar cells. J. Cryst. Growth 2008, 310, 2192–2197. [CrossRef] 5. Li, H.D.; Wang, Z.Y.; Kan, X.; Guo, X.; He, H.T.; Wang, Z.; Wang, J.N.; Wong, T.L.; Wang, N.; Xie, M.H. The Van der Waals epitaxy of Bi2Se3 on Si(111) vicinal surface: An approach to prepare high-quality thin ﬁlms of topological insulator. New J. Phys. 2010, 12, 103038. [CrossRef] 5. Li, H.D.; Wang, Z.Y.; Kan, X.; Guo, X.; He, H.T.; Wang, Z.; Wang, J.N.; Wong, T.L.; Wang, N.; Xie, M.H. The Van der Waals epitaxy of Bi2Se3 on Si(111) vicinal surface: An approach to prepare high-quality thin ﬁlms of topological insulator. New J. Phys. 2010, 12, 103038. [CrossRef] 6. Wang, Z.Y.; Li, H.D.; Kan, X.; Guo, X.; He, H.T.; Wang, Z.; Wang, J.N.; Wong, T.L.; Wang, N.; Xie, M.H. Growth of topological insulator Bi2Se3 thin ﬁlms by the van-der-Waals epitaxy on vicinal Si(111) substrate. Conﬂicts of Interest: The authors declare no conﬂict of interest. Conﬂicts of Interest: The authors declare no conﬂict of interest. 5. Conclusions In summary, single-crystalline Bi2Se3 and Sb2Te3 were grown on a H-passivated P–Si (111) successfully, and the ﬁlm quality was analyzed by XRD. The heterostructure of the Bi2Se3/H–Si and Sb2Te3/H–Si photodetectors showed an obviously photoelectric eﬀect when the light was incident from the substrate side. Meanwhile, the heterostructure of Bi2Se3/H–Si and Sb2Te3/H–Si possessed high quantum eﬃciency between 800 nm and 1200 nm and obvious response speeds. This indicated that photodetectors based on Bi2Se3/H–Si and Sb2Te3/H–Si have bright prospects in the future. Author Contributions: X.P. analysed the data and wrote the manuscript. J.H. characterized SPM. L.G. contributed to materials preparation and characterize the material properties. H.L. provided guidance for the experiment, and all authors participated in the discussion of the paper. Funding: This paper is supported by the National Key R&D Program of China (Grant Nos. 2018YFA0306703 and 2018YFA0306102), the National Natural Science Foundation of China (Grant No. 61474014), and the Sichuan Science and Technology Program under Grant No. 2019YJ0202. Acknowledgments: We thank Ali for helping polishing the manuscript. 7 of 8 Nanomaterials 2019, 9, 1771 References Hricovini, K.; Günther, R.; Thiry, P.; Taleb-Ibrahimi, A.; Indlekofer, G.; Bonnet, J.E.; Dumas, P.; Petroﬀ, Y.; Blase, X.; Zhu, X.; et al. Electronic structure and its dependence on local order for H/Si(111)-(1×1) surfaces. Phys. Rev. Lett. 1993, 70, 1992. [CrossRef] 16. Bludau, W.; Onton, A.; Heinke, W. Temperature dependence of the band gap of silicon. J. Appl. Phys. 2003, 45, 1846–1848. [CrossRef] 17. Neuwald, U.; Hessel, H.E.; Feltz, A.; Memmert, U.; Behm, R.J. Initial stages of native oxide growth on hydrogen passivated Si(111) surfaces studied by scanning tunneling microscopy. Appl. Phys. Lett. 1992, 60, 1307–1309. [CrossRef] 18. Northrup, J.E.; Cohen, M.L. Reconstruction Mechanism and Surface-State Dispersion for Si(111)-(2×1). Phys. Rev. Lett. 1982, 49, 1349–1352. [CrossRef] 19. Wachs, A.L.; Shapiro, A.P.; Hsieh, T.C.; Chiang, T.-C. Observation of ﬁlm states and surface-state precursors for Ag ﬁlms on Si(111). Phys. Rev. B Condens. Matter 1986, 33, 1460. [CrossRef] 20. Goldstein, B.; Szostak, D.J. Work function and Auger measurements of the initial oxidation of hydrogenated amorphous Si and of single-crystal Si. Appl. Phys. Lett. 1978, 33, 85–87. [CrossRef] 8 of 8 Nanomaterials 2019, 9, 1771 21. Kim, D.; Cho, S.; Butch, N.P.; Syers, P.; Kirshenbaum, K.; Adam, S.; Paglione, J.; Fuhrer, M.S. Surface conduction of topological Dirac electrons in bulk insulating Bi2Se3. Nat. Phys. 2012, 8, 459–463. [CrossRef] 21. Kim, D.; Cho, S.; Butch, N.P.; Syers, P.; Kirshenbaum, K.; Adam, S.; Paglione, J.; Fuhrer, M.S. Surface conduction of topological Dirac electrons in bulk insulating Bi2Se3. Nat. Phys. 2012, 8, 459–463. [CrossRef] 22. Zelikson, M.; Weiser, K.; Salzman, J.; Kanicki, J. Determination of electron and hole mobilities in an a-Si:H 22. Zelikson, M.; Weiser, K.; Salzman, J.; Kanicki, J. Determination of electron and hole mobilities in an a-Si:H from photo-electric eﬀects in a waveguide structure. J. Non Cryst. Solids 1991, 137, 455–458. [CrossRef] from photo-electric eﬀects in a waveguide structure. J. Non Cryst. Solids 1991, 137, 455–458. [CrossRef] 23. Hajlaoui, M.; Papalazarou, E.; Mauchain, J.; Perfetti, L.; Taleb-Ibrahimi, A.; Navarin, F.; Monteverde, M.; Auban-Senzier, P.; Pasquier, C.R.; Moisan, N.; et al. Tuning a Schottky barrier in a photoexcited topological insulator with, transient Dirac cone electron-hole asymmetry. Nat. Commun. 2014, 5, 3003. [CrossRef] 23. Hajlaoui, M.; Papalazarou, E.; Mauchain, J.; Perfetti, L.; Taleb-Ibrahimi, A.; Navarin, F.; Monteverde, M.; Auban-Senzier, P.; Pasquier, C.R.; Moisan, N.; et al. Tuning a Schottky barrier in a photoexcited topological insulator with, transient Dirac cone electron-hole asymmetry. Nat. Commun. 2014, 5, 3003. © 2019 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/). References [CrossRef] y y 24. Miwa, R.H.; Schmidt, T.M.; Srivastava, G.P. Bi covered Si(111) surface revisited. J. Phys. Condens. Matter 2003, 15, 2441–2447. [CrossRef] 25. Johansson, L.S.O.; Landemark, E.; Karlsson, C.J.; Uhrberg, R.I.G. Fermi-level pinning and surface-state band structure of the Si(111)-(√3×√3) R 30◦-Ag surface. Phys. Rev. Lett. 1989, 63, 2092–2095. [CrossRef] [PubMed] © 2019 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).
https://openalex.org/W2905007658	https://sistemascmc.ifam.edu.br/educitec/index.php/educitec/article/download/526/191	Portuguese	null	Projeto sustentável de escola flutuante ribeirinha	Educitec	2,018	cc-by	7,387	Sustainable project of riverside floating school Thiago Felix da Silva Instituto Federal de Educação, Ciência e Tecnologia do Amazonas thiagofelixdasilva099@gmail.com Vinicius dos Santos Albuquerque Instituto Federal de Educação, Ciência e Tecnologia do Amazonas viniciusalbuquerque.ptr@gmail.com LIOTTO, P. F.; BARROS, L. M.; SILVA, T. F.; ALBUQUERQUE, V. S. Educitec, Manaus, v. 04, n. 08, p. 179-194, nov. 2018. Edição especial. Abstract The variation of the level of the waters in the Amazon region influences the community activities, being necessary adaptations for the Amazon way of life. Thus we can highlight the interference in school activities of the riverside. With school education as the only door to progress, the riparian seeks to adapt to the environment. With the variation of water levels, the structure of the school is affected drastically causing interruptions in the school calendar. To mitigate these actions the research seeks patterns that are adaptable to this environment and that can still have a sustainable character provoking reinforcement in the formation of the riverside. The present research seeks to present a way to prevent interruption of school classes in the school period, even in unfavorable environmental conditions. Together with the model, some basic comfort standards for academic activity are proposed. The materials used in the proposed solutions constitute an adaptation without great ruptures in the riverside mode of living. Thus the use of wood in a sustainable way and other material, quite frequent in the industrialized cities, are allied forming a composite able to have the structural resistance of the wood and the thermal property of the plates made with residues of Expanded Polystyrene (Styrofoam). As a result, it was demonstrated the viability of a sustainable riverside school project, able to meet the needs of riverside schools, as comfort in the teaching-learning process and at the same time it was possible to adapt to hydrological variability in areas of difficult access such as the Amazon region. Key words: Amazon region. School structure. Materials. Sustainable. sustentável de escola ribeirinha, capaz de atender as necessidades escolas ribeirinhas, como conforto no processo ensino-aprendizagem e ao mesmo tempo foi possível a adaptação à variabilidade hidrológica em áreas de difícil acesso como a região amazônica. Palavras-chave: Região amazônica. Estrutura escolar. Materiais. Sustentável. Resumo A variação do nível das águas na região amazônica influencia as atividades comunitárias, sendo necessárias adaptações para o modo de vida amazônida. Assim podemos ressaltar a interferência nas atividades escolares dos ribeirinhos. Tendo a formação escolar como única porta para o progresso, o ribeirinho procura adaptar-se com o meio. Com a variação dos níveis das águas, a estrutura da escola é afetada drasticamente provocando interrupções no calendário letivo. Para mitigar essas ações a pesquisa busca padrões que sejam adaptáveis para este meio e que ainda possa ter caráter sustentável provocando reforços na formação dos ribeirinhos. A presente pesquisa busca apresentar uma forma de impedir a interrupção das aulas letivas no período escolar, ainda que em condições ambientais desfavoráveis. Aliada ao modelo é proposto alguns padrões de conforto elementar para a atividade acadêmica. Os materiais utilizados nas soluções propostas constituem uma adaptação sem grandes rupturas no modo ribeirinho de viver. Assim o uso da madeira de maneira sustentável e outro material, bastante frequente nas cidades industrializadas, são aliados formando um compósito capaz de ter a resistência estrutural da madeira e a propriedade térmica das placas feitas com resíduos de Poliestireno Expandido (Isopor). Como resultado, foi demonstrada a viabilidade de um projeto 179 LIOTTO, P. F.; BARROS, L. M.; SILVA, T. F.; ALBUQUERQUE, V. S. Educitec, Manaus, v. 04, n. 08, p. 179-194, nov. 2018. Edição especial. desenha ciclos hidrológicos, bem definidos, apresentando uma estação de águas altas (cheias) e uma estação de águas baixas (seca) por ano (PIEDADE, 2012). Nestes ciclos hidrológicos, acontece um trabalho complexo de permuta com o entorno, seja nas estações cheias ou secas, tornando-se quase impossível prever o comportamento dos níveis dos rios em cada período anual (FARIAS, 2011). Esse entrave reflete na dificuldade de se realizar um planejamento das atividades que envolvam a comunidade ribeirinha em especial as atividades escolares. A educação, bem como outras atividades realizadas pelas comunidades instaladas em áreas alagáveis, é afetada por essa variação sazonal nos níveis das águas. Para se adaptarem a esta realidade, as escolas nas comunidades rurais, utilizam um calendário especial, programando as atividades escolares no período de vazante que é o período onde se inicia a estação de baixa das águas (SCHERER, 2004). Nesse contexto as escolas ribeirinhas desenvolvem processos educacionais diferenciados para os dois grandes períodos: o de seca, onde se caracteriza por grandes erosões nas margens afetando a estrutura das escolas, e de inundações, que paralisam as atividades (GOMES, 1977). Por conta disto, o tempo mínimo de estudo é afetado. É estimado que no Brasil o tempo mínimo de estudo seja de 12 anos incluindo ensino fundamental incluindo ensino fundamental e médio. Porem, pesquisas apontam para números ainda mais baixos. No Brasil o tempo mínimo é da ordem de 7,1 anos enquanto que apenas na Amazônia Legal é menor ficando na casa dos 6,6 anos (SOUSA et al., 2016). Outro fator observado ainda é a distorção idade-série-elevada que na Amazônia Legal, em 2008, os alunos do ensino fundamental com idade superior a recomendada representavam 26%, sendo essa distorção maior no ensino médio que atingia 39% dos alunos (SOUSA et al., 2016). O regime especial de ensino dessas escolas, é formado em geral, por classes multisseriadas que atendem nos dois períodos, matutino e vespertino. Nas escolas de região alagável, em algumas localidades, o regime diferenciado começa no mês de agosto e prossegue até abril do ano seguinte, no entanto os dias letivos são obrigatórios e devem ser rigorosamente cumpridos. Algumas dessas escolas mudam de lugar por causa das cheias ou por causa do fenômeno provocado pela erosão dos rios para garantir a continuidade no ensino (GOMES, 1977). Para essas necessidades, as escolas flutuantes já são alternativas para o desenvolvimento da formação técnica-profissional de ribeirinhos. LIOTTO, P. F.; BARROS, L. M.; SILVA, T. F.; ALBUQUERQUE, V. S. Educitec, Manaus, v. 04, n. 08, p. 179-194, nov. 2018. Edição especial. Introdução Começamos por observar que o clima da Terra esteve, sempre, em mudanças, caracterizadas por períodos longos e curtos (BARCELLOS, 2009). Isso pode ser visto nos últimos relatórios emitidos por cientistas do Painel Intergovernamental sobre Mudanças Climáticas (IPCC), que avaliam a atuação do fenômeno conhecido com El Niño. Esse fenômeno é apontado pelos pesquisadores como produtor de um padrão de inundações e secas na região amazônica (FEARNSIDE, 2009). Devido a essas variabilidades climáticas e hidrográficas, as cheias e secas e a amplitude média do nível dos rios apresentam variações significativas. Considerando as cabeceiras dos rios e também a sua extensão, a sazonalidade LIOTTO, P. F.; BARROS, L. M.; SILVA, T. F.; ALBUQUERQUE, V. S. Educitec, Manaus, v. 04, n. 08, p. 179-194, nov. 2018. Edição especial. 180 Sua finalidade é desenvolver programas de Formação Profissional nos setores secundários e terciário, suprindo a necessidade de mão-de-obra qualificada que demanda os municípios da Bacia Hidrográfica do Amazonas. A modalidade de ensino e os cursos que são ofertados pela UMFS estão focados em Qualificação Profissional de nível Básico e Aperfeiçoamento Profissional e são administrados pelo SENAI (ALBUQUERQUE, 2013). Figura 01 – Barco escola Samaúma I Fonte: Conselho Nacional do SESI (2014) Figura 01 – Barco escola Samaúma I Fonte: Conselho Nacional do SESI (2014) Madhyapara é uma comunidade ribeirinha, no interior da região oeste, do país de Bangladesh. A organização não-governamental Shidhulai Swanirvar Sangstha, desenvolveu uma solução criativa que foi intitulado de Escola Barco de Madhyapara. A ideologia abordada foi a de “levar a escola até os alunos quando os alunos não podem ir à escola”, o que garante o maior alcance do ensino. A organização administra uma frota de barcos que atuam como bibliotecas, centro de educação de adultos e oficinas escolares e são levados em épocas de inundações onde escolas ficam impossibilitadas de serem utilizadas. Os barcos são equipados com painéis solares que alimentam computadores, luzes e outros equipamentos (VEIGA, 2018). Figura 02 – Escola barco de Madhyapara Fonte: National Geographic (2010) Figura 02 – Escola barco de Madhyapara Fonte: National Geographic (2010) Todos esses dois projetos, que inspiram a busca por diminuir fronteiras no ensino levando aos mais distantes pontos a educação, são na verdade soluções de tempo determinado que ampliam o conhecimento dos ribeirinhos em cursos de curta duração, deixando ainda uma lacuna para a estrutura de apoio ao ensino regular. Trata-se de escolas-barcos. Essas edificações se diferem das outras edificações, pois são definidas como sendo barcos com propulsão e construídas para navegação, servindo ocasionalmente de escolas. Sendo assim, quando observamos as tipologias de edificações flutuantes vemos que se dividem em dois tipos: Anfíbias – que acompanham o movimento do nível de água, mantendo-se na mesma posição – e habitação de base flutuante – que navegam autonomamente na água (NETO, 2015). Figura 02 – Escola barco de Madhyapara Fonte: National Geographic (2010) Todos esses dois projetos, que inspiram a busca por diminuir fronteiras no ensino levando aos mais distantes pontos a educação, são na verdade soluções de tempo determinado que ampliam o conhecimento dos ribeirinhos em cursos de curta duração, deixando ainda uma lacuna para a estrutura de apoio ao ensino regular. Seja na região Amazônica ou em outras regiões tropicais, barcos escolas levam um pedaço dos centros técnico-financeiro para as regiões menos desenvolvidas. A Unidade Móvel Fluvial Samaúma (UFMS), pode ser citada como exemplo de escola flutuante, porém com a característica de ser móvel, podendo ser autotransportada para qualquer comunidade ribeirinha do Estado do Amazonas. LIOTTO, P. F.; BARROS, L. M.; SILVA, T. F.; ALBUQUERQUE, V. S. Educitec, Manaus, v. 04, n. 08, p. 179-194, nov. 2018. Edição especial. 181 LIOTTO, P. F.; BARROS, L. M.; SILVA, T. F.; ALBUQUERQUE, V. S. Educitec, Manaus, v. 04, n. 08, p. 179-194, nov. 2018. Edição especial. As escolas, de ensino regular, necessitam de uma estrutura física mais próxima da fixa que venha se fixar na região proporcionando a regularidade que os ribeirinhos necessitam. Normalmente, essas escolas são construídas em tipos de palafitas, com estruturas e vedações em madeira local. Esse tipo de edificação está condicionado ao processo de deterioração da própria variação do nível das águas dos rios. Fenômenos como as aluviões e os desbarrancamentos das encostas podem ser alguns dos empecilhos que se sujeitam esse tipo de escola (HERMES, 2012). Aliada a essa necessidade é evidente que essas escolas precisam buscar meios eficazes de construção e preservação do ambiente, métodos alternativos de preservação e conservação do meio ambiente escolar formal afim de que os alunos tenham condições de desenvolver as suas atividades educacionais. Quanto mais tempo os alunos estejam em sala de aula confortável e de boa estrutura, tende-se a resolver muitos problemas como a defasagem do ano letivo e o alto índice de analfabetismo (HERMES, 2012). Alguns pesquisadores sensibilizados com as variáveis envolvidas em arquitetura das escolas, vêm buscando aperfeiçoar os espaços, principalmente os ambientes internos, propondo áreas mais agradáveis, mais adequadas e com melhores relações de ensino-aprendizagem aos alunos. Podemos de antemão verificar que o ser humano tem melhores condições de vida e saúde quando o mesmo vive em um ambiente adequado e que seu organismo funcione sem ser submetido à fadiga e estresse. A falta de conforto térmico é uma das mais recorrentes queixas dentre os fatores que englobam o equilíbrio ambiental, a saber tratam-se de três: o conforto térmico, como já citado, conforto acústico, visual e ergonômico (GRZYBOWSKI, 2004). As condições climáticas consideradas para este trabalho são as da realidade Amazônica e se caracterizam por bastante umidade e temperaturas acentuadas na maior parte do ano, exigindo a adoção de proteção no interior das edificações. A madeira, que é um excelente isolante térmico e presente em abundância na região, muitas das vezes é imaginada como a grande precursora na degradação ambiental de florestas. Porém, trata-se de um material renovável e que com técnicas de manejo sustentável e reflorestamento corretos tornasse uma excelente solução (BARROS; ALQUERQUE, 2016). Outro elemento que apresenta potencialidade na aplicação como isolamento térmico são os resíduos de Poliestireno Expandido (EPS). T t d l b E difi õ dif d t Figura 02 – Escola barco de Madhyapara Fonte: National Geographic (2010) Fonte: National Geographic (2010) Fonte: National Geographic (2010) Todos esses dois projetos, que inspiram a busca por diminuir fronteiras no ensino levando aos mais distantes pontos a educação, são na verdade soluções de tempo determinado que ampliam o conhecimento dos ribeirinhos em cursos de curta duração, deixando ainda uma lacuna para a estrutura de apoio ao ensino regular. Trata-se de escolas-barcos. Essas edificações se diferem das outras edificações, pois são definidas como sendo barcos com propulsão e construídas para navegação, servindo ocasionalmente de escolas. Sendo assim, quando observamos as tipologias de edificações flutuantes vemos que se dividem em dois tipos: Anfíbias – que acompanham o movimento do nível de água, mantendo-se na mesma posição – e habitação de base flutuante – que navegam autonomamente na água (NETO, 2015). LIOTTO, P. F.; BARROS, L. M.; SILVA, T. F.; ALBUQUERQUE, V. S. Educitec, Manaus, v. 04, n. 08, p. 179-194, nov. 2018. Edição especial. 182 LIOTTO, P. F.; BARROS, L. M.; SILVA, T. F.; ALBUQUERQUE, V. S. Educitec, Manaus, v. 04, n. 08, p. 179-194, nov. 2018. Edição especial. Esses resíduos, também em abundância, são originados das atividades industriais desenvolvidas pelo Polo Industrial de Manaus (PIM) polo este que foi implantado com o objetivo primordial de desenvolvimento da Amazônia evitando que a floresta fosse depredada. A produção de resíduo de EPS no PIM ultrapassa a ordem de 5.169 T/ano (BARROS; ALQUERQUE, 2016). Pesquisadores, ainda apontam os painéis de EPS, referente à construção modular, como sistema construtivo capaz de trazer muitos benefícios, pois possuem um leque de vantagens tais como versatilidade, rapidez e inovação (OLIVEIRA; VILELA, 2015). Outras pesquisas apontam os painéis sanduiche LIOTTO, P. F.; BARROS, L. M.; SILVA, T. F.; ALBUQUERQUE, V. S. Educitec, Manaus, v. 04, n. 08, p. 179-194, nov. 2018. Edição especial. 183 com miolo de EPS como mais um elemento no acondicionamento térmico ( LOURENSINI, 2017). Painéis sanduiches são caracterizados por estruturas compostas de duas camadas externas, estas por sua vez são materiais de densidade e resistência elevada e, uma camada interna de material com menor densidade e resistência mas que contribui para a eficiência térmica e redução do peso do painel (SILVA, 2012). Outro item importante de ser observado é a questão de destinação final para os esgotamentos sanitários. A poluição é tudo aquilo que altere o meio ambiente prejudicando suas características originais, afetando a saúde, o bem-estar, e a segurança. A poluição cria condições adversas às atividades sociais e econômicas, ocasionando danos relevantes à flora, à fauna, a qualquer recurso natural, e ainda aos acervos históricos, culturais e paisagísticos. O tratamento de esgoto previne ou atenua inúmeras doenças associadas aos poluentes, além de equilibrarem o meio ambiente. Existem vários métodos para o tratamento de efluentes poluidores. Para escolher o método mais adequado deve-se observar as condições do curso d´água, estudo de autodepuração e dos limites definidos pela legislação ambiental, das características do efluente gerado, a disponibilidade financeira, a eficiência e a área disponível. A preocupação com o grau de tratamento e ao destino final dos esgotos é crescente. As tecnologias apontam tratamentos de esgotos tendo como principal objetivo o lançamento em corpos d´água. Esses processos de tratamento biológico da matéria orgânica biodegradável constituem a alternativa mais interessante sob os pontos de vista técnico e econômico para a efetiva redução de concentração dos compostos preponderantes nos esgotos de escolas. A partir dessas observações é que se vê claramente uma grande barreira imposta pela natureza aos moradores ribeirinhos que se localizam em áreas de várzeas para dali obterem seus sustentos. O desenvolvimento de uma escola capaz de ser adaptável a essa realidade promovendo assim o aprimoramento dos indivíduos ribeirinhos sem que haja um conflito com a natureza é que demostra a importância deste tipo de estudo. Após a compreensão das dificuldades vivenciadas à educação e as possíveis propostas já vivenciadas pelas comunidades ribeirinhas no concernente à educação e as possíveis propostas já vividas por outras atividades semelhantes é que podemos fixar os objetivos deste trabalho. Portanto, o fruto deste trabalho é um projeto padrão de uma escola sustentável para localidades ribeirinhas que utilizarão madeiras locais e EPS como materiais compósitos para proporcionarem bons resultados térmicos. Para isso o projeto será composto por plantas arquitetônicas e seus respectivos detalhamentos, e ainda, os projetos complementares de instalações hidrossanitárias, responsáveis por modelar os sistemas de captação e tratamento de águas pluviais. Materiais e Métodos LIOTTO, P. F.; BARROS, L. M.; SILVA, T. F.; ALBUQUERQUE, V. S. Educitec, Manaus, v. 04, n. 08, p. 179-194, nov. 2018. Edição especial. LIOTTO, P. F.; BARROS, L. M.; SILVA, T. F.; ALBUQUERQUE, V. S. Educitec, Manaus, v. 04, n. 08, p. 179-194, nov. 2018. Edição especial. Materiais e Métodos 184 184 Resultados e discussões A partir da pesquisa de campo pode-se constar com clareza a problemática sofrida pelas escolas ribeirinhas em áreas alagáveis. Na visita realizada em período de cheia foram encontradas várias escolas interditadas, pois os níveis das águas fluviais já estavam perto dos níveis dos pisos e outras até com as águas cobrindo o nível dos pisos, como é apresentada nas figuras 03a e 03b. Figura 03 – Escolas em palafitas interditadas: (a) Escola Municipal Aldinei dos Santos Barroso, localizada no Município de Careiro da Várzea; (b) Escola Municipal São Cristovão, localizada no Município de Iranduba. Figura 03 – Escolas em palafitas interditadas: (a) Escola Municipal Aldinei dos Santos Barroso localizada no Município de Careiro da Várzea; (b) Escola Municipal São Cristovão, localizada no Município de Iranduba. Figura 03 – Escolas em palafitas interditadas: (a) Escola Municipal Aldinei dos Santos Barroso, localizada no Município de Careiro da Várzea; (b) Escola Municipal São Cristovão, localizada no Município de Iranduba. (a) (b) Fonte: Próprios autores (2017) A visita em áreas ribeirinhas também foi capaz de apresentar a tipificação comum utilizada pelos ribeirinhos para fugir das variações das águas. Estas casas que são denominadas de palafitas e construídas pelos proprietários, têm a madeira como elemento estrutural e elemento de vedação, telhas de alumínio e são elevadas além dos níveis historicamente mais altos, assim como é apresentado na figura 04a e 04b. Figura 04 – Construções típicas de ribeirinhos: (a) Edificação em madeira tipo palafita em período de cheias; (b) Edificação em madeira tipo palafita em período de vazante. (a) (b) Fonte: Próprios autores (2017) Figura 03 – Escolas em palafitas interditadas: (a) Escola Municipal Aldinei dos Santos Barroso, localizada no Município de Careiro da Várzea; (b) Escola Municipal São Cristovão, localizada no Município de Iranduba. (a) (b) Fonte: Próprios autores (2017) A visita em áreas ribeirinhas também foi capaz de apresentar a tipificação comum utilizada pelos ribeirinhos para fugir das variações das águas. Estas casas que são denominadas de palafitas e construídas pelos proprietários, têm a madeira como elemento estrutural e elemento de vedação, telhas de alumínio e são elevadas além dos níveis historicamente mais altos, assim como é apresentado na figura 04a e 04b. Para se idealizar uma escola padrão, partiu-se inicialmente das necessidades comuns encontradas em escolas ribeirinhas. A implantação e os detalhamentos foram caracterizados considerando a sua instalação no Lago de Janauacá. Este localiza-se a 150 km de Manaus entre os municípios do Careiro da Várzea e Manaquiri, no Estado do Amazonas. A comunidade detém a prática da produção agrícola com o cultivo da macaxeira junto com a atividade pesqueira de espécies Surubins e Pirararas, mas o lago também proporciona a pesca de outras espécies ícticas. Para nortear os trabalhos de elaboração do programa de necessidades e levantamento de dados diversos, foram realizadas análises socioculturais a partir de informações obtidas com entrevistas com alguns moradores, agentes comunitários e professores que trabalhavam em escolas próximas. As pesquisas de campos abordaram os seguintes aspectos: levantamento de edificações – analisando as tipologias, sistemas construtivos, questões socioambientais; uso da madeira aplicada à arquitetura – observando a adequação a realidade local aliada ao emprego com planejamento econômico e sustentável; levantamento fotográfico – caracterizando o problema e seus diversos empecilhos. A concepção da edificação foi realizada considerando três módulos: o primeiro concentrado as atividades administrativas escolares; e os outros dois, englobando as atividades pedagógicas escolares (salas de aula). O objetivo dessa divisão está em considerar a flexibilidade da construção. Quando considerada toda a edificação sobre uma base única, surgem problemas com áreas grandes em relação as variações de nível. Uma vez sendo divididas, as diversas subidas e descidas podem ser consideradas como diferenciais permitindo menor impacto na estrutura. Ainda, na Região Norte por conta da baixa densidade populacional, condições financeiras e falta de políticas públicas, o sistema de coleta e transporte de esgotos, formado por rede coletora, poços de visita, estações elevatórias e de tratamento, torna-se inviável. Diferentemente o que acontece nas grandes cidades, uma vez que esse custo se paga em pouco tempo com a contribuição dos impostos. Sendo assim, as estações de tratamento locais ou compactas surgem como ótima alternativa para atender pequenas demandas isoladas como o das comunidades ribeirinhas. Estas estações são sistemas descentralizados, caracterizados por sistemas coletivos de tratamento de esgoto simplificados, que apresentaram diversas vantagens, como por exemplo, baixo custo de implantação, necessidade de pequenas áreas para acomodações, elevada sustentabilidade do sistema, devido ao baixo consumo de energia e produtos químicos. Ainda possuem uma variedade de configurações de estações compactas que combinam sistemas anaeróbicos e sistemas anaeróbico/aeróbicos. 185 LIOTTO, P. F.; BARROS, L. M.; SILVA, T. F.; ALBUQUERQUE, V. S. Educitec, Manaus, v. 04, n. 08, p. 179-194, nov. 2018. Edição especial. Resultados e discussões Figura 04 Construções típicas de ribeirinhos: (a) Edificação em madeira tipo palafita em período no Município (a) (b) autores (2017) (a) (b) Fonte: Próprios autores (2017) (a) (b) Fonte: Próprios autores (2017) A visita em áreas ribeirinhas também foi capaz de apresentar a tipificação comum utilizada pelos ribeirinhos para fugir das variações das águas. Estas casas que são denominadas de palafitas e construídas pelos proprietários, têm a madeira como elemento estrutural e elemento de vedação, telhas de alumínio e são elevadas além dos níveis historicamente mais altos, assim como é apresentado na figura 04a e 04b. Figura 04 – Construções típicas de ribeirinhos: (a) Edificação em madeira tipo palafita em períod de cheias; (b) Edificação em madeira tipo palafita em período de vazante. Figura 04 – Construções típicas de ribeirinhos: (a) Edificação em madeira tipo palafita em período de cheias; (b) Edificação em madeira tipo palafita em período de vazante. Figura 04 – Construções típicas de ribeirinhos: (a) Edificação em madeira tipo palafita em período de cheias; (b) Edificação em madeira tipo palafita em período de vazante. (a) (b) Fonte: Próprios autores (2017) As escolas de base flutuante, portanto, tornam-se uma solução para atender o objetivo de propor estruturas modernas e eficientes, adaptáveis e confortáveis e que conseguem, com programa arquitetônico coerente, manter os padrões das escolas tais como conhecemos. Com isso o método de construção tipo palafitas ou os barcos-escola serão reformulados (HERMES, 2012). (a) (b) Fonte: Próprios autores (2017) (b) autores (2017) (a) (b) Fonte: Próprios autores (2017) Fonte: Próprios autores (2017) As escolas de base flutuante, portanto, tornam-se uma solução para atender o objetivo de propor estruturas modernas e eficientes, adaptáveis e confortáveis e que conseguem, com programa arquitetônico coerente, manter os padrões das escolas tais como conhecemos. Com isso o método de construção tipo palafitas ou os barcos-escola serão reformulados (HERMES, 2012). As escolas de base flutuante, portanto, tornam-se uma solução para atender o objetivo de propor estruturas modernas e eficientes, adaptáveis e confortáveis e que conseguem, com programa arquitetônico coerente, manter os padrões das escolas tais como conhecemos. Com isso o método de construção tipo palafitas ou os barcos-escola serão reformulados (HERMES, 2012). 186 Isso é possível, pois as edificações flutuantes combinam as características das escolas-barcos e o sistema de estacas das palafitas: quando ocorre a variação do nível da água a base fornece a flutuabilidade necessária (NETO, 2015). A escola ribeirinha padrão, portanto, foi projetada com três bases flutuantes e independentes, interligas, onde a estrutura será feita de aço naval, o mesmo usado confecção das balsas utilizadas na região. As bases flutuantes terão posicionamento fixo, variando apenas o nível conforme o fenômeno de subida e descida das águas. Os blocos foram então divididos em três tipos, sendo um administrativo que abriga a secretaria, diretoria, biblioteca, sala dos professores refeitório e cozinha, conforme é apresentado na figura 05b, o segundo bloco tipo, trata-se de um conjunto de quatro salas de aula com capacidade para 20 alunos cada, este bloco ainda conta com instalações sanitárias, o terceiro bloco tipo, trata-se de outro bloco de salas, mas com seis salas de aula, podendo uma ser adaptada para um laboratório de informática, como apresentado na figura 05a. Os blocos de sala de aula assim divididos proporcionam uma implantação padrão para atender a comunidade. Para as comunidades que necessitam de maior quantidade de salas de aula terão os dois blocos e, aquela comunidade que tem demanda menor, pode apenas ter um bloco. Importante verificarmos que por ser um projeto padrão de escolas flutuantes, deve ser levantado antes da construção a demanda da comunidade. Para a implantação no local de origem do projeto, localizado no Lago Janauacá, a escola flutuante contará com o complexo completo, sendo edificado os três blocos. LIOTTO, P. F.; BARROS, L. M.; SILVA, T. F.; ALBUQUERQUE, V. S. 187 Educitec, Manaus, v. 04, n. 08, p. 179-194, nov. 2018. Edição especial. Figura 05 – Bases flutuantes: (a) Plantas Baixas – Blocos de Sala de Aula; (b) Planta Baixa – Bloco Administrativo (a) Figura 05 – Bases flutuantes: (a) Plantas Baixas – Blocos de Sala de Aula; (b) Planta Baixa – Bloco Administrativo (a) LIOTTO, P. F.; BARROS, L. M.; SILVA, T. F.; ALBUQUERQUE, V. S. 187 Educitec, Manaus, v. 04, n. 08, p. 179-194, nov. 2018. Edição especial. LIOTTO, P. F.; BARROS, L. M.; SILVA, T. F.; ALBUQUERQUE, V. S. 187 LIOTTO, P. F.; BARROS, L. M.; SILVA, T. F.; ALBUQUERQUE, V. S. Educitec, Manaus, v. 04, n. 08, p. 179-194, nov. 2018. Edição especial. LIOTTO, P. F.; BARROS, L. M.; SILVA, T. LIOTTO, P. F.; BARROS, L. M.; SILVA, T. F.; ALBUQUERQUE, V. S. Educitec, Manaus, v. 04, n. 08, p. 179-194, nov. 2018. Edição especial. LIOTTO, P. F.; BARROS, L. M.; SILVA, T. F.; ALBUQUERQUE, V. S. LIOTTO, P. F.; BARROS, L. M.; SILVA, T. F.; ALBUQUERQUE, V. S. Educitec Manaus v 04 n 08 p 179-194 nov 2018 Edição especial F.; ALBUQUERQUE, V. S. Educitec, Manaus, v. 04, n. 08, p. 179-194, nov. 2018. Edição especial. (b) Fonte: Próprios autores (2018) Sistemas de ventilação Uma série de elementos foram inseridos no projeto para que fosse atingido o conforto térmico e, vão desde a base até a cobertura. As bases flutuantes terão aberturas no piso, devidamente sinalizadas e protegidas, que permitirão a troca massa de ar quente armazenada dentro da estrutura metálica com o meio externo com a ação dos ventos. As salas de aulas e todos os ambientes serão vedados com paredes sanduíches, sendo compostas por duas peças de madeira e no meio uma peça de isopor, conforme a figura 6a. Os forros dos tetos de cada ambiente serão compostos por EPS e, no teto dos ambientes térreos do bloco administrativo, será fixado EPS, conforme figura 6b e 6c. Figura 06 – Elementos de tratamento térmico: (a) Parede sanduiche; (b) Reforço térmico para o pavimento do bloco administrativo; (c) Instalação de telha sanduiche. (a) (a) (b) (c) Fonte: Próprios autores (2017) A madeira é um material que aliada à condição de sustentabilidade é adequada a vários usos na construção civil por ser renovável e de fácil emprego e reciclagem. Para o seu processamento não precisa de processos industriais complicados e poluidores, além da facilidade de ser trabalhada. (b) (c) (c) (b) (a) Fonte: Próprios autores (2017) Fonte: Próprios autores (2017) A madeira é um material que aliada à condição de sustentabilidade é adequada a vários usos na construção civil por ser renovável e de fácil emprego e reciclagem. Para o seu processamento não precisa de processos industriais complicados e poluidores, além da facilidade de ser trabalhada. A madeira é um material que aliada à condição de sustentabilidade é adequada a vários usos na construção civil por ser renovável e de fácil emprego e reciclagem. Para o seu processamento não precisa de processos industriais complicados e poluidores, além da facilidade de ser trabalhada. Trabalhar com madeira significa trabalhar com detalhes, com uma construção mais artesanal. Como material de construção, apresenta uma série de vantagens dificilmente reunidas em outro material, mas que destacamos para este trabalho as boas condições naturais de isolamento térmico e absorção acústica. Esse uso da madeira busca não apenas a facilidade de seu uso, existe, porém, com uso eficiente de recursos naturais e valores culturais, a contribuição para o pensamento da construção sustentável, adicionando informações do passado na arquitetura contemporâneo. acompanharão as subidas e descidas além de combater alguns movimentos prejudiciais. Nesses pilares ainda são projetados os dentes que formam o conjunto de cremalheiras que em cada pilar são instalados quatro na sua face lateral. Quando em níveis baixos a base é assentada em alicerces fixos semelhantes aos pilares de amarração. acompanharão as subidas e descidas além de combater alguns movimentos prejudiciais. Nesses pilares ainda são projetados os dentes que formam o conjunto de cremalheiras que em cada pilar são instalados quatro na sua face lateral. Quando em níveis baixos a base é assentada em alicerces fixos semelhantes aos pilares de amarração. (b) Fonte: Próprios autores (2018) As edificações flutuantes são projetadas para locais de permanente ligação com a terra propensos a inundações e utilizam sistemas para elevar-se e acompanhar a subida das águas. Quando o nível de água voltar ao normal a edificação assenta em radier e fica semelhante a tipologia tradicional (NETO, 2015). Essas subidas e descidas são guiadas por pilares de concreto armado que possuem pequenos dentes que a base da edificação é capaz de percorrer. Isso faz com que a edificação suba suavemente com a inundação e devido ao formato dos dentes é projetado um limite mínimo inferior para que seja regulado no período de vazantes. (b) Fonte: Próprios autores (2018) As edificações flutuantes são projetadas para locais de permanente ligação com a terra propensos a inundações e utilizam sistemas para elevar-se e acompanhar a subida das águas. Quando o nível de água voltar ao normal a edificação assenta em radier e fica semelhante a tipologia tradicional (NETO, 2015). Essas subidas e descidas são guiadas por pilares de concreto armado que possuem pequenos dentes que a base da edificação é capaz de percorrer. Isso faz com que a edificação suba suavemente com a inundação e devido ao formato dos dentes é projetado um limite mínimo inferior para que seja regulado no período de vazantes. Para garantir que a edificação não fique à deriva e avanço, que são movimentos próprios de flutuantes, e ainda, diminuição dos movimentos de arfagem que em águas fluviais são menores, esses pilares servirão também como ancoragem. Sendo assim, o sistema de ancoragem tem papel fundamental na estabilidade da edificação, sendo composto por quatro pilares para cada base flutuante que LIOTTO, P. F.; BARROS, L. M.; SILVA, T. F.; ALBUQUERQUE, V. S. 188 LIOTTO, P. F.; BARROS, L. M.; SILVA, T. F.; ALBUQUERQUE, V. S. Educitec, Manaus, v. 04, n. 08, p. 179-194, nov. 2018. Edição especial. LIOTTO, P. F.; BARROS, L. M.; SILVA, T. F.; ALBUQUERQUE, V. S. Educitec, Manaus, v. 04, n. 08, p. 179-194, nov. 2018. Edição especial. Sistemas de ventilação O sistema de ventilação da cobertura será composto por aberturas nas laterais do telhado que permitirão a entrada de ventos, possibilitando que haja a troca de ar quente armazenado na estrutura por ar mais resfriado. A saída do ar quente LIOTTO, P. F.; BARROS, L. M.; SILVA, T. F.; ALBUQUERQUE, V. S. Educitec, Manaus, v. 04, n. 08, p. 179-194, nov. 2018. Edição especial. 189 será por abertura na parte superior do telhado. Pode ser demonstrado esse sistema na figura 07 que apresenta uma vista lateral dos blocos de salas de aula. Figura 07 – Fachadas dos blocos de salas de aula. Fonte: Próprios autores (2017) A ventilação por meio de cavidade instalada no teto da edificação é um método eficaz de substituir o ar quente, acumulado no interior, por ar frio utilizando a convecção, assim é mostrado na figura 08. Esse método também auxilia na redução do calor irradiado do teto para as partes internas da edificação. Estes sistemas são proporcionados por instalações de aberturas no telhado e em outra posição podendo ser na região da empena ou no beiral dos telhados (COUNCIL, 2011). será por abertura na parte superior do telhado. Pode ser demonstrado esse sistema na figura 07 que apresenta uma vista lateral dos blocos de salas de aula. Figura 07 – Fachadas dos blocos de salas de aula. Fonte: Próprios autores (2017) Fonte: Próprios autores (2017) Fonte: Próprios autores (2017) A ventilação por meio de cavidade instalada no teto da edificação é um método eficaz de substituir o ar quente, acumulado no interior, por ar frio utilizando a convecção, assim é mostrado na figura 08. Esse método também auxilia na redução do calor irradiado do teto para as partes internas da edificação. Estes sistemas são proporcionados por instalações de aberturas no telhado e em outra posição podendo ser na região da empena ou no beiral dos telhados (COUNCIL, 2011). Figura 08 – Abertura de ventilação do telhado. Fonte: Sustainable Tropical Building Design Figura 08 – Abertura de ventilação do telhado. Fonte: Sustainable Tropical Building Design Para a cobertura foi considerada a telha termoacústica com isolamento de EPS, como apresentado na figura 09. Está telha é indicada pois a exigência de desempenho térmico se faz presente. Pelo fato de a edificação ser flutuante a cobertura não poderá contar com o auxílio de árvores para a diminuição da temperatura. Estas telhas contam com o bom desempenho não apenas térmico, mas acústico também e com menor custo quando comparadas com as de isolamento de PU (poliuretano). 190 LIOTTO, P. F.; BARROS, L. M.; SILVA, T. F.; ALBUQUERQUE, V. S. Educitec, Manaus, v. 04, n. 08, p. 179-194, nov. 2018. Edição especial. 190 Figura 09 – Telha sanduiche trapezoidal com isolamento de EPS. Fonte: Calha Forte (2016) p Fonte: Calha Forte (2016) Fonte: Calha Forte (2016) As telhas com isolamento EPS são fabricadas com placas de isopor nas medidas das telhas e coladas entre telhas metálicas com cola PU, formando uma espécie de sanduiche, com densidade de 15 a 25 g/m³ e com condutividade K=0,039 a 0,032 Kcal/M.H°C e espessura de 50 mm. Pelo fato dessas telhas serem mais rígidas, espera-se, também, reduzir o consumo de madeira na estrutura do telhado, pois assim não apenas estaria diminuindo a necessidade de extração da madeira como também o peso próprio da estrutura sobre as bases flutuantes. LIOTTO, P. F.; BARROS, L. M.; SILVA, T. F.; ALBUQUERQUE, V. S. Educitec, Manaus, v. 04, n. 08, p. 179-194, nov. 2018. Edição especial. Conclusões O projeto de uma escola flutuante é uma ideia inovadora para a região amazônica que luta anualmente com as interferências geradas pelas subidas e descidas das águas dos rios. As escolas necessitam comprimir o período letivo completo sem interrupções para que seja garantida a qualidade da aprendizagem. As comunidades ribeirinhas têm suas atividades locadas normalmente em torno dessas escolas, o que deposita sobre elas a responsabilidade da melhor condução da formação. Essa formação torna-se melhor quando apoia a comunidade em atividades que busquem a sustentabilidade e a não degradação permanente. Com isso a necessidade da escola buscar dentro de níveis alcançáveis modelos que proporcionem relação com o meio ambiente é apresentada no projeto, tornando esse projeto modelo em construção escolar para ribeirinhos da região amazônica. como também na superfície filtrante. O processo completo, quando bem instalado, é eficiente na redução de cargas orgânicas elevadas, melhorando ainda mais a qualidade dos despejos. Figura 09 – Esquema de instalação Fossa Filtro. Fonte: Próprios autores (2017) O sistema, que é locado no porão de cada base flutuante, foi dimensionado para receber anualmente uma extração do lodo da fossa enquanto que o filtro é capaz de se manter sem substituição dos elementos filtrantes por até 5 anos. Passado o tempo de substituição, o elemento filtrante, no caso a brita nº4 ou nº5 deve ser toda retirada e colocada nova. como também na superfície filtrante. O processo completo, quando bem instalado, é eficiente na redução de cargas orgânicas elevadas, melhorando ainda mais a qualidade dos despejos. Figura 09 – Esquema de instalação Fossa Filtro. Fonte: Próprios autores (2017) igura 09 – Esquema de instalação Fossa Filtro. Fonte: Próprios autores (2017) Fonte: Próprios autores (2017) O sistema, que é locado no porão de cada base flutuante, foi dimensionado para receber anualmente uma extração do lodo da fossa enquanto que o filtro é capaz de se manter sem substituição dos elementos filtrantes por até 5 anos. Passado o tempo de substituição, o elemento filtrante, no caso a brita nº4 ou nº5 deve ser toda retirada e colocada nova. O sistema, que é locado no porão de cada base flutuante, foi dimensionado para receber anualmente uma extração do lodo da fossa enquanto que o filtro é capaz de se manter sem substituição dos elementos filtrantes por até 5 anos. Passado o tempo de substituição, o elemento filtrante, no caso a brita nº4 ou nº5 deve ser toda retirada e colocada nova. LIOTTO, P. F.; BARROS, L. M.; SILVA, T. F.; ALBUQUERQUE, V. S. Educitec, Manaus, v. 04, n. 08, p. 179-194, nov. 2018. Edição especial. Sistemas de esgoto e drenagem O sistema de drenagem de águas pluviais composto por calhas de drenagem e tubulações servirá também para um abastecimento reserva que suprirá a necessidade hídrica dos vasos sanitários e áreas de lavagens de piso. O sistema de disposição final dos esgotos, para atender a escola e também buscar melhores condições de despejo dos dejetos no rio onde será locada a escola, constituirá o conjunto fossa séptica e filtro anaeróbio, como apresentado na figura 09. Esse sistema prevê a utilização de dois sistemas interligados de tratamento dos dejetos. É previsto também que cada base flutuante deva ser dotada do seu sistema, permitindo a independência dos sistemas. A fossa tem um funcionamento dividido em quatro funções básicas: a retenção, decantação, digestão e redução de volumes. Na retenção está contabilizado o período de 12 horas para seu funcionamento completo. A decantação processa uma sedimentação de 60 a 70% dos sólidos em suspensão finalizando-se em lodo. A outra parte que não decanta como óleos e gorduras são retidos na superfície livre do líquido sendo chamado de escuma. A digestão, parte principal do funcionamento da fossa, consiste no ataque das bactérias anaeróbias ao lodo e a escuma provocando uma destruição de organismos patogênicos. O resultado dessa etapa é a redução de volume, finalizando em gases, líquidos e reduzido volume dos sólidos retidos e digeridos. Estes sólidos assumem uma característica mais estável permitindo que o efluente líquido seja lançado com melhores condições. Para a passagem deste material para o filtro, que irá realizar outra etapa no tratamento, é previsto um filtro na saída superior. O conjunto filtro anaeróbio ainda é utilizado como mais um elemento nesse processo de tratamento. Este mecanismo trata-se de um reator biológico onde o efluente que já recebeu um melhoramento tem a oportunidade de ser depurado por meio de organismos anaeróbio, dispersos tanto no espaço vazio do reator 191 Referências ALBUQUERQUE, Regina Lucia Azevedo e colab. Projeto Samaúma: navegando para levar educação aos ribeirinhos da amazônia. Igapó - Revista de educação, 192 LIOTTO, P. F.; BARROS, L. M.; SILVA, T. F.; ALBUQUERQUE, V. S. Educitec, Manaus, v. 04, n. 08, p. 179-194, nov. 2018. Edição especial. LIOTTO, P. F.; BARROS, L. M.; SILVA, T. F.; ALBUQUERQUE, V. S. Educitec, Manaus, v. 04, n. 08, p. 179-194, nov. 2018. Edição especial. LIOTTO, P. F.; BARROS, L. M.; SILVA, T. F.; ALBUQUERQUE, V. S. Ciência e Tecnologia do IFAM, v. 07, n. 2238–4286, p. 50–55, 2013. Ciência e Tecnologia do IFAM, v. 07, n. 2238–4286, p. 50–55, 2013. LOURENSINI, Luise. Desenvolvimento de painel pré-fabricado em concreto armado para vedação, com núcleo composto por material para isolamento térmico. 2017. Monografia (Graduação em Engenharia Civil) – Universidade do Vale do Taquari - Univates, Lajeado, 07 dez. 2017. Disponível em: <http://hdl.handle.net/10737/1927>. LOURENSINI, Luise. Desenvolvimento de painel pré-fabricado em concreto armado para vedação, com núcleo composto por material para isolamento térmico. 2017. Monografia (Graduação em Engenharia Civil) – Universidade do Vale do Taquari - Univates, Lajeado, 07 dez. 2017. Disponível em: <http://hdl.handle.net/10737/1927>. BARCELLOS, Christovam e MONTEIRO, Antonio Miguel Vieira e CORVALAN, Carlos. Mudanças climáticas e ambientais e as doenças infecciosas: cenários e incertezas para o Brasil. Epidemiol. Serv. Saúde, v. 18, n. 3, p. 285– 304, 2009. Disponível em: p <http://scielo.iec.pa.gov.br/scielo.php?script=sci_arttext&pid=S1679- 49742009000300011&lng=pt&nrm=iso%3E>. BARROS, Laerte Melo; ALQUERQUE, Vinicius dos Santos. Projeto sustentável de uma escola ribeirinha em madeira com associação de materiais acústicos e térmicos. p. 1–7, 2016. COUNCIL, Cairns Regional. Sustainable Tropical Building Design: Guidlines for Commercial Buildings. p. 48, 2011. COUNCIL, Cairns Regional. Sustainable Tropical Building Design: Guidlines for Commercial Buildings. p. 48, 2011. FARIAS, Soad e colab. Preservação da paisagem ribeirinha na Amazônia Resumo. In: SEMINÁRIO DOCOMOMO BRASIL INTERDISCIPLINARIDADE E EXPERIÊNCIAS EM DOCUMENTAÇÃO E PRESERVAÇÃO DO PATRIMÔNIO RECENTE, 9., Anais... 2011. FEARNSIDE, Philip M. A Vulnerabilidade Da Floresta Amazônica Perante As Mudanças Climáticas. Oecologia Australis, v. 13, n. 04, p. 609–618, 2009. Disponível em: <http://www.oecologiaaustralis.org/ojs/index.php/oa/article/view/oeco.2009.130 4.05/110>. Disponível em: <http://www.oecologiaaustralis.org/ojs/index.php/oa/article/view/oeco.2009.130 4.05/110>. http://www.oecologiaaustralis.org/ojs/index.php/oa/article/view/oeco.2009.130 05/110>. GOMES, Aguinaldo Rodrigues. Escolas da várzea: à margem dos rios e da sociedade brasileira. p. 1–11, 1977. LIOTTO, P. F.; BARROS, L. M.; SILVA, T. F.; ALBUQUERQUE, V. S. Educitec Manaus v 04 n 08 p 179-194 nov 2018 Edição especial LIOTTO, P. F.; BARROS, L. M.; SILVA, T. F.; ALBUQUERQUE, V. S. Educitec, Manaus, v. 04, n. 08, p. 179-194, nov. 2018. Edição especial. OMES, Aguinaldo Rodrigues. Escolas da várzea: à margem dos rios e da ociedade brasileira. p. 1–11, 1977. GRZYBOWSKI, G. T. Conforto Térmico nas Escolas Públicas em Cuiabá- MT: estudo de caso. Cuiabá, 2004. 2004. Tese de Doutorado. Dissertação (Mestrado em Física e Meio Ambiente)-Escola de Engenharia, Universidade Federal de Mato Grosso, Cuiabá, 2004. HERMES, Fernando e OLIVEIRA, Dênio Ramam Carvalho e FERREIRA, Maurício de Pina. Elaboração de um projeto padrão de escola para as regiões ribeirinhas. Revista do centro de ciências exatas e tecnologia, p. 37–56, 2012. NETO, Tiago Sanule da Costa. Arquitetura Flutuante: Projetar uma habitação-tipo para um ambiente em transformação. 2015. 112 f. Universdade do Minho, 2015. OLIVEIRA, Mayane Guedes De e VILELA, Felipe Oliveira. A construção modular com utilização de painéis eps. Revista Episteme Transversalis, v. 08, 2015. PIEDADE, M T F e colab. Impactos ecológicos da inundação e seca na vegetação das áreas alagáveis amazônicas. Eventos climáticos extremos na 193 SCHERER, Elenise. Mosaico Terra-Água : a Vulnerabilidade Social Ribeirinha Na Amazônia – Brasil. In: CONGRESSO LUSO-AFRO-BRASILEIRO DE CIÊNCIAS SOCIAIS, 8., Anais... p. 1–18, 2004. SILVA, Pedro Daniel Moreira Da. Aplicação de Técnicas BIM à Construção Modular com Painéis Sandwich. 2012. SILVA, Pedro Daniel Moreira Da. Aplicação de Técnicas BIM à Construção Modular com Painéis Sandwich. 2012. ILVA, Pedro Daniel Moreira Da. Aplicação de Técnicas BIM à Construção odular com Painéis Sandwich. 2012. SOUSA, Luciana Cristina Romeo e SANTOS, Ricardo Bruno Nascimento e SOUSA, David Silva Pereira. Desenvolvimento e pobreza multidimensional na Amazônia Legal. Revista Espacios, v. 37, n. No 21, p. 1–19, 2016. SOUSA, Luciana Cristina Romeo e SANTOS, Ricardo Bruno Nascimento e SOUSA, David Silva Pereira. Desenvolvimento e pobreza multidimensional na Amazônia Legal. Revista Espacios, v. 37, n. No 21, p. 1–19, 2016. SOUSA, Luciana Cristina Romeo e SANTOS, Ricardo Bruno Nascimento e SOUSA, David Silva Pereira. Desenvolvimento e pobreza multidimensional na Amazônia Legal. Revista Espacios, v. 37, n. No 21, p. 1–19, 2016. VEIGA, Adriana Almeida. Práticas inovadoras de currículo na escola barco de madhyapara - Chalanbeel, Bangladesh. Revista Eletrônica de Ciências da Educação, v. 17, n. 1677–3098, p. 1–12, 2018. VEIGA, Adriana Almeida. Práticas inovadoras de currículo na escola barco de madhyapara - Chalanbeel, Bangladesh. Revista Eletrônica de Ciências da Educação, v. 17, n. 1677–3098, p. 1–12, 2018. VEIGA, Adriana Almeida. Práticas inovadoras de currículo na escola barco de madhyapara - Chalanbeel, Bangladesh. Revista Eletrônica de Ciências da Educação, v. 17, n. 1677–3098, p. 1–12, 2018. Submetido em 25/08/2018. Aceito em 25/10/2018. Submetido em 25/08/2018. Aceito em 25/10/2018. 194
https://openalex.org/W2032442722	https://angeo.copernicus.org/articles/23/1565/2005/angeo-23-1565-2005.pdf	English	null	Zonal wave numbers 1-5 in planetary waves from the TOMS total ozone at 65° S	Annales geophysicae	2,005	cc-by	6,362	Zonal wave numbers 1–5 in planetary waves from the TOMS total ozone at 65◦S A. Grytsai1, Z. Grytsai1, A. Evtushevsky1, G. Milinevsky1,2, and N. Leonov1,2 1National Taras Shevchenko University of Kyiv, Ukraine 2Ukrainian Antarctic Center, Kyiv, Ukraine Received: 5 November 2004 – Revised: 12 January 2005 – Accepted: 8 February 2005 – Published: 27 July 2005 Part of Special Issue “Atmospheric studies by optical methods” A. Grytsai1, Z. Grytsai1, A. Evtushevsky1, G. Milinevsky1,2, and N. Leonov1,2 1National Taras Shevchenko University of Kyiv, Ukraine 2Ukrainian Antarctic Center, Kyiv, Ukraine National Taras Shevchenko University of Kyiv, Ukraine 2Ukrainian Antarctic Center, Kyiv, Ukraine Received: 5 November 2004 – Revised: 12 January 2005 – Accepted: 8 February 2005 – Published: 27 July 2005 Received: 5 November 2004 – Revised: 12 January 2005 – Accepted: 8 February 2005 – Publishe Part of Special Issue “Atmospheric studies by optical methods” Part of Special Issue “Atmospheric studies by optical methods” Abstract. Planetary waves in the total ozone at the southern latitude of 65◦S are studied to obtain the main characteristics of the zonal wave numbers 1–5. The TOMS total ozone data were used to analyze the amplitude and periodicity variations of the ﬁve spectral components during August−December of 1979–2003. A presence of the shorter period of waves 1–3 in 1996 (7 days) in comparison with 2002 (8–12 days) is re- vealed which can be attributed to the distinction in conditions of typical and anomalously weak stratospheric polar vortex, probably, a strong and weak mean zonal wind. The inter- annual variations of the monthly and 5-month mean ampli- tudes of the zonal wave numbers 1–5 are described. Wave 1 has the largest amplitude in October (up to 139 DU in 2000) and increasing amplitude trend (15 DU/decade for October 1979–2003). The 5-month mean amplitudes averaged over 1979–2003 are 53.6, 29.9, 15.5, 10.5, and 7.8 DU for the wave number sequence 1, 2, 3, 4 and 5, respectively. For the stationary components the amplitudes are 38.3, 4.8, 1.8, 1.2, 0.7 DU, respectively. Thus, the stationary component of wave 1 and the traveling one of waves 2–5 are predominant. The tendencies in a long-term change in the wave number amplitude can be explained by taking into account the degree of wave deformation of the stratospheric polar vortex edge, net meridional displacements of the lower stratosphere air, and the difference between the total ozone loss and negative trends in the polar and mid-latitude regions. satellite data obtained from the Total Ozone Mapping Spec- trometer (TOMS). 1 Introduction Planetary waves in the stratosphere inﬂuence the meridional and zonal circulation and affect the spatial distribution and time variations of the total ozone. Since 1979 the highly dy- namic wave processes in total ozone have been studied by the Zonal wave numbers 1–5 in planetary waves from the TOMS total ozone at 65◦S It was found that during Antarctic winter- spring the zonal wave number 1 is the most important in for- mation of quasi-stationary ozone distribution (Wirth, 1993). High interannual variability of planetary waves in the strato- sphere of the Southern Hemisphere has been established (Hio and Hirota, 2002; Vargin, 2003; Hio and Yoden, 2004). The higher zonal wave numbers are observed predominantly as traveling structures in ozone and other stratospheric param- eters (Fishbein et al., 1993; Hio and Yoden, 2004). The pe- riods of ozone variations are registered in the range of 4–30 days (Fishbein et al., 1993; Vargin, 2003). Large variability of total ozone on these time scales is connected with the dynamical atmospheric processes. Ozone is concentrated in the lower stratosphere at altitudes of 10–20 km, where the photochemical lifetime of ozone is several weeks and its distribution is controlled by dy- namical inﬂuences (Salby, 1996). The origin of the total ozone ﬂuctuations was investigated in detail by Salby and Callaghan (1993). It was shown that a large component of total ozone variability is explained by a quasi-columnar mo- tion of air in the lower stratosphere. Planetary waves make the dominant contribution to this variability, because they in- duce large meridional excursions of air. The aim of this work is to analyze of the total ozone variations in the vortex edge region of the Southern Hemi- sphere as a consequence of the meridional displacement of the stratospheric air due to vortex edge deformation by the planetary waves. Statistical analysis of the mean monthly amplitude of the zonal wave numbers 1–5 at the latitude cir- cle of 65◦S is carried out using the TOMS total ozone data for August−December of 1979–2003. Keywords. General circulation – Middle atmosphere dy- namics – Waves and tides Annales Geophysicae Annales Geophysicae Annales Geophysicae, 23, 1565–1573, 2005 SRef-ID: 1432-0576/ag/2005-23-1565 © European Geosciences Union 2005 Correspondence to: G. Milinevsky (science@uac.gov.ua) 3 Zonal wave number 1–5 in 1996 and 2002 The variations of the daily zonal distributions of the total ozone in each of the wave numbers 1–5 during August- December were analyzed. The TOMS measurements for the southern winter at 65◦S are possible from late July; there- fore, August is the only winter month which is suited for the monthly data analysis. December was chosen as the last month of the indicated time interval, because the wave ac- tivity in total ozone, as it will be shown below, can be ob- served at the latitude 65◦S in early Antarctic summer. This agrees with the long-term tendency in the change in the polar stratospheric vortex duration. For example, during the last decades the date of the ozone hole disappearance, which is closely related to the breakdown of the polar vortex, shifted from November in the 1980s, to early December in the 1990s (Figs. 3–5 in WMO, 2003). location of the Ukrainian Antarctic station Akademik Ver- nadsky. The ﬁrst results concerning the variations of plan- etary waves in total ozone at this latitude were obtained in Grytsai et al. (2005). For 1979–2003 the time interval of August-December covering the Antarctic late winter, spring and early summer was considered. The years of 1994 and 1995 are excluded because of the gaps in the satellite ozone data. Expansion in series of the longitudinal distribution of the total ozone values f (λ) for the zonal wave number m is: f (λ) = ∞ X m=−∞ ameimλ, (1) (1) As examples, the years 1996 and 2002 are presented in Figs. 2a–e, and f–j, respectively. It is well known that during Antarctic spring 2002 the ﬁrst recorded major stratospheric warming and splitting of the polar vortex in late September was observed (Baldwin et al., 2003). The vortex develop- ment in 2002 was very different from the preceding years and the vortex of 1996 was chosen as a typical event. At the latitude 65◦S the maximum speed of the zonal mean zonal wind in October 1996 exceeded 50 m/s at 20 hPa level (Hio and Yoden, 2004), whereas in late August and early Septem- ber 2002 the wind was ∼20–30 m/s weaker than normal and where λ is the longitude and the coefﬁcient am is where λ is the longitude and the coefﬁcient am is am = 1 2π π Z −π f (λ) e−imλdλ. 2 Analysis method We analyze the longitudinal distribution of the total ozone along the 65.5◦S latitude circle based on the TOMS Ver- sion 7 data (TOMS, 2004). This latitude is the closest to the Correspondence to: G. Milinevsky (science@uac.gov.ua) Correspondence to: G. Milinevsky (science@uac.gov.ua) Correspondence to: G. Milinevsky (science@uac.gov.ua) A. Grytsai et al.: Planetary waves in total ozone 1566 Fig. 1. The distribution of the TOMS total ozone values along 65◦S, 15 October 1996 (a), the spectral components of the wave numbers 1–5 (b) and comparison of the observed and the restored distributions (c). 1–5 (b) is 6.9 DU or 2.5% in this example and is typically 7–11 DU that can be attributed to the wave numbers higher than 5. To describe the main properties of the zonal wave num- bers in each of the years quantitatively, the average ampli- tudes were calculated. First, the daily amplitude modules were averaged to obtain the amplitude Amgen, which charac- terizes the general wave event both in a stationary and trav- eling form: Am gen = 1 πN N X j=1 π Z −π fj (λ) e−imλdλ , (3) (3) where j is the day number, N is the total day quantity and λ is the longitude. The monthly and 5-month averages were obtained, which are presented in Sect. 4. Then the amplitude Amst of the wave number stationary part was found by averaging the daily longitudinal proﬁle f (λ): Am st = 1 πN N X j=1 π Z −π fj (λ) e−imλdλ . (4) (4) The second procedure removes the anti-phase values of total ozone and accumulates in-phase ones, giving the amplitude Amst and phase λst of the stationary wave component (see below in Sect. 5). Fig. 1. The distribution of the TOMS total ozone values along 65◦S, 15 October 1996 (a), the spectral components of the wave numbers 1–5 (b) and comparison of the observed and the restored distributions (c). 3 Zonal wave number 1–5 in 1996 and 2002 (2) (2) Figure 1 shows the example of the analysis made using Eqs. (1) and (2) to the TOMS distribution of the total ozone values at 65◦S. The standard deviation between the observed distribution (a) and restored one based on the wave numbers A. Grytsai et al.: Planetary waves in total ozone 1567 abruptly weakened from 17 September (Allen et al., 2003; Baldwin et al., 2003; the data for 60◦S). the wavelength is constant for each of the wave numbers at the ﬁxed latitude circle. Similarly, the longer periods of the wave numbers 1–3 in 2002 characterize the lower phase ve- locity. In general, this agrees with the conditions of a strong and weak polar vortex, i.e. stronger and weaker zonal wind in 1996 and 2002, respectively. In the same way, an increase in the wave 1 period during August-September 2002 can be concerned with an intense deceleration of the zonal wind, which took place beginning in the second half of August (Baldwin et al., 2003). It is seen from Fig. 2 that the individual wave numbers show the highest activity during September-November, al- though a strong timing is not observed. Wave 1 exhibits a quasi-stationary behaviour, which is observed in 1996 up to early December (Fig. 2a). The latest interval of a wave ac- tivity one can see in the middle of December 1996 from the wave number 3 plot (Fig. 2c). The eastward motion is seen as an increase in the east- ern longitude of the wave ridges and troughs in time, which becomes apparent in the slope of the color stripes in the time- longitude plots in Fig. 2. The intervals of relative phase sta- bility are sometimes observed. It is evident that each of the wave numbers can be observed in a standing form. A noticeable increase in the wave 1 period from 8 to 15 days during August-September (Fig. 3f) corresponds to the phase velocity decrease from 24 to 13 m/s (estimation was made for the length of wave 1 at the 65◦S latitude). The wave 2 and 3 velocities are equal to 8–9 and 5–6 m/s, respectively. The 7-day periods in 1996 give the higher phase velocities of 27, 14 and 9 m/s for waves 1–3, respectively. A time evolution of the spectral components in 2002 is very different (Figs. 2f–j; note the gap in the TOMS data during 2–12 August 2002). 3 Zonal wave number 1–5 in 1996 and 2002 Figure 3 presents the distinc- tions between 1996 and 2002 as a change in periodicity of the traveling components. Figure 3 shows the result of wavelet-transform applied to the time-longitude distributions presented in Fig. 2. We have selected the 150-day time series of the wave number amplitude at the ﬁxed longitude. The se- ries is an analogue of the single-point measurements which give the 5-month time variations of the individual wave num- ber. The periods shorter than 7 days do not appear in waves 1–3 at all (Figs. 3a–c, f–h). But the periods in the range of 4–7 days exist in wave numbers 4–5, both in 1996 and 2002 (Figs. 3d–e and i–j, respectively). In the event of 2002 wave 5 shows the 7-day oscillation during October- November (Fig. 3j), whereas the polar vortex breakdown oc- curred in late September. Obviously, the wave 4–5 activity can exist independently of the polar vortex strength and du- ration. At least this situation is observed in the two events of 1996 and 2002. As it is seen from Figs. 2b–e and g–j, the longitude can be selected arbitrarily for each of the wave numbers 2–5 be- cause of their regular zonal structure. Wave 1 has one lon- gitudinal cycle with a deviated phase of the stationary com- ponent. Since the time-longitude distribution in Figs. 2a,f gives the superposition of the traveling and quasi-stationary components of wave 1, a choice of longitude can correspond only to the intensity time change but not to the periodicity. We have selected the longitude of Vernadsky station 64◦W to analyze a periodicity in all of the wave numbers. It should be noted that the periods at the level of ∼5 days are absent in all analyzed years and the 7–20-day periods are predominant. As it is evident from spectral comparisons by Lawrence and Jarvis (2001) made for the Antarctic station Halley (76◦S, 26◦W) the 5-day periods in August are not observed and the 12–20 day periods are the most intense. This result is obtained for the 30-km level in the stratosphere for 1996. More detailed quantitative analysis is necessary to deter- mine the interrelation between the individual wave number periodicity changes and polar vortex parameters. In this pa- per we focus further on the amplitude variations of the zonal wave numbers 1–5. 3 Zonal wave number 1–5 in 1996 and 2002 Figure 3 presents the changes in periodicity under condi- tions of a strong and weak polar vortex of 1996 and 2002, respectively, for the same spectral components as in Fig. 2. The range of periods is truncated at 20 days. The main distinction between the two years is the presence of the quasi 7-day periods in wave 1–3 in 1996 (Figs. 3a- c), which is not observed in these wave numbers in 2002 (Figs. 3f–h). Note that the 7-day periods in the wave num- bers 1–3 during October 1996 have been considered by Hio and Yoden (2004) in more detail with respect to wave-wave interaction. Here we note the distinctive features in the char- acteristic periods of the spectral components. 4 Variability of the amplitude of wave numbers 1–5 The interannual variations of the monthly mean amplitude of waves 1–5 during 1979–2003 are shown in Figs. 4a–e. Each of the months from August to December is presented on sep- arate plots. On the right the interannual variations of the total ozone zonal mean at 65◦S are presented for the same month sequence (Figs. 4f-j), to compare the relative contribution of waves 1–5 to the total ozone variability. Some individual features are peculiar to the development of wave 1–5 activity during each of months. In comparison to 1996, the wave numbers 1–3 in 2002 have the longer 8–12-day periods (Figs. 3f–h). The tendency of a systematic increase in the lower value of the wave 1 pe- riod is observed. The period is changed from ∼7 days in August to 12–15 days in September and up to ∼20 days in early October. Wave number 1 has the largest amplitude and increasing trend in all of 5 months (Figs. 4a–e, black line). In August the largest anomalies in the wave 1 amplitude relative to the mean tendency are observed in 1988 and 2002 (wave am- The distinction in the wave 1–3 periods indicates different zonal phase velocity in 1996 and 2002. The shorter peri- ods in 1996 correspond to the higher zonal velocity, because A. Grytsai et al.: Planetary waves in total ozone 1568 Fig. 2. Time-longitude variations of the daily deviation from the zonal mean at 65◦S for the wave numbers 1–5 during August-December 1996 (a-e) and 2002 (f-j). Fig. 2. Time-longitude variations of the daily deviation from the zonal mean at 65◦S for the wave numbers 1–5 during August-December 1996 (a-e) and 2002 (f-j). Fig. 2. Time-longitude variations of the daily deviation from the zonal mean at 65◦S for the wave numbers 1–5 during August-December 1996 (a-e) and 2002 (f-j). Fig. 3. The periodicity variations of wave numbers 1–5 during August-December 1996 (a–e) and 2002 (f–j). Fig. 3. The periodicity variations of wave numbers 1–5 during August-December 1996 (a–e) and 2002 (f–j). Fig. 3. The periodicity variations of wave numbers 1–5 during August-December 1996 (a–e) and 2002 (f–j). Fig. 3. The periodicity variations of wave numbers 1–5 during August-December 1996 (a–e) and 2002 (f–j). plitude of about 69 and 93 DU, respectively, see Fig. 4a). 4 Variability of the amplitude of wave numbers 1–5 In September the wave 1 peak in 1988 (112 DU) is some- what higher than in 2002 (107 DU), and much higher relative to the mean tendency (about 60 and 30 DU, respectively). The quasi-biennial oscillations in the wave 1 amplitude are clearly seen in September. In October wave 1 has the largest amplitude with a peak value of 139 DU in 2000. to month is relatively small in the range of 20–40 DU, except for December, with an average level of about 15 DU. Wave 2 is the closest to wave 1 in amplitude in August. Wave number 3 has again a statistically signiﬁcant in- crease of amplitude (at the 95% conﬁdence level) in Septem- ber (Fig. 4b, blue line). During August-December its ampli- tude varies at the level of 10–20 DU. Wave number 2 shows some increasing trend in September only (Fig. 4b, red line). The change in amplitude from month Wave numbers 4 and 5 with the amplitudes of about 5– 15 DU (Figs. 4a–e, green and yellow lines, respectively) have A. Grytsai et al.: Planetary waves in total ozone A. Grytsai et al.: Planetary waves in total ozone 1569 A. Grytsai et al.: Planetary waves in total ozone 1569 Fig. 4. Interannual variations of the monthly mean amplitude of the wave numbers 1–5 for August-December (a–e) and zonal mean total ozone at 65◦S (f–j) for the same months. Colors are black (wave 1), red (wave 2), blue (wave 3), green (wave 4) and yellow (wave 5). Fig. 4. Interannual variations of the monthly mean amplitude of the wave numbers 1–5 for August-December (a–e) and zonal mean total ozone at 65◦S (f–j) for the same months. Colors are black (wave 1), red (wave 2), blue (wave 3), green (wave 4) and yellow (wave 5). a statistically signiﬁcant level of 4–6% relative to the zonal mean during August-October in the late 1990s −early 2000s, when the zonal mean values are the lowest (240–280 DU, Figs. 4f–h). In early 1980s these wave numbers have a rel- ative amplitude of 2–3%, which is close to the noise level, when taking into account an error in the TOMS total ozone data of 1–2% (WMO, 1999). (Fig. 4h), and in early 1990s in November and December (Figs. 4i–j). Obviously, this is an indication of a global change in ozone dynamics. Similar tendencies, which allow one to predict the future ozone layer recovery, are widely re- ported and discussed in the last years (WMO, 2003; Fioletov, 2004; Randel, 2004). It is interesting to compare the distinctions in the ampli- tude trends for wave numbers 1–3. In Fig. 5 the results for the three months are presented, namely for August (red), September (blue) and October (black). Increasing trends in the wave 1 amplitude (Fig. 5a) are 11, 14 and 15 DU/decade in August, September and October, respectively. A tendency in the interannual variations of the total ozone zonal mean at 65◦S should be noted separately. It is seen from Fig. 4 that the decreasing trend in August and Septem- ber is observed up to the early 2000s (Figs. 4f–g) and it ceases earlier and earlier: in the late 1990s in October A. Grytsai et al.: Planetary waves in total ozone 1570 able 1. The values of the monthly mean amplitude of the wave numbers 1–5 and zonal mean total ozone at 65◦S (in Dobson Units), veraged for 1979–2003. A. Grytsai et al.: Planetary waves in total ozone August September October November December Wave 1 30.9 62.7 94.3 61.5 20.6 Wave 2 22.1 38.8 42.4 30.5 13.2 Wave 3 11.4 16.7 21.3 18.3 9.5 Wave 4 9.4 10.2 13.6 11.9 7.1 Wave 5 7.6 7.8 9.0 8.9 5.5 Zonal mean 286.9 275.2 310.2 334.8 331.9 Number of values 23 24 23 24 22 ig. 5. Comparison of long-term trends in amplitude of the wave umbers 1–3 during August (red line), September (blue line) and ctober (black line). Fig. 6. Monthly mean amplitude of the wave numbers 1–5 averaged for the 1979–2003 period (a) and the interannual change of the 5- month mean amplitude (b). In Fig. 6a, the monthly mean amplitudes 1–5 averaged through 1979–2003 are shown for each of the wave num- bers. These values are also summarized in Table 1. The nearly symmetric change in the wave 1 amplitude value rela- tive to the peak value of 94 DU in October is seen. This value e values of the monthly mean amplitude of the wave numbers 1–5 and zonal mean total ozone at 65◦S (in 1979–2003. Table 1. The values of the monthly mean amplitude of the wave numbers 1–5 and zonal mean total ozone at 65◦S (in Dobson Units), averaged for 1979–2003. onthly mean amplitude of the wave numbers 1–5 and zonal mean total ozone at 65◦S (in Dobson Units), August September October November December Wave 1 30.9 62.7 94.3 61.5 20.6 Wave 2 22.1 38.8 42.4 30.5 13.2 Wave 3 11.4 16.7 21.3 18.3 9.5 Wave 4 9.4 10.2 13.6 11.9 7.1 Wave 5 7.6 7.8 9.0 8.9 5.5 Zonal mean 286.9 275.2 310.2 334.8 331.9 Number of values 23 24 23 24 22 Number of values 23 24 23 24 22 g. 5. Comparison of long-term trends in amplitude of the wave mbers 1–3 during August (red line), September (blue line) and tober (black line). Statistically signiﬁcant trends of about 8 and 5 DU/decade e observed in waves 2 and 3 respectively in September Fig. 6. Monthly mean amplitude of the wave numbers 1–5 averag for the 1979–2003 period (a) and the interannual change of the month mean amplitude (b). In Fig. 6a, the monthly mean amplitudes 1–5 averag through 1979–2003 are shown for each of the wave num bers. These values are also summarized in Table 1. 5 Stationary component October’s peak in the wave 1 amplitude (Fig. 6a) displays the nature of wave disturbance of total ozone in the vortex edge region. Deformation of the stratospheric polar vortex by the planetary wave 1 results in a meridional displacement of stratospheric air. Because the vortex edge prevents one from mixing the mid-latitude and polar air masses during Antarctic winter and spring (Lee at al., 2001), wave deformation results in a displacement of the midlatitudinal ozone rich air, which accumulates outside the polar vortex, toward the pole and low ozone polar air toward the equator. In particular, this process is seen well visually from the time sequences of the TOMS total ozone ﬁelds (TOMS, 2004b). From the averaging procedure Eq. (4) the 5-month mean lon- gitudinal distributions of total ozone are obtained. Figure 7 shows 23 curves (1979–2003) for each of the wave numbers 1–5 in order to present a degree of the interannual variabil- ity of the stationary component amplitude Amst and phase λst. It is shown that wave number 1 is inherent in the most stable longitudinal distribution of the total ozone (Fig. 7a). The wave numbers 2 and 4 also show some indications of the long-term phase stability (Figs. 7b and d). The stationary components of the wave numbers 3 and 5 are very change- able in amplitude and phase from year to year (Figs. 7c and e), but wave 4 exhibits enough stable longitudinal distribu- tion, although at the low amplitude level. The interannual variations of the 5-month average amplitude Amst are pre- sented separately in Fig. 8. An effect of such meridional air displacements in the op- posite directions is that the low and high ozone air which originated from different latitudinal bands appears at the lati- tude circle of 65◦S. This forms the waved distribution of total ozone along the circle of 65◦S, which displays the charac- ter and degree of vortex edge deformation: the more merid- ional displacement in wave structure, the more contrast in total ozone at the ﬁxed latitude circle in this region. To- tal wave numbers contribute to a deformation of the vortex edge, but wave 1 is predominant. As it was shown in Grytsai et al. A. Grytsai et al.: Planetary waves in total ozone T nearly symmetric change in the wave 1 amplitude value rel tive to the peak value of 94 DU in October is seen. This val is three times more than amplitude value in August. No that wave 1 has almost an equal amplitude in September an Fig. 5. Comparison of long-term trends in amplitude of the wave numbers 1–3 during August (red line), September (blue line) and October (black line). Fig. 6. Monthly mean amplitude of the wave numbers 1–5 averaged for the 1979–2003 period (a) and the interannual change of the 5- month mean amplitude (b). Fi 6 M hl li d f h b 1 5 Fig. 6. Monthly mean amplitude of the wave numbers 1–5 averaged for the 1979–2003 period (a) and the interannual change of the 5- month mean amplitude (b). In Fig. 6a, the monthly mean amplitudes 1–5 averaged through 1979–2003 are shown for each of the wave num- bers. These values are also summarized in Table 1. The nearly symmetric change in the wave 1 amplitude value rela- tive to the peak value of 94 DU in October is seen. This value is three times more than amplitude value in August. Note that wave 1 has almost an equal amplitude in September and November. Even in December the wave 1 activity exists at the mean amplitude level of 20 DU. Higher wave numbers do not exhibit the sharp change in the monthly means from month to month. Fig. 5. Comparison of long-term trends in amplitude of the wave numbers 1–3 during August (red line), September (blue line) and October (black line). Statistically signiﬁcant trends of about 8 and 5 DU/decade are observed in waves 2 and 3, respectively, in September only (Figs. 5b and c, blue line). So, a distinguishing feature of September is the clear tendency of an increasing long- term trend of waves 2–3 amplitude, which is not observed for these wave numbers in August and October. Another summarizing result is given in Fig. 6b, where the long-term change in the 5-month mean amplitude during A. Grytsai et al.: Planetary waves in total ozone 1571 Callaghan, 1993), caused by the planetary wave deformation of the polar vortex, obviously gives a main contribution to the observed variations of the zonal wave number amplitude. 1979–2003 is shown. 5 Stationary component (2005), the long-term negative trends in the values of the quasi-stationary minimum and maximum in total ozone at 65◦S (by zonal distribution of the total wave numbers) are in agreement with the mean total ozone trends observed in the adjusted latitude bands. Finally, the average amplitude Amst and Amgen for the 1979–2003 period can be compared (Fig. 9). Each of the wave numbers 1–5 shows the presence of the stationary com- ponent, and it is predominant in the wave 1 behavior. For the waves 2 to 5 the traveling form is predominant. The 23-year average zonal distribution gives the amplitude sequence of 38.3, 4.8, 1.8, 1.2 and 0.7 DU for the stationary components of the wave numbers 1 to 5, respectively. It is seen that the wave 2 contribution to the long-term sta- tionary distribution of total ozone is eight times less than the contribution of wave 1. This is close to the value of ∼1:6 ob- tained for the Southern Hemisphere October from the TOMS data of 1979–1986 in Wirth (1993). The lowest total ozone and largest negative trend over Antarctica that are observed in October (WMO, 2003) can explain October’s maximum of the wave 1 amplitude (Fig. 6a, and Table 1), as well as its highest increasing trends indicated above. The vortex edge deformation and air dis- placement across the latitudinal circle, which arises from the zonal wave numbers 2 and 3, has the relatively smaller spa- tial scale. The stratospheric air mass, which forms the zonal distribution of the total ozone by waves 2 and 3, is limited by the narrower latitudinal band in the vortex edge region. Then the meridional difference in the mean ozone loss and negative trend is smaller than in the wave 1 case. It is possi- ble that this circumstance has an effect in the relatively lower amplitude and trend for waves 2–3 (Figs. 4a–e). The largest trend in the amplitude for waves 2–3 is observed in Septem- ber (Figs. 5b–c), which can be concerned with a minimum value of the zonal mean at 65◦S just in September (see Ta- ble 1). The amplitudes Amst of waves 2–5 at the level of 1–5 DU do not noticeably inﬂuence the stationary ozone distribution on the many years scale, although they can increase tem- porarily on the monthly or weekly scale. A. Grytsai et al.: Planetary waves in total ozone The most evident increase in the 5- month mean amplitudes is observed for the wave numbers 1 and 2, but this tendency is not kept in the last years. The am- plitudes averaged over 1979–2003 are 53.6, 29.9, 15.5, 10.5, and 7.8 DU for the wave number sequence 1, 2, 3, 4 and 5, respectively. 6 Conclusion Planetary waves in the total ozone at the southern latitude of 65◦S have been analyzed in this work. The TOMS data spectral analysis was made to study the behaviour of zonal wave numbers 1-5 during August-December of 1979–2003. The ﬁrst task was to establish whether the wave number pe- riodicity plays a part in the stratospheric polar vortex devel- opment. The activity of wave numbers 1–5 was considered under the conditions of the strong (1996) and weak (2002) vortex. The ﬁrst three spectral components exhibit a distinc- tion in periodicity, with the lowest period of about 7 days in 1996, which is absent in 2002. Under conditions of the vor- tex 2002 the periods of 8–12 days predominate. The presence of a shorter period in 1996 in comparison with 2002 probably corresponds to the distinction in the zonal wind velocity, i.e. to the typically higher (lower) velocity in the strong (weak) Then the long-term change in the wave number ampli- tude in the southern vortex edge region is substantially deter- mined by the ozone loss in the polar stratosphere and increas- ing contrast between the total ozone at the polar and equa- torial side of the vortex edge. The quasi-columnar merid- ional displacement of air in the lower stratosphere (Salby and A. Grytsai et al.: Planetary waves in total ozone 1572 ig. 7. Interannual changes of the stationary component of the wa umbers 1–5 (a–e). The August-December averaging of the da eviation from the zonal mean at 65◦S was used. The period 979 2003 i t d Fig. 8. Interannual variations of the stationary component ampli- tude Amst of the wave numbers 1–5 (the colors are the same as in Fig. 6). Fig. 8. Interannual variations of the stationary component ampli- tude Amst of the wave numbers 1–5 (the colors are the same as in Fig. 6). Fig. 9. Relationship between the wave number amplitude Amst (sta- tionary) and Amgen (general), averaged for the 1979–2003 period. Fig. 9. Relationship between the wave number amplitude Amst (sta- tionary) and Amgen (general), averaged for the 1979–2003 period. vortex. A noticeable increase in the wave 1 period from 8 to 15 days during August-September (Fig. 3f) can also indicate a deceleration of the zonal wind in the second half of August (Baldwin et al., 2003). A. Grytsai et al.: Planetary waves in total ozone 5. Wave numbers 4 and 5 show the monthly mean ampli- tudes of about 5–15 DU, which can be estimated as sta- tistically signiﬁcant (4–6% relative to the zonal mean) during August-October in the late 1990s–early 2000s; in other time intervals they are close to the noise level due to the higher level of zonal mean total ozone at 65◦S; Fioletov, V. E.: Total ozone variations over midlatitudes and on the global scale, OZONE, Volume I, XX Quadrennial Ozone Sym- posium, Proceedings, 25–26, 2004. Fishbein, E. F., Elson, L. S., Froidevaux, L., Manney, G. L., Read, W. G., Waters, J. W., and Zurek, R. W.: MLS observations of stratospheric waves in temperature and O3 during 1992 Southern winter, Geophys Res. Lett., 20, 1255–1258, 1993. Grytsai, A. V., Evtushevsky, A. M., and Milinevsky, G. P.: Interan- nual variations of planetary waves in ozone layer at 65◦S, Int. J. Rem. Sens., issue on XX Quadrennial Ozone Symposium, in press, 2005. 6. The monthly mean amplitudes of wave 1 in August- December, averaged on the 25-year time scale, is nearly symmetric relative to the peak value of 94 DU in Octo- ber. In December the wave 1 activity exists at the am- plitude level of 20 DU. The higher wave numbers do not show the sharp change in the monthly means from month to month. Hio, Y. and Hirota, I.: Interannual variations of planetary waves in the Southern Hemisphere stratosphere, J. Met. Soc. Jap., 80(4B), 1013–1027, 2002. Hio, Y. and Yoden, S.: Quasi-periodic variations of the polar vor- tex in the Southern Hemisphere due to wave-wave interaction, J. Atmos. Sci., 61, 2510–2527, 2004. The long-term change in the 5-month mean amplitude shows the most evident increase in amplitude for the wave numbers 1 and 2, but this tendency is not kept in the last years. The 5-month mean amplitudes averaged over 1979– 2003 are 53.6, 29.9, 15.5, 10.5, and 7.8 DU for the wave number sequence 1, 2, 3, 4 and 5, respectively. For the sta- tionary components the amplitudes are 38.3, 4.8, 1.8, 1.2, 0.7 DU, respectively. Thus, the wave 1 stationary component and the traveling component of waves 2–5 are predominant. A. Grytsai et al.: Planetary waves in total ozone The tendencies for a long-term change in the wave number amplitude can be explained by taking into account the degree of wave deformation of the polar vortex edge, net meridional displacements of the lower stratosphere air, and the contrast between total ozone on each side of the vortex edge (i.e. in polar and middle latitudes), which is determined by the dif- ference in the ozone loss and negative trends in these regions. Lawrence, A. R. and Jarvis, M. J.: Initial comparisons of plane- tary waves in the stratosphere, mesosphere and ionosphere over Antarctica, Geophys. Res. Lett., 28, 203–206, 2001. Lee, A. M., Roscoe, H. K., Jones, A. E., Haynes, P. H., Shuck- burgh, E. F., Morrey, M. W., and Pumphrey, H. C.: The impact of the mixing properties within the Antarctic stratospheric vortex on ozone loss in spring, J. Geophys. Res., 106(D3), 3203–3211, 2001. Randel, W. J.: Challenges for understanding global ozone vari- ability, OZONE, Volume I, XX Quadrennial Ozone Symposium, Proceedings, 17–18, 2004. Salby, M. L.: Fundamentals of Atmospheric Physics, (Eds.) Dmowska, R. and Holton, R., Academic Press, 627, 1996. Salby, M. L. and Callaghan, P. F.: Fluctuations of total ozone and their relationship to stratospheric air motions, J. Geophys. Res., 98(D2), 2715–2727, 1993. TOMS ozone data, http://toms.gsfc.nasa.gov/ftpdata.html, 2004. Acknowledgements. We thank two anonymous reviewers for their constructive comments and helpful suggestions. This work was partly supported by grants FFD F7/362-2001 and STCU Gr-50(J). TOMS ozone image, http://toms.gsfc.nasa.gov/ftpimage.html, 2004b. Vargin, P. N.: Analysis of an eastward-propagating planetary waves from satellite data on the total ozone content, Izv. RAS. Phy. Atm. Ocean, (in Russian), 39, 327–334, 2003. Topical Editor U.-P. Hoppe thanks two referees for their help in evaluating this paper. Wirth, V.: Quasi-stationary planetary waves in total ozone and their correlation with lower stratospheric temperature, J. Geophys. Res., 98(D5), 8873–8882, 1993. 6 Conclusion The two last spectral components are observed in the range of periods 5–20 days, both in 1996 and 2002, i.e. regardless of vortex strength. The second task of this work was to describe the monthly and interannual variations of the amplitude of zonal wave numbers 1–5. On a monthly scale the spectral components’ behaviour has individual features: 1. During August-December wave 1 has the largest ampli- tude (up to 139 DU in October 2000) and an increasing trend (up to 15 DU/decade for October 1979–2003); 2. The largest anomalies in the wave 1 amplitude relative to the mean tendency is observed in 1988 and 2002 in August (Fig. 4a); 3. The quasi-biennial oscillations in wave 1 amplitude are clearly seen in September only (Fig. 4b); Fig. 7. Interannual changes of the stationary component of the wave numbers 1–5 (a–e). The August-December averaging of the daily deviation from the zonal mean at 65◦S was used. The period of 1979–2003 is presented. 4. In waves 2 and 3 the statistically signiﬁcant trends of about 8 and 5 DU/decade, respectively, are observed in September only (Figs. 5b and c); 1573 A. Grytsai et al.: Planetary waves in total ozone References WMO: Scientiﬁc Assessment of Ozone Depletion: 1998, World Meteorological Organization, Report No. 44, 1999. Allen, D. R., Bevilacqua, R. M., Nedoluha, G. E., Randall, C. E., and Manney, G. L.: Unusual stratospheric transport and mixing during the 2002 Antarctic winter, Geophys. Res. Lett., 30, 1599, doi:10.1029/2003GL017117, 2003. Meteorological Organization, Report No. 44, 1999. WMO: Scientiﬁc Assessment of Ozone Depletion: 2002, World Meteorological Organization, Report No. 47, 2003. Meteorological Organization, Report No. 47, 2003. Baldwin, M., Hirooka, T., O’Neill, A., and Yoden, S.: Major strato- spheric warming in the Southern Hemisphere in 2002: dynamical aspects of the ozone hole split, SPARC Newsletter, 20, 24–26, 2003.
https://openalex.org/W3100909419	https://www.research-collection.ethz.ch/bitstream/20.500.11850/114054/2/Eden-Sfondrini2016_Article_Three-pointFunctionsInN4Mathca.pdf	English	null	Three-point functions in N = 4 $$ \mathcal{N}=4 $$ SYM: the hexagon proposal at three loops	The Journal of high energy physics/The journal of high energy physics	2,016	cc-by	9,170	ETH Library ETH Library Open Access, c⃝The Authors. Article funded by SCOAP3. Journal Article g y p Journal of High Energy Physics 2016(2), https://doi.org/10.1007/JHEP02(2016)165 This page was generated automatically upon download from the ETH Zurich Research Collection. For more information, please consult the Terms of use. This page was generated automatically upon download from the ETH Zurich Research Collection. For more information, please consult the Terms of use. Published for SISSA by Springer Received: December 21, 2015 Accepted: February 3, 2016 Published: February 24, 2016 Received: December 21, 2015 Accepted: February 3, 2016 Published: February 24, 2016 1See refs. [6–10] for reviews and a list of references. Three-point functions in N = 4 SYM: the hexagon proposal at three loops JHEP02(2016)165 Burkhard Edena and Alessandro Sfondrinib aInstitut f¨ur Mathematik & Institut f¨ur Physik, Humboldt-Universit¨at zu Berlin, Zum großen Windkanal 6, D-12489 Berlin, Germany bInstitut f¨ur Theoretische Physik, ETH Z¨urich, Wolfgang-Pauli-Str. 27, CH-8093 Z¨urich, Switzerland E-mail: eden@math.hu-berlin.de, sfondria@itp.phys.ethz.ch Abstract: Basso, Komatsu and Vieira recently proposed an all-loop framework for the computation of three-point functions of single-trace operators of N = 4 super-Yang-Mills, the “hexagon program”. This proposal results in several remarkable predictions, including the three-point function of two protected operators with an unprotected one in the SU(2) and SL(2) sectors. Such predictions consist of an “asymptotic” part — similar in spirit to the asymptotic Bethe Ansatz of Beisert and Staudacher for two-point functions — as well as additional ﬁnite-size “wrapping” L¨uscher-like corrections. The focus of this paper is on such wrapping corrections, which we compute at three-loops in the SL(2) sector. The resulting structure constants perfectly match the ones obtained in the literature from four-point correlators of protected operators. Keywords: AdS-CFT Correspondence, Integrable Field Theories, Supersymmetric gauge theory ArXiv ePrint: 1510.01242 ArXiv ePrint: 1510.01242 ArXiv ePrint: 1510.01242 Open Access, c⃝The Authors. Article funded by SCOAP3. doi:10.1007/JHEP02(2016)165 doi:10.1007/JHEP02(2016)165 Contents 1 Introduction 1 2 The hexagon proposal 2 2.1 Asymptotic three-point function 3 2.2 Wrapping eﬀects 4 3 Three-point functions at three loops 5 3.1 Wrapping in the opposite channel 6 3.2 Wrapping in the adjacent channels 7 4 Results, conclusions and outlook 8 A Telescoping the transfer matrix 10 B Evaluation of the opposite-channel wrapping 10 B.1 Corrections to the measure 11 B.2 Corrections to the transfer matrix 11 B.3 Measure and residues from the ﬁrst order mirror dressing phase 12 C Evaluation of the adjacent-channel wrapping 13 Contents 1 Introduction 1 2 The hexagon proposal 2 2.1 Asymptotic three-point function 3 2.2 Wrapping eﬀects 4 3 Three-point functions at three loops 5 3.1 Wrapping in the opposite channel 6 3.2 Wrapping in the adjacent channels 7 4 Results, conclusions and outlook 8 A Telescoping the transfer matrix 10 B Evaluation of the opposite-channel wrapping 10 B.1 Corrections to the measure 11 B.2 Corrections to the transfer matrix 11 B.3 Measure and residues from the ﬁrst order mirror dressing phase 12 C Evaluation of the adjacent-channel wrapping 13 1 2 3 4 5 6 7 8 10 10 11 11 12 13 JHEP02(2016)165 C Evaluation of the adjacent-channel wrapping 1 Introduction A remarkable advance in the study of holographic [1], or gauge/string correspondence [2–4] was the discovery of integrability in the planar limit [5] of the correspondence: the duality’s dynamics is severely constrained by inﬁnitely many hidden symmetries. These powerful symmetries can be used to compute the energies of string states, or equivalently the two- point functions of operators of the dual CFT. This was ﬁrstly done for the most supersym- metric case of AdS5/CFT4, and more recently for other less supersymmetric dual pairs.1 The missing ingredient to fully describe generic (non-protected) states in the planar limit through integrability is to exploit symmetries to compute three-point functions. A great deal of eﬀort has been devoted to this problem in AdS5/CFT4, both from the point of view of string theory and of gauge theory [11–39]. Recently, a crucial development was the proposal from Basso, Komatsu and Vieira (BKV) of an all-loop framework for the computation of three-point functions using in- tegrability: the hexagon program [34]. This framework automatically incorporates the weakly-coupled “tailoring” procedure [16–18, 21], but it is “all-loop” in nature. It is in fact – 1 – a generalisation of the integrable bootstrap to three-point functions, with a new fundamen- tal object — the hexagon amplitude — playing the role of the scattering matrix. This can be quite easily used to construct the asymptotic part of the structure constant, in the sense of the asymptotic Bethe Ansatz of Beisert and Staudacher for two-point functions [40]. For short operators, corrections due to wrapping eﬀects similar to those familiar from the spectral problem [41] should be added. BKV propose an explicit recipe for doing so in a manner reminiscent of L¨uscher corrections [42, 43]. In ref. [34] BKV put their proposal to several tests, including the direct comparison of certain structure constants against known weak-coupling [44, 45] and strong-coupling [31] results. At weak-coupling, BKV explicitly compute the two-loop, three-point function of one non-BPS operator with two BPS ones, in the SU(2) and SL(2) sectors. This is matched to the ﬁeld theory results, which have been independently obtained from computing the four-point correlators of BPS operators [44–51]. Since the operators appearing in the three- point function are very short, this two-loop calculation already probes the ﬁrst “wrapping” correction to the BKV asymptotic formula — the ones coming from the edge of the hexagon opposite to the non-BPS state, see also ﬁgure 1 below. 1 Introduction JHEP02(2016)165 The aim of this paper is to apply the hexagon approach to compute three-loop three- point functions in the SL(2) sector. This calculation will for the ﬁrst time simultaneously probe the wrapping corrections on all the hexagon edges that have a “mirror” kinemat- ics. What is more, a prediction for the structure constant of these operator is known in the literature [45]. This yields a further quite non-trivial (and successful!) check of the hexagon proposal. This paper is structured as follows. In section 2 we brieﬂy summarise the BKV pro- posal, working out explicitly some formulae which we will need later — namely, the ones for the wrapping corrections in the two “adjacent” channels. In section 3 we specialise these formulae to the aforementioned three-loop computation and discuss how to evaluate them. In section 4 we present our results and conclusions. We relegate the technical details concerning the evaluation of the wrapping corrections to the appendices. Note added. Shortly after the submission of this pre-print we became aware of an up- coming work by Basso, Goncalves, Komatsu and Vieira where, among other things, these three-loop structure constants are computed using the hexagon approach and successfully matched to gauge theory [52]. 2 The hexagon proposal Let us brieﬂy review the BKV proposal [34]. One starts by cutting a three-string interaction (which has the topology of a pair of pants at leading order) into two pieces “along the pants’ seams”. This results into two patches, each having six distinguished edges — three corresponding to the cuts, and three to half of a “cuﬀ” each, see ﬁgure 1. These hexagons are the central objects of the proposal. The hexagon can be decorated with six sets of particles. Along the edges corresponding to “cuﬀs”, we will have closed-string excitations. Along the edges which will be glued back – 2 – u1 u2 v u3 v Figure 1. A stringy interaction with the topology of a pair of pants can be cut into two hexagons by cutting along the pants’ seams. In the ﬁgure the dashed edges (seams) of the two hexagons are glued, while the solid lines represent the cuﬀs of the pants. Closed-string excitations live on the “cuﬀs”: here we depicted excitations with rapidities u1, u2, u3 on the upper cuﬀ. Mirror excitations live on the dashed edges, and are necessary to properly account for wrapping processes. Here we depict one mirror excitation v on the mirror edge opposite to the excited string edge. Generically, the rapidities of each closed string will be the Bethe roots of the corresponding state, while one should integrate over all possible mirror rapidities. u1 u2 v u3 v Figure 1. A stringy interaction with the topology of a pair of pants can be cut into two hexagons by cutting along the pants’ seams. In the ﬁgure the dashed edges (seams) of the two hexagons are glued, while the solid lines represent the cuﬀs of the pants. Closed-string excitations live on the “cuﬀs”: here we depicted excitations with rapidities u1, u2, u3 on the upper cuﬀ. Mirror excitations live on the dashed edges, and are necessary to properly account for wrapping processes. Here we depict one mirror excitation v on the mirror edge opposite to the excited string edge. Generically, the rapidities of each closed string will be the Bethe roots of the corresponding state, while one should integrate over all possible mirror rapidities. JHEP02(2016)165 to yield the three-point functions, we will have excitations in the “mirror” kinematics, similarly to what happens when computing ﬁnite-size corrections to the energy of string states. 2 The hexagon proposal Given a three-point function by specifying the physical closed-string excitations at each of its cuﬀs, we can compute the relative structure constant by summing hexagon amplitudes over all possible ways of distributing the physical excitations over the two pieces of the cuﬀs, as well as summing over all possible mirror states and integrating over the mirror rapidities. Cleverly using the SU(2\|2) super-symmetry of the hexagon, as well the crossing trans- formation and imposing scattering factorisation `a la Zamolodchikov [53], the hexagon am- plitude h was ﬁxed exactly in [34], at least up to a scalar factor h(x, y). This is constrained by crossing symmetry to satisfy h(1/x, y) h(x, y) = c(x, y) , c(x, y) = x−−y− x−−y+ 1 −1/x+y− 1 −1/x+y+ , (2.1) (2.1) where we crossed the Zhukovski variables as x± →1/x±, corresponding to crossing the rapidity u →u2γ [34]. The last ingredient of the BKV proposal is then to set h(x, y) = x−−y− x−−y+ 1 −1/x−y+ 1 −1/x+y+ 1 σ(x, y) , (2.2) (2.2) where σ is the dressing factor of Beisert, Eden and Staudacher [54]. where σ is the dressing factor of Beisert, Eden and Staudacher [54]. 2.1 Asymptotic three-point function Using these ingredients, BKV predict the asymptotic part of the three-point function of two protected and one non-protected operator in the SL(2) sector to be C•◦◦ 123 C◦◦◦ 123 2 = QS k=1 µ(uk) det∂ujφk Q j<k Sjk   X α∪¯α={u} A(α,¯α)   2 . (2.3) (2.3) – 3 – – 3 – Here Sjk is the SL(2) diagonal scattering element of the S matrix by Beisert [55], µ is a measure deﬁned by the residue of the pole in the transition hD\|D of an excitation from one physical edge to another [34], Here Sjk is the SL(2) diagonal scattering element of the S matrix by Beisert [55], µ is a measure deﬁned by the residue of the pole in the transition hD\|D of an excitation from one physical edge to another [34], µ(u) = i resv=u hD\|D(u\|v) −1 , (2.4) (2.4) and the determinant is the Gaudin norm deﬁned in terms of φj which satisﬁes and the determinant is the Gaudin norm deﬁned in terms of φj which satisﬁes eiφj = eipjL Y k̸=j Sjk . (2.5) (2.5) JHEP02(2016)165 Finally, we should sum over the partitions α and ¯α the expression A(α,¯α) = (−1)\|¯α\| Y j<k j,k∈α∪¯α hjk Y k∈¯α eipkℓY j∈α k∈¯α 1 hjk . (2.6) (2.6) Here ℓis the separation between the non-protected operator and the others. Denoting the length of the non-protected operator as L = L1, we have ℓ= ℓ12 = ℓ31 with ℓij = 1 2(Li + Lj −Lk), all indices being distinct. 2.2 Wrapping eﬀects The contributions of the two adjacent channels (the two lateral dashed edges in ﬁgure 1)can be easily found and are given by JHEP02(2016)165 intγ a(u, {ui}) = Aα,¯α(−1)aTa(u−γ) Y j∈α ha(uγ, uj) ca(u−γ, uj) Y j∈¯α ha(uγ, uj) ca(u+γ, uj), (2.12) (2.12) and and int5γ a (u, {ui}) = Aα,¯α(−1)aTa(u−γ) Y j∈α ha(uγ, uj) ca(u+γ, uj) Y j∈¯α ha(uγ, uj) ca(u−γ, uj), (2.13) (2.13) where ca(u, v) can be found from eq. (2.1) by fusion. There are two important diﬀerences between eq. (2.11) and (2.12)–(2.13). Firstly, the contributions of the adjacent channels are sub-leading since Ta(uγ) = O(1) while Ta(u−γ) = O(g2); for this reason, (2.12)–(2.13) did not contribute in the evaluation of the two-loop SL(2) three-point functions in ref. [34]. Secondly, in (2.12)–(2.13) the sum over partitions and the integration over u do not factor, making the evaluation of these contributions somewhat more involved. It is worth noting that at higher-loop level there are additional wrapping eﬀects con- tributing, namely the ones described by ordinary L¨uscher corrections for each of the single- trace operators in the three-point function. Of course in this set-up such corrections only appear for the non-protected operator. In the SL(2) sector, as it is well known [59–61], such wrapping corrections ﬁrst appear at four loops, and therefore go beyond the scope of this work. 2.2 Wrapping eﬀects So far we have not accounted for the presence of mirror particles on the edges of the hexagon to be glued. This can be done in a L¨uscher-like approach, where the leading ﬁnite-volume contribution is given by allowing at most a single mirror particle per edge. Then one has to correct the asymptotic expression by A(α,¯α) →A(α,¯α) + δA(α,¯α) 12 + δA(α,¯α) 23 + δA(α,¯α) 31 . (2.7) (2.7) Each of the δA is related to one of the mirror channels. They are given by δA(α,¯α) jk = X a>0 Z du 2π µγ a(u) 1 x[+a]x[−a] ℓjkint(2j−1)γ a (u\|{ui}). (2.8) (2.8) Here and in what follows a denotes the bound-state number, and x[±a] are the bound-state Zhukovski variables, which depend on the shifted rapidities u ± a 2i. Following ref. [34], we indicate mirror-like shifts of the rapidity, corresponding to shifting an excitation to a neighbouring edge of the hexagon n times, by nγ. The mirror measure is the same in all channels and reads µγ a(u) = a (x[+a]x[−a])2 g2(x[+a]x[−a] −1)2(x[+a]2 −1)(x[−a]2 −1). (2.9) (2.9) The integrand depends on which mirror channel we consider. Schematically intnγ a (u, {ui}) = (−1)\|¯α\| Y j∈¯α eipjℓ Y k>j j∈¯α,k∈α Sjk X Xa (−1)fXahXaD...D(unγ, α)hD...D ¯ Xa(¯α, u−nγ), (2.10) intnγ a (u, {ui}) = (−1)\|¯α\| Y j∈¯α eipjℓ Y k>j j∈¯α,k∈α Sjk (2.10) (2.10) X Xa (−1)fXahXaD...D(unγ, α)hD...D ¯ Xa(¯α, u−nγ), ( ) – 4 – which involves the scattering of the mirror (bound-state) particles Xa with all the physical particles in the α partitions, and similarly for their conjugates ¯ Xa with the ¯α partition. The channel opposite to the non-protected state (the bottom edge of the hexagon in ﬁgure 1), corresponding to a shift of 3γ, has been computed in ref. [34] int3γ a (u, {ui}) = Aα,¯α (−1)aTa(uγ) Q j∈α∪¯α ha(uγ, uj), (2.11) (2.11) where Ta is the transfer matrix in the anti-symmetric representation (see appendix H in ref. [34]), and ha is bound-state scalar factor, which can be found by fusion [56–58]. 3.1 Wrapping in the opposite channel The computation of the opposite (3γ) channel follows the one of ref. [34], but must include next-to-leading order corrections. Firstly, it is useful to massage a bit the transfer matrix Ta(uγ) from ref. [34] — see appendix A for details. Then, we ﬁnd it convenient to strip a denominator out of Ta(uγ), introducing eTa(uγ) = (−1)a Ta(uγ) Qs j=1 dena(uγ, uj) , dena(uγ, uj) = (x[−a] −x+ j )(1 −1/x[−a]x− j ). (3.1) (3.1) The reason for doing so is that, when plugging Ta into (2.11), this denominator will simplify against the rational part of the mirror-string dressing factor as we will see. We need to expand eTa up to O(g2). On top of the tree-level term computed in [34], we have a correction coming from the one-loop rapidities and one coming from the explicit g-dependence of eTa, JHEP02(2016)165 eTa(uγ) = 1 + g2 s X j=1 u(1) i ∂ui eT (0) a (uγ) + g2E(1) eT (1) a (uγ) + O(g4). (3.2) (3.2) In terms of the Baxter polynomials Q(u) = Qs j=1(u −uj) we then have2 s of the Baxter polynomials Q(u) = Q j=1(u −uj) we then have eT (0) a (uγ) = Q(u[a+1]) + Q(u[−a−1]) −Q(u[a−1]) −Q(u[−a+1]), eT (1) a (uγ) = Q(u[−a−1]) u[−a] −Q(u[a+1]) u[+a] + a−1 X k=1 Q(u[2k−1−a]) −Q(u[2k+1−a]) u[2k−a] . (3.3) (3.3) Note that the one-loop transfer matrix is multiplied by the one-loop energy E(1) = s X j=1 −i v2 j + 1/4 = −2iS1(s) , (3.4) (3.4) where the last expression is the well-known representation of the one-loop, spin-s energy as a harmonic sum [59]. In (3.3) one might worry that eT (1) a seems to have a pole on the real-line when 2k = a. However, it is easy to see that the summand is regular at u = 0 when one imposes the zero-momentum or level-matching condition Q(+i/2) = Q(−i/2). As we mentioned, the denominator dena simpliﬁes drastically against the rational part of the dressing factor ha. In fact, for the purpose of our calculation we can write dena(uγ, v)ha(uγ, v) ≈1 x+ v 1 + ig2E(1)Ψ(u) , Ψa(u) = 4γ + ψ 1 + iu[−a] + ψ 1 −iu[−a] + ψ 1 + iu[+a] + ψ 1 −iu[+a] , (3.5) (3.5) where the digamma functions ψ come from the expansion of the mirror-string dressing factor [58, 62]. 3 Three-point functions at three loops Using these ingredients, we can now compute three-loop three-point functions involving twist-two, spin-s operators. Speciﬁcally, we pick one twist-two (non-protected) operator O1 = tr(DsZ2), and two protected operators O2 = tr( ¯ZY ) and O3 = tr( ¯Z ¯Y ). This will probe the wrapping contributions to the adjacent channels at γ, 5γ, as well as the next- to-leading-order contribution of the opposite (3γ) channel. Of course the ﬁnal result will also depend on the three-loop expansion of the asymptotic term described in section 2.1, as well as on the loop corrections to the rapidities from the Bethe Ansatz. Accounting for these corrections is straightforward, and we will therefore focus our attention on the wrapping eﬀects. – 5 – 2The Zhukovski variables x± are particularly convenient when discussing crossing transformations, but for our perturbative computations it is convenient to expand them as customary x± → √ 2/g u± + . . .. 3.1 Wrapping in the opposite channel The last ingredient for the evaluation of the opposite channel contribution is the prod- uct of the mirror measure µ(uγ) and of the mirror energy 1/x[+a]x[−a] from eq. (2.8). Since in this case ℓ= 1, we ﬁnd that this is g4 a (u[+a]u[−a])3 1 + g2 2 (u[+a])2 + 1 u[+a]u[−a] + 2 (u[−a])2 + . . . . (3.6) (3.6) 2The Zhukovski variables x± are particularly convenient when discussing crossing transformations, but for our perturbative computations it is convenient to expand them as customary x± → √ 2/g u± + . . .. – 6 – In order to evaluate this expression we have to perform the integration over u, sum over the bound-state number a and an additional sum from eq. (3.3). To this end, it is necessary to massage the expressions we have found, in a way reminiscent of [34]. We comment on the necessary manipulations in appendix B. Eventually, we obtain an explicit expression for the integrand to be integrated over the real line, as well as some residues due to manipulations of the integration contours. However, just as it was the case in [34], the integral cannot be evaluated analytically — to our knowledge. Numeric integration is however possible to good accuracy, even if some care is necessary in the estimate of the numerical errors.3 Nonetheless we could always reduce the result to rational numbers and ζ3, ζ5 by the pslq algorithm [63]. It is interesting to note that the residues subtract all terms involving even ζ values, while sometimes not contributing to the pure ζ3 or ζ5. Remarkably, the integral arising from corrections to mirror dressing phase (see appendix B.3) is apparently always purely rational once the residues are extracted. JHEP02(2016)165 3In particular, we have encountered some issues in this sense when using the NIntegrate algorithms of Wolfram Mathematica. 3.2 Wrapping in the adjacent channels Let us proceed as in the previous subsection. We ﬁrstly deﬁne eTa(u−γ) = (−1)a Ta(u−γ) Qs j=1 dena(u−γ, uj) , dena(u−γ, uj) = (x[−a] −x− j )(1 −1/x[−a]x+ j ). (3.7) (3.7) xpanding eTa(u−γ), we ﬁnd that the ﬁrst non-vanishing contribution appears at order g2: eTa(u−γ) = g2E(1) " −Q(u[−a+1]) u[−a] + Q(u[a−1]) u[+a] + a−1 X k=1 Q(u[2k−1−a]) −Q(u[2k+1−a]) u[2k−a] # . (3.8) As we have discussed, the remaining expressions in eqs. (2.12)–(2.13) contain terms which are both u-dependent and partition dependent. Let us focus on one channel, say 1γ (2.12), and introduce the incomplete Baxter polynomial Qα(u) = Q j∈α(u −uj). Using that Q(u) = Qα(u)Q¯α(u) we can e.g. eliminate the products over the ¯α partition. The price is to introduce a product over the α partition As we have discussed, the remaining expressions in eqs. (2.12)–(2.13) contain terms which are both u-dependent and partition dependent. Let us focus on one channel, say 1γ (2.12), and introduce the incomplete Baxter polynomial Qα(u) = Q j∈α(u −uj). Using that Q(u) = Qα(u)Q¯α(u) we can e.g. eliminate the products over the ¯α partition. The price is to introduce a product over the α partition Mα =  Y j∈α e−2ipj  Qα(u[a−1])Qα(u[−a−1]) Qα(u[a+1])Qα(u[−a+1]), (3.9) (3.9) which modiﬁes the partition dependent term so that it diﬀers from the asymptotic one A = P α∪¯α Aα,¯α. Instead we now have e A = X α∪¯α Aα,¯αMα, (3.10) (3.10) which depends on a and u. Clearly the 5γ channel can be found by exchanging α ↔¯α. As it turns out, e A is identical in the two cases. which depends on a and u. Clearly the 5γ channel can be found by exchanging α ↔¯α. As it turns out, e A is identical in the two cases. 3In particular, we have encountered some issues in this sense when using the NIntegrate algorithms of Wolfram Mathematica. – 7 – Given that we are interested in integrating this expression, it is convenient to make all possible poles manifest. To this end, we introduce a function eQ(u, a, s) that satisﬁes e A Q(u[a+1]) Q(u[−a+1]) = A eQ(u, a, s). (3.11) (3.11) By construction, eQ(u, a, s) is a polynomial in u, i.e. it has no poles. Such a polynomial can be explicitly evaluated for every given spin s, see appendix C. 4 Results, conclusions and outlook Summing up the various contributions described in the last two sections we ﬁnally obtain the structure constants of two half-BPS operators of length-2 going into a twist-2 operator: s C•◦◦ C◦◦◦ 2 for twist L = 2, bridge ℓ= 1 and spin s 2 1 6 −g2 + (7 + 3ζ3)g4 −(48 + 8ζ3 + 25ζ5)g6 + . . . 4 1 70 −205 1764g2 + 76393 74088 + 5 14ζ3 g4 − 242613655 28005264 + 1315 1323ζ3 + 125 42 ζ5 g6 + . . . 6 1 924 − 553 54450g2 + 880821373 8624880000 + 7 220ζ3 g4 − 1364275757197 1423105200000 + 520093 6534000ζ3 + 35 132ζ5 g6 + . . . 8 1 12870 − 14380057 18036018000g2 + 5944825782678337 682443241880400000 + 761 300300ζ3 g4 − 758072803634287465765957 8607383632540733040000000 + 15248925343 2840672835000ζ3 + 761 36036ζ5 g6 + . . . 10 1 184756 − 3313402433 55983859495200g2 + 171050793565932326659 248804677619932936320000 + 671 3527160ζ3 g4 − 9135036882706194334305789554347 1243961012766985364412864576000000 + 11482697774339 35269831481976000ζ3 + 3355 2116296ζ5 g6 + . . . s C•◦◦ C◦◦◦ 2 for twist L = 2, bridge ℓ= 1 and spin s 2 1 6 −g2 + (7 + 3ζ3)g4 −(48 + 8ζ3 + 25ζ5)g6 + . . . 4 1 70 −205 1764g2 + 76393 74088 + 5 14ζ3 g4 − 242613655 28005264 + 1315 1323ζ3 + 125 42 ζ5 g6 + . . . 6 1 924 − 553 54450g2 + 880821373 8624880000 + 7 220ζ3 g4 − 1364275757197 1423105200000 + 520093 6534000ζ3 + 35 132ζ5 g6 + . . . 8 1 12870 − 14380057 18036018000g2 + 5944825782678337 682443241880400000 + 761 300300ζ3 g4 − 758072803634287465765957 8607383632540733040000000 + 15248925343 2840672835000ζ3 + 761 36036ζ5 g6 + . . . 10 1 184756 − 3313402433 55983859495200g2 + 171050793565932326659 248804677619932936320000 + 671 3527160ζ3 g4 − 9135036882706194334305789554347 1243961012766985364412864576000000 + 11482697774339 35269831481976000ζ3 + 3355 2116296ζ5 g6 + . . . The O(g6) values in the table are in exact agreement with the conformal partial-wave analysis of the four-point function of stress energy tensor multiplets [44, 51]. This test of the hexagon conjecture [34] was the main motivation for our work. 3.2 Wrapping in the adjacent channels Using the expansion of the remaining terms in (2.12) we ﬁnd that the integrand takes the form A Q(i/2) eQ(u, a, s) Q(u[a+1])Q(u[−a−1])Q(u[a−1])Q(u[−a+1]) Ta(u−γ). (3.12) (3.12) JHEP02(2016)165 Even if eQ(u, a, s) in general is not a Baxter polynomial, it has degree s. It follows that the integrand decays fast for large \|u\|, and can be evaluated by residues. We further comment on the related technicalities in appendix C. Even if eQ(u, a, s) in general is not a Baxter polynomial, it has degree s. It follows that the integrand decays fast for large \|u\|, and can be evaluated by residues. We further comment on the related technicalities in appendix C. 4 Results, conclusions and outlook s C•◦◦ C◦◦◦ 2 for twist L = 2, bridge ℓ12 = ℓ31 = 1, ℓ23 > 1 and spin s 2 1 6 −g2 + 7g4 + (10ζ5η −10ζ5 + 7ζ3 −48)g6 + . . . 4 1 70 −205 1764g2 + 36653 37044g4 + 1 6ζ3 + 25 21ζ5 η −25 21ζ5 + 193 216ζ3 −442765625 56010528 g6 + . . . 6 1 924 − 553 54450g2 + 826643623 8624880000g4 + − 1 1440 + 7 264ζ3 + 7 66ζ5 η −7 66ζ5 + 24143 297000ζ3 −1183056555847 1423105200000 g6 + . . . 8 1 12870 − 14380057 18036018000g2 + 2748342985341731 341221620940200000g4 + − 79 604800 + 3 1040ζ3 + 761 90090ζ5 η −761 90090ζ5 + 1039202363 158918760000ζ3 −1270649655622342732745039 17214767265081466080000000 g6 + . . . 10 1 184756 − 3313402433 55983859495200g2 + 156422034186391633909 248804677619932936320000g4 + − 45071 2813045760 + 781 2930256ζ3+ 671 1058148ζ5 η − 671 1058148ζ5 + 8295615163 16799157648000ζ3 − 7465848687069712820911408164847 1243961012766985364412864576000000 g6 + . . . s C•◦◦ C◦◦◦ 2 for twist L = 2, bridge ℓ12 = ℓ31 = 1, ℓ23 > 1 and spin s 2 1 6 −g2 + 7g4 + (10ζ5η −10ζ5 + 7ζ3 −48)g6 + . . . 4 1 70 −205 1764g2 + 36653 37044g4 + 1 6ζ3 + 25 21ζ5 η −25 21ζ5 + 193 216ζ3 −442765625 56010528 g6 + . . . 6 1 924 − 553 54450g2 + 826643623 8624880000g4 + − 1 1440 + 7 264ζ3 + 7 66ζ5 η −7 66ζ5 + 24143 297000ζ3 −1183056555847 1423105200000 g6 + . . . 8 1 12870 − 14380057 18036018000g2 + 2748342985341731 341221620940200000g4 + − 79 604800 + 3 1040ζ3 + 761 90090ζ5 η −761 90090ζ5 + 1039202363 158918760000ζ3 −1270649655622342732745039 17214767265081466080000000 g6 + . . . 10 1 184756 − 3313402433 55983859495200g2 + 156422034186391633909 248804677619932936320000g4 + − 45071 2813045760 + 781 2930256ζ3+ 671 1058148ζ5 η − 671 1058148ζ5 + 8295615163 16799157648000ζ3 − 7465848687069712820911408164847 1243961012766985364412864576000000 g6 + . . . JHEP02(2016)165 It would be very interesting to extend this analysis to higher-order corrections. A ﬁrst question is how one can make sense of wrapping corrections to the operators of the three- point functions. While this is a familiar problem in the context of two-point functions, this issue still to be explored in the hexagon program. 4 Results, conclusions and outlook As we mentioned, these eﬀects would ﬁrst appear at order O(g8), and therefore are accessible to gauge-theoretical computations, which would provide another crucial check of the hexagon approach. In the long run one could try to constrain a putative octagon operator [34] for four- point functions by perturbative data. Ideally we obtain a machinery that will directly furnish non-trivial kinematics, so which will allow us to reach out beyond the computation of sets of constants. Another diﬃcult but very interesting question is whether the hexagon approach can be promoted to a truly non-perturbative formalism. While as we saw L¨uscher-like cor- rections work remarkably well, one should also account account for more than one virtual particle at a time4 within the hexagon approach [34]. Ideally one would hope that a sort of “Thermodynamic Bethe Ansatz” formulation could be constructed for the hexagon. We are conﬁdent to witness remarkable developments in this direction in the near future. 4For advances in this direction in the context of two-point functions see ref. [64]. 4 Results, conclusions and outlook It probes the correctness of the new approach to ﬁnite size corrections already rather deeply, as is well illustrated by the multitude of eﬀects we had to take into account. The O(g6) values in the table are in exact agreement with the conformal partial-wave analysis of the four-point function of stress energy tensor multiplets [44, 51]. This test of the hexagon conjecture [34] was the main motivation for our work. It probes the correctness of the new approach to ﬁnite size corrections already rather deeply, as is well illustrated by the multitude of eﬀects we had to take into account. An obvious extension to this work is to predict structure constants for the fusion of two higher-length single-trace half-BPS operators into twist 2 operators. In that case the bridge length for the adjacent channels would stay put at ℓ12 = ℓ31 = 1, while ℓ23 for the opposite channel would rise. For length-3 operators we have ℓ23 = 2, so that the leading order analysis of [34] now applies to the three-loop correction. As the mirror measure now starts on 1/(u+u−)4 we obtain a modiﬁed eﬀective integration measure ˆµ = 16π3 3(1 + 4u2)4 cosh2(πu) π(−1 + 8u2 + 48u4) cosh2(πu) (4.1) + 48u(−1 + 4u2) tanh(πu) −2π(−1 + 8u2 + 48u4) tanh2(πu) . ˆµ = 16π3 3(1 + 4u2)4 cosh2(πu) π(−1 + 8u2 + 48u4) cosh2(πu) (4.1) (4.1) + 48u(−1 + 4u2) tanh(πu) −2π(−1 + 8u2 + 48u4) tanh2(πu) . + 48u(−1 + 4u2) tanh(πu) −2π(−1 + 8u2 + 48u4) tanh2(πu) . – 8 – – 8 – with the subtraction of residues following the by now standard path. The contribution of the opposite channel for length 3 is marked in the table by the coeﬃcient η, which ought to be put to 1/2 in this case. If the length of the BPS is greater or equal 4, we have bridge length ℓ23 > 2 and the leading contribution in the opposite channel moves out to O(g8) or higher. The approach of [34] then predicts the result in the table below at η = 0, so notably perfect universality of the structure constants up to three loops. We will check what constraints these results can impose on Ans¨atze for higher-charge planar correlation functions [44]. A Telescoping the transfer matrix The transfer matrix in the antisymmetric bound state representation as deﬁned in formula (H1) in ref. [34] can be substantially simpliﬁed when the level-matching condition, or, in gauge theory parlance, the zero momentum condition Q j x+ j /x− j = 1 is satisﬁed. Using JHEP02(2016)165 Y j (x[a] −x∓ j ) 1 − 1 x[a]x∓ j ! = Q(u[a±1]) (A.1) (A.1) it follows that R+(u[a]) B+(u[a]) R−(u[a]) B−(u[a]) = Q(u[a+1]) Q(u[a−1]), (A.2) (A.2) where we followed the notation of [34] R±(u) = Y j (x(u) −x∓ j ) , B±(u) = Y j 1 x(u) −x∓ j . (A.3) (A.3) In every summand in Ta(uγ) and Ta(u−γ) the Q factors “telescope”, so that all terms but the ﬁrst in the denominator and the last in the numerator cancel. In terms of the function eTa deﬁned in (3.1) eTa(uγ) = R−(u[−a])B+(u[−a]) + R+(u[a])B−(u[a]) −2R−(u[a])B−(u[a]) + a−1 X k=1 R+(u[2k−a])B−(u[2k−a]) + R−(u[2k−a])B+(u[2k−a]) −2R−(u[2k−a])B−(u[2k−a]) . (A.4) (A.4) A similar expression for eTa(uγ) can be found immediately by crossing x[±a] →1/x[±a], i.e. by swapping B± ↔R±. A similar expression for eTa(uγ) can be found immediately by crossing x[±a] →1/x[±a], i.e. by swapping B± ↔R±. Acknowledgments We would like to thank G. Arutyunov and S. van Tongeren for discussions. B.E. is sup- ported by the DFG, “eigene Stelle” ED 78/4-2 and acknowledges partial support by the 4For advances in this direction in the context of two-point functions see ref. [64]. We would like to thank G. Arutyunov and S. van Tongeren for discussions. B.E. is sup- ported by the DFG, “eigene Stelle” ED 78/4-2 and acknowledges partial support by the – 9 – Marie Curie network GATIS under REA Grant Agreement No 317089. A.S. would like to thank the group for Mathematical Physics of Space, Time and Matter at Humboldt Uni- versity where part of this research was carried out. A.S.’s research was partially supported by the NCCR SwissMAP, funded by the Swiss National Science Foundation. B Evaluation of the opposite-channel wrapping With respect to the computation detailed in ref. [34] we now have to include sub-leading order O(g6) contributions. These may come from corrections to diﬀerent bits of the leading- order expansion of eq. (2.8): 1. From corrections to the rapidities, through Ps j=1 u(1) j ∂uj. This does not substantially alter the analytic form of leading-order expression, and can be easily evaluated by shifting the integration variable to reduce the integral to take values over a single Q-function [34]. 1. From corrections to the rapidities, through Ps j=1 u(1) j ∂uj. This does not substantially alter the analytic form of leading-order expression, and can be easily evaluated by shifting the integration variable to reduce the integral to take values over a single Q-function [34]. – 10 – 2. From corrections to the integration measure µ(uγ), cf. eq. (3.6). 3. From corrections eT (1) a (uγ) to the transfer matrix eTa(uγ), cf. eq. (3.3). 4. From corrections coming from the dressing factor, cf. eq. (3.5). Below we will discuss in more detail these last three contributions. Below we will discuss in more detail these last three contributions. B.1 Corrections to the measure The strategy here is once again to perform shifts in the integration variable u in such a way as to end up with a a single Q-function Q(u). These shifts produce the total eﬀective measure 2 JHEP02(2016)165 µc = 16π2 (3(1 + 4u2)5) cosh2(πu) h 24(1 −40u2 + 80u4) + 48πu(1 −16u4) tanh(πu) −4π3u(1 + 4u2)3 tanh3(πu) + π2(1 + 4u2)2 cosh−2(πu) (−1 + 12u2)(−2 + cosh(2πu)) + 8πu(1 + 4u2) tanh(πu) i . (B.1) (B.1) Like the leading measure this falls oﬀexponentially for large \|u\| so that integration against the Baxter polynomials Q(u) is possible for any spin. The shifts u →u ± a+1 2 i must again be accompanied by the subtraction of residues from crossing the poles at ∓a 2i. B.3 Measure and residues from the ﬁrst order mirror dressing phase B.3 Measure and residues from the ﬁrst order mirror dressing phase B.3 Measure and residues from the ﬁrst order mirror dressing phase Beyond a rational factor the mirror dressing phase contains the very special combination of digamma functions Ψa(u) of eq. (3.5). This results in an integrand of the form JHEP02(2016)165 yond a rational factor the mirror dressing phase contains the very special combination digamma functions Ψa(u) of eq. (3.5). This results in an integrand of the form S1(s) X a>0 a Ψa(u) (u[+a] u[−a])3 Q(u[a+1]) + Q(u[−a−1]) −Q(u[a−1]) −Q(u[−a+1]) . (B.5) (B.5) Once again the strategy is to shift each term in the integrand in such a way as to extract an overall factor of Q(u). The presence of the digamma functions mandates some extra care. For each shift, two of the four digamma functions ψ in Ψa become independent of the bound state number a. The sum over a then simply yields a second polygamma factor. As for the digamma functions that do depend on the bound state number even after shifting, we recall that ψ(1 + x) = ∞ X k=1 1 k − 1 k + x −γ, (B.6) (B.6) which leads to a second summation over terms of the type 1/(a−k +y) or 1/(a+k +y), as well as formally some ζ(1) terms. One of the factors (u[+]a)−3 or (u[−a])−3 from the bound state measure yields 1/(a ± 1/2 ± iu)3, too. The idea is now to use partial fractions w.r.t. a to decompose products of the two factors involving a into single (if higher order) poles. We then swap the sums as before and shift to a′ = a −k and a′′ = a + k. The sums from 1 −k, k + 1 to inﬁnity, respectively, are conveniently split into an inﬁnite part with bounds 1 . . . ∞and a ﬁnite bit −a′ ∈{0 . . . k −1} and a′′ ∈{1 . . . k}. After some algebra it is found that all ζ(1) terms cancel and what is more, the ﬁnite sums “telescope” whereby the nested double sum totally disappears. The net result of this complicated looking exercise is surprisingly concise γ (u−)3 h 2iψ′(1−iu−)+2iψ′(1+iu−)+u−ψ′′(1−iu−)−u−ψ′′(1+iu−) i −(u−↔u+) . (B. B.2 Corrections to the transfer matrix Let us consider eT (1)) a (uγ) (3.3). Once again, we want to shift u in such a way as to integrate only on Q(u). For the boundary terms with Q(u[−a−1]), Q(u[a+1]) this is quite as before. The sum over k can be concisely rewritten: originally one has Q(u) ∞ X a=1 a a−1 X k=1 1 u+ ((a −k) −i u+)3 (k + i u+)3 −(u+ ↔u−). (B.2) (B.2) We can swap the order of summations and rewrite the sum in terms of a′ = a −k. In this way, the sums decouple and we can evaluate in terms of polygamma functions. The “eﬀective” integration measure emerging form these shifts is µp = 32iπ2 (1 + 4u2)4 cosh2(πu) 1 −24u2 + 16u4 + πu(3 + 8u2 −16u4) tanh(πu) + 2i 1 + 4u2 ψ′ iu− ψ′′ −iu+ + ψ′ −iu+ ψ′′ iu− . (B.3) (B.3) The terms in the second line cannot straightforwardly be rewritten in terms of trigonometric functions; nonetheless this part also has the desired asymptotic behaviour e−2π\|u\| for large values of the argument. Residues from the boundary terms in the second line of (3.3) are computed as usual. A further remark concerns the subtraction for the sum part of (3.3): to begin with let a be odd. Then the shift of Q(u−a+2k−1) →Q(u) crosses the denominator pole only if −a + 2k < 0, likewise for −a + 2k > 0 only the other term picks up a residue. For even – 11 – a, k = a/2 we look at (Q(u−)−Q(u+))/u. In either term the shift approaches the pole but does not cross it. As in a principle value prescription both terms contribute half a residue. Therefore in both cases — even and odd bound state number — we have to subtract a, k = a/2 we look at (Q(u−)−Q(u+))/u. In either term the shift approaches the pole but does not cross it. As in a principle value prescription both terms contribute half a residue. Therefore in both cases — even and odd bound state number — we have to subtract i Q (i/2) ∞ X a=1 a−1 X k=1 a (a −k)3 k3 = 2i ζ2ζ3 Q (i/2) , (B.4) (B.4) where the right-hand-side can be found by the same manipulations on the double sum as above. B.3 Measure and residues from the ﬁrst order mirror dressing phase B.3 Measure and residues from the ﬁrst order mirror dressing phase (B.7) Adding it to the contribution from the a independent digamma functions we obtain the complete eﬀective measure µm = 32π2 (1 + 4u2)4 cosh2(πu) h −1 + 24u2 −16u4 + πu(−3 −8u2 + 16u4) tanh(πu) i − 16π2 (1 + 4u2)3 cosh2(πu) h 1 −12u2 + 2πu(1 + 4u2) tanh(πu) i × h 2γ + ψ(−iu+) + ψ(iu−) i , (B.8) µm = 32π2 (1 + 4u2)4 cosh2(πu) h −1 + 24u2 −16u4 + πu(−3 −8u2 + 16u4) tanh(πu) i (B.8) × h 2γ + ψ(−iu+) + ψ(iu−) i , × h 2γ + ψ(−iu+) + ψ(iu−) i , induced by the mirror dressing phase. – 12 – Here the subtraction of residues deserves further attention. To begin with note that Ψa has simple poles at ±i(a/2+k). Shifting thus needs to be complemented by the subtraction of residues only at ±ia/2. Let us focus on the Q(u[a+1]) polynomial, i.e. on the residue at u = +ia/2. We decompose Ψa(u) = Ψreg a (u) + i u −i 2a (B.9) (B.9) by adding and subtracting the pole. Now, the pole part of Ψ(u) will combine with the (u[−a])3 factor from the bound-state mirror measure to form a fourth-order pole. This residue is then given by a third derivative acting on Q(u[a+1])/(u[+a])3. by adding and subtracting the pole. Now, the pole part of Ψ(u) will combine with the (u[−a])3 factor from the bound-state mirror measure to form a fourth-order pole. This residue is then given by a third derivative acting on Q(u[a+1])/(u[+a])3. Secondly, the bound-state measure itself has the familiar third-order pole, with now a residue Ψreg a (u) Q(u[a+1])/(u[a])3. It is not hard to work out that JHEP02(2016)165 Ψreg a \| i 2 a = 2S1(a) −1 a , Ψreg a ′\| i 2 a = i 2S2(a) −1 a2 , Ψreg a ′′\| i 2 a = −2 2S3(a) −1 a3 −4ζ(3) . (B.10) Ψreg a ′\| i 2 a = i 2S2(a) −1 a2 , Ψreg a ′′\| i 2 a = −2 2S3(a) −1 a3 −4ζ(3) . (B.10) The bound state sum therefore creates all the double ζ-values {ζ21, ζ22, ζ23, ζ31, ζ32, ζ41} up to transcendentality weight 5. All of these can be recast in terms of ordinary zeta-values. B.3 Measure and residues from the ﬁrst order mirror dressing phase The bound state sum therefore creates all the double ζ-values {ζ21, ζ22, ζ23, ζ31, ζ32, ζ41} up to transcendentality weight 5. All of these can be recast in terms of ordinary zeta-values. C Evaluation of the adjacent-channel wrapping The crucial ingredient in the evaluation of the adjacent-channel wrapping is the polyno- mial eQ(u, a, s) introduced in (3.11). This is a degree-s polynomial that can be explicitly evaluated for each given s. Interestingly, if we extract a normalisation pre-factor eQ = n eQ′ so that the highest-degree monomial in eQ′ has unit coeﬃcient, we ﬁnd n = Q(i/2). We collect the ﬁrst few expression for eQ in this normalisation in the table below. Curiously, when a = 1, the ˜Qs are in fact Baxter polynomials. spin eQ(u, a, s)/Q(i/2) 2 −1 3 + 1 4a2 + u2 4 12 35 −23 84a2 −1 48a4 −23 21u2 + 1 6a2u2 + u4 6 −60 77 + 139 220a2 + 3 44a4 + 1 320a6 + 139 55 u2 −6 11a2u2 −9 80a4u2 −36 11u4 −1 4a2u4 + u6 Let us now see how to integrate (3.12). As had been mentioned above, in the sum part of (3.8) the poles 1/u−a+2k are absent because the diﬀerence in the respective numerator factors out a power of u−a+2k. Since we are not interested in shifting u it is best to cancel these factors and to work with the remaining total numerator polynomials. In these it is of course trivial to execute the sum over k. Note that eQ times this polynomial or Q(u[a−1]), Q(u[−a+1]) contains powers up to a2s, to be augmented by the explicit factor a in the bound state measure. – 13 – Any residue has various denominator terms (a + x)n where x may contain ±1/2, ±1 and/or one or two Bethe roots. We may now use partial fractions w.r.t. to a to reduce to single (generically higher order) poles which can be summed over the bound state counter to yield polygamma functions. The procedure is well-behaved in that 1. all potentially divergent sums ζ0, ζ−1 . . . ζ−2s−1 cancel, 1. all potentially divergent sums ζ0, ζ−1 . . . ζ−2s−1 cancel, 2. ζ1 also cancels when the sums 1/(a + x) are expressed as ψ(1 + x) by eq. (B.6), 2. ζ1 also cancels when the sums 1/(a + x) are expressed as ψ(1 + x) by eq. (B.6), 2. ζ1 also cancels when the sums 1/(a + x) are expressed as ψ(1 + x) by eq. (B.6) 3. and the transcendentality level does not increase with the spin. C Evaluation of the adjacent-channel wrapping One then ﬁnds some ζ-values with rational coeﬃcients and a sum over polygamma functions containing the Bethe roots in their arguments. For spin 2 this can immediately be simpliﬁed to rational numbers and ζ3, ζ5 due to the property ψ(1+x) = ψ(x)+1/x and its derivatives. In general, we have not tried to analytically solve this — which may well be possible — but rather numerically evaluated to high precision and reconverted by the pslq algorithm. The adjacent channel calculation has a deﬁnite advantage on the opposite channel case because no integration is needed to arrive at the ﬁnal result whereby the numerical precision can be much better. JHEP02(2016)165 Open Access. This article is distributed under the terms of the Creative Commons Attribution License (CC-BY 4.0), which permits any use, distribution and reproduction in any medium, provided the original author(s) and source are credited. References [1] G. ’t Hooft, Dimensional reduction in quantum gravity, in Salamfest (1993) 0284 [gr-qc/9310026] [INSPIRE]. [2] J.M. Maldacena, The large-N limit of superconformal ﬁeld theories and supergravity, Int. J. Theor. Phys. 38 (1999) 1113 [hep-th/9711200] [INSPIRE]. [3] E. Witten, Anti-de Sitter space and holography, Adv. Theor. Math. Phys. 2 (1998) 253 [hep-th/9802150] [INSPIRE]. [4] S.S. Gubser, I.R. Klebanov and A.M. Polyakov, Gauge theory correlators from noncritical string theory, Phys. Lett. B 428 (1998) 105 [hep-th/9802109] [INSPIRE]. [5] G. ’t Hooft, A Planar Diagram Theory for Strong Interactions, Nucl. Phys. B 72 (1974) 461 [INSPIRE]. [5] G. ’t Hooft, A Planar Diagram Theory for Strong Interactions, Nucl. Phys. B 72 (1974) 461 [INSPIRE]. [6] G. Arutyunov and S. Frolov, Foundations of the AdS5 × S5 Superstring. Part I, J. Phys. A 42 (2009) 254003 [arXiv:0901.4937] [INSPIRE]. [7] N. Beisert et al., Review of AdS/CFT Integrability: An Overview, Lett. Math. Phys. 99 (2012) 3 [arXiv:1012.3982] [INSPIRE]. [8] T. Klose, Review of AdS/CFT Integrability, Chapter IV.3: N = 6 Chern-Simons and Strings on AdS4 × CP 3, Lett. Math. Phys. 99 (2012) 401 [arXiv:1012.3999] [INSPIRE]. [9] A. Sfondrini, Towards integrability for AdS3/CFT2, J. Phys. A 48 (2015) 023001 [arXiv:1406.2971] [INSPIRE]. – 14 – [10] S.J. van Tongeren, Integrability of the AdS5 × S5 superstring and its deformations, J. Phys. A 47 (2014) 433001 [arXiv:1310.4854] [INSPIRE]. [11] K. Okuyama and L.-S. Tseng, Three-point functions in N = 4 SYM theory at one-loop, JHEP 08 (2004) 055 [hep-th/0404190] [INSPIRE]. [12] L.F. Alday, J.R. David, E. Gava and K.S. Narain, Structure constants of planar N = 4 Yang-Mills at one loop, JHEP 09 (2005) 070 [hep-th/0502186] [INSPIRE]. [13] R. Roiban and A. Volovich, Yang-Mills correlation functions from integrable spin chains, JHEP 09 (2004) 032 [hep-th/0407140] [INSPIRE]. [14] M.S. Costa, R. Monteiro, J.E. Santos and D. Zoakos, On three-point correlation functions in the gauge/gravity duality, JHEP 11 (2010) 141 [arXiv:1008.1070] [INSPIRE]. JHEP02(2016)165 [15] K. Zarembo, Holographic three-point functions of semiclassical states, JHEP 09 (2010) 030 [arXiv:1008.1059] [INSPIRE]. [16] J. Escobedo, N. Gromov, A. Sever and P. Vieira, Tailoring Three-Point Functions and Integrability, JHEP 09 (2011) 028 [arXiv:1012.2475] [INSPIRE]. [17] J. Escobedo, N. Gromov, A. Sever and P. Vieira, Tailoring Three-Point Functions and Integrability II. Weak/strong coupling match, JHEP 09 (2011) 029 [arXiv:1104.5501] [INSPIRE]. [18] N. Gromov, A. Sever and P. Vieira, Tailoring Three-Point Functions and Integrability III. Classical Tunneling, JHEP 07 (2012) 044 [arXiv:1111.2349] [INSPIRE]. [19] R.A. References Janik and A. Wereszczynski, Correlation functions of three heavy operators: The AdS contribution, JHEP 12 (2011) 095 [arXiv:1109.6262] [INSPIRE]. [20] Y. Kazama and S. Komatsu, On holographic three point functions for GKP strings from integrability, JHEP 01 (2012) 110 [Erratum ibid. 1206 (2012) 150] [arXiv:1110.3949] [INSPIRE]. [21] N. Gromov and P. Vieira, Tailoring Three-Point Functions and Integrability IV. Theta-morphism, JHEP 04 (2014) 068 [arXiv:1205.5288] [INSPIRE]. [22] I. Kostov, Classical Limit of the Three-Point Function of N = 4 Supersymmetric Yang-Mills Theory from Integrability, Phys. Rev. Lett. 108 (2012) 261604 [arXiv:1203.6180] [INSPIRE]. [23] I. Kostov, Three-point function of semiclassical states at weak coupling, J. Phys. A 45 (2012) 494018 [arXiv:1205.4412] [INSPIRE]. [24] I. Kostov and Y. Matsuo, Inner products of Bethe states as partial domain wall partition functions, JHEP 10 (2012) 168 [arXiv:1207.2562] [INSPIRE]. [25] O. Foda, Y. Jiang, I. Kostov and D. Serban, A tree-level 3-point function in the SU(3)-sector of planar N = 4 SYM, JHEP 10 (2013) 138 [arXiv:1302.3539] [INSPIRE]. [26] Y. Kazama and S. Komatsu, Wave functions and correlation functions for GKP strings from integrability, JHEP 09 (2012) 022 [arXiv:1205.6060] [INSPIRE]. [27] Y. Kazama and S. Komatsu, Three-point functions in the SU(2) sector at strong coupling, JHEP 03 (2014) 052 [arXiv:1312.3727] [INSPIRE]. [28] T. Klose and T. McLoughlin, Worldsheet Form Factors in AdS/CFT, Phys. Rev. D 87 (2013) 026004 [arXiv:1208.2020] [INSPIRE]. – 15 – [29] Y. Jiang, I. Kostov, F. Loebbert and D. Serban, Fixing the Quantum Three-Point Function, JHEP 04 (2014) 019 [arXiv:1401.0384] [INSPIRE]. [30] Z. Bajnok, R.A. Janik and A. Wereszczy´nski, HHL correlators, orbit averaging and form factors, JHEP 09 (2014) 050 [arXiv:1404.4556] [INSPIRE]. [31] Y. Kazama, S. Komatsu and T. Nishimura, Novel construction and the monodromy relation for three-point functions at weak coupling, JHEP 01 (2015) 095 [Erratum ibid. 1508 (2015) 145] [arXiv:1410.8533] [INSPIRE]. [32] Y. Jiang, I. Kostov, A. Petrovskii and D. Serban, String Bits and the Spin Vertex, Nucl. Phys. B 897 (2015) 374 [arXiv:1410.8860] [INSPIRE]. JHEP02(2016)165 [33] Z. Bajnok and R.A. Janik, String ﬁeld theory vertex from integrability, JHEP 04 (2015) 042 [arXiv:1501.04533] [INSPIRE]. [34] B. Basso, S. Komatsu and P. Vieira, Structure Constants and Integrable Bootstrap in Planar N = 4 SYM Theory, arXiv:1505.06745 [INSPIRE]. [35] L. Hollo, Y. Jiang and A. Petrovskii, Diagonal Form Factors and Heavy-Heavy-Light Three-Point Functions at Weak Coupling, JHEP 09 (2015) 125 [arXiv:1504.07133] [INSPIRE]. [36] I. Balitsky, V. Kazakov and E. References Sobko, Three-point correlator of twist-2 operators in BFKL limit, arXiv:1506.02038 [INSPIRE]. [37] Y. Jiang, S. Komatsu, I. Kostov and D. Serban, The hexagon in the mirror: the three-point function in the SoV representation, arXiv:1506.09088 [INSPIRE]. [38] C. Kristjansen, S. Mori and D. Young, On the Regularization of Extremal Three-point Functions Involving Giant Gravitons, Phys. Lett. B 750 (2015) 379 [arXiv:1507.03965] [INSPIRE]. [39] D. Arnaudov and R.C. Rashkov, Three-point correlation functions from pulsating strings in AdS5 × S5, Phys. Rev. D 92 (2015) 126001 [arXiv:1509.02831] [INSPIRE]. [40] N. Beisert and M. Staudacher, Long-range PSU(2, 2\|4) Bethe Ansatze for gauge theory and strings, Nucl. Phys. B 727 (2005) 1 [hep-th/0504190] [INSPIRE]. [40] N. Beisert and M. Staudacher, Long-range PSU(2, 2\|4) Bethe Ansatze for gauge theory and strings, Nucl. Phys. B 727 (2005) 1 [hep-th/0504190] [INSPIRE]. [41] J. Ambjørn, R.A. Janik and C. Kristjansen, Wrapping interactions and a new source of corrections to the spin-chain/string duality, Nucl. Phys. B 736 (2006) 288 [hep-th/0510171] [INSPIRE]. [42] M. L¨uscher, Volume Dependence of the Energy Spectrum in Massive Quantum Field Theories. 1. Stable Particle States, Commun. Math. Phys. 104 (1986) 177 [INSPIRE]. [43] M. L¨uscher, Volume Dependence of the Energy Spectrum in Massive Quantum Field Theories. 2. Scattering States, Commun. Math. Phys. 105 (1986) 153 [INSPIRE]. [44] B. Eden, P. Heslop, G.P. Korchemsky and E. Sokatchev, Hidden symmetry of four-point correlation functions and amplitudes in N = 4 SYM, Nucl. Phys. B 862 (2012) 193 [arXiv:1108.3557] [INSPIRE]. [45] B. Eden, P. Heslop, G.P. Korchemsky and E. Sokatchev, Constructing the correlation function of four stress-tensor multiplets and the four-particle amplitude in N = 4 SYM, Nucl. Phys. B 862 (2012) 450 [arXiv:1201.5329] [INSPIRE]. [46] G. Arutyunov, B. Eden, A.C. Petkou and E. Sokatchev, Exceptional nonrenormalization properties and OPE analysis of chiral four point functions in N = 4 SYM(4), Nucl. Phys. B 620 (2002) 380 [hep-th/0103230] [INSPIRE]. – 16 – [47] G. Arutyunov, F.A. Dolan, H. Osborn and E. Sokatchev, Correlation functions and massive Kaluza-Klein modes in the AdS/CFT correspondence, Nucl. Phys. B 665 (2003) 273 [hep-th/0212116] [INSPIRE]. [48] G. Arutyunov, S. Penati, A. Santambrogio and E. Sokatchev, Four point correlators of BPS operators in N = 4 SYM at order g4, Nucl. Phys. B 670 (2003) 103 [hep-th/0305060] [INSPIRE]. [49] G. Arutyunov and E. Sokatchev, On a large-N degeneracy in N = 4 SYM and the AdS/CFT correspondence, Nucl. Phys. B 663 (2003) 163 [hep-th/0301058] [INSPIRE]. References [50] F.A. Dolan and H. Osborn, Conformal partial wave expansions for N = 4 chiral four point functions, Annals Phys. 321 (2006) 581 [hep-th/0412335] [INSPIRE]. JHEP02(2016)165 JHEP02(2016)165 [51] B. Eden, Three-loop universal structure constants in N = 4 SUSY Yang-Mills theory, arXiv:1207.3112 [INSPIRE]. [52] B. Basso, V. Goncalves, S. Komatsu and P. Vieira, Gluing Hexagons at Three Loops, arXiv:1510.01683 [INSPIRE]. [53] A.B. Zamolodchikov and A.B. Zamolodchikov, Factorized S-matrices in two-dimensions as the exact solutions of certain relativistic quantum ﬁeld models, Annals Phys. 120 (1979) 253 [INSPIRE]. [54] N. Beisert, B. Eden and M. Staudacher, Transcendentality and Crossing, J. Stat. Mech. (2007) P01021 [hep-th/0610251] [INSPIRE]. [55] N. Beisert, The SU(2\|2) dynamic S-matrix, Adv. Theor. Math. Phys. 12 (2008) 948 [hep-th/0511082] [INSPIRE]. [56] H.-Y. Chen, N. Dorey and K. Okamura, On the scattering of magnon boundstates, JHEP 11 (2006) 035 [hep-th/0608047] [INSPIRE]. [57] R. Roiban, Magnon Bound-state Scattering in Gauge and String Theory, JHEP 04 (2007) 048 [hep-th/0608049] [INSPIRE]. [58] G. Arutyunov and S. Frolov, The Dressing Factor and Crossing Equations, J. Phys. A 42 (2009) 425401 [arXiv:0904.4575] [INSPIRE]. [59] A.V. Kotikov, L.N. Lipatov, A. Rej, M. Staudacher and V.N. Velizhanin, Dressing and wrapping, J. Stat. Mech. (2007) P10003 [arXiv:0704.3586] [INSPIRE]. [60] Z. Bajnok and R.A. Janik, Four-loop perturbative Konishi from strings and ﬁnite size eﬀects for multiparticle states, Nucl. Phys. B 807 (2009) 625 [arXiv:0807.0399] [INSPIRE]. [61] Z. Bajnok, R.A. Janik and T. Lukowski, Four loop twist two, BFKL, wrapping and strings, Nucl. Phys. B 816 (2009) 376 [arXiv:0811.4448] [INSPIRE]. [62] Z. Bajnok, A. Hegedus, R.A. Janik and T. Lukowski, Five loop Konishi from AdS/CFT, Nucl. Phys. B 827 (2010) 426 [arXiv:0906.4062] [INSPIRE]. [63] H.R.P. Ferguson, D.H. Bailey and S. Arno, Analysis of pslq, an integer relation ﬁnding algorithm, Math. Comput. 68 (1999) 351. [64] D. Bombardelli, A next-to-leading Luescher formula, JHEP 01 (2014) 037 [arXiv:1309.4083] [INSPIRE]. – 17 –
https://openalex.org/W4255179045	https://peerj.com/articles/3025v0.5/submission	English	null	Peer Review #2 of "Neogene amphibians and reptiles (Caudata, Anura, Gekkota, Lacertilia, and Testudines) from the south of Western Siberia, Russia, and Northeastern Kazakhstan (v0.1)"	null	2,017	cc-by	56,124	Neogene amphibians and reptiles (Caudata, Anura, Gekkota, Lacertilia, Testudines) from south of Western Siberia, Russia and Northeastern Kazakhstan 1 Department of Geosciences, Eberhard-Karls-University Tübingen, Tübingen, German 2 JURASSICA Museum, Porrentruy, Switzerland 3 Department of Geosciences, University of Fribourg, Fribourg, Switzerland 4 Institute of Geology, Russian Academy of Sciences, Moscow, Russia 5 Senckenberg Center for Human Evolution and Palaeoecology, Tübingen, Germany Corresponding Author: Davit Vasilyan Email address: davit.vasilyan@jurassica.ch Corresponding Author: Davit Vasilyan Email address: davit.vasilyan@jurassica.ch Background. The present-day amphibian and reptile fauna of Western Siberia is the least diverse of the Palaearctic Realm, as a consequence of the unfavourable climatic conditions that predominate in this region. The origin and emergence of these herpetofaunal groups are poorly understood. Aside from the better-explored European Neogene localities yielding amphibian and reptile fossil remains, the Neogene herpetofauna of Western Asia is understudied. The few available data need critical reviews and new interpretations, taking into account the more recent records of the European herpetofauna. The comparison of this previous data with that of the European fossil record would provide data on palaeobiogeographic affiliations of the region as well as on the origin and emergence of the present-day fauna of Western Siberia. An overview of the earliest occurrences of certain amphibian lineages is still needed. In addition, studies that address such knowledge gaps can be useful for molecular biologists in their calibration of molecular clocks. Methods and Results. In this study, we considered an critically review available data from amphibian and reptile fauna from over 40 Western Siberian, Russian and Northeastern Kazakhstan localities, ranging from the Middle Miocene to Early Pleistocene. Herein, we provided new interpretations that arose from our assessment of the previously published and new data. More than 50 amphibian and reptile taxa were identified belonging to families Hynobiidae, Cryptobranchidae, Salamandridae, Palaeobatrachidae, Bombinatoridae, Pelobatidae, Hylidae, Bufonidae, Ranidae, Gekkonidae, Lacertidae and Emydidae. Palaeobiogeographic analyses were performed for these groups and palaeoprecipitation values were estimated for 12 localities, using the bioclimatic analysis of herpetofaunal assemblages. Conclusions. The Neogene assemblage of Western Siberia was found to be dominated by groups of European affinities, such as Palaeobatrachidae, Bombina, Hyla, Bufo bufo, and a small part of this assemblage included Eastern Palaearctic taxa (e.g. Salamandrella, Tylototriton, Bufotes viridis). For several taxa (e.g. Mioproteus, Hyla, Bombina, Rana temporaria), the Western Siberian occurrences represented their most eastern Eurasian records. The most diverse collection of fossil remains was found in the Middle Miocene. Manuscript to be reviewed Manuscript to be reviewed Neogene amphibians and reptiles (Caudata, Anura, Gekkota, Lacertilia, Testudines) from south of Western Siberia, Russia and Northeastern Kazakhstan Less diversity has been registered towards the Early Pleistocene, potentially due to the progressive cooling of the climate in the Northern Hemisphere. The results of our study showed higher-amplitude changes of precipitation development in Western Siberia from the Early Miocene to the Pliocene, than previously assumed. PeerJ reviewing PDF \| (2016:08:12606:4:0:NEW 23 Jan 2017) Manuscript to be reviewed Manuscript to be reviewed 1 Neogene amphibians and reptiles (Caudata, Anura, 2 Gekkota, Lacertilia, Testudines) from south of Western 3 Siberia, Russia and Northeastern Kazakhstan 4 5 Davit Vasilyan1, 2, 3, Vladimir S. Zazhigin4, and Madelaine Böhme1,5 6 7 1 Department of Geosciences, Eberhard-Karls-University Tübingen, Sigwartstraße 10, 72076 8 Tübingen, Germany. 9 2 JURASSICA Museum, Route de Fontenais 21, 2900 Porrentruy, Switzerland, 10 davit.vasilyan@jurassica.ch. 11 3 Department of Geosciences, University of Fribourg, Chemin du musée 6, 1700 Fribourg, 12 Switzerland. 13 4 Institute of Geology, Russian Academy of Sciences, Pyzhevsky per. 7, 119017 Moscow, Russia, 14 zazhvol@gmail.com. 15 5 Senckenberg Center for Human Evolution and Palaeoecology, Sigwartstraße 10, 72076 16 Tübingen, Germany, m.boehme@uni-tuebingen.de. 17 18 Corresponding author: 19 Davit Vasilyan 20 Current address: JURASSICA Museum, Route de Fontenais 21, 2900 Porrentruy, Switzerland 21 Email address: davit.vasilyan@jurassica.ch 22 1 Neogene amphibians and reptiles (Caudata, Anura, 2 Gekkota, Lacertilia, Testudines) from south of Western 3 Siberia, Russia and Northeastern Kazakhstan 22 PeerJ reviewing PDF \| (2016:08:12606:4:0:NEW 23 Jan 2017) 23 Abstract 23 Abstract 24 Background. The present-day amphibian and reptile fauna of Western Siberia is the least 25 diverse of the Palaearctic Realm, as a consequence of the unfavourable climatic conditions that 26 predominate in this region. The origin and emergence of these herpetofaunal groups are poorly 27 understood. Aside from the better-explored European Neogene localities yielding amphibian 28 and reptile fossil remains, the Neogene herpetofauna of Western Asia is understudied. The few 29 available data need critical reviews and new interpretations, taking into account the more 30 recent records of the European herpetofauna. The comparison of this previous data with that 31 of the European fossil record would provide data on palaeobiogeographic affiliations of the 32 region as well as on the origin and emergence of the present-day fauna of Western Siberia. An 33 overview of the earliest occurrences of certain amphibian lineages is still needed. In addition, 34 studies that address such knowledge gaps can be useful for molecular biologists in their 35 calibration of molecular clocks. 23 Abstract 24 Background. The present-day amphibian and reptile fauna of Western Siberia is the least 25 diverse of the Palaearctic Realm, as a consequence of the unfavourable climatic conditions that 26 predominate in this region. The origin and emergence of these herpetofaunal groups are poorly 27 understood. Aside from the better-explored European Neogene localities yielding amphibian 28 and reptile fossil remains, the Neogene herpetofauna of Western Asia is understudied. The few 29 available data need critical reviews and new interpretations, taking into account the more 30 recent records of the European herpetofauna. The comparison of this previous data with that 31 of the European fossil record would provide data on palaeobiogeographic affiliations of the 32 region as well as on the origin and emergence of the present-day fauna of Western Siberia. An 33 overview of the earliest occurrences of certain amphibian lineages is still needed. In addition, 34 studies that address such knowledge gaps can be useful for molecular biologists in their 35 calibration of molecular clocks. 36 Methods and Results. In this study, we considered an critically review available data from 37 amphibian and reptile fauna from over 40 Western Siberian, Russian and Northeastern 38 Kazakhstan localities, ranging from the Middle Miocene to Early Pleistocene. Herein, we 39 provided new interpretations that arose from our assessment of the previously published and 40 new data. Manuscript to be reviewed 45 Conclusions. The Neogene assemblage of Western Siberia was found to be dominated by 46 groups of European affinities, such as Palaeobatrachidae, Bombina, Hyla, Bufo bufo, and a small 47 part of this assemblage included Eastern Palaearctic taxa (e.g. Salamandrella, Tylototriton, 48 Bufotes viridis). For several taxa (e.g. Mioproteus, Hyla, Bombina, Rana temporaria), the 49 Western Siberian occurrences represented their most eastern Eurasian records. The most 50 diverse collection of fossil remains was found in the Middle Miocene. Less diversity has been 51 registered towards the Early Pleistocene, potentially due to the progressive cooling of the 52 climate in the Northern Hemisphere. The results of our study showed higher-amplitude changes 53 of precipitation development in Western Siberia from the Early Miocene to the Pliocene, than 54 previously assumed. 55 56 Introduction 57 Western Siberia is a geographic region restricted to the territories of Russia and parts of 58 northern Kazakhstan. It includes the region between the Ural Mountains in the west, Central 59 Siberian Plateau in the east, and the Kazakh Plain and Altay Mountains, including the Zaisan 60 Lake in the south (Fig. 1). Western Siberia region incorporates the drainage basin of the major 61 Siberian rivers such as the Irtysh and Ob rivers, both flowing into the Kara Sea of the Arctic 62 Ocean. The region is characterised by a highly continental climate, under the influence of the 63 Westerlies (winds). The mean annual precipitation (MAP) is relatively uniform and varies from 64 400 mm in the north (415 mm at Omsk) to 200 mm in the south (255 mm at Pavlodar). The 65 region has a high relative humidity in summer due to labile convective heating and frequent 66 torrential rainfalls. The mean annual range of temperature reaches 4 °C and more (Omsk: cold 45 Conclusions. The Neogene assemblage of Western Siberia was found to be dominated by 46 groups of European affinities, such as Palaeobatrachidae, Bombina, Hyla, Bufo bufo, and a small 47 part of this assemblage included Eastern Palaearctic taxa (e.g. Salamandrella, Tylototriton, 48 Bufotes viridis). For several taxa (e.g. Mioproteus, Hyla, Bombina, Rana temporaria), the 49 Western Siberian occurrences represented their most eastern Eurasian records. The most 50 diverse collection of fossil remains was found in the Middle Miocene. Less diversity has been 51 registered towards the Early Pleistocene, potentially due to the progressive cooling of the 52 climate in the Northern Hemisphere. PeerJ reviewing PDF \| (2016:08:12606:4:0:NEW 23 Jan 2017) 23 Abstract More than 50 amphibian and reptile taxa were identified belonging to families 41 Hynobiidae, Cryptobranchidae, Salamandridae, Palaeobatrachidae, Bombinatoridae, 42 Pelobatidae, Hylidae, Bufonidae, Ranidae, Gekkonidae, Lacertidae and Emydidae. 43 Palaeobiogeographic analyses were performed for these groups and palaeoprecipitation values 44 were estimated for 12 localities, using the bioclimatic analysis of herpetofaunal assemblages. 43 Palaeobiogeographic analyses were performed for these groups and palaeoprecipitation values 44 were estimated for 12 localities, using the bioclimatic analysis of herpetofaunal assemblages. PeerJ reviewing PDF \| (2016:08:12606:4:0:NEW 23 Jan 2017) Manuscript to be reviewed 67 month temperature – CMT -19 °C, warm month temperature – WMT 20 °C, mean annual 68 temperature – MAT 0.4 °C; Semipalatinsk: CMT -16 °C, WMT 22 °C, MAT 3.1 °C; Lake Zaisan: 69 CMT up to -27 °C, WMT 23 °C; after Müller & Hennings (2000). The area is covered by diverse 70 biomes, namely the tundra (ʽcold steppeʼ) and taiga (coniferous forests) biomes, which are 71 replaced by open landscapes in the north (tundra) and in the south (steppe). The region that 72 contains the studied Neogene outcrops is located in the transitional zone between the dry and 73 the more humid temperate biomes, where taiga, forest-steppe and steppe biomes are 74 distributed (Ravkin et al., 2008). 75 Due to the strong continental climate, the present-day herpetofauna in the territory of Western 76 Siberia is comparatively far less diverse,represented only by six to ten amphibian species and 77 seven reptile species (Table 1). It is assumed that the present distribution of amphibians and 78 reptiles in Western Siberia was strongly influenced by Quaternary climatic fluctuation (Ravkin, 79 Bogomolova & Chesnokova, 2010). According to Borkin (Borkin, 1999), the present-day 80 amphibian fauna of Western Siberia belongs to the Siberian region of amphibian distribution in 81 the Palaearctic Realm. According to different authors (e.g. Kuzmin, 1995; Amphibiaweb, 2016), 82 the region is inhabited by few amphibians, namely two species of salamanders and four to eight 83 species of anurans, belonging to five genera and five families (Table 1). This is the poorest 84 regional diversity of fauna in the Palaearctic Realm, without any endemic species. Only 85 Salamandrella keyserlingii and Rana amurensis are characteristic of the territory, but they are 86 widely distributed and are also found in smaller areas in the neighbouring regions (Borkin, 87 1999). The Western Siberian reptile fauna listing includes few species: Natrix natrix, Elaphe 67 month temperature – CMT -19 °C, warm month temperature – WMT 20 °C, mean annual 68 temperature – MAT 0.4 °C; Semipalatinsk: CMT -16 °C, WMT 22 °C, MAT 3.1 °C; Lake Zaisan: 69 CMT up to -27 °C, WMT 23 °C; after Müller & Hennings (2000). The area is covered by diverse 70 biomes, namely the tundra (ʽcold steppeʼ) and taiga (coniferous forests) biomes, which are 71 replaced by open landscapes in the north (tundra) and in the south (steppe). Manuscript to be reviewed The results of our study showed higher-amplitude changes 53 of precipitation development in Western Siberia from the Early Miocene to the Pliocene, than 54 previously assumed. 55 56 Introduction 57 Western Siberia is a geographic region restricted to the territories of Russia and parts of 58 northern Kazakhstan. It includes the region between the Ural Mountains in the west, Central 59 Siberian Plateau in the east, and the Kazakh Plain and Altay Mountains, including the Zaisan 60 Lake in the south (Fig. 1). Western Siberia region incorporates the drainage basin of the major 61 Siberian rivers such as the Irtysh and Ob rivers, both flowing into the Kara Sea of the Arctic 62 Ocean. The region is characterised by a highly continental climate, under the influence of the 63 Westerlies (winds). The mean annual precipitation (MAP) is relatively uniform and varies from 64 400 mm in the north (415 mm at Omsk) to 200 mm in the south (255 mm at Pavlodar). The 65 region has a high relative humidity in summer due to labile convective heating and frequent 66 torrential rainfalls. The mean annual range of temperature reaches 4 °C and more (Omsk: cold PeerJ reviewing PDF \| (2016:08:12606:4:0:NEW 23 Jan 2017) Manuscript to be reviewed The region that 72 contains the studied Neogene outcrops is located in the transitional zone between the dry and 73 the more humid temperate biomes, where taiga, forest-steppe and steppe biomes are 74 distributed (Ravkin et al., 2008). 67 month temperature – CMT -19 °C, warm month temperature – WMT 20 °C, mean annual 68 temperature – MAT 0.4 °C; Semipalatinsk: CMT -16 °C, WMT 22 °C, MAT 3.1 °C; Lake Zaisan: 69 CMT up to -27 °C, WMT 23 °C; after Müller & Hennings (2000). The area is covered by diverse 70 biomes, namely the tundra (ʽcold steppeʼ) and taiga (coniferous forests) biomes, which are 71 replaced by open landscapes in the north (tundra) and in the south (steppe). The region that 72 contains the studied Neogene outcrops is located in the transitional zone between the dry and 73 the more humid temperate biomes, where taiga, forest-steppe and steppe biomes are 74 distributed (Ravkin et al., 2008). 75 Due to the strong continental climate, the present-day herpetofauna in the territory of Western 76 Siberia is comparatively far less diverse,represented only by six to ten amphibian species and 77 seven reptile species (Table 1). It is assumed that the present distribution of amphibians and 78 reptiles in Western Siberia was strongly influenced by Quaternary climatic fluctuation (Ravkin, 79 Bogomolova & Chesnokova, 2010). According to Borkin (Borkin, 1999), the present-day 80 amphibian fauna of Western Siberia belongs to the Siberian region of amphibian distribution in 81 the Palaearctic Realm. According to different authors (e.g. Kuzmin, 1995; Amphibiaweb, 2016), 82 the region is inhabited by few amphibians, namely two species of salamanders and four to eight 83 species of anurans, belonging to five genera and five families (Table 1). This is the poorest 84 regional diversity of fauna in the Palaearctic Realm, without any endemic species. Only 85 Salamandrella keyserlingii and Rana amurensis are characteristic of the territory, but they are 86 widely distributed and are also found in smaller areas in the neighbouring regions (Borkin, 87 1999). The Western Siberian reptile fauna listing includes few species: Natrix natrix, Elaphe 75 Due to the strong continental climate, the present-day herpetofauna in the territory of Western 76 Siberia is comparatively far less diverse,represented only by six to ten amphibian species and 77 seven reptile species (Table 1). Manuscript to be reviewed It is assumed that the present distribution of amphibians and 78 reptiles in Western Siberia was strongly influenced by Quaternary climatic fluctuation (Ravkin, 79 Bogomolova & Chesnokova, 2010). According to Borkin (Borkin, 1999), the present-day 80 amphibian fauna of Western Siberia belongs to the Siberian region of amphibian distribution in 81 the Palaearctic Realm. According to different authors (e.g. Kuzmin, 1995; Amphibiaweb, 2016), 82 the region is inhabited by few amphibians, namely two species of salamanders and four to eight 83 species of anurans, belonging to five genera and five families (Table 1). This is the poorest 84 regional diversity of fauna in the Palaearctic Realm, without any endemic species. Only 85 Salamandrella keyserlingii and Rana amurensis are characteristic of the territory, but they are 86 widely distributed and are also found in smaller areas in the neighbouring regions (Borkin, 87 1999). The Western Siberian reptile fauna listing includes few species: Natrix natrix, Elaphe PeerJ reviewing PDF \| (2016:08:12606:4:0:NEW 23 Jan 2017) Manuscript to be reviewed 88 dione, Vipera berus, Vipera renardi, Gloydius halys, Zootoca vivipara, Lacerta agilis and Eremias 89 arguta (Ananjeva et al., 2006; Ravkin, Bogomolova & Chesnokova, 2010). 90 91 Geology and stratigraphy 92 The Neogene sediments in Western Siberia have a wide distribution. Over many decades, 93 through systematic palaeontological studies and research in the Neogene and Quaternary 94 sediments of this area, rich fossil deposits of molluscan and small and large mammalian faunas 95 have been discovered (e.g. Zykin, 1979; Zykin & Zazhigin, 2008; Zykin, 2012). Based on the 96 studies of the small fossil mammals, the Neogene stratigraphy of the area is complemented 97 with biochronologic data. Continental sedimentation in the western part of the Siberian Plain 98 began in the Oligocene, after regression of the Turgai Strait in the Late Eocene, and continued 99 until the Quaternary period (e.g. Chkhikvadze, 1984, 1989; Tleuberdina et al., 1993; Malakhov, 100 2005). The sedimentary basin is surrounded by the Ural Mountains in the west, the Central 101 Kazakh Steppe and Altai-Sayan Mountains in the south, and the western margin of the Siberian 102 Plateau in the east. The surrounding regions deliver clastic material to the basin. Some 103 researchers include the Zaisan Basin, located to the west of the Altai-Sayan Mountains in this 104 territory (Borisov, 1963). The Neogene sediments are represented by lacustrine, fluvial, alluvial 105 and other continental depositions, overlying marine Eocene sediments. The thickest section 106 (300 m) of the Neogene and early Quaternary sediments occurs in the Omsk Basin. Neogene 88 dione, Vipera berus, Vipera renardi, Gloydius halys, Zootoca vivipara, Lacerta agilis and Eremias 89 arguta (Ananjeva et al., 2006; Ravkin, Bogomolova & Chesnokova, 2010). 91 Geology and stratigraphy 92 The Neogene sediments in Western Siberia have a wide distribution. Over many decades, 93 through systematic palaeontological studies and research in the Neogene and Quaternary 94 sediments of this area, rich fossil deposits of molluscan and small and large mammalian faunas 95 have been discovered (e.g. Zykin, 1979; Zykin & Zazhigin, 2008; Zykin, 2012). Based on the 96 studies of the small fossil mammals, the Neogene stratigraphy of the area is complemented 97 with biochronologic data. Continental sedimentation in the western part of the Siberian Plain 98 began in the Oligocene, after regression of the Turgai Strait in the Late Eocene, and continued 99 until the Quaternary period (e.g. Manuscript to be reviewed Gnibitenko, 2006; Gnibidenko et al., 2011), in which the recovered polarity 120 signals are combined with biochronologic data and correlated to the geomagnetic polarity time 121 scale (GPTS) (Fejfar et al., 1997; Vangengeim, Pevzner & Tesakov, 2005; Zykin, Zykina & 122 Zazhigin, 2007). The biozonation is based on fast-evolving lineages of small mammals, mainly 123 jerboas (Dipodidae), hamsters (Cricetidae) and voles (Arvicolidae). Owing to these bio- 124 magnetostratigraphic data the mean temporal resolution of the late Neogene faunal record 125 from the Ob-Irtysh Interfluve is estimated to be approximately 200 kyr (Fig. 2, Table S1, Data 126 S2). The main sections of these vertebrate fossil localities are referred to certain svitas (e.g. 127 Kalkaman, Pavlodar, Irtysh Svitas), however, the stratigraphic assignment of three localities 128 Olkhovka 1A, 1B, 1C to svitas is not available (Fig. 2, Table S1). No fossils are available in the 129 initial deposits of the early Late Miocene. Manuscript to be reviewed Chkhikvadze, 1984, 1989; Tleuberdina et al., 1993; Malakhov, 100 2005). The sedimentary basin is surrounded by the Ural Mountains in the west, the Central 101 Kazakh Steppe and Altai-Sayan Mountains in the south, and the western margin of the Siberian 102 Plateau in the east. The surrounding regions deliver clastic material to the basin. Some 103 researchers include the Zaisan Basin, located to the west of the Altai-Sayan Mountains in this PeerJ reviewing PDF \| (2016:08:12606:4:0:NEW 23 Jan 2017) 131 State-of-art in palaeoherpetological studies in Western Siberia 131 State-of-art in palaeoherpetological studies in Western Siberia Manuscript to be reviewed 110 sediments are found in several areas along the Irtysh River and its tributaries – Petropavlovsk- 111 Ishim (e.g. Petropavlovsk 1, Biteke 1A), Omsk (e.g. Novaya Stanitsa 1, Cherlak), Pavlodar (e.g. 112 Pavlodar, Baikadam) and the Novosibirsk areas (e.g. Kamen-na-Obi) (Fig. 1). Detailed geological 113 descriptions of the stratigraphic sections and fossil localities are summarised in Zykin (1979); 114 Zykin & Zazhigin (2004); Gnibitenko (2006); Zykin (2012). 110 sediments are found in several areas along the Irtysh River and its tributaries – Petropavlovsk- 111 Ishim (e.g. Petropavlovsk 1, Biteke 1A), Omsk (e.g. Novaya Stanitsa 1, Cherlak), Pavlodar (e.g. 112 Pavlodar, Baikadam) and the Novosibirsk areas (e.g. Kamen-na-Obi) (Fig. 1). Detailed geological 113 descriptions of the stratigraphic sections and fossil localities are summarised in Zykin (1979); 114 Zykin & Zazhigin (2004); Gnibitenko (2006); Zykin (2012). 115 The stratigraphic subdivision is based mainly on the Russian concept of svitas. A svita has 116 lithologic, biochronologic and genetic (sedimentologic) significance and has no precise 117 equivalent in Western stratigraphic theory and terminology (Lucas et al., 2012). The 118 stratigraphy of Neogene sediments in Western Siberia is supported by magnetostratigraphic 119 investigations (e.g. Gnibitenko, 2006; Gnibidenko et al., 2011), in which the recovered polarity 120 signals are combined with biochronologic data and correlated to the geomagnetic polarity time 121 scale (GPTS) (Fejfar et al., 1997; Vangengeim, Pevzner & Tesakov, 2005; Zykin, Zykina & 122 Zazhigin, 2007). The biozonation is based on fast-evolving lineages of small mammals, mainly 123 jerboas (Dipodidae), hamsters (Cricetidae) and voles (Arvicolidae). Owing to these bio- 124 magnetostratigraphic data the mean temporal resolution of the late Neogene faunal record 125 from the Ob-Irtysh Interfluve is estimated to be approximately 200 kyr (Fig. 2, Table S1, Data 126 S2). The main sections of these vertebrate fossil localities are referred to certain svitas (e.g. 127 Kalkaman, Pavlodar, Irtysh Svitas), however, the stratigraphic assignment of three localities 128 Olkhovka 1A, 1B, 1C to svitas is not available (Fig. 2, Table S1). No fossils are available in the 129 initial deposits of the early Late Miocene. 115 The stratigraphic subdivision is based mainly on the Russian concept of svitas. A svita has 116 lithologic, biochronologic and genetic (sedimentologic) significance and has no precise 117 equivalent in Western stratigraphic theory and terminology (Lucas et al., 2012). The 118 stratigraphy of Neogene sediments in Western Siberia is supported by magnetostratigraphic 119 investigations (e.g. Manuscript to be reviewed 132 The fossil record of amphibians and reptiles in Western Siberia, including the Zaisan Basin 133 record, remain largely unknown. There are very few works devoted to the studies of the 134 Western Siberian late Paleogene and Neogene herpetofaunal assemblages (e.g. Chkhikvadze, 135 1984, 1989; Tleuberdina et al., 1993; Malakhov, 2005). The vast majority of data on fossil 136 amphibians and reptiles are represented as short notes or are mentioned in faunal lists (e.g. 137 Bendukidze & Chkhikvadze, 1976; Chkhikvadze, 1985; Malakhov, 2005). In this present 138 contribution we analysed the available data from specimens described below and from new 139 generated data as well. 132 The fossil record of amphibians and reptiles in Western Siberia, including the Zaisan Basin 133 record, remain largely unknown. There are very few works devoted to the studies of the 134 Western Siberian late Paleogene and Neogene herpetofaunal assemblages (e.g. Chkhikvadze, 135 1984, 1989; Tleuberdina et al., 1993; Malakhov, 2005). The vast majority of data on fossil 136 amphibians and reptiles are represented as short notes or are mentioned in faunal lists (e.g. 137 Bendukidze & Chkhikvadze, 1976; Chkhikvadze, 1985; Malakhov, 2005). In this present 138 contribution we analysed the available data from specimens described below and from new 139 generated data as well. 140 The earliest report on Neogene fossil amphibians was compiled by Iskakova (1969), wherein 141 she described amphibian faunas from two Priirtyshian localities, Gusiniy Perelet and Karashigar. 142 Gusiniy Perelet is a well-renovated Late Miocene vertebrate fossil locality, situated on the 143 riverbank of the Irtysh River, within the town of Pavlodar. The sedimentary sequence in this 144 locality contains layers of different ages from the late Late Miocene until the late Early Pliocene. 145 Three localities (also ‘horizons’) within the town of Pavlodar (Pavlodar 1A, 1B, 3B) are grouped 146 into several svitas and can be distinguished from the Gusiniy Perelet vertebrate locality. The 147 fossil content of the Gusiniy Perelet locality comes from the lower horizon – Pavlodar 1A. 148 Iskakova (1969) described an amphibian fauna from this layer. 140 The earliest report on Neogene fossil amphibians was compiled by Iskakova (1969), wherein 141 she described amphibian faunas from two Priirtyshian localities, Gusiniy Perelet and Karashigar. 142 Gusiniy Perelet is a well-renovated Late Miocene vertebrate fossil locality, situated on the 143 riverbank of the Irtysh River, within the town of Pavlodar. Manuscript to be reviewed 154 temporaria) were identified based mainly on the vertebrae (cervical, dorsal and sacral) 155 morphology, which is not diagnostic in frogs at that taxonomic level. Chkhikvadze (1984) 156 restudied the material from the Pavlodar 1A (= Gusiniy Perelet) locality and identified Bufo cf. 157 raddei, Bufo sp., Pelophylax cf. ridibundus, Eremias sp., and Coluber sp. In this study, we did not, 158 however; assess the material from the above-mentioned works in order to verify Chkhikvadze 159 (1984) taxonomic identifications. Our sample from this locality (Pavlodar 1A) (Table S1), did not 160 reveal any element listed in these earlier studies (Chkhikvadze, 1984). 161 Chkhikvadze (1984) summarised all known fossil amphibians and reptiles from the former 162 Union of Soviet Socialist Republics (USSR), including those from Western Siberia. Accurate 163 descriptions are not yet available for many of these species. The Middle Miocene Kalkaman 164 locality (Tleuberdina, 1993), presently known as Malyi Kalkaman 1 (Zykin, 2012), has provided a 165 diverse record of fossil herpetofauna. The fossil record of this locality was partially restudied 166 and amended by us, which included the collection of new material. 167 Over the last decade, fresh attempts has been made to study the herpetofauna from the 168 Western Siberian localities (Malakhov, 2003, 2004, 2005, 2009). In the resultant works, 169 undescribed material from several Neogene localities of Kazakhstan were summarised, revised, 154 temporaria) were identified based mainly on the vertebrae (cervical, dorsal and sacral) 155 morphology, which is not diagnostic in frogs at that taxonomic level. Chkhikvadze (1984) 156 restudied the material from the Pavlodar 1A (= Gusiniy Perelet) locality and identified Bufo cf. 157 raddei, Bufo sp., Pelophylax cf. ridibundus, Eremias sp., and Coluber sp. In this study, we did not, 158 however; assess the material from the above-mentioned works in order to verify Chkhikvadze 159 (1984) taxonomic identifications. Our sample from this locality (Pavlodar 1A) (Table S1), did not 160 reveal any element listed in these earlier studies (Chkhikvadze, 1984). 154 temporaria) were identified based mainly on the vertebrae (cervical, dorsal and sacral) 155 morphology, which is not diagnostic in frogs at that taxonomic level. Chkhikvadze (1984) 156 restudied the material from the Pavlodar 1A (= Gusiniy Perelet) locality and identified Bufo cf. 157 raddei, Bufo sp., Pelophylax cf. ridibundus, Eremias sp., and Coluber sp. Manuscript to be reviewed The sedimentary sequence in this 144 locality contains layers of different ages from the late Late Miocene until the late Early Pliocene. 145 Three localities (also ‘horizons’) within the town of Pavlodar (Pavlodar 1A, 1B, 3B) are grouped 146 into several svitas and can be distinguished from the Gusiniy Perelet vertebrate locality. The 147 fossil content of the Gusiniy Perelet locality comes from the lower horizon – Pavlodar 1A. 148 Iskakova (1969) described an amphibian fauna from this layer. 149 The age of the Karashigar locality is unclear. In a study by Tleuberdina et al. (1989), this locality 150 has been estimated to date back to the Late Oligocene; however, Lychev (1990) placed it in the 151 Middle Miocene, Kalkaman Svita (the list of the small mammal fauna; see Data S2). The 152 amphibian taxa described by Iskakova (1969) in the Priirtyshian localities (Bombina cf. 153 bombina, Pelobates cf. fuscus, Bufo cf. viridis, Bufo cf. bufo, Rana cf. ridibunda, Rana cf. PeerJ reviewing PDF \| (2016:08:12606:4:0:NEW 23 Jan 2017) PeerJ reviewing PDF \| (2016:08:12606:4:0:NEW 23 Jan 2017) Manuscript to be reviewed In this study, we did not, 158 however; assess the material from the above-mentioned works in order to verify Chkhikvadze 159 (1984) taxonomic identifications. Our sample from this locality (Pavlodar 1A) (Table S1), did not 160 reveal any element listed in these earlier studies (Chkhikvadze, 1984). 161 Chkhikvadze (1984) summarised all known fossil amphibians and reptiles from the former 162 Union of Soviet Socialist Republics (USSR), including those from Western Siberia. Accurate 163 descriptions are not yet available for many of these species. The Middle Miocene Kalkaman 164 locality (Tleuberdina, 1993), presently known as Malyi Kalkaman 1 (Zykin, 2012), has provided a 165 diverse record of fossil herpetofauna. The fossil record of this locality was partially restudied 166 and amended by us, which included the collection of new material. 167 Over the last decade, fresh attempts has been made to study the herpetofauna from the 168 Western Siberian localities (Malakhov, 2003, 2004, 2005, 2009). In the resultant works, 169 undescribed material from several Neogene localities of Kazakhstan were summarised, revised, 170 and studied, thereby providing critical overviews. In spite of the advances of the recent years, 171 however, the Neogene herpetofauna from Western Asia remains largely unknown, with 172 available fossil material continuing to be insufficiently studied. The main goals of the present 173 study were, therefore, to assess the descriptions and taxonomic classifications of the new 174 amphibian and reptile fossil material collected by Vladimir Zazhigin (co-author), as well as 175 already published data so as to provide a comprehensive faunistic analysis and PeerJ reviewing PDF \| (2016:08:12606:4:0:NEW 23 Jan 2017) Manuscript to be reviewed 176 palaeobiogeographic and environmental interpretations. To avoid confusion around the nam 177 used by different authors in the Russian literature to describe the localities, we have provide 178 all known names for these studied fossil localities. 179 180 Materials & Methods 181 The new materials used in the present study were collected by V. Zazhigin (co-author) using t 182 screen-washing technique during his long-term excavations in different Western Siberian 183 localities from the 1960’s to 2008. These localities outcrop along the riverbanks of the Irtysh, 184 Ishim and Ob rivers. This fossil material is stored in the Institute of Geology, Russian Academ 185 of Sciences under the collection numbers: 186 GIN 950/2001 (Baikadam), GIN 1107/1001 (Malyi Kalkaman 1), GIN 1107/2001 (Malyi Kalkam 187 2), GIN 1106/1001 (Shet Irgyz 1) GIN 952/1001 (Petropavlovsk 1), GIN 1109/1001 (Znamenka 188 GIN 640/5001 (Pavlodar 1A), GIN 951/1001 (Selety 1A), GIN 951/2001 (Kedey), GIN 948/2001 189 (Novaya Stanitsa 1A), GIN 1115/1001 (Borki 1A), GIN 1110/2001 (Cherlak), GIN 945/2001 190 (Beteke 1A), GIN 640/6001 (Pavlodar 1B), GIN 1130/1001 (Lezhanka 2A), GIN 1130/2001 191 (Lezhanka 2B), GIN 1111/1001 (Olkhovka 1A), GIN 1111/2001 (Olkhovka 1B), GIN 1111/3001 192 (Olkhovka 1C), GIN 1118/3001 (Peshniovo 3), GIN 1131/2001 (Isakovka 2), GIN 1131/1001 193 (Isakovka 1A), GIN 1131/3001 (Isakovka 1B), GIN 1117/1001 (Kamyshlovo), GIN 945/2001 194 (Beteke 1B), GIN 945/3001 (Beteke 1C), GIN 1112/1001 (Andreievka – Speranskoe), GIN 195 1108/2001 (Pavlodar 2B), GIN 1112/2001 (Andreievka 1), GIN 1129/2001 (Livenka), GIN 196 1129/1001 (Lezhanka 1), GIN 1108/3001 (Pavlodar 3A), GIN 950/3001 (Lebiazhie 1A), GIN 176 palaeobiogeographic and environmental interpretations. To avoid confusion around the names 177 used by different authors in the Russian literature to describe the localities, we have provided 178 all known names for these studied fossil localities. 176 palaeobiogeographic and environmental interpretations. To avoid confusion around the names 177 used by different authors in the Russian literature to describe the localities, we have provided 178 all known names for these studied fossil localities. 176 palaeobiogeographic and environmental interpretations. To avoid confusion around the names 177 used by different authors in the Russian literature to describe the localities, we have provided 178 all known names for these studied fossil localities. 179 180 Materials & Methods 181 The new materials used in the present study were collected by V. Zazhigin (co-author) using the 182 screen-washing technique during his long-term excavations in different Western Siberian 183 localities from the 1960’s to 2008. Manuscript to be reviewed 197 950/4001 (Lebiazhie 1B), GIN 950/5001 (Podpusk 1), GIN 945/60001 (Beteke 2), GIN 946/2001 198 (Kamen-na-Obi), GIN 945/8001 (Beteke 4), GIN 664/2001 (Razdole). 197 950/4001 (Lebiazhie 1B), GIN 950/5001 (Podpusk 1), GIN 945/60001 (Beteke 2), GIN 946/2001 198 (Kamen-na-Obi), GIN 945/8001 (Beteke 4), GIN 664/2001 (Razdole). 199 Various groups of amphibians and reptiles are represented in the available material. A report of 200 part of this material, i.e. of the snakes and anguine lizards, has been published in a separate 201 paper (e.g. Vasilyan, Böhme & Klembara, 2016). The present study included an assessment of 202 the materials collected from four fossil sites in Kazakhstan: Akyspe (also known as Agyspe), Aral 203 Horizon, leg. by Bendukidze in 1977; Kentyubek, Turgai Basin; Ryzhaya II (Ryzhaya Sopka), 204 Zaisan Svita, Zaisan Basin, leg. in 1970; Ayakoz (known also as Ayaguz), Zaisan Basin, leg. in 205 1970-1971; Petropavlovsk 1/21, leg. 1972 (Table S1). In addition, the few available data from 206 the literature were included in this study (after critical revision) to amend the record of 207 herpetofaunal assemblages of some localities as well as to reassign and revise the stratigraphic 208 position of these localities using biochronologic information of small and large mammalian 209 fauna (see full list in the Datas S2, S3). 210 The photographs of the fossil material were taken using a digital microscope Leica DVM5000 211 (Tübingen) and inspected with a scanning electron microscope, FEI Inspect S (Madrid). The 212 figures and tables were produced using Adobe Photoshop and Illustrator programs. The 213 osteological nomenclature of this study followed that of Vasilyan et al. (2013) for the 214 salamander remains, that of Sanchíz (1998a) for frogs, that of Daza, Aurich & Bauer (2011) and 215 Daza & Bauer (2010) was used for Gekkota, and the lepidosaurian terminology of Evans (2008). 1 In the town of Petropavlovsk two fossil sites (Petropavlovsk 1 (MN12) and Petropavlovsk 2 (MN14)) having different ages are known, see Zykin (2012). Since the enclosed collection label to the material indicates only ʻlocality Petropavlovsk, leg. 1972ʼ any stratigraphic allocation of the fossils to one of those layers is impossible. 1 In the town of Petropavlovsk two fossil sites (Petropavlovsk 1 (MN12) and Petropavlovsk 2 (MN14)) having different ages are known, see Zykin (2012). Since the enclosed collection label to the material indicates only ʻlocality Petropavlovsk, leg. Manuscript to be reviewed These localities outcrop along the riverbanks of the Irtysh, 184 Ishim and Ob rivers. This fossil material is stored in the Institute of Geology, Russian Academy 185 of Sciences under the collection numbers: 179 180 Materials & Methods 181 The new materials used in the present study were collected by V. Zazhigin (co-author) using the 182 screen-washing technique during his long-term excavations in different Western Siberian 183 localities from the 1960’s to 2008. These localities outcrop along the riverbanks of the Irtysh, 184 Ishim and Ob rivers. This fossil material is stored in the Institute of Geology, Russian Academy 185 of Sciences under the collection numbers: 186 GIN 950/2001 (Baikadam), GIN 1107/1001 (Malyi Kalkaman 1), GIN 1107/2001 (Malyi Kalkaman 187 2), GIN 1106/1001 (Shet Irgyz 1) GIN 952/1001 (Petropavlovsk 1), GIN 1109/1001 (Znamenka), 188 GIN 640/5001 (Pavlodar 1A), GIN 951/1001 (Selety 1A), GIN 951/2001 (Kedey), GIN 948/2001 189 (Novaya Stanitsa 1A), GIN 1115/1001 (Borki 1A), GIN 1110/2001 (Cherlak), GIN 945/2001 190 (Beteke 1A), GIN 640/6001 (Pavlodar 1B), GIN 1130/1001 (Lezhanka 2A), GIN 1130/2001 191 (Lezhanka 2B), GIN 1111/1001 (Olkhovka 1A), GIN 1111/2001 (Olkhovka 1B), GIN 1111/3001 192 (Olkhovka 1C), GIN 1118/3001 (Peshniovo 3), GIN 1131/2001 (Isakovka 2), GIN 1131/1001 193 (Isakovka 1A), GIN 1131/3001 (Isakovka 1B), GIN 1117/1001 (Kamyshlovo), GIN 945/2001 194 (Beteke 1B), GIN 945/3001 (Beteke 1C), GIN 1112/1001 (Andreievka – Speranskoe), GIN 195 1108/2001 (Pavlodar 2B), GIN 1112/2001 (Andreievka 1), GIN 1129/2001 (Livenka), GIN 196 1129/1001 (Lezhanka 1), GIN 1108/3001 (Pavlodar 3A), GIN 950/3001 (Lebiazhie 1A), GIN PeerJ reviewing PDF \| (2016:08:12606:4:0:NEW 23 Jan 2017) Manuscript to be reviewed For the localities considered to be ʻpoorʼ in 219 amphibian and reptile taxa, the range-through approach (Barry et al., 2002) was used, in which 220 the faunas of two or more localities with age differences less than 100 kyr and/or belonging to 221 a single stratigraphic unit – svita, were considered as one. The taxa, that were added to the 222 herpetofaunal assemblage using the range-through approach, are indicated in grey in Table S4. 223 Manuscript to be reviewed 1972ʼ any stratigraphic allocation of the fossils to one of those layers is impossible. PeerJ reviewing PDF \| (2016:08:12606:4:0:NEW 23 Jan 2017) Manuscript to be reviewed 216 Based on the herpetofaunal assemblages, the palaeoprecipitation values for the fossil localities 217 were estimated using the method of bioclimatic analysis of the ecophysiologic groups of 218 amphibian and reptile taxa (Böhme et al., 2006). For the localities considered to be ʻpoorʼ in 219 amphibian and reptile taxa, the range-through approach (Barry et al., 2002) was used, in which 220 the faunas of two or more localities with age differences less than 100 kyr and/or belonging to 221 a single stratigraphic unit – svita, were considered as one. The taxa, that were added to the 222 herpetofaunal assemblage using the range-through approach, are indicated in grey in Table S4. 223 224 Institutional/collection abbreviations. 225 GPIT: Paläontologische Sammlung der Universität Tübingen, Tübingen, Germany; HC: collection 226 of Marcela Hodrova (Prague University), now stored in GPIT; MNCN: Museo Nacional de 227 Ciencias Naturales, Madrid, Spain; NMNHK: National Museum of Natural History, Kiev, Ukraine; 228 PIN: Palaeontological Institute, Russian Academy of Sciences, Moscow, Russia; GNM: National 229 Museum of Georgia, Tbilisi, Georgia; GIN: Geologic Institute, National Academy of Russia, 230 Moscow, Russia. 231 232 Anatomical abbreviations. 233 ao: antrum olfactorium; alo: antrum pro lobo olfactorio; dl: dental lamina; ds: dental shelf; hl: 234 horizontal lamella; is: incisura semielliptical; ff: frontoparietal facet; fcpr: facial process of 216 Based on the herpetofaunal assemblages, the palaeoprecipitation values for the fossil localities 217 were estimated using the method of bioclimatic analysis of the ecophysiologic groups of 218 amphibian and reptile taxa (Böhme et al., 2006). For the localities considered to be ʻpoorʼ in 219 amphibian and reptile taxa, the range-through approach (Barry et al., 2002) was used, in which 220 the faunas of two or more localities with age differences less than 100 kyr and/or belonging to 221 a single stratigraphic unit – svita, were considered as one. The taxa, that were added to the 222 herpetofaunal assemblage using the range-through approach, are indicated in grey in Table S4. 223 216 Based on the herpetofaunal assemblages, the palaeoprecipitation values for the fossil localities 217 were estimated using the method of bioclimatic analysis of the ecophysiologic groups of 218 amphibian and reptile taxa (Böhme et al., 2006). 224 Institutional/collection abbreviations. 225 GPIT: Paläontologische Sammlung der Universität Tübingen, Tübingen, Germany; HC: collection 226 of Marcela Hodrova (Prague University), now stored in GPIT; MNCN: Museo Nacional de 227 Ciencias Naturales, Madrid, Spain; NMNHK: National Museum of Natural History, Kiev, Ukraine; 228 PIN: Palaeontological Institute, Russian Academy of Sciences, Moscow, Russia; GNM: National 229 Museum of Georgia, Tbilisi, Georgia; GIN: Geologic Institute, National Academy of Russia, 230 Moscow, Russia. 225 GPIT: Paläontologische Sammlung der Universität Tübingen, Tübingen, Germany; HC: collection 226 of Marcela Hodrova (Prague University), now stored in GPIT; MNCN: Museo Nacional de 227 Ciencias Naturales, Madrid, Spain; NMNHK: National Museum of Natural History, Kiev, Ukraine; 228 PIN: Palaeontological Institute, Russian Academy of Sciences, Moscow, Russia; GNM: National 229 Museum of Georgia, Tbilisi, Georgia; GIN: Geologic Institute, National Academy of Russia, 230 Moscow, Russia. 231 232 Anatomical abbreviations. 233 ao: antrum olfactorium; alo: antrum pro lobo olfactorio; dl: dental lamina; ds: dental shelf; hl: 234 horizontal lamella; is: incisura semielliptical; ff: frontoparietal facet; fcpr: facial process of 235 maxilla; fMx5: foramina for mandibular division of the fifth cranial (trigeminal) nerve; hfr: 236 haemal foramen; hl: horizontal lamella; lf: lacrimal facet; lg: longitudinal groove; lh: lamina 237 horizontalis; lp: lateral processes; ls: lamina supraorbitalis; mc: Meckelian canal; na: neural 232 Anatomical abbreviations. PeerJ reviewing PDF \| (2016:08:12606:4:0:NEW 23 Jan 2017) Manuscript to be reviewed Manuscript to be reviewed 238 arch; nc: neural canal; nf: nasal facet; on 239 parasphenoid facet; pfc: palatine facet; 240 postzygapophysis; pv: paries verticalis; p 241 opening of superior alveolar canal; sg: s 242 process. 243 244 RESULTS 245 Systematic palaeontology 246 Class Amphibia Gray, 1825 247 Order Caudata Scopoli, 1777 248 Family Hynobiidae Cope, 1859 249 Genus Salamandrella Dybowski, 1870 250 Salamandrella sp. 251 252 (Fig. 3D-3G) 253 Localities and material examined. Maly 254 Selety 1A, GIN 951/1001-AM01 – -AM03 255 GIN 951/1001-AM05, 1 distal end of bo 256 AM01 – -AM11, 11 trunk vertebrae; Lez 257 GIN 1130/1001-AM27 – -AM28, 2 cauda 258 trunk vertebrae; Lezhanka 2B, GIN 1130 259 extremity bone; Olkhovka 1B, GIN 1111/ 238 arch; nc: neural canal; nf: nasal facet; onf: orbitonasal foramina; olf: olfactory foramina; pf: 239 parasphenoid facet; pfc: palatine facet; ph: paries horizontalis; prz: prezygapophysis; psz: 240 postzygapophysis; pv: paries verticalis; pxp: premaxillary process; pyp: pterygapophysis; sac: 241 opening of superior alveolar canal; sg: symphyseal groove; sf: splenial facet; tpr: transverse 242 process. 243 244 RESULTS 245 Systematic palaeontology 246 Class Amphibia Gray, 1825 247 Order Caudata Scopoli, 1777 248 Family Hynobiidae Cope, 1859 249 Genus Salamandrella Dybowski, 1870 250 Salamandrella sp. 251 252 (Fig. 3D-3G) 253 Localities and material examined. Malyi Kalkamana 1, GIN 1107/1001-AM12, 1 right femur; 254 Selety 1A, GIN 951/1001-AM01 – -AM03, 3 trunk and GIN 951/1001-AM04, 1 caudal vertebra; 255 GIN 951/1001-AM05, 1 distal end of bone (humerus?); Novaya Stanitsa 1A, GIN 948/2001- 256 AM01 – -AM11, 11 trunk vertebrae; Lezhanka 2A, GIN 1130/1001-AM01 – -AM26, 26 trunk and 257 GIN 1130/1001-AM27 – -AM28, 2 caudal vertebrae; Cherlak, GIN 1110/2001-AM01 – -AM12, 12 238 arch; nc: neural canal; nf: nasal facet; onf: orbitonasal foramina; olf: olfactory foramina; pf: 239 parasphenoid facet; pfc: palatine facet; ph: paries horizontalis; prz: prezygapophysis; psz: 240 postzygapophysis; pv: paries verticalis; pxp: premaxillary process; pyp: pterygapophysis; sac: 241 opening of superior alveolar canal; sg: symphyseal groove; sf: splenial facet; tpr: transverse 242 process. 242 process. 243 244 RESULTS 245 Systematic palaeontology 246 Class Amphibia Gray, 1825 247 Order Caudata Scopoli, 1777 248 Family Hynobiidae Cope, 1859 249 Genus Salamandrella Dybowski, 1870 250 Salamandrella sp. 251 252 (Fig. 3D-3G) 253 Localities and material examined. Manuscript to be reviewed 260 1131/2001-AM01, 1 trunk vertebra; Andreievka – Speransko, GIN 1112/1001-AM01, 1 trunk 261 vertebra; Lezhanka 1, GIN 1129/1001-AM01 – -AM02, 2 trunk and GIN 1129/1001-AM03, 1 262 caudal vertebrae; Beteke 1C, GIN 945/3001-AM01 – -AM02, 2 trunk vertebrae. 263 Description and comments. The vertebrae have an elongated to nearly slender form. The 264 vertebral centrum is amphicoelous. The basapophyses at the vertebral centrum are either 265 absent or are present in the form of a small protuberance at the laterodorsal corners of the 266 anterior portion of the vertebral centrum (Fig. 3G). A pair of subcentral foramina is situated at 267 the basis of the transverse processes. The neural arch is tall in lateral view (Fig. 3F) and 268 relatively broad in dorsal view (Fig. 3D). The posterior edge of the pterygapophysis is 269 bifurcated. Sometimes the neural spine is present but in general the dorsal surface of the 270 neural arch is flat. The pre- and postzygapophyses have an elongated oval shape. In anterior 271 view, the neural canal has an outline of a regular pentagon. The transverse process is 272 unicapitate (Fig. 3D, 3G). The anterior and posterior alar processes are absent. The vertebrae 273 can be assigned to the family Hynobiidae based on: (1) the small size and their amphicoelous 274 centrum with circular articular surfaces; (2) the lack of or being weakly pronounced 275 basapophyses; (3) the lack of neural spine; (4) the notch on the posterior margin of neural arch 276 (5) the fused rib-bearers; and (6) the intervertebrally exiting spinal nerve in both trunk and 277 caudal vertebrae (e.g. Edwards, 1976; Venczel, 1999a, 1999b). Further, characteristic features 278 can be observed on the vertebrae of representatives of the genus Salamandrella, namely the 279 absence of the subcentral foramen and the concave anterior margin of the neural arch that 280 reaches the middle part of the prezygapophyseal articular facets (Venczel, 1999b; Ratnikov & 281 Litvinchuk, 2009; Syromyatnikova, 2014) (Fig. 3D-3G). The detailed description of hynobiid p 260 1131/2001-AM01, 1 trunk vertebra; Andreievka – Speransko, GIN 1112/1001-AM01, 1 trunk 261 vertebra; Lezhanka 1, GIN 1129/1001-AM01 – -AM02, 2 trunk and GIN 1129/1001-AM03, 1 262 caudal vertebrae; Beteke 1C, GIN 945/3001-AM01 – -AM02, 2 trunk vertebrae. Manuscript to be reviewed Malyi Kalkamana 1, GIN 1107/1001-AM12, 1 right femur; 254 Selety 1A, GIN 951/1001-AM01 – -AM03, 3 trunk and GIN 951/1001-AM04, 1 caudal vertebra; 255 GIN 951/1001-AM05, 1 distal end of bone (humerus?); Novaya Stanitsa 1A, GIN 948/2001- 256 AM01 – -AM11, 11 trunk vertebrae; Lezhanka 2A, GIN 1130/1001-AM01 – -AM26, 26 trunk and 257 GIN 1130/1001-AM27 – -AM28, 2 caudal vertebrae; Cherlak, GIN 1110/2001-AM01 – -AM12, 12 258 trunk vertebrae; Lezhanka 2B, GIN 1130/2001-AM01, 1 trunk vertebra, GIN 1130/2001-AM02, 1 259 extremity bone; Olkhovka 1B, GIN 1111/2001-AM01, 1 trunk vertebra; Iskakovka 2A, GIN 253 Localities and material examined. Malyi Kalkamana 1, GIN 1107/1001-AM12, 1 right femur; 254 Selety 1A, GIN 951/1001-AM01 – -AM03, 3 trunk and GIN 951/1001-AM04, 1 caudal vertebra; 255 GIN 951/1001-AM05, 1 distal end of bone (humerus?); Novaya Stanitsa 1A, GIN 948/2001- 256 AM01 – -AM11, 11 trunk vertebrae; Lezhanka 2A, GIN 1130/1001-AM01 – -AM26, 26 trunk and 257 GIN 1130/1001-AM27 – -AM28, 2 caudal vertebrae; Cherlak, GIN 1110/2001-AM01 – -AM12, 12 258 trunk vertebrae; Lezhanka 2B, GIN 1130/2001-AM01, 1 trunk vertebra, GIN 1130/2001-AM02, 1 259 extremity bone; Olkhovka 1B, GIN 1111/2001-AM01, 1 trunk vertebra; Iskakovka 2A, GIN PeerJ reviewing PDF \| (2016:08:12606:4:0:NEW 23 Jan 2017) Manuscript to be reviewed 282 material from the Western Siberian localities and comparison with recent and fossil hynobiids 283 provided in a forthcoming paper. 284 285 Family Cryptobranchidae Fitzinger, 1826 286 Cryptobranchidae indet. 287 288 (Fig. 3A-3C) 289 Localities and material examined. Pavlodar 1A (=Gusiniy Perelet), 1 fragmentary right dentar 290 and 2 fragments of jaw bones, for details about the stratigraphic allocation see the section 291 ‘Cryptobranchidae’, unnr. PIN specimen. 292 Description and comments. Among the fragments, a posterodorsal portion of a large right 293 dentary, 27 mm in length, is present. In lingual view, the pars dentalis is composed entirely of 294 dental lamina and the subdental lamina is present, but reduced. The pars dentalis possesses 3 295 pedicels of pleurodont teeth. The subdental shelf inclines slightly ventrally. The lamina 296 horizontalis is prominent. The corpus dentalis above the Meckelian groove has a concave 297 surface. Ventrally, this surface possesses a ridge running parallel to the lamina horzontalis. Th 298 cross section of the dentary shows a relatively low portion of cancellous bone and a dominan 299 of compact bone. The size of the bones, the form and structure of the pars dentalis and the 300 cross section of the bone are characteristic of giant salamanders (Vasilyan et al., 2013). 301 302 Family Proteidae Gray, 1825 303 Genus Mioproteus Estes & Darevsky, 1977 282 material from the Western Siberian localities and comparison with recent and fossil hynobiids is 283 provided in a forthcoming paper. 289 Localities and material examined. Pavlodar 1A (=Gusiniy Perelet), 1 fragmentary right dentary 290 and 2 fragments of jaw bones, for details about the stratigraphic allocation see the section 291 ‘Cryptobranchidae’, unnr. PIN specimen. 289 Localities and material examined. Pavlodar 1A (=Gusiniy Perelet), 1 fragmentary right dentary 290 and 2 fragments of jaw bones, for details about the stratigraphic allocation see the section 291 ‘Cryptobranchidae’, unnr. PIN specimen. 292 Description and comments. Among the fragments, a posterodorsal portion of a large right 293 dentary, 27 mm in length, is present. In lingual view, the pars dentalis is composed entirely of 294 dental lamina and the subdental lamina is present, but reduced. The pars dentalis possesses 30 295 pedicels of pleurodont teeth. The subdental shelf inclines slightly ventrally. The lamina 296 horizontalis is prominent. The corpus dentalis above the Meckelian groove has a concave 297 surface. Ventrally, this surface possesses a ridge running parallel to the lamina horzontalis. Manuscript to be reviewed 260 1131/2001-AM01, 1 trunk vertebra; Andreievka – Speransko, GIN 1112/1001-AM01, 1 trunk 261 vertebra; Lezhanka 1, GIN 1129/1001-AM01 – -AM02, 2 trunk and GIN 1129/1001-AM03, 1 262 caudal vertebrae; Beteke 1C, GIN 945/3001-AM01 – -AM02, 2 trunk vertebrae. PeerJ reviewing PDF \| (2016:08:12606:4:0:NEW 23 Jan 2017) Manuscript to be reviewed The 298 cross section of the dentary shows a relatively low portion of cancellous bone and a dominance 299 of compact bone. The size of the bones, the form and structure of the pars dentalis and the 300 cross section of the bone are characteristic of giant salamanders (Vasilyan et al., 2013). 301 302 Family Proteidae Gray, 1825 303 Genus Mioproteus Estes & Darevsky, 1977 Manuscript to be reviewed The pars dentalis of the premaxilla 314 is located on the anterior side of the bone. The crowns of pleurodont teeth are missing and 315 only their pedicellar portions are preserved. The bone surface is slightly rough in dorsal view. 316 The lamelliform anterolateral ridge of the posterior process is high at the middle part of the 317 bone. The amphicoelous vertebrae are flat and wide. The centrum is dumb-bell in shape and 318 narrows to the middle region. The basapophyses, if present, are small and weakly developed. 319 Two subcentral foramina are present at the central part of the vertebral centrum. In lateral 320 view, the vertebra is low; the anterior and posterior zygapophyseal crests are pointed, forming 321 the dorsal border of the deep depressions anteriorly and posteriorly to the transverse process. 322 The middle part of the neural arch is lower than its cranial and caudal margins. The posterior 323 edge of the neural arch is forked (Fig. 3H) (not visible at Fig. 3M). The neural spine extends as 324 far as the preserved anterior margins of the neural arch, whereas posteriorly, it terminates 312 Description. The preserved left premaxilla is fragmentary (Fig. 3P-3Q) and the posterior process 313 is broken off. In ventral view, the bone has a rough surface. The pars dentalis of the premaxilla 314 is located on the anterior side of the bone. The crowns of pleurodont teeth are missing and 315 only their pedicellar portions are preserved. The bone surface is slightly rough in dorsal view. 316 The lamelliform anterolateral ridge of the posterior process is high at the middle part of the 317 bone. The amphicoelous vertebrae are flat and wide. The centrum is dumb-bell in shape and 318 narrows to the middle region. The basapophyses, if present, are small and weakly developed. 319 Two subcentral foramina are present at the central part of the vertebral centrum. In lateral 320 view, the vertebra is low; the anterior and posterior zygapophyseal crests are pointed, forming 321 the dorsal border of the deep depressions anteriorly and posteriorly to the transverse process. 322 The middle part of the neural arch is lower than its cranial and caudal margins. The posterior 323 edge of the neural arch is forked (Fig. 3H) (not visible at Fig. 3M). Manuscript to be reviewed Manuscript to be reviewed 304 Mioproteus sp. 305 306 (Fig. 3H-3S) 307 Localities and material examined. Ryzhaya II (known also Ryzhaya Sopka), GNM unnr. 308 specimen, 2 trunk vertebrae; Malyi Kalkaman 2, GIN 1107/2001-AM01, 1 right premaxilla; Bor 309 1A, GIN 1115/1001-AM01, 1 trunk vertebra; Ayakoz, GNM unnr. specimen, 1 trunk vertebra; 310 Akespe, unnr. HC specimens, 3 vertebrae; Petropavlovsk 1/2, GNM unnr. specimen, 22 311 vertebrae. 312 Description. The preserved left premaxilla is fragmentary (Fig. 3P-3Q) and the posterior proce 313 is broken off. In ventral view, the bone has a rough surface. The pars dentalis of the premaxilla 314 is located on the anterior side of the bone. The crowns of pleurodont teeth are missing and 315 only their pedicellar portions are preserved. The bone surface is slightly rough in dorsal view. 316 The lamelliform anterolateral ridge of the posterior process is high at the middle part of the 317 bone. The amphicoelous vertebrae are flat and wide. The centrum is dumb-bell in shape and 318 narrows to the middle region. The basapophyses, if present, are small and weakly developed. 319 Two subcentral foramina are present at the central part of the vertebral centrum. In lateral 320 view, the vertebra is low; the anterior and posterior zygapophyseal crests are pointed, forming 321 the dorsal border of the deep depressions anteriorly and posteriorly to the transverse process 322 The middle part of the neural arch is lower than its cranial and caudal margins. The posterior 323 edge of the neural arch is forked (Fig. 3H) (not visible at Fig. 3M). The neural spine extends as 307 Localities and material examined. Ryzhaya II (known also Ryzhaya Sopka), GNM unnr. 308 specimen, 2 trunk vertebrae; Malyi Kalkaman 2, GIN 1107/2001-AM01, 1 right premaxilla; Borki 309 1A, GIN 1115/1001-AM01, 1 trunk vertebra; Ayakoz, GNM unnr. specimen, 1 trunk vertebra; 310 Akespe, unnr. HC specimens, 3 vertebrae; Petropavlovsk 1/2, GNM unnr. specimen, 22 311 vertebrae. 308 specimen, 2 trunk vertebrae; Malyi Kalkaman 2, GIN 1107/2001-AM01, 1 right premaxilla; Borki 309 1A, GIN 1115/1001-AM01, 1 trunk vertebra; Ayakoz, GNM unnr. specimen, 1 trunk vertebra; 310 Akespe, unnr. HC specimens, 3 vertebrae; Petropavlovsk 1/2, GNM unnr. specimen, 22 311 vertebrae. 312 Description. The preserved left premaxilla is fragmentary (Fig. 3P-3Q) and the posterior process 313 is broken off. In ventral view, the bone has a rough surface. The neural spine extends as 324 far as the preserved anterior margins of the neural arch, whereas posteriorly, it terminates PeerJ reviewing PDF \| (2016:08:12606:4:0:NEW 23 Jan 2017) Manuscript to be reviewed 325 before the posterior margin of the neural arch. The preserved right pre- and postzygapophyseal 326 articular facets are ellipsoid. 325 before the posterior margin of the neural arch. The preserved right pre- and postzygapophyseal 326 articular facets are ellipsoid. 325 before the posterior margin of the neural arch. The preserved right pre- and postzygapophyseal 326 articular facets are ellipsoid. 327 Comparison and comments. A direct comparison with Mioproteus specimens from previous 328 reports was not possible due to the extremely scarce description of the skull elements 329 attributed to this taxon (e.g. Estes & Darevsky, 1977; Miklas, 2002). We therefore used the 330 material of Mioproteus sp. from the Grytsiv locality (Ukraine, earliest Late Miocene) (Fig. 3R-3S) 331 for the taxonomic identification of the fossil premaxilla from Malyi Kalkaman 2 (Fig. 3P-3Q). Our 332 comparison founds no differences in the premaxilla morphology between the Kazakhstan and 333 Ukrainian Mioproteus sp. The vertebrae from the Borki 1A and Ayakoz localities can be easily 334 assigned to the genus Mioproteus based on following characters: (1) robust vertebra with an 335 amphicoelous centrum; (2) a tall cranial margin of the neural arch; (3) the presence of the 336 basapophyses; (4) a distinct wide depression at the anterior base of the transverse process; (5) 337 intervertebrally exiting spinal nerves; and (6) a forked neural spine (Edwards, 1976; Estes & 338 Darevsky, 1977; Ivanov, 2008). 339 340 Family Salamandridae Goldfuss, 1820 341 Subfamily Pleurodelinae Tschudi, 1838 342 Genus Chelotriton Pomel, 1853 343 Chelotriton sp. 344 345 (Fig. 3T-3Y) 325 before the posterior margin of the neural arch. The preserved right pre- and postzygapophyseal 326 articular facets are ellipsoid. 327 Comparison and comments. A direct comparison with Mioproteus specimens from previous 328 reports was not possible due to the extremely scarce description of the skull elements 329 attributed to this taxon (e.g. Estes & Darevsky, 1977; Miklas, 2002). We therefore used the 330 material of Mioproteus sp. from the Grytsiv locality (Ukraine, earliest Late Miocene) (Fig. 3R-3S) 331 for the taxonomic identification of the fossil premaxilla from Malyi Kalkaman 2 (Fig. 3P-3Q). Our 332 comparison founds no differences in the premaxilla morphology between the Kazakhstan and 333 Ukrainian Mioproteus sp. PeerJ reviewing PDF \| (2016:08:12606:4:0:NEW 23 Jan 2017) Manuscript to be reviewed The vertebrae from the Borki 1A and Ayakoz localities can be easily 334 assigned to the genus Mioproteus based on following characters: (1) robust vertebra with an 335 amphicoelous centrum; (2) a tall cranial margin of the neural arch; (3) the presence of the 336 basapophyses; (4) a distinct wide depression at the anterior base of the transverse process; (5) 337 intervertebrally exiting spinal nerves; and (6) a forked neural spine (Edwards, 1976; Estes & 338 Darevsky, 1977; Ivanov, 2008). 327 Comparison and comments. A direct comparison with Mioproteus specimens from previous 328 reports was not possible due to the extremely scarce description of the skull elements 329 attributed to this taxon (e.g. Estes & Darevsky, 1977; Miklas, 2002). We therefore used the 330 material of Mioproteus sp. from the Grytsiv locality (Ukraine, earliest Late Miocene) (Fig. 3R-3S) 331 for the taxonomic identification of the fossil premaxilla from Malyi Kalkaman 2 (Fig. 3P-3Q). Our 332 comparison founds no differences in the premaxilla morphology between the Kazakhstan and 333 Ukrainian Mioproteus sp. The vertebrae from the Borki 1A and Ayakoz localities can be easily 334 assigned to the genus Mioproteus based on following characters: (1) robust vertebra with an 335 amphicoelous centrum; (2) a tall cranial margin of the neural arch; (3) the presence of the 336 basapophyses; (4) a distinct wide depression at the anterior base of the transverse process; (5) 337 intervertebrally exiting spinal nerves; and (6) a forked neural spine (Edwards, 1976; Estes & 338 Darevsky, 1977; Ivanov, 2008). PeerJ reviewing PDF \| (2016:08:12606:4:0:NEW 23 Jan 2017) Manuscript to be reviewed 346 Localities and material examined. Malyi Kalkaman 1, GNM unnr. specimen, 1 trunk vertebra; 347 Ayakoz, GNM unnr. specimen, 1 trunk vertebra. 346 Localities and material examined. Malyi Kalkaman 1, GNM unnr. specimen, 1 trunk vertebra; 347 Ayakoz, GNM unnr. specimen, 1 trunk vertebra. 348 Description. The single fragmentary trunk vertebra of Chelotriton from the Malyi Kalkaman 1 349 locality has been scantily described (Tleuberdina et al., 1993, pp. 133-134). The centrum of the 350 vertebra is ophistocoelous and dorsally curved. Both the posterior one-third of the vertebra 351 and cotyle are broken. The condyle is dorsoventrally slightly compressed and oval in shape. The 352 middle part of the ventral surface of the centrum bears a pair of the foramina subcentrale. The 353 ventral bases of both transverse processes are pierced by a foramen (potentially the ventral 354 foramen for the spinal nerve). 348 Description. The single fragmentary trunk vertebra of Chelotriton from the Malyi Kalkaman 1 349 locality has been scantily described (Tleuberdina et al., 1993, pp. 133-134). The centrum of the 350 vertebra is ophistocoelous and dorsally curved. Both the posterior one-third of the vertebra 351 and cotyle are broken. The condyle is dorsoventrally slightly compressed and oval in shape. The 352 middle part of the ventral surface of the centrum bears a pair of the foramina subcentrale. The 353 ventral bases of both transverse processes are pierced by a foramen (potentially the ventral 354 foramen for the spinal nerve). 355 The neural spine is tall, long, and almost equal in length to the vertebral centrum. The dorsal 356 surface of the neural spine has the form of an elongated isosceles triangle and it is covered by a 357 distinct pustular sculpture. The anterior margin of the neural spine is concave in outline. The 358 posterior half of the spine is wider than the anterior one (Fig. 3Y). In anterior view, the neural 359 arch and the neural canal have a triangular form. The roof of the neural canal is flat, on both 360 sides of the neural spine. 355 The neural spine is tall, long, and almost equal in length to the vertebral centrum. The dorsal 356 surface of the neural spine has the form of an elongated isosceles triangle and it is covered by a 357 distinct pustular sculpture. The anterior margin of the neural spine is concave in outline. PeerJ reviewing PDF \| (2016:08:12606:4:0:NEW 23 Jan 2017) Manuscript to be reviewed The 358 posterior half of the spine is wider than the anterior one (Fig. 3Y). In anterior view, the neural 359 arch and the neural canal have a triangular form. The roof of the neural canal is flat, on both 360 sides of the neural spine. 361 The pre- and postzygapophyses are damaged. The anterior portion of the left postzygapophysis 362 is present and it shows a horizontal surface. The anterior bases of both prezygapophyses at the 363 contact with the centrum possess small subprezygapophyseal foramina. Behind the left 364 prezygapophysis, the accessory alar process exhibits a marked step (Fig. 3Y), projects 365 posteroventrally and connects caudally with the anterior alar process. The contact point of the 366 accessory and anterior alar processes probably corresponds to the base of the parapophysis. 367 Both transverse processes are broken, but the bases are preserved. Apparently, two rounded 361 The pre- and postzygapophyses are damaged. The anterior portion of the left postzygapophysis 362 is present and it shows a horizontal surface. The anterior bases of both prezygapophyses at the 363 contact with the centrum possess small subprezygapophyseal foramina. Behind the left 364 prezygapophysis, the accessory alar process exhibits a marked step (Fig. 3Y), projects 365 posteroventrally and connects caudally with the anterior alar process. The contact point of the 366 accessory and anterior alar processes probably corresponds to the base of the parapophysis. 367 Both transverse processes are broken, but the bases are preserved. Apparently, two rounded 367 Both transverse processes are broken, but the bases are preserved. Apparently, two rounded PeerJ reviewing PDF \| (2016:08:12606:4:0:NEW 23 Jan 2017) Manuscript to be reviewed 368 upper and lower prominences, seen in left lateral view, correspond to the dia- and 369 parapophysis. The parapophysis is located anteriorly and dorsally to the level of the 370 diapophysis; thus, the transverse process becomes a bent projection. The arterial canal runs 371 behind the base of the transverse process. Anteriorly, its dorsal and ventral walls are built by 372 the accessory and anterior alar processes. 373 The vertebra from the Ayakoz locality (Fig. 3T-3X) is fragmentary, its neural arch and left 374 transverse process are lost, the centrum is compact, short and wide, and it possesses an 375 elliptical central foramen. The diapophysis of the preserved right transverse process is broken, 376 but it can be assumed that the dia- and parapophysis were separated from each other. The 377 accessory alar process runs from the prezygapophysis to the dorsal edge of the diapophysis. 378 The posterior and anterior alar processes run from the cotyle and condyle straight along the 379 transverse process to the parapophysis. This morphology is characteristic of the first trunk 380 vertebrae. 381 Comparison and comments. This vertebra was previously described by Tleuberdina et al. 382 (1993). Here we have assigned this specimen to the genus Chelotriton owing to the presence of 383 a triangular and well-sculptured plate on the top of the neurapophysis. This character, 368 upper and lower prominences, seen in left lateral view, correspond to the dia- and 369 parapophysis. The parapophysis is located anteriorly and dorsally to the level of the 370 diapophysis; thus, the transverse process becomes a bent projection. The arterial canal runs 371 behind the base of the transverse process. Anteriorly, its dorsal and ventral walls are built by 372 the accessory and anterior alar processes. 373 The vertebra from the Ayakoz locality (Fig. 3T-3X) is fragmentary, its neural arch and left 374 transverse process are lost, the centrum is compact, short and wide, and it possesses an 375 elliptical central foramen. The diapophysis of the preserved right transverse process is broken, 376 but it can be assumed that the dia- and parapophysis were separated from each other. The 377 accessory alar process runs from the prezygapophysis to the dorsal edge of the diapophysis. 378 The posterior and anterior alar processes run from the cotyle and condyle straight along the 379 transverse process to the parapophysis. This morphology is characteristic of the first trunk 380 vertebrae. Manuscript to be reviewed 373 The vertebra from the Ayakoz locality (Fig. 3T-3X) is fragmentary, its neural arch and left 374 transverse process are lost, the centrum is compact, short and wide, and it possesses an 375 elliptical central foramen. The diapophysis of the preserved right transverse process is broken, 376 but it can be assumed that the dia- and parapophysis were separated from each other. The 377 accessory alar process runs from the prezygapophysis to the dorsal edge of the diapophysis. 378 The posterior and anterior alar processes run from the cotyle and condyle straight along the 379 transverse process to the parapophysis. This morphology is characteristic of the first trunk 380 vertebrae. 381 Comparison and comments. This vertebra was previously described by Tleuberdina et al. 382 (1993). Here we have assigned this specimen to the genus Chelotriton owing to the presence of 383 a triangular and well-sculptured plate on the top of the neurapophysis. This character, 381 Comparison and comments. This vertebra was previously described by Tleuberdina et al. 382 (1993). Here we have assigned this specimen to the genus Chelotriton owing to the presence of 383 a triangular and well-sculptured plate on the top of the neurapophysis. This character, 384 however, is not a unique feature of Chelotriton and is also seen in other salamanders, e.g. 385 Recent species of Tylototriton and Echinotriton, and in Cynops pyrrhogaster, Lissotriton boscai 386 (unnr. GPIT specimen), Paramesotriton (MNCN 23557, 13645), as well as the fossil taxa 387 Archaeotriton (Böhme, 1998), aff. Tylototriton sp. (Baikadam locality, this paper), 388 Carpathotriton (Venczel, 2008). The vertebra from the Malyi Kalkaman 1 resembles the species 389 of Chelotriton, Paramesotriton, Tylototriton, Echinotriton, Cynops pyrrhogaster, and 388 Carpathotriton (Venczel, 2008). The vertebra from the Malyi Kalkaman 1 resembles the species 389 of Chelotriton, Paramesotriton, Tylototriton, Echinotriton, Cynops pyrrhogaster, and PeerJ reviewing PDF \| (2016:08:12606:4:0:NEW 23 Jan 2017) Manuscript to be reviewed Manuscript to be reviewed 390 Carpathotriton in their mutual presence of a subprezygapophyseal foramen. The vertebra can, 391 however, be justified as Chelotriton sp. and distinguished from other salamanders by: (1) its 392 longer length (vs. Echinotriton, Cynops and Carpathotriton); (2) a longer neural spine with a 393 rugose sculptured and triangular dorsal surface (vs. aff. Tylototriton sp., Baikadam locality, this 394 paper); and (3) a well-pronounced accessory alar process (vs. Tylototriton). 395 The fragmentary vertebra from the Ayakoz locality can be assigned also to this group because 396 of the presence of massive rib-bearers and large dimensions (Ivanov, 2008). Its vertebra is 397 identical to that of vertebra of Chelotriton sp. type II described from the Mokrá-Western 398 Quarry, 2/2003 Reptile Joint (Early Miocene, Czech Republic) (Ivanov, 2008). 399 The abundant European Cenozoic record of the genus Chelotriton, however, showed that 400 vertebral morphology is insufficient for taxonomic identification as Chelotriton (Böhme, 2008). 401 This genus has an unknown higher diversity, which can be uncovered by the study of complete 402 skeletons of those species. We hence classified the vertebrae from studied localities as aff. 403 Chelotriton sp. 404 405 Genus Tylototriton Anderson, 1871 (Anderson, 1871) 406 aff. Tylototriton sp. 407 408 (Fig. 4A-4K) 409 L lit d t i l i d B ik d GIN 950/2001 AM01 A14 A17 5 t k t b 390 Carpathotriton in their mutual presence of a subprezygapophyseal foramen. The vertebra can, 391 however, be justified as Chelotriton sp. and distinguished from other salamanders by: (1) its 392 longer length (vs. Echinotriton, Cynops and Carpathotriton); (2) a longer neural spine with a 393 rugose sculptured and triangular dorsal surface (vs. aff. Tylototriton sp., Baikadam locality, this 394 paper); and (3) a well-pronounced accessory alar process (vs. Tylototriton). 395 The fragmentary vertebra from the Ayakoz locality can be assigned also to this group because 396 of the presence of massive rib-bearers and large dimensions (Ivanov, 2008). Its vertebra is 397 identical to that of vertebra of Chelotriton sp. type II described from the Mokrá-Western 398 Quarry, 2/2003 Reptile Joint (Early Miocene, Czech Republic) (Ivanov, 2008). 405 Genus Tylototriton Anderson, 1871 (Anderson, 1871) 409 Locality and material examined. Baikadam, GIN 950/2001-AM01, -A14–-A17 5 trunk vertebrae; 410 Ayakoz, GNM unnr. specimen, 2 trunk vertebrae. PeerJ reviewing PDF \| (2016:08:12606:4:0:NEW 23 Jan 2017) Manuscript to be reviewed 411 Description. All preserved vertebrae are opisthocoelous. The condyle and cotyle are 412 dorsoventrally compressed. The vertebrae are slender, slightly narrow, and high. The neural 413 canal is round, but in anterior view, the ventral margin of the neural canal is flat. The same 414 occurs with the dorsal wall of the vertebral centrum. The centrum is dorsally curved in lateral 415 view (Fig. 4A, 4F, 4G). The neural spine was most probably high but does not reach the level of 416 the pustular region. The neural spine begins behind the cranial margin of the neural arch. The 417 neural arch is tilted dorsally and does not extend beyond the posterior edge of the 418 postzygapophysis. The dorsal plate of the neural spine is short, poorly developed, and covered 419 with rugosities. It has the form of an isosceles triangle. Due to the concave shape of the 420 posterior margin of the caudal border, we suggest that the neural spine was probably 421 bifurcated. The length of the neural spine, without the sculptured structure, is the same in all 422 preserved vertebrae and corresponds nearly to almost half of the entire vertebral length (Fig. 423 4A, 4F, 4G). 424 Th d t h h i t l d l t t th l l ( Fi 4A) Th 411 Description. All preserved vertebrae are opisthocoelous. The condyle and cotyle are 412 dorsoventrally compressed. The vertebrae are slender, slightly narrow, and high. The neural 413 canal is round, but in anterior view, the ventral margin of the neural canal is flat. The same 414 occurs with the dorsal wall of the vertebral centrum. The centrum is dorsally curved in lateral 415 view (Fig. 4A, 4F, 4G). The neural spine was most probably high but does not reach the level of 416 the pustular region. The neural spine begins behind the cranial margin of the neural arch. The 417 neural arch is tilted dorsally and does not extend beyond the posterior edge of the 418 postzygapophysis. The dorsal plate of the neural spine is short, poorly developed, and covered 419 with rugosities. It has the form of an isosceles triangle. Due to the concave shape of the 420 posterior margin of the caudal border, we suggest that the neural spine was probably 421 bifurcated. Manuscript to be reviewed 433 process connects the prezygapophysis caudally with the base of the parapophysis (e.g. 4F). The 434 dorsal lamina starts from the diapophysis and extends to the postzygapophyses, whereas the 435 lamelliform posterior alar process starts at the parapophysis and terminates directly before the 436 cotyle. Subparallel to the accessory alar process, a thin anterior alar process runs along the 437 cranial half of the centrum. Behind and in front of the transverse process two ‘cavities’ (a 438 shallow anterior and a deep posterior) are present. These ʻcavitiesʼ are connected by a canal 439 (possibly an arterial canal), that runs through the transverse process. In ventral view, the 440 vertebral centrum does not possess a ventral keel. The centrum is flattened and nearly plane in 441 the middle portion. Its surface is rough and pierced by numerous foramina. Two large 442 subcentral foramina are located at the posterior corner between the centrum and transverse 443 process (Fig. 4C, 4L). 433 process connects the prezygapophysis caudally with the base of the parapophysis (e.g. 4F). The 434 dorsal lamina starts from the diapophysis and extends to the postzygapophyses, whereas the 435 lamelliform posterior alar process starts at the parapophysis and terminates directly before the 436 cotyle. Subparallel to the accessory alar process, a thin anterior alar process runs along the 437 cranial half of the centrum. Behind and in front of the transverse process two ‘cavities’ (a 438 shallow anterior and a deep posterior) are present. These ʻcavitiesʼ are connected by a canal 439 (possibly an arterial canal), that runs through the transverse process. In ventral view, the 440 vertebral centrum does not possess a ventral keel. The centrum is flattened and nearly plane in 441 the middle portion. Its surface is rough and pierced by numerous foramina. Two large 442 subcentral foramina are located at the posterior corner between the centrum and transverse 443 process (Fig. 4C, 4L). 444 Comparison and comments. The vertebrae resemble the morphology of pleurodeline 445 salamanders Echinotriton, Tylototriton, Cynops, Chelotriton, Paramesotriton and Tylototriton 446 and Carpathotriton in characteristics such as: (1) the presence of rugosities on the neural arch; 447 (2) the connection of the prezygapophysis and parapophysis with the accessory alar process, 448 except in Carpathotriton, Cynops and cf. Tylototriton sp. from Möhren 13 (Böhme, 2010: p. 11, 449 fig. Manuscript to be reviewed The length of the neural spine, without the sculptured structure, is the same in all 422 preserved vertebrae and corresponds nearly to almost half of the entire vertebral length (Fig. 423 4A, 4F, 4G). 424 The pre- and postzygapophyses are horizontal and almost at the same level (e.g. Fig. 4A). The 425 pre- and postzygapophyseal articular facets are oval in shape. Small subprezygapophyseal 426 foramina are present at the level of the connection between the anterior bases of both 427 prezygapophyses with the vertebral centrum. The posterolaterally directed transverse process 428 is horizontally flattened and displays a bicapitate articulation surface with the rip. The 429 diapophysis and parapophysis are separated, with the former being smaller than the latter. A 430 low and moderately deep notch is developed at the posterior edge of the neural arch. The 431 transverse process has an anterior (accessory alar process) and posterior laminar edges (i.e. the 432 posterior alar process and dorsal lamina). The straight, posteroventrally directed accessory alar PeerJ reviewing PDF \| (2016:08:12606:4:0:NEW 23 Jan 2017) Manuscript to be reviewed 6f), where this process connects prezygapophysis with diapophysis; (3) a moderately 450 developed posterior ʻcavityʼ behind the transverse process; and (4) the presence of 451 subprezygapophyseal foramen (for collection references see subsection ʻComparisonʼ of 452 Chelotriton sp. in this report). In terms of the general morphology, the vertebrae mainly 453 resemble the genus Tylototriton and differ from the compared genera in having: (1) a low, 454 elongate, narrow and lesser flattened vertebrae; (2) a weakly developed pustular structure of 444 Comparison and comments. The vertebrae resemble the morphology of pleurodeline 445 salamanders Echinotriton, Tylototriton, Cynops, Chelotriton, Paramesotriton and Tylototriton 446 and Carpathotriton in characteristics such as: (1) the presence of rugosities on the neural arch; 447 (2) the connection of the prezygapophysis and parapophysis with the accessory alar process, 448 except in Carpathotriton, Cynops and cf. Tylototriton sp. from Möhren 13 (Böhme, 2010: p. 11, 449 fig. 6f), where this process connects prezygapophysis with diapophysis; (3) a moderately 450 developed posterior ʻcavityʼ behind the transverse process; and (4) the presence of 451 subprezygapophyseal foramen (for collection references see subsection ʻComparisonʼ of 452 Chelotriton sp. in this report). In terms of the general morphology, the vertebrae mainly 453 resemble the genus Tylototriton and differ from the compared genera in having: (1) a low, 454 elongate, narrow and lesser flattened vertebrae; (2) a weakly developed pustular structure of 454 elongate, narrow and lesser flattened vertebrae; (2) a weakly developed pustular structure of PeerJ reviewing PDF \| (2016:08:12606:4:0:NEW 23 Jan 2017) Manuscript to be reviewed However, we do not consider it reasonable to describe a new form unless 467 cranial material of this salamander is available. Manuscript to be reviewed 455 the neural arch (similar character as seen in Paramesotriton); (3) a low and long neural spine 456 without the sculptured structure; (4) a dorsoventrally compressed cotyle and condyle; (5) a 457 deep posterior ʻcavityʼ behind the transverse process, and an extended dorsal lamina and 458 posterior alar process; (6) a low and shallow posterior notch of the neural arch; and (7) in 459 having an accessory alar process that reaches the parapophysis, which differs from specimens 460 of the genus Cynops wherein it reaches the diapophysis. The Siberian Tylototriton differs from 461 the European Oligocene cf. Tylototriton (see Böhme, 2010: p. 11, fig. 6f) by having: (1) a 462 ventrally deflected accessory alar process that terminates ventrally to the parapophysis; (2) a 463 shorter and lower neural spine; and (3) a shorter dorsal plate of the neural spine. 457 deep posterior ʻcavityʼ behind the transverse process, and an extended dorsal lamina and 458 posterior alar process; (6) a low and shallow posterior notch of the neural arch; and (7) in 459 having an accessory alar process that reaches the parapophysis, which differs from specimens 460 of the genus Cynops wherein it reaches the diapophysis. The Siberian Tylototriton differs from 461 the European Oligocene cf. Tylototriton (see Böhme, 2010: p. 11, fig. 6f) by having: (1) a 462 ventrally deflected accessory alar process that terminates ventrally to the parapophysis; (2) a 463 shorter and lower neural spine; and (3) a shorter dorsal plate of the neural spine. 464 Taking into account the above-mentioned differences, we suggest that the described vertebrae 465 should be assigned to a new pleurodeline salamander genus that shows affinities with the 466 genus Tylototriton. However, we do not consider it reasonable to describe a new form unless 467 cranial material of this salamander is available. 468 469 Order Anura Fischer von Waldheim, 1813 470 Family Palaeobatrachidae (Cope, 1865) 471 Palaeobatrachidae sp. indet. 472 473 (Fig. 5A-5D) 474 Locality and material examined. Novaya Stanitsa 1A, GIN 948/2001-AM12, 1 sphenethmoid. 475 Description. This specimen is represented by a very robust sphenethmoid that lacks the 476 posterior region. The two anterior cavities corresponding to the antrum olfactorium are 464 Taking into account the above-mentioned differences, we suggest that the described vertebrae 465 should be assigned to a new pleurodeline salamander genus that shows affinities with the 466 genus Tylototriton. Manuscript to be reviewed 455 the neural arch (similar character as seen in Paramesotriton); (3) a low and long neural spine 456 without the sculptured structure; (4) a dorsoventrally compressed cotyle and condyle; (5) a 457 deep posterior ʻcavityʼ behind the transverse process, and an extended dorsal lamina and 458 posterior alar process; (6) a low and shallow posterior notch of the neural arch; and (7) in 459 having an accessory alar process that reaches the parapophysis, which differs from specimens 460 of the genus Cynops wherein it reaches the diapophysis. The Siberian Tylototriton differs from 461 the European Oligocene cf. Tylototriton (see Böhme, 2010: p. 11, fig. 6f) by having: (1) a 462 ventrally deflected accessory alar process that terminates ventrally to the parapophysis; (2) a 463 shorter and lower neural spine; and (3) a shorter dorsal plate of the neural spine. 464 Taking into account the above-mentioned differences, we suggest that the described vertebrae 465 should be assigned to a new pleurodeline salamander genus that shows affinities with the 466 genus Tylototriton. However, we do not consider it reasonable to describe a new form unless 467 cranial material of this salamander is available. 468 469 Order Anura Fischer von Waldheim, 1813 470 Family Palaeobatrachidae (Cope, 1865) 471 Palaeobatrachidae sp. indet. 472 473 (Fig. 5A-5D) 474 Locality and material examined Novaya Stanitsa 1A GIN 948/2001-AM12 1 sphenethmoid 455 the neural arch (similar character as seen in Paramesotriton); (3) a low and long neural spine 456 without the sculptured structure; (4) a dorsoventrally compressed cotyle and condyle; (5) a 457 deep posterior ʻcavityʼ behind the transverse process, and an extended dorsal lamina and 458 posterior alar process; (6) a low and shallow posterior notch of the neural arch; and (7) in 459 having an accessory alar process that reaches the parapophysis, which differs from specimens 460 of the genus Cynops wherein it reaches the diapophysis. The Siberian Tylototriton differs from 461 the European Oligocene cf. Tylototriton (see Böhme, 2010: p. 11, fig. 6f) by having: (1) a 462 ventrally deflected accessory alar process that terminates ventrally to the parapophysis; (2) a 463 shorter and lower neural spine; and (3) a shorter dorsal plate of the neural spine. Manuscript to be reviewed 477 anteroposteriorly shallow. The posterior cavity, antrum pro lobo olfactorio, is deep and narrow 478 (Fig. 5A, 5B). The olfactory foramen is larger than the orbitonasal foramen (Fig. 5C). The 479 processus rostralis is elongated and projects anteriorly. Anteriorly, dorsal face of the bone, two 480 sharply marked crescentic depressions (nasal facets) correspond to the contacts with the nasal 481 bones (Fig. 5A). In dorsal view, the frontoparietal facet (contacting with the frontoparietal 482 cranial bones) shows a slightly striated surface. The lateral processes protrude laterally. The 483 lamina supraorbitalis is well developed. The most anterior part of the incisura semielliptical is 484 preserved on the specimen. The remaining part of this structure demonstrates that it 485 approaches cranially to the anterior border of the bone. The ventral face of the sphenethmoid 486 possesses a narrow and long depression corresponding to the contact area with the cultriform 487 process of the parasphenoid (the parasphenoid facet) (Fig. 5B). 477 anteroposteriorly shallow. The posterior cavity, antrum pro lobo olfactorio, is deep and narrow 478 (Fig. 5A, 5B). The olfactory foramen is larger than the orbitonasal foramen (Fig. 5C). The 479 processus rostralis is elongated and projects anteriorly. Anteriorly, dorsal face of the bone, two 480 sharply marked crescentic depressions (nasal facets) correspond to the contacts with the nasal 481 bones (Fig. 5A). In dorsal view, the frontoparietal facet (contacting with the frontoparietal 482 cranial bones) shows a slightly striated surface. The lateral processes protrude laterally. The 483 lamina supraorbitalis is well developed. The most anterior part of the incisura semielliptical is 484 preserved on the specimen. The remaining part of this structure demonstrates that it 485 approaches cranially to the anterior border of the bone. The ventral face of the sphenethmoid 486 possesses a narrow and long depression corresponding to the contact area with the cultriform 487 process of the parasphenoid (the parasphenoid facet) (Fig. 5B). 477 anteroposteriorly shallow. The posterior cavity, antrum pro lobo olfactorio, is deep and narrow 478 (Fig. 5A, 5B). The olfactory foramen is larger than the orbitonasal foramen (Fig. 5C). The 479 processus rostralis is elongated and projects anteriorly. Anteriorly, dorsal face of the bone, two 480 sharply marked crescentic depressions (nasal facets) correspond to the contacts with the nasal 481 bones (Fig. 5A). In dorsal view, the frontoparietal facet (contacting with the frontoparietal 482 cranial bones) shows a slightly striated surface. The lateral processes protrude laterally. PeerJ reviewing PDF \| (2016:08:12606:4:0:NEW 23 Jan 2017) PeerJ reviewing PDF \| (2016:08:12606:4:0:NEW 23 Jan 2017) Manuscript to be reviewed The 483 lamina supraorbitalis is well developed. The most anterior part of the incisura semielliptical is 484 preserved on the specimen. The remaining part of this structure demonstrates that it 485 approaches cranially to the anterior border of the bone. The ventral face of the sphenethmoid 486 possesses a narrow and long depression corresponding to the contact area with the cultriform 487 process of the parasphenoid (the parasphenoid facet) (Fig. 5B). 488 Comparison and comments. The bone has strong similarities to that of palaeobatrachids in 488 Comparison and comments. The bone has strong similarities to that of palaeobatrachids in 489 having: (1) a long sphenethmoid with a frontoparietal fenestra corresponding to more than half 490 of the bone length; (2) the articulation area of the parasphenoid delimited by two parallel 491 ridges, in ventral view; (3) a very short septum nasi and lateral process (Vergnaud-Grazzini & 492 Młynarski, 1969; Sanchíz & Młynarski, 1979). The palaeobatrachid from the Novaya Stanitsa 1A 493 locality exhibits all these characters aside from the short septum nasi, which is longer in the 494 fossil bone. We presume that the frontoparietal fenestra was longer more than half of the 495 sphenethmoid length because the overall length of the frontoparietal and nasal facets has 496 similar proportions as these seen in other palaeobatrachids. Furthermore, according to Venczel, 497 Codrea & Fărcaş (2012), the sphenethmoidal ossification forms the anterior margin of 488 Comparison and comments. The bone has strong similarities to that of palaeobatrachids in 489 having: (1) a long sphenethmoid with a frontoparietal fenestra corresponding to more than half 490 of the bone length; (2) the articulation area of the parasphenoid delimited by two parallel 491 ridges, in ventral view; (3) a very short septum nasi and lateral process (Vergnaud-Grazzini & 492 Młynarski, 1969; Sanchíz & Młynarski, 1979). The palaeobatrachid from the Novaya Stanitsa 1A 493 locality exhibits all these characters aside from the short septum nasi, which is longer in the 494 fossil bone. We presume that the frontoparietal fenestra was longer more than half of the 495 sphenethmoid length because the overall length of the frontoparietal and nasal facets has 496 similar proportions as these seen in other palaeobatrachids. Furthermore, according to Venczel, 497 Codrea & Fărcaş (2012), the sphenethmoidal ossification forms the anterior margin of 488 Comparison and comments. Manuscript to be reviewed 498 frontoparietal fontanelle in palaeobatrachid frogs (Palaeobatrachus + Albionbatrachus), which 499 can also be observed in the studied specimen. 500 501 Family Bombinatoridae Gray, 1825 502 Genus Bombina Oken, 1816 503 Bombina sp. / Bombina cf. bombina (Linnaeus, 1761) 504 505 (Fig. 6A-6F) 506 Localities and material examined. Malyi Kalkaman 2, GIN 1107/2001-AM02, 1 ilium; Selety 1A, 507 GIN 1107/2001-AM06, 1 ilium; Cherlak, GIN 1107/2001-AM06, 1 ilium. 508 Description. The bone description is based on the ilium from the Selety 1A locality, since the 509 specimens from the Malyi Kalkaman 2 and Cherkal localities are greatly damaged. In lateral 510 view, the iliac shaft is almost straight and lacks the dorsal crest. The tuber superior is a weakly 511 pronounced tubercle. In dorsal view, a spiral groove is observable and continues on the medial 512 surface of the shaft. The acetabulum is round and strongly extended (Fig. 6A). The junction 513 between the iliac shaft and corpus ossi is slightly constricted and the ventral base of the corpus 514 ossi possesses a preacetabular fossa. The ventral ridge of the acetabulum is high. In lateral and 515 posterior views, the pars descendens is reduced and wide, whereas the pars ascendens is high 516 but narrow (Fig. 6A, 6B). In ventral view, the pars descendens is broad and nearly flat. In medial 517 view, the acetabular area is bordered by shallow ridges, between which there is, a triangular 518 and medially prominent interiliac tubercle (Fig. 6B, 6C). 498 frontoparietal fontanelle in palaeobatrachid fro 499 can also be observed in the studied specimen. 500 501 Family Bombinatoridae Gray, 1825 502 Genus Bombina Oken, 1816 503 Bombina sp. / Bombina cf. bombina (Linnaeus, 504 505 (Fig. 6A-6F) 506 Localities and material examined. Malyi Kalkam 507 GIN 1107/2001-AM06, 1 ilium; Cherlak, GIN 11 508 Description. The bone description is based on t 509 specimens from the Malyi Kalkaman 2 and Che 510 view, the iliac shaft is almost straight and lacks 511 pronounced tubercle. In dorsal view, a spiral gr 512 surface of the shaft. The acetabulum is round a 513 between the iliac shaft and corpus ossi is slight 514 ossi possesses a preacetabular fossa. The ventr 515 posterior views, the pars descendens is reduce 516 but narrow (Fig. 6A, 6B). In ventral view, the pa 517 view, the acetabular area is bordered by shallo 518 and medially prominent interiliac tubercle (Fig. Manuscript to be reviewed The bone has strong similarities to that of palaeobatrachids in 489 having: (1) a long sphenethmoid with a frontoparietal fenestra corresponding to more than half 490 of the bone length; (2) the articulation area of the parasphenoid delimited by two parallel 491 ridges, in ventral view; (3) a very short septum nasi and lateral process (Vergnaud-Grazzini & 492 Młynarski, 1969; Sanchíz & Młynarski, 1979). The palaeobatrachid from the Novaya Stanitsa 1A 493 locality exhibits all these characters aside from the short septum nasi, which is longer in the 494 fossil bone. We presume that the frontoparietal fenestra was longer more than half of the 495 sphenethmoid length because the overall length of the frontoparietal and nasal facets has 496 similar proportions as these seen in other palaeobatrachids. Furthermore, according to Venczel, 497 Codrea & Fărcaş (2012), the sphenethmoidal ossification forms the anterior margin of PeerJ reviewing PDF \| (2016:08:12606:4:0:NEW 23 Jan 2017) Manuscript to be reviewed Manuscript to be reviewed 519 Comparison and comments. The lack of the vexillum and a poorly developed tuber superior is 520 characteristic of the genus Bombina (Böhme, 1977). The ilium differs from Bombina orientalis 521 by a poorly developed tuber superior. The ilium from the Selety 1A locality is distinguishable 522 from Bombina variegata and resembles Bombina bombina in having: (1) a developed pars 523 descendens; (2) a posteroventral ridge of the pars descendens projecting ventrally rather than 524 posteriorly (Böhme, 1977); and (3) a developed preacetabular fossa (Sanchíz & Młynarski, 525 1979). We, therefore, tentatively assign the bone to B. bombina due to the absence of well- 526 preserved material of the fire-bellied toads from the Selety 1A locality. The specific assignment 527 of the ilia from the Malyi Kalkaman 2 locality is impossible due to their fragmentary 528 preservation; therefore we describe them as Bombina sp. 529 The specimen from the Cherlak locality (Fig. 6D-6F) is greatly damaged with only a few 530 observable characters remaining that allow for its identification within Bombinatoridae. The 531 identifying characters are: (1) a large pars descendens at its anterior section, but dorsally 532 reduced; (2) a present but larger tuber superior than that of the Maly Kalkaman 2 and Selety 1A 533 specimens (within the family, larger tuber superior are present in the Barbatula (Folie et al., 534 2013)); and (3) although the ventral wall of the acetabulum is not preserved, the remaining part 535 of its base allows for the assumption that it was markedly pronounced. Due to the incomplete 536 preservation, the important characters needed for taxonomic identification, e.g. interiliac 537 tubercle and junctura ilioischiadica, cannot be observed. The ilium from the Cherlak locality 538 can, therefore, be tentatively referred to the family Bombinatoridae. 539 540 Family Pelobatidae Bonaparte, 1850 519 Comparison and comments. The lack of the vexillum and a poorly developed tuber superior is 520 characteristic of the genus Bombina (Böhme, 1977). The ilium differs from Bombina orientalis 521 by a poorly developed tuber superior. The ilium from the Selety 1A locality is distinguishable 522 from Bombina variegata and resembles Bombina bombina in having: (1) a developed pars 523 descendens; (2) a posteroventral ridge of the pars descendens projecting ventrally rather than 524 posteriorly (Böhme, 1977); and (3) a developed preacetabular fossa (Sanchíz & Młynarski, 525 1979). We, therefore, tentatively assign the bone to B. Manuscript to be reviewed 498 frontoparietal fontanelle in palaeobatrachid frogs (Palaeobatrachus + Albionbatrachus), which 499 can also be observed in the studied specimen. 502 Genus Bombina Oken, 1816 503 Bombina sp. / Bombina cf. bombina (Linnaeus, 1761) 504 505 (Fig. 6A-6F) 506 Localities and material examined. Malyi Kalkaman 2, GIN 1107/2001-AM02, 1 ilium; Selety 1A, 507 GIN 1107/2001-AM06, 1 ilium; Cherlak, GIN 1107/2001-AM06, 1 ilium. 508 Description. The bone description is based on the ilium from the Selety 1A locality, since the 509 specimens from the Malyi Kalkaman 2 and Cherkal localities are greatly damaged In lateral 506 Localities and material examined. Malyi Kalkaman 2, GIN 1107/2001-AM02, 1 ilium; Selety 1A, 507 GIN 1107/2001-AM06, 1 ilium; Cherlak, GIN 1107/2001-AM06, 1 ilium. PeerJ reviewing PDF \| (2016:08:12606:4:0:NEW 23 Jan 2017) PeerJ reviewing PDF \| (2016:08:12606:4:0:NEW 23 Jan 2017) Manuscript to be reviewed bombina due to the absence of well- 526 preserved material of the fire-bellied toads from the Selety 1A locality. The specific assignment 527 of the ilia from the Malyi Kalkaman 2 locality is impossible due to their fragmentary 528 preservation; therefore we describe them as Bombina sp. 529 The specimen from the Cherlak locality (Fig. 6D-6F) is greatly damaged with only a few 530 observable characters remaining that allow for its identification within Bombinatoridae. The 531 identifying characters are: (1) a large pars descendens at its anterior section, but dorsally 532 reduced; (2) a present but larger tuber superior than that of the Maly Kalkaman 2 and Selety 1A 533 specimens (within the family, larger tuber superior are present in the Barbatula (Folie et al., 534 2013)); and (3) although the ventral wall of the acetabulum is not preserved, the remaining part 535 of its base allows for the assumption that it was markedly pronounced. Due to the incomplete 536 preservation, the important characters needed for taxonomic identification, e.g. interiliac 537 tubercle and junctura ilioischiadica, cannot be observed. The ilium from the Cherlak locality 538 can, therefore, be tentatively referred to the family Bombinatoridae. 529 The specimen from the Cherlak locality (Fig. 6D-6F) is greatly damaged with only a few 530 observable characters remaining that allow for its identification within Bombinatoridae. The 531 identifying characters are: (1) a large pars descendens at its anterior section, but dorsally 532 reduced; (2) a present but larger tuber superior than that of the Maly Kalkaman 2 and Selety 1A 533 specimens (within the family, larger tuber superior are present in the Barbatula (Folie et al., 534 2013)); and (3) although the ventral wall of the acetabulum is not preserved, the remaining part 535 of its base allows for the assumption that it was markedly pronounced. Due to the incomplete 536 preservation, the important characters needed for taxonomic identification, e.g. interiliac 537 tubercle and junctura ilioischiadica, cannot be observed. The ilium from the Cherlak locality 538 can, therefore, be tentatively referred to the family Bombinatoridae. PeerJ reviewing PDF \| (2016:08:12606:4:0:NEW 23 Jan 2017) Manuscript to be reviewed Manuscript to be reviewed 541 Genus Pelobates Wagler, 1830 542 Pelobates sp. 543 544 (Fig. 6G-6I) 545 Localities and material examined. Selet 546 Description. The corpus ossi and distal 547 descendes and pars ascendes are broke 548 superior. An oblique posterolaterally-an 549 dorsal surface. Laterally, the high and lo 550 (Fig. 6I). The junction between the iliac 551 subacetabular groove is shallow and bro 552 well-marked rim. In medial view, the co 553 surface. The interiliac facet is composed 554 developed interiliac tubercle is visible b 555 ventroposteriorly oblique, whereas the 556 The rugose surface of the facet indicate 557 The acetabular dorsal tuber is higher th 558 Comparison and comments. The ilium c 559 absence of a dorsal crest, the absence o 541 Genus Pelobates Wagler, 1830 542 Pelobates sp. 543 544 (Fig. 6G-6I) 545 Localities and material examined. Selety 1A, GIN 1110/2001-AM13, 1 right ilium. 546 Description. The corpus ossi and distal portion of the iliac shaft are present. The tips of the pars 547 descendes and pars ascendes are broken. The bone surface is smooth and there is no tuber 548 superior. An oblique posterolaterally-anteromedially directed spiral groove extends on the 549 dorsal surface. Laterally, the high and long pars ascendens possesses a supraacetabular fossa 550 (Fig. 6I). The junction between the iliac shaft and corpus ossi is not constricted. The 551 subacetabular groove is shallow and broad. The acetabulum has a nearly triangular form, with a 552 well-marked rim. In medial view, the corpus ilii possesses an interiliac facet with a rugose 553 surface. The interiliac facet is composed of a large lower and a small upper portions. A well- 554 developed interiliac tubercle is visible between these portions (Fig. 6G). The lower portion is 555 ventroposteriorly oblique, whereas the upper one is flat, less rugose and has a concave surface. 556 The rugose surface of the facet indicates an extensive contact between two ilia (Fig. 6G, 6H). 557 The acetabular dorsal tuber is higher than the ventral one. 558 Comparison and comments. The ilium can be assigned to the family Pelobatidae based on the 559 absence of a dorsal crest, the absence of a dorsal tubercle and the presence of an oblique spiral 545 Localities and material examined. Selety 1A, GIN 1110/2001-AM13, 1 right ilium. 546 Description. The corpus ossi and distal portion of the iliac shaft are present. The tips of the pars 547 descendes and pars ascendes are broken. The bone surface is smooth and there is no tuber 548 superior. An oblique posterolaterally-anteromedially directed spiral groove extends on the 549 dorsal surface. Laterally, the high and long pars ascendens possesses a supraacetabular fossa 550 (Fig. 6I). The junction between the iliac shaft and corpus ossi is not constricted. The 551 subacetabular groove is shallow and broad. The acetabulum has a nearly triangular form, with a 552 well-marked rim. In medial view, the corpus ilii possesses an interiliac facet with a rugose 553 surface. The interiliac facet is composed of a large lower and a small upper portions. A well- 554 developed interiliac tubercle is visible between these portions (Fig. 6G). The lower portion is 555 ventroposteriorly oblique, whereas the upper one is flat, less rugose and has a concave surface. 556 The rugose surface of the facet indicates an extensive contact between two ilia (Fig. 6G, 6H). 557 The acetabular dorsal tuber is higher than the ventral one. 558 Comparison and comments. The ilium can be assigned to the family Pelobatidae based on the 559 absence of a dorsal crest, the absence of a dorsal tubercle and the presence of an oblique spiral 560 groove on the dorsal surface (Roček et al., 2014). The bone has the same characters of the 561 genus Pelobates: (1) a high and long pars ascendes; (2) a well-developed spiral groove (Böhme, 562 2010); (3) the lack of a dorsal crest of the iliac shaft (Folie et al., 2013); and (4) a rugose surface 558 Comparison and comments. The ilium can be assigned to the family Pelobatidae based on the 559 absence of a dorsal crest, the absence of a dorsal tubercle and the presence of an oblique spiral 560 groove on the dorsal surface (Roček et al., 2014). The bone has the same characters of the 561 genus Pelobates: (1) a high and long pars ascendes; (2) a well-developed spiral groove (Böhme, 562 2010); (3) the lack of a dorsal crest of the iliac shaft (Folie et al., 2013); and (4) a rugose surface PeerJ reviewing PDF \| (2016:08:12606:4:0:NEW 23 Jan 2017) Manuscript to be reviewed 563 of the interiliac facet (Rage & Hossini, 2000). However, further identification of the ilium is 564 impossible, as it does not show relevant differences at the specific level. 565 566 Family Hylidae Rafinesque, 1815 567 Genus Hyla Laurenti, 1768 568 Hyla savignyi Audouin, 1827 569 Hyla gr. H. savignyi 570 571 (Fig. 6J-6O) 572 Localities and material examined. Shet-Irgyz 1, GIN 1106/1001-AM01, 1 left ilium; Novaya 573 Stanitsa 1A, GIN 948/2001-AM20, 1 maxilla, GIN 948/2001-AM13, 1 scapula and GIN 948/2001- 574 AM14, 1 sacral vertebra; Lezhanka 2A, GIN 1130/1001-AM29 – -AM32, 4 ilia, GIN 1130/1001- 575 AM33 – -AM36, 4 scapulae and GIN 1130/1001-AM41, 1 trunk vertebra; Cherlak, GIN 576 1130/1001-AM14 – -AM15, 2 ilia; Olkhovka 1B, GIN 1111/2001-AM02, 1 fragmentary ilium; 577 Pavlodar 2B, GIN 1108/2001-AM01 – -AM03, 3 ilia. 578 Description. The ilia from all localities resemble the same morphology, i.e. the tuber superior is 579 dorsally prominent and slightly laterally. The tuber superior is located at the anterior corner of 580 the acetabulum. The preserved iliac shaft is nearly cylindrical, slightly mediolaterally 581 compressed and is devoid of crista dorsalis. The supraacetabular part of the ilium is smaller 563 of the interiliac facet (Rage & Hossini, 2000). However, further identification of the ilium is 564 impossible, as it does not show relevant differences at the specific level. 572 Localities and material examined. Shet-Irgyz 1, GIN 1106/1001-AM01, 1 left ilium; Novaya 573 Stanitsa 1A, GIN 948/2001-AM20, 1 maxilla, GIN 948/2001-AM13, 1 scapula and GIN 948/2001- 574 AM14, 1 sacral vertebra; Lezhanka 2A, GIN 1130/1001-AM29 – -AM32, 4 ilia, GIN 1130/1001- 575 AM33 – -AM36, 4 scapulae and GIN 1130/1001-AM41, 1 trunk vertebra; Cherlak, GIN 576 1130/1001-AM14 – -AM15, 2 ilia; Olkhovka 1B, GIN 1111/2001-AM02, 1 fragmentary ilium; 577 Pavlodar 2B, GIN 1108/2001-AM01 – -AM03, 3 ilia. 578 Description. The ilia from all localities resemble the same morphology, i.e. the tuber superior is 579 dorsally prominent and slightly laterally. The tuber superior is located at the anterior corner of 580 the acetabulum. The preserved iliac shaft is nearly cylindrical, slightly mediolaterally 581 compressed and is devoid of crista dorsalis. The supraacetabular part of the ilium is smaller 582 than the preacetabular. The ventroposterior margin of the iliac shaft is connected with the pars 583 descendes by an expanded preacetabular zone, building a broad and thin lamina. The 584 acetabulum has a nearly triangular form. Manuscript to be reviewed 585 ventroanterior edge. The posterodorsal corner of the acetabulum ascends and builds a small 586 and prominent acetabular tuber (Fig. 6L). In medial view, the bone surface is smooth, 587 sometimes with a shallow depression in the middle part of the corpus ossi. In distal view, the 588 junctura ilioischiadica is slender, the acetabulum is high and the interiliac facet displays a well- 589 pronounced ventromedial expansion. The acetabular dorsal tuber is higher than the ventral one 590 (Fig. 6K). 591 The scapula, a triradiate element of the pectoral girdle, is comparatively long (Fig. 6M-6O). The 592 bone surface is relatively smooth and is pierced by several foramina. The corpus scapulae, the 593 middle part of the bone, is slender and long. The pars suprascapularis is preserved in a 594 fragmentary state and most probably was not high. In dorsal view, the elongate pars acromialis 595 is narrow and almost equal in length (Fig. 6M). The shorter and flattened processus glenoidalis 596 is slightly broad. The processus glenoidalis and pars acromialis are separated by relatively deep 597 sinus interglenoidalis (Fig. 6N). The margo posterior, at the corner of the processus gleinoidalis 598 and corpus scapula, possesses an oval to elongated-oval angular fossa (Fig. 6N, 6O). The tear- 599 shaped glenoid fossa reaches the posterior corner of the processus glenoidalis. The crista 600 supraglenoidalis is slightly pronounced. 585 ventroanterior edge. The posterodorsal corner of the acetabulum ascends and builds a small 586 and prominent acetabular tuber (Fig. 6L). In medial view, the bone surface is smooth, 585 ventroanterior edge. The posterodorsal corner of the acetabulum ascends and builds a small 586 and prominent acetabular tuber (Fig. 6L). In medial view, the bone surface is smooth, 594 fragmentary state and most probably was not high. In dorsal view, the elongate pars acromialis 595 is narrow and almost equal in length (Fig. 6M). The shorter and flattened processus glenoidalis 596 is slightly broad. The processus glenoidalis and pars acromialis are separated by relatively deep 597 sinus interglenoidalis (Fig. 6N). The margo posterior, at the corner of the processus gleinoidalis 598 and corpus scapula, possesses an oval to elongated-oval angular fossa (Fig. 6N, 6O). The tear- 599 shaped glenoid fossa reaches the posterior corner of the processus glenoidalis. The crista 600 supraglenoidalis is slightly pronounced. 601 Comparison and comments. Manuscript to be reviewed The acetabular rim is prominent at its high PeerJ reviewing PDF \| (2016:08:12606:4:0:NEW 23 Jan 2017) PeerJ reviewing PDF \| (2016:08:12606:4:0:NEW 23 Jan 2017) Manuscript to be reviewed The Siberian fossil tree frog differs from already described fossils 602 and some recent species of the genus Hyla. The following recent material is available for 603 comparison: Hyla savignyi, Armenia (four individuals, unnr. GPIT specimen), Hyla orientalis, 604 Armenia (two individuals, unnr. GPIT specimen) and Hyla arborea, Germany ? (one individual, 605 unnr. GPIT specimen). The Siberian forms can be distinguished from Hyla sp. (Rudabánya 606 locality in Hungary, middle Late Miocene (Roček, 2005); Bois Roche Cave in France, early Late PeerJ reviewing PDF \| (2016:08:12606:4:0:NEW 23 Jan 2017) Manuscript to be reviewed 629 superior and the form of the junctura ilioischiadica. Because of the observed differences in both 630 the Recent and fossil forms, as well as the similarities to H. savignyi, we assume that the fossil 631 tree frogs from Western Siberian and the Russian Platform, probably represent a new form 632 related to the group of Hyla savignyi. 633 634 Family Bufonidae Gray, 1825 635 Genus Bufo Laurenti, 1768 636 Bufo bufo (Linnaeus, 1758) 637 638 (Fig. 6P-6W) 639 Localities and material examined. Novaya Stanitsa 1A, GIN 948/2001-AM15, 1 left and GIN 640 948/2001-AM16 – -17, 2 right ilia, GIN 948/2001-AM18 – -19, 2 trunk vertebrae; Borki 1A, GIN 641 1115/1001-AM02, 1 sacral vertebra, GIN 1115/1001-AM03, 1 left ilium; Olkhovka 1B, GIN 642 1111/2001-AM04, 1 left, GIN 1111/2001-AM03, 2 right ilia and GIN 1111/2001-AM05, 1 trunk 643 vertebra; Olkhovka 1C, GIN 1111/3001-AM01, 1 left scapula, GIN 1111/3001-AM02, 1 trunk 644 vertebra and GIN 1111/3001-AM03, 1 urostyle; Lezhanka 2A, GIN 1130/1001-AM37, 1 left ilia, 645 GIN 1130/1001-AM38, 1 left scapula, GIN 1130/1001-AM39, 1 sacral and GIN 1130/1001-AM40, 646 1 trunk vertebrae; Isakovka 1B, GIN 1131/3001-AM01, 1 left ilium; Isakovka 1A: GIN 1131/1001- 647 AM01, -AM05, 2 right ilia; Peshniovo 3, GIN 1118/3001-AN01, 1 sacral vertebra; Lezhanka 1, 648 GIN 1129/1001-AM04, 1 trunk vertebra; Andreievka 1, GIN 1112/2001-AM01 1 right scapula. 649 Description and comments. The ilia are large and have a robust corpus ossi. The spiral groove is 650 broad and very shallow. The tuber superior is broad, low, covered with irregular tubercles, and Manuscript to be reviewe 629 superior and the form of the junctura 630 the Recent and fossil forms, as well a 631 tree frogs from Western Siberian and 632 related to the group of Hyla savignyi. 633 634 Family Bufonidae Gray, 1825 635 Genus Bufo Laurenti, 1768 636 Bufo bufo (Linnaeus, 1758) 637 638 (Fig. 6P-6W) 639 Localities and material examined. No 640 948/2001-AM16 – -17, 2 right ilia, GIN 641 1115/1001-AM02, 1 sacral vertebra, G 642 1111/2001-AM04, 1 left, GIN 1111/20 643 vertebra; Olkhovka 1C, GIN 1111/300 644 vertebra and GIN 1111/3001-AM03, 1 645 GIN 1130/1001-AM38, 1 left scapula, 646 1 trunk vertebrae; Isakovka 1B, GIN 1 647 AM01, -AM05, 2 right ilia; Peshniovo 648 GIN 1129/1001-AM04, 1 trunk verteb 649 Description and comments. The ilia a 650 broad and very shallow. The tuber su 629 superior and the form of the junctura ilioischiadica. Manuscript to be reviewed 607 Pleistocene (Blain & Villa, 2006)), Hyla arborae (TD8 locality in Spain, early Middle Pleistocene 608 (Blain, 2009)), Hyla cf. arborea (Mátraszőlős 2 locality in Hungary, middle Middle Miocene 609 (Venczel, 2004)), Hyla gr. H. arborea (Capo Mannu D1 Local Fauna in Italy, Late Pliocene 610 (Delfino, Bailon & Pitruzzella, 2011)), Hyla aff. japonica (Tologoy 38x, Baikal Lake, Russia, late 611 Late Pleistocene (Ratnikov, 1997)) and Recent Hyla japonica (Nokariya, 1983) in having: (a) a 612 fossa supragleinoidalis; (b) a slenderer and lower corpus scapula and pars suprascapularis; and 613 (c) a shorter and broader processus glenoidalis. Apart from these differences, the Siberian fossil 614 tree frogs resemble Hyla sp. from the Bois Roche Cave, France (Blain & Villa, 2006), and H. 615 arborea (one individual, unnr. GPIT specimen) in possessing a low and broad processus 616 glenoidalis. The Recent H. savignyi is the only tree frog showing a fossa supragleinoidalis like 617 the one present in the studied remains. The Recent H. savignyi also possesses other similarities 618 to the fossil tree frog, namely: (1) a slender junctura ilioischiadica; (2) the same position of the 619 tuber superior; (3) comparable acetabular tubers; and (4) a similar slightly curved pars 620 ascendens. There are, however, also differences between the Recent H. savignyi and the fossil 621 tree frog. The fossil tree frog has: (1) a dorsally and slightly laterally prominent tuber superior; 622 (2) a deeper and larger fossa supragleinoidalis; and (3) a ventromedial expansion of the 623 interiliac facet; whereas H. savigyni has: (1) a dorsally and laterally significantly prominent 624 tuber superior; (2) a shallow and small angular fossa; and (3) the interiliac facet devoid of 625 ventromedial expansion. Among other fossil tree frogs, the Western Siberian Hyla sp. has the 626 lowest and broadest processus glenoidalis. Another fossil tree frog Hyla sp. reported from the 627 Russian Platform in the Kuznetsovka locality (0.5-0.65 Ma) (Ratnikov, 2002: fig. 2), displays a 628 similar morphology of the ilium as in the Siberian fossil, i.e. the orientation of the tuber PeerJ reviewing PDF \| (2016:08:12606:4:0:NEW 23 Jan 2017) PeerJ reviewing PDF \| (2016:08:12606:4:0:NEW 23 Jan 2017) Manuscript to be reviewed Because of the observed differences in both 630 the Recent and fossil forms, as well as the similarities to H. savignyi, we assume that the fossil 631 tree frogs from Western Siberian and the Russian Platform, probably represent a new form 632 related to the group of Hyla savignyi. 629 superior and the form of the junctura ilioischiadica. Because of the observed differences in both 630 the Recent and fossil forms, as well as the similarities to H. savignyi, we assume that the fossil 631 tree frogs from Western Siberian and the Russian Platform, probably represent a new form 632 related to the group of Hyla savignyi. 639 Localities and material examined. Novaya Stanitsa 1A, GIN 948/2001-AM15, 1 left and GIN 640 948/2001-AM16 – -17, 2 right ilia, GIN 948/2001-AM18 – -19, 2 trunk vertebrae; Borki 1A, GIN 641 1115/1001-AM02, 1 sacral vertebra, GIN 1115/1001-AM03, 1 left ilium; Olkhovka 1B, GIN 642 1111/2001-AM04, 1 left, GIN 1111/2001-AM03, 2 right ilia and GIN 1111/2001-AM05, 1 trunk 643 vertebra; Olkhovka 1C, GIN 1111/3001-AM01, 1 left scapula, GIN 1111/3001-AM02, 1 trunk 644 vertebra and GIN 1111/3001-AM03, 1 urostyle; Lezhanka 2A, GIN 1130/1001-AM37, 1 left ilia, 645 GIN 1130/1001-AM38, 1 left scapula, GIN 1130/1001-AM39, 1 sacral and GIN 1130/1001-AM40, 646 1 trunk vertebrae; Isakovka 1B, GIN 1131/3001-AM01, 1 left ilium; Isakovka 1A: GIN 1131/1001- 647 AM01, -AM05, 2 right ilia; Peshniovo 3, GIN 1118/3001-AN01, 1 sacral vertebra; Lezhanka 1, 648 GIN 1129/1001-AM04, 1 trunk vertebra; Andreievka 1, GIN 1112/2001-AM01 1 right scapula. 649 Description and comments. The ilia are large and have a robust corpus ossi. The spiral groove is 650 broad and very shallow. The tuber superior is broad, low, covered with irregular tubercles, and PeerJ reviewing PDF \| (2016:08:12606:4:0:NEW 23 Jan 2017) Manuscript to be reviewed 651 it is situated above the acetabulum (Fig. 6P). The smooth and concave pars descendens is more 652 developed than the pars ascendens. The ventral edge of the pars descendens is thin and 653 lamelliform. The preacetabular fossa is absent (Fig. 6P). In posterior view, the anterolateral 654 edge of the acetabular is strongly curved. The junctura ilioischiadica shows a higher acetabular 655 ventral tuber than the dorsal tuber, and the ventral half of the corpus ossi projects 656 ventromedially (Fig. 6P). 657 The scapula is a robust bone and is longer than it is high. The material is represented by all 658 ontogenetic series. The angular fossa is absent; a shallow groove on the ventral side of the pars 659 acromialis is present and well pronounced in larger individuals. The pars acromialis and corpus 660 scapula have nearly the same height. The pars suprascapularis laterally increases in height. The 661 pars suprascapularis and corpus scapulae (anterior) have smooth surfaces. The base of the 662 lateral edge of the fossa glenoidalis is elevated but does not project laterally. The crista 663 supraglenoidalis is well developed in larger individuals. The anterior margin is concave. The 664 base of the pars acromalis is high and thin (Fig. 6R).There is a shallow and expanded depression 665 in ventral view. The anteromadial margin of the pars acromalis possesses a low tubercle. The 666 transition from the corpus scapula to the pars acromialis is nearly straight and the wall is thin 667 (Fig 6S, 6T). 668 In several localities, the large-sized frog vertebrae and urostyle (Fig. 6U-6W) are present in 669 association with diagnostic elements (ilia and scapula) (e.g. Olkhovka 1C locality) or are isolated 651 it is situated above the acetabulum (Fig. 6P). The smooth and concave pars descendens is more 652 developed than the pars ascendens. The ventral edge of the pars descendens is thin and 653 lamelliform. The preacetabular fossa is absent (Fig. 6P). In posterior view, the anterolateral 654 edge of the acetabular is strongly curved. The junctura ilioischiadica shows a higher acetabular 655 ventral tuber than the dorsal tuber, and the ventral half of the corpus ossi projects 656 ventromedially (Fig. 6P). 651 it is situated above the acetabulum (Fig. 6P). The smooth and concave pars descendens is more 652 developed than the pars ascendens. The ventral edge of the pars descendens is thin and 653 lamelliform. Manuscript to be reviewed The preacetabular fossa is absent (Fig. 6P). In posterior view, the anterolateral 654 edge of the acetabular is strongly curved. The junctura ilioischiadica shows a higher acetabular 655 ventral tuber than the dorsal tuber, and the ventral half of the corpus ossi projects 656 ventromedially (Fig. 6P). 657 The scapula is a robust bone and is longer than it is high. The material is represented by all 658 ontogenetic series. The angular fossa is absent; a shallow groove on the ventral side of the pars 659 acromialis is present and well pronounced in larger individuals. The pars acromialis and corpus 660 scapula have nearly the same height. The pars suprascapularis laterally increases in height. The 661 pars suprascapularis and corpus scapulae (anterior) have smooth surfaces. The base of the 662 lateral edge of the fossa glenoidalis is elevated but does not project laterally. The crista 663 supraglenoidalis is well developed in larger individuals. The anterior margin is concave. The 664 base of the pars acromalis is high and thin (Fig. 6R).There is a shallow and expanded depression 665 in ventral view. The anteromadial margin of the pars acromalis possesses a low tubercle. The 666 transition from the corpus scapula to the pars acromialis is nearly straight and the wall is thin 667 (Fig 6S, 6T). 668 In several localities, the large-sized frog vertebrae and urostyle (Fig. 6U-6W) are present in 669 association with diagnostic elements (ilia and scapula) (e.g. Olkhovka 1C locality) or are isolated 670 (e.g. Pehsniovo 3 locality). These individuals of the same size can be assigned to the large Bufo 671 bufo. The morphological traits described above (e.g. lack of angular fossa on the scapula and 672 preacetabular fossa on ilium, general outline, form, and size of the scapula and ilium) as well as 668 In several localities, the large-sized frog vertebrae and urostyle (Fig. 6U-6W) are present in 669 association with diagnostic elements (ilia and scapula) (e.g. Olkhovka 1C locality) or are isolated 670 (e.g. Pehsniovo 3 locality). These individuals of the same size can be assigned to the large Bufo 671 bufo. The morphological traits described above (e.g. lack of angular fossa on the scapula and 672 preacetabular fossa on ilium, general outline, form, and size of the scapula and ilium) as well as 668 In several localities, the large-sized frog vertebrae and urostyle (Fig. Manuscript to be reviewed 6U-6W) are present in 669 association with diagnostic elements (ilia and scapula) (e.g. Olkhovka 1C locality) or are isolated 670 (e.g. Pehsniovo 3 locality). These individuals of the same size can be assigned to the large Bufo 671 bufo. The morphological traits described above (e.g. lack of angular fossa on the scapula and 672 preacetabular fossa on ilium, general outline, form, and size of the scapula and ilium) as well as PeerJ reviewing PDF \| (2016:08:12606:4:0:NEW 23 Jan 2017) Manuscript to be reviewed 673 the bone dimensions are the same as those found in the common toad Bufo bufo (Blain, Gibert 674 & Ferràndez-Cañadell, 2010). 675 676 Genus Bufotes Rafinesque, 1815 677 Bufotes viridis Laurenti, 1768 678 679 (Fig. 6X-6Z) 680 Localities and material examined. Baikadam, GIN 950/2001-AM02 – -AM04, 3 left and GIN 681 950/2001-AM05 – -AM09, 5 right ilia; Malyi Kalkaman 1, GIN 1107/1001-AM02 and -AM03, 1 682 left and 1 right scapulae; Malyi Kalkaman 2, GIN 1107/2001-AM03, 1 right scapula; Znamenka, 683 GIN 1109/1001-AM01 and -AM02, 1 left and 1 right scapulae, GIN 1109/1001-AM03 – -AM07, 5 684 left and GIN 1109/1001-AM08 – -AM11, 4 right ilia; Pavlodar 1A, GIN 640/5001-AM01 – -AM24, 685 240 left and GIN 640/5001-AM25 – -AM58, 34 right ilia, GIN 640/5001-AM63 – -AM77, 15 left 686 and GIN 640/5001-AM78 – -AM88, 11 right scapulae; Cherlak, GIN 1110/2001-AM16, 1 right 687 ilium; Selety 1A, GIN 951/1001-AM08 – -AM10, 3 left and GIN 951/1001-AM11 – -AM14, 4 right 688 ilia; Isakovka 1A, GIN 1131/1001-AM02 – -AM04, 3 left ilia; Kedey, GIN 951/2001-AM01 and – - 689 AM02, 1 left and 1 right ilia; Lebiazhie 1A, GIN 950/3001-AM01, 1 left scapula, GIN 950/3001- 690 AM01 2 left ilia; Lebiazhie 1B, GIN 950/4001-AM01, -AM02, 2 right ilia. 691 Description and comments. The iliac shaft is slightly lateromedially compressed and bears a 692 weakly pronounced depression along the middle section The spiral groove between the corpus 673 the bone dimensions are the same as those found in the common toad Bufo bufo (Blain, Gibert 674 & Ferràndez-Cañadell, 2010). 680 Localities and material examined. Baikadam, GIN 950/2001-AM02 – -AM04, 3 left and GIN 681 950/2001-AM05 – -AM09, 5 right ilia; Malyi Kalkaman 1, GIN 1107/1001-AM02 and -AM03, 1 682 left and 1 right scapulae; Malyi Kalkaman 2, GIN 1107/2001-AM03, 1 right scapula; Znamenka, 683 GIN 1109/1001-AM01 and -AM02, 1 left and 1 right scapulae, GIN 1109/1001-AM03 – -AM07, 5 684 left and GIN 1109/1001-AM08 – -AM11, 4 right ilia; Pavlodar 1A, GIN 640/5001-AM01 – -AM24, 685 240 left and GIN 640/5001-AM25 – -AM58, 34 right ilia, GIN 640/5001-AM63 – -AM77, 15 left 686 and GIN 640/5001-AM78 – -AM88, 11 right scapulae; Cherlak, GIN 1110/2001-AM16, 1 right 687 ilium; Selety 1A, GIN 951/1001-AM08 – -AM10, 3 left and GIN 951/1001-AM11 – -AM14, 4 right 688 ilia; Isakovka 1A, GIN 1131/1001-AM02 – -AM04, 3 left ilia; Kedey, GIN 951/2001-AM01 and – - 689 AM02, 1 left and 1 right ilia; Lebiazhie 1A, GIN 950/3001-AM01, 1 left scapula, GIN 950/3001- 690 AM01 2 left ilia; Lebiazhie 1B, GIN 950/4001-AM01, -AM02, 2 right ilia. 691 Description and comments. The iliac shaft is slightly lateromedially compressed and bears a 692 weakly pronounced depression along the middle section. The spiral groove between the corpus 693 ossi and iliac shaft is weakly developed. The tuber superior is low and possesses a uni- or 691 Description and comments. The iliac shaft is slightly lateromedially compressed and bears a 692 weakly pronounced depression along the middle section. The spiral groove between the corpus 693 ossi and iliac shaft is weakly developed. The tuber superior is low and possesses a uni- or PeerJ reviewing PDF \| (2016:08:12606:4:0:NEW 23 Jan 2017) Manuscript to be reviewed 694 bilabiate protuberance in its central part. The angular fossa is well pronounced. The 695 anteroventral edge of the acetabular rim is straight. The pars descendens projects sharply 696 ventrally. There is no observable ʻcalamitaʼ ridge (Fig. 6X). The remains show typical features for 697 Bufotes viridis: i.e. the form and shape of the tuber superior and acetabulum (Böhme, 1977; 698 Blain, Gibert & Ferràndez-Cañadell, 2010). Due to the absence of well-preserved material, we 699 prefer to tentatively assign the remains to the Bufotes viridis group. 700 701 Bufo sp. 702 Localities and material examined. Cherlak, GIN 1110/2001-AM17, 1 left scapula; Olkhovka 1A, 703 GIN 1111/1001-AM01, -AM02, 2 left ilia; Pavlodar 2B, GIN 1108/2001-AM04 – -AM06, 3 left ilia. 704 Description and comments. The greatly damaged ilia exhibits the typical morphology of the 705 genus Bufo, i.e. the iliac shaft without the dorsal crest and a spiral groove between the shaft 706 and corpus ilii (Böhme, 1977). There is a preacetabular fossa in the caudoventral corner of the 707 acetabulum. The tuber superior is eroded. In medial view, the pars descendens is 708 ventromedially directed. 709 710 Family Ranidae Batsch, A. J. G. K., 1796 711 Genus Pelophylax Fitzinger, 1843 712 Pelophylax sp. 713 694 bilabiate protuberance in its central part. The angular fossa is well pronounced. The 695 anteroventral edge of the acetabular rim is straight. The pars descendens projects sharply 696 ventrally. There is no observable ʻcalamitaʼ ridge (Fig. 6X). The remains show typical features for 697 Bufotes viridis: i.e. the form and shape of the tuber superior and acetabulum (Böhme, 1977; 698 Blain, Gibert & Ferràndez-Cañadell, 2010). Due to the absence of well-preserved material, we 699 prefer to tentatively assign the remains to the Bufotes viridis group. 694 bilabiate protuberance in its central part. The angular fossa is well pronounced. The 695 anteroventral edge of the acetabular rim is straight. The pars descendens projects sharply 696 ventrally. There is no observable ʻcalamitaʼ ridge (Fig. 6X). The remains show typical features for 697 Bufotes viridis: i.e. the form and shape of the tuber superior and acetabulum (Böhme, 1977; 698 Blain, Gibert & Ferràndez-Cañadell, 2010). Due to the absence of well-preserved material, we 699 prefer to tentatively assign the remains to the Bufotes viridis group. 702 Localities and material examined. PeerJ reviewing PDF \| (2016:08:12606:4:0:NEW 23 Jan 2017) Manuscript to be reviewed 715 Localities and material examined. Malyi Kalkaman 1, GIN 1107/1001-AM04, 1 left ilium, GIN 716 1107/1001-AM13, 1 left articular; Malyi Kalkaman 2, GIN 1107/2001-AM04, -AM05, 2 right ilia , 717 and GIN 1107/2001-AM06, 1 left ilium; Petropavlovsk 1, GIN 952/1001-AM01, 1 left ilium; 718 Olkhovka 1C, GIN 1111/3001-AM04, 1 right ilium; Kamyshovo, GIN 1107/1001-AM01, 1 right 719 scapula; Lezhanka 1, GIN 1129/1001-AM05, 1 left and GIN 1129/1001-AM06, 1 right ilia, GIN 720 1129/1001-AM07, 1 left scapula; Andreevka 1, GIN 1112/2001-AM02, 1 right and GIN 721 1112/2001-AM03, 1 left ilia; Livenka, GIN 1129/2001-AM01, 1 right ilium. 722 Description and comments. The ilia have a strong, oval, nearly vertically oriented and ventrally 723 well-defined high tuber superior. The dorsal crest is high; anteriorly it is often broken. The tuber 724 superior is high and slightly more S-shaped than the crest; a well-developed supraacetabular 725 fossa is present. Posterior to the tuber, the dorsal margin of the bone is bent ventrally towards 726 the acetabulum. In posterior view, the tuber superior is curved ventromedially (Fig. 6AA). The 727 junctura ilioschiadica is damaged but based on the preserved structures we speculate that it 728 was tall (Fig. 6AB). 729 The scapula is an elongated and short bone. In ventral view, a weakly-developed crista 730 supraglenoidalis is observable. It runs subparallel to the margo posterior and reaches the 731 middle part of the pars suprascapulars (Fig. 6AC, 6AD). 715 Localities and material examined. Malyi Kalkaman 1, GIN 1107/1001-AM04, 1 left ilium, GIN 716 1107/1001-AM13, 1 left articular; Malyi Kalkaman 2, GIN 1107/2001-AM04, -AM05, 2 right ilia , 717 and GIN 1107/2001-AM06, 1 left ilium; Petropavlovsk 1, GIN 952/1001-AM01, 1 left ilium; 718 Olkhovka 1C, GIN 1111/3001-AM04, 1 right ilium; Kamyshovo, GIN 1107/1001-AM01, 1 right 719 scapula; Lezhanka 1, GIN 1129/1001-AM05, 1 left and GIN 1129/1001-AM06, 1 right ilia, GIN 720 1129/1001-AM07, 1 left scapula; Andreevka 1, GIN 1112/2001-AM02, 1 right and GIN 721 1112/2001-AM03, 1 left ilia; Livenka, GIN 1129/2001-AM01, 1 right ilium. 722 Description and comments. The ilia have a strong, oval, nearly vertically oriented and ventrally 723 well-defined high tuber superior. The dorsal crest is high; anteriorly it is often broken. The tuber 724 superior is high and slightly more S-shaped than the crest; a well-developed supraacetabular 725 fossa is present. Posterior to the tuber, the dorsal margin of the bone is bent ventrally towards 726 the acetabulum. Manuscript to be reviewed Cherlak, GIN 1110/2001-AM17, 1 left scapula; Olkhovka 1A, 703 GIN 1111/1001-AM01, -AM02, 2 left ilia; Pavlodar 2B, GIN 1108/2001-AM04 – -AM06, 3 left ilia. 704 Description and comments. The greatly damaged ilia exhibits the typical morphology of the 705 genus Bufo, i.e. the iliac shaft without the dorsal crest and a spiral groove between the shaft 706 and corpus ilii (Böhme, 1977). There is a preacetabular fossa in the caudoventral corner of the 707 acetabulum. The tuber superior is eroded. In medial view, the pars descendens is 708 ventromedially directed. Manuscript to be reviewed In posterior view, the tuber superior is curved ventromedially (Fig. 6AA). The 727 junctura ilioschiadica is damaged but based on the preserved structures we speculate that it 728 was tall (Fig. 6AB). 729 The scapula is an elongated and short bone. In ventral view, a weakly-developed crista 730 supraglenoidalis is observable. It runs subparallel to the margo posterior and reaches the 731 middle part of the pars suprascapulars (Fig. 6AC, 6AD). 729 The scapula is an elongated and short bone. In ventral view, a weakly-developed crista l id li i b bl I b ll l h i d h h 730 supraglenoidalis is observable. It runs subparallel to the margo posterior and reaches the 731 middle part of the pars suprascapulars (Fig. 6AC, 6AD). p g p g p 731 middle part of the pars suprascapulars (Fig. 6AC, 6AD). 732 The characters listed above, i.e. like the form and orientation of bones, tuber superior and 733 crista supraglenoidalis, allow for the attribution of the fossils to the genus of the green (water) 734 frogs Pelophylax (Böhme, 1977; Sanchíz, Schleich & Esteban, 1993; Bailon, 1999; Blain, Bailon & 735 Agustí, 2007). Any further identification is impossible due to the fragmentary preservation of 736 the material. 732 The characters listed above, i.e. like the form and orientation of bones, tuber superior and 733 crista supraglenoidalis, allow for the attribution of the fossils to the genus of the green (water) 734 frogs Pelophylax (Böhme, 1977; Sanchíz, Schleich & Esteban, 1993; Bailon, 1999; Blain, Bailon & 735 Agustí, 2007). Any further identification is impossible due to the fragmentary preservation of 736 the material. 732 The characters listed above, i.e. like the form and orientation of bones, tuber superior and 733 crista supraglenoidalis, allow for the attribution of the fossils to the genus of the green (water) 734 frogs Pelophylax (Böhme, 1977; Sanchíz, Schleich & Esteban, 1993; Bailon, 1999; Blain, Bailon & 735 Agustí, 2007). Any further identification is impossible due to the fragmentary preservation of 736 the material. PeerJ reviewing PDF \| (2016:08:12606:4:0:NEW 23 Jan 2017) Manuscript to be reviewed 737 738 Genus Rana Linnaeus, 1758 (Linnaeus, 1758) 739 Rana sp. / Rana temporaria Nilsson, 1842 (Nilsson, 1842) 740 741 (Fig. 6AE-6AH) 742 Localities and material examined. Ayakoz, unnr. HC specimens, numerous ilia, Baikadam, GIN 743 950/2001-AM10, 1 left, GIN 950/2001-AM11 – -AM13, and 3 right ilia; Malyi Kalkaman 1, GIN 744 1107/1001-AM05 – -AM09, 5 left ilia, GIN 1107/1001-AM10, 1 right ilia, GIN 1107/1001-AM01, - 745 AM11, 2 right scapula; Malyi Kalkaman 2, GIN 1107/2001-AM07, 1 right ilium, GIN 1107/2001- 746 AM08 – -AM13, 6 left ilia; Olkhovka 1C, GIN 1111/3001-AM05, 1 right ilium; Lezhanka 1, GIN 747 1129/1001-AM08, 1 left ilium; Kentyubek, unnr. HC specimens, 2 left ilia. 748 Description. The ilia have a reduced, compact, anteriorly directed and low tuber superior. The 738 Genus Rana Linnaeus, 1758 (Linnaeus, 1758) 742 Localities and material examined. Ayakoz, unnr. HC specimens, numerous ilia, Baikadam, GIN 743 950/2001-AM10, 1 left, GIN 950/2001-AM11 – -AM13, and 3 right ilia; Malyi Kalkaman 1, GIN 744 1107/1001-AM05 – -AM09, 5 left ilia, GIN 1107/1001-AM10, 1 right ilia, GIN 1107/1001-AM01, - 745 AM11, 2 right scapula; Malyi Kalkaman 2, GIN 1107/2001-AM07, 1 right ilium, GIN 1107/2001- 746 AM08 – -AM13, 6 left ilia; Olkhovka 1C, GIN 1111/3001-AM05, 1 right ilium; Lezhanka 1, GIN 747 1129/1001-AM08, 1 left ilium; Kentyubek, unnr. HC specimens, 2 left ilia. 748 Description. The ilia have a reduced, compact, anteriorly directed and low tuber superior. The 749 lateral surface is rough. The dorsal crest is low. The pars descendens is more developed than 750 the pars ascendens (Fig.6AE). In posterior view, the junctura ilioschiadica Is low (Fig. 6AF) in 751 comparison to the ilium of Pelophylax sp. (Fig. 6AA). The tuber superior projects dorsolaterally. 752 The pars descendens projects medially (Fig. 6AE). 748 Description. The ilia have a reduced, compact, anteriorly directed and low tuber superior. The 749 lateral surface is rough. The dorsal crest is low. The pars descendens is more developed than 750 the pars ascendens (Fig.6AE). In posterior view, the junctura ilioschiadica Is low (Fig. 6AF) in 751 comparison to the ilium of Pelophylax sp. (Fig. 6AA). The tuber superior projects dorsolaterally. 752 The pars descendens projects medially (Fig. 6AE). 753 The middle portions of both scapulae are preserved without the proximal parts of the pars 754 acromialis and suprascapularis. In dorsal view, a crista supraglenoidalis is observable at the 755 processus glenoidalis, which continues until the pars suprascapularis along the longitudinal axis 756 of the bone. It is very prominent and forms a lamelliform convex ridge. The base of the 757 processus glenoidalis is high and straight (Fig. 6AG, 6AH). 753 The middle portions of both scapulae are preserved without the proximal parts of the pars 754 acromialis and suprascapularis. In dorsal view, a crista supraglenoidalis is observable at the 755 processus glenoidalis, which continues until the pars suprascapularis along the longitudinal axis 756 of the bone. It is very prominent and forms a lamelliform convex ridge. The base of the 757 processus glenoidalis is high and straight (Fig. 6AG, 6AH). PeerJ reviewing PDF \| (2016:08:12606:4:0:NEW 23 Jan 2017) PeerJ reviewing PDF \| (2016:08:12606:4:0:NEW 23 Jan 2017) Manuscript to be reviewed The teeth are slender, unicuspid, and not narrowly arranged. All maxillaries 786 and dentary teeth are straight, except the most anterior ones on the dentary, which are 787 anteriorly lightly oblique. The central teeth on dental lamina of both the maxilla and dentary 788 are larger than the anterior and posterior ones (Fig. 7C, 7G). The cusps of maxilla teeth are 789 rarely posteriorly oriented. The most complete dentary bone contains at least 17 (in total, 790 probably 20) teeth, counted by both teeth and their alveoles (Fig. 7B-7D). 785 Tooth morphology. The teeth are slender, unicuspid, and not narrowly arranged. All maxillaries 786 and dentary teeth are straight, except the most anterior ones on the dentary, which are 787 anteriorly lightly oblique. The central teeth on dental lamina of both the maxilla and dentary 788 are larger than the anterior and posterior ones (Fig. 7C, 7G). The cusps of maxilla teeth are 789 rarely posteriorly oriented. The most complete dentary bone contains at least 17 (in total, 790 probably 20) teeth, counted by both teeth and their alveoles (Fig. 7B-7D). 785 Tooth morphology. The teeth are slender, unicuspid, and not narrowly arranged. All maxillaries 786 and dentary teeth are straight, except the most anterior ones on the dentary, which are 787 anteriorly lightly oblique. The central teeth on dental lamina of both the maxilla and dentary 788 are larger than the anterior and posterior ones (Fig. 7C, 7G). The cusps of maxilla teeth are 789 rarely posteriorly oriented. The most complete dentary bone contains at least 17 (in total, 790 probably 20) teeth, counted by both teeth and their alveoles (Fig. 7B-7D). 785 Tooth morphology. The teeth are slender, unicuspid, and not narrowly arranged. All maxillaries 786 and dentary teeth are straight, except the most anterior ones on the dentary, which are 787 anteriorly lightly oblique. The central teeth on dental lamina of both the maxilla and dentary 788 are larger than the anterior and posterior ones (Fig. 7C, 7G). The cusps of maxilla teeth are 789 rarely posteriorly oriented. The most complete dentary bone contains at least 17 (in total, 790 probably 20) teeth, counted by both teeth and their alveoles (Fig. 7B-7D). 791 Dentaries. The dentary is a slender and elongated. In the symphyseal region, the bone is slightly 792 medially curved. Manuscript to be reviewed 758 Comments. The ilia and scapulae morphology strongly resembles that of brown frogs genus, 759 Rana (Böhme, 1977). Due to the fragmentary preservation of the bone material, any precise 760 taxonomic identification of the frogs from nearly all localities was impossible. The comparison 761 with Recent species (e.g. Rana temporaria (unnr. GPIT specimen), Rana dalamtina (unnr. GPIT 762 specimen; Bailon, 1999), Rana graeca (unnr. GPIT specimen), Rana arvalis (unnr. GPIT 763 specimen), Rana dybowskii (MNCN 40459), Rana amurensis (unnr. GPIT specimen) etc.) 764 revealed more similarities with the European and Western Asiatic species rather than to 765 Eastern Asiatic brown frogs. 758 Comments. The ilia and scapulae morphology strongly resembles that of brown frogs genus, 759 Rana (Böhme, 1977). Due to the fragmentary preservation of the bone material, any precise 760 taxonomic identification of the frogs from nearly all localities was impossible. The comparison 761 with Recent species (e.g. Rana temporaria (unnr. GPIT specimen), Rana dalamtina (unnr. GPIT 762 specimen; Bailon, 1999), Rana graeca (unnr. GPIT specimen), Rana arvalis (unnr. GPIT 763 specimen), Rana dybowskii (MNCN 40459), Rana amurensis (unnr. GPIT specimen) etc.) 764 revealed more similarities with the European and Western Asiatic species rather than to 765 Eastern Asiatic brown frogs. 766 Only the Malyi Kalkaman 1 locality provided adequate material for specific identification. The 767 ilia and scapulae from this locality’s material resembled the Recent species, Rana temporaria, 768 which has the widest distribution among the brown frogs in Eurasia. The fossil bones of brown 769 frogs from other Western Siberian localities are described here as Rana sp. Due to the poor 770 preservation of the ilia from the Kentyubek locality, it can be only identified at the family level. 771 772 Class Reptilia Laurenti, 1768 773 Order Squamata Oppel, 1811 774 Suborder Gekkota Cuvier, 1817 775 Family Gekkonidae Gray, 1825 776 Genus Alsophylax Fitzinger, 1843 777 Alsophylax sp. 778 779 (Fig. 7) PeerJ reviewing PDF \| (2016:08:12606:4:0:NEW 23 Jan 2017) Manuscript to be reviewed 780 Locality and material examined. Cherlak, GIN 1110/2001-RE01 – -RE10, 10 right dentaries, GIN 781 1110/2001-RE11 – -RE24, 14 left dentaries, GIN 1110/2001-RE26 – -RE38, 13 left maxillae, GIN 782 1110/2001-RE39 – -RE43, 5 right maxillae, GIN 1110/2001-RE44, 1 cervical trunk vertebra; 783 Mynsualmas-MSA 3: 1 right maxilla, unnr. GPIT specimen. 784 Description. 785 Tooth morphology. The teeth are slender, unicuspid, and not narrowly arranged. All maxillaries 786 and dentary teeth are straight, except the most anterior ones on the dentary, which are 787 anteriorly lightly oblique. The central teeth on dental lamina of both the maxilla and dentary 788 are larger than the anterior and posterior ones (Fig. 7C, 7G). The cusps of maxilla teeth are 789 rarely posteriorly oriented. The most complete dentary bone contains at least 17 (in total, 790 probably 20) teeth, counted by both teeth and their alveoles (Fig. 7B-7D). 791 Dentaries. The dentary is a slender and elongated. In the symphyseal region, the bone is slightly 792 medially curved. The pars ventralis is assumed to be enlarged, due to the bone posteriorly 793 increasing in height. The dentary is characterised by a completely closed Meckelian canal, 794 which runs along approximately two-thirds of the bone length (Fig. 7B). The symphyseal 795 articulation surface is reduced. It does not build a pronounced articulation surface. The ventral 796 surface of the symphysis bears a longitudinal, posteriorly deepening symphyseal groove, visible 780 Locality and material examined. Cherlak, GIN 1110/2001-RE01 – -RE10, 10 right dentaries, GIN 781 1110/2001-RE11 – -RE24, 14 left dentaries, GIN 1110/2001-RE26 – -RE38, 13 left maxillae, GIN 782 1110/2001-RE39 – -RE43, 5 right maxillae, GIN 1110/2001-RE44, 1 cervical trunk vertebra; 783 Mynsualmas-MSA 3: 1 right maxilla, unnr. GPIT specimen. 780 Locality and material examined. Cherlak, GIN 1110/2001-RE01 – -RE10, 10 right dentaries, GIN 781 1110/2001-RE11 – -RE24, 14 left dentaries, GIN 1110/2001-RE26 – -RE38, 13 left maxillae, GIN 782 1110/2001-RE39 – -RE43, 5 right maxillae, GIN 1110/2001-RE44, 1 cervical trunk vertebra; 783 Mynsualmas-MSA 3: 1 right maxilla, unnr. GPIT specimen. 780 Locality and material examined. Cherlak, GIN 1110/2001-RE01 – -RE10, 10 right dentaries, GIN 781 1110/2001-RE11 – -RE24, 14 left dentaries, GIN 1110/2001-RE26 – -RE38, 13 left maxillae, GIN 782 1110/2001-RE39 – -RE43, 5 right maxillae, GIN 1110/2001-RE44, 1 cervical trunk vertebra; 783 Mynsualmas-MSA 3: 1 right maxilla, unnr. GPIT specimen. 784 Description. 785 Tooth morphology. Manuscript to be reviewed 802 direction, also changing in form from a more rounded outline to an oval appearance. The 803 position of the last mental foramen is arranged lingually in front of the posterior opening of the 804 Meckelian canal. The cavity of the Meckelian canal is divided in two, i.e. the upper and lower 805 subcanals, by a distinct horizontal lamella (Fig. 7E). The horizontal lamella runs parallel to the 806 lamina horizontalis and can be observed posteriorly behind the opening of the Meckelian canal. 807 The upper subcanal opens to the labial surface of the bone near to the mental foramina. The 808 symphyseal groove corresponds to the anterior opening of the lower subcanal. In lingual view, 809 the lamina horizontalis is situated in a low position. Its margin is rounded but not prominent. A 810 shallow and anteriorly extending dental shelf divides the lamina horizontalis from the dental 811 lamina (Fig. 7C). Posteriorly, the bone is nearly L-shaped in the transverse section. In all 812 observed specimens, the pars horizontalis is destroyed in the preserved bone. The caudal 813 portion of the paries verticalis shows bifurcation (Fig. 7E), which corresponds to the coronoid 814 insertion. 802 direction, also changing in form from a more rounded outline to an oval appearance. The 803 position of the last mental foramen is arranged lingually in front of the posterior opening of the 804 Meckelian canal. The cavity of the Meckelian canal is divided in two, i.e. the upper and lower 805 subcanals, by a distinct horizontal lamella (Fig. 7E). The horizontal lamella runs parallel to the 806 lamina horizontalis and can be observed posteriorly behind the opening of the Meckelian canal. 807 The upper subcanal opens to the labial surface of the bone near to the mental foramina. The 808 symphyseal groove corresponds to the anterior opening of the lower subcanal. In lingual view, 809 the lamina horizontalis is situated in a low position. Its margin is rounded but not prominent. A 810 shallow and anteriorly extending dental shelf divides the lamina horizontalis from the dental 811 lamina (Fig. 7C). Posteriorly, the bone is nearly L-shaped in the transverse section. In all 812 observed specimens, the pars horizontalis is destroyed in the preserved bone. The caudal 813 portion of the paries verticalis shows bifurcation (Fig. 7E), which corresponds to the coronoid 814 insertion. 815 Maxilla. Manuscript to be reviewed The pars ventralis is assumed to be enlarged, due to the bone posteriorly 793 increasing in height. The dentary is characterised by a completely closed Meckelian canal, 794 which runs along approximately two-thirds of the bone length (Fig. 7B). The symphyseal 795 articulation surface is reduced. It does not build a pronounced articulation surface. The ventral 796 surface of the symphysis bears a longitudinal, posteriorly deepening symphyseal groove, visible 797 in both the lingual and ventral views (Fig. 7B-7D). The Meckelian canal is open posteriorly at 798 about the 15th-16th tooth position. The splenial facet on the dentary, the anterior margin of 799 Meckelian opening, shows a light concave and elongated surface (Fig. 7B-7E). In lateral view, 800 the bone is smooth, and the only complete dentary possesses five foramina that are arranged 801 in a longitudinal row (Fig. 7A). The size of the foramina increases slightly in the anteroposterior 791 Dentaries. The dentary is a slender and elongated. In the symphyseal region, the bone is slightly 792 medially curved. The pars ventralis is assumed to be enlarged, due to the bone posteriorly 793 increasing in height. The dentary is characterised by a completely closed Meckelian canal, 794 which runs along approximately two-thirds of the bone length (Fig. 7B). The symphyseal 795 articulation surface is reduced. It does not build a pronounced articulation surface. The ventral 796 surface of the symphysis bears a longitudinal, posteriorly deepening symphyseal groove, visible 797 in both the lingual and ventral views (Fig. 7B-7D). The Meckelian canal is open posteriorly at 798 about the 15th-16th tooth position. The splenial facet on the dentary, the anterior margin of 799 Meckelian opening, shows a light concave and elongated surface (Fig. 7B-7E). In lateral view, 800 the bone is smooth, and the only complete dentary possesses five foramina that are arranged 801 in a longitudinal row (Fig. 7A). The size of the foramina increases slightly in the anteroposterior PeerJ reviewing PDF \| (2016:08:12606:4:0:NEW 23 Jan 2017) PeerJ reviewing PDF \| (2016:08:12606:4:0:NEW 23 Jan 2017) Manuscript to be reviewed The preserved posterior part of the maxillary possesses a relatively low lacrimal facet 816 of the facial process of the maxilla (pars nasalis sensu Estes (1969)), while the latter is always 817 not preserved. The internal wall of the maxilla posteriorly bears a small distinct longitudinal 818 groove, running parallel to the lamina horizontalis (Fig. 7F-7H). The groove begins at the 819 posterior basis of the lacrimal facet and continuous until the preserved posterior tip of the 820 bone. The groove narrows at the middle section of the bone (at the position of the 3rd-4th last 821 tooth), where the lacrimal facet terminates. The lamina horizontalis is clearly visible, expands 822 laterally just under the tip of the lacrimal facet and builds a palatine facet (Fig. 7F-7H). The 823 lamina horizontalis becomes distinctly and posteriorly narrower but does not diminish fully at 815 Maxilla. The preserved posterior part of the maxillary possesses a relatively low lacrimal facet 816 of the facial process of the maxilla (pars nasalis sensu Estes (1969)), while the latter is always 817 not preserved. The internal wall of the maxilla posteriorly bears a small distinct longitudinal 818 groove, running parallel to the lamina horizontalis (Fig. 7F-7H). The groove begins at the 819 posterior basis of the lacrimal facet and continuous until the preserved posterior tip of the 820 bone. The groove narrows at the middle section of the bone (at the position of the 3rd-4th last 821 tooth), where the lacrimal facet terminates. The lamina horizontalis is clearly visible, expands 822 laterally just under the tip of the lacrimal facet and builds a palatine facet (Fig. 7F-7H). The 823 lamina horizontalis becomes distinctly and posteriorly narrower but does not diminish fully at PeerJ reviewing PDF \| (2016:08:12606:4:0:NEW 23 Jan 2017) Manuscript to be reviewed 824 the posterior end of the bone. The jugal process of the maxilla is bifurcated at its distal end (Fig. 825 7H). The maxillary lappet is damaged, but its base is preserved. The internal wall surface of the 826 maxilla contains few rugosities. Here an anteroposteriorly directed, fairly well-pronounced, 824 the posterior end of the bone. The jugal process of the maxilla is bifurcated at its distal end (Fig. 825 7H). The maxillary lappet is damaged, but its base is preserved. The internal wall surface of the 826 maxilla contains few rugosities. Here an anteroposteriorly directed, fairly well-pronounced, 827 median ridge, is observed. In labial view, several foramina occur above the dental row. Some of 828 these foramina are arranged in a longitudinal line that corresponds to the foramina for the 829 mandibular division of the fifth cranial (trigeminal) nerve. This line runs parallel to the lamina 830 horizontalis. The last foramen in the row pierces the maxilla at the base of the lacrimal facet 831 under its tip. The bases of the facial process and maxillary lappet lay a relatively large superior 832 alveolar canal (sac, Fig. 7J, 7L) for the maxillary nerve and its accompanying blood vessel. The 833 remaining foramina at the maxilla are dispersed irregularly on the bone surface. The 834 premaxillary process is present, but it is highly damaged. The anterior basis of the lacrimal facet 835 is pierced by a foramen. 824 the posterior end of the bone. The jugal process of the maxilla is bifurcated at its distal end (Fig. 825 7H). The maxillary lappet is damaged, but its base is preserved. The internal wall surface of the 826 maxilla contains few rugosities. Here an anteroposteriorly directed, fairly well-pronounced, 827 median ridge, is observed. In labial view, several foramina occur above the dental row. Some of 828 these foramina are arranged in a longitudinal line that corresponds to the foramina for the 829 mandibular division of the fifth cranial (trigeminal) nerve. This line runs parallel to the lamina 830 horizontalis. The last foramen in the row pierces the maxilla at the base of the lacrimal facet 831 under its tip. The bases of the facial process and maxillary lappet lay a relatively large superior 832 alveolar canal (sac, Fig. 7J, 7L) for the maxillary nerve and its accompanying blood vessel. Manuscript to be reviewed The 833 remaining foramina at the maxilla are dispersed irregularly on the bone surface. The 834 premaxillary process is present, but it is highly damaged. The anterior basis of the lacrimal facet 835 is pierced by a foramen. 836 Vertebra. A single cervical vertebra of a gecko specimen shows an elongate amphicoelous 837 centrum (Fig. 7N-7P). The cotyles are approximately circular. In anterior view, the vertebra has 838 a semi-circular outline. In lateral view, the vertebra is anteroposteriorly compressed. The neural 839 arch is concave on both sides. The transverse processes are high, extremely short, and vertically 840 aligned. The distal end of the process is round. The haemal foramina are present at the lower 841 base of the transverse processes. The prezygapophyses are small and slightly prominent. The 842 neural arch is plane and triangular in outline. It possesses a slender and low neural crest. The 843 postzygapophyses are small, nearly invisible and are situated on the ventrolateral edges of the 844 pterygapophysis. 836 Vertebra. A single cervical vertebra of a gecko specimen shows an elongate amphicoelous 837 centrum (Fig. 7N-7P). The cotyles are approximately circular. In anterior view, the vertebra has 838 a semi-circular outline. In lateral view, the vertebra is anteroposteriorly compressed. The neural 839 arch is concave on both sides. The transverse processes are high, extremely short, and vertically 840 aligned. The distal end of the process is round. The haemal foramina are present at the lower 841 base of the transverse processes. The prezygapophyses are small and slightly prominent. The 842 neural arch is plane and triangular in outline. It possesses a slender and low neural crest. The 843 postzygapophyses are small, nearly invisible and are situated on the ventrolateral edges of the 844 pterygapophysis. PeerJ reviewing PDF \| (2016:08:12606:4:0:NEW 23 Jan 2017) Manuscript to be reviewed 845 Comparison and comments. Numerous characters allow for the identification of the material as 846 a member of the family Gekkonidae. These characters are namely: (1) the amphicoelous 847 condition of the vertebra; (2) the maxillae and dentaries bearing numerous pleurodont, 848 isodont, densely packed, cylindrical, and slender monocuspid teeth; (3) the presence of a 849 medially extended dental shelf of the maxilla; and (4) the lingually closed Meckelian canal 850 (Hoffstetter & Gasc, 1969; Daza, Alifanov & Bauer, 2012). The gekkonid remains from the 851 Cherlak locality display a low number of teeth on the dentary (up to 20) and a rounded tooth 852 apex (making the teeth digitiform), which are diagnostic characters for the genus Alsophylax 853 (Nikitina & Ananjeva, 2009). Within the gekkonids, a low number of teeth (up to 20) is also 854 characteristic of Mediodactylus russowi, Phelsuma laticauda, and Ph. serraticauda (Nikitina, 855 2009). The Siberian fossil geckos can be distinguished from Mediodactylus by peculiarities of 856 the maxilla (i.e. the presence of a lingual longitudinal groove and a reduced row of foramina of 857 the trigeminal nerve), the dentary with a distinct and longer horizontal lamella, plus a reduced 858 symphyseal groove. The Recent genus Phelsuma can be excluded from consideration since 859 these geckos are restricted to the islands of the southwest part of the Indian Ocean and belong 860 to another zoogeographic zone. The fossil geckos resemble the Recent species Alsophylax 861 pipiens (see in Estes (1969); tab. 2C) in the presence of the prefrontal process and their short 862 row of foramina of the trigeminal nerve, which terminates below the prefrontal process. 863 Further comparison with the Recent genus Alsophylax is, however, impossible due to the lack of 864 available comparative osteological material of the Recent Alsophylax species. 865 Fossil geckos were present in the Early Miocene of Kazakhstan, as is evident from the 866 Mynsualmas-MSA 3 locality (unnr. GPIT specimen) (Böhme & Ilg, 2003). The re-studying of the 863 Further comparison with the Recent genus Alsophylax is, however, impossible due to the lack of 864 available comparative osteological material of the Recent Alsophylax species. 863 Further comparison with the Recent genus Alsophylax is, however, impossible due to the lack of 864 available comparative osteological material of the Recent Alsophylax species. Manuscript to be reviewed 865 Fossil geckos were present in the Early Miocene of Kazakhstan, as is evident from the 866 Mynsualmas-MSA 3 locality (unnr. GPIT specimen) (Böhme & Ilg, 2003). The re-studying of the PeerJ reviewing PDF \| (2016:08:12606:4:0:NEW 23 Jan 2017) Manuscript to be reviewed 867 material revealed that the posterior fragment of a right maxilla shows morphology similar to 868 Alsophylax sp. from the Cherlak locality in the presence of a lingual longitudinal groove, the 869 absence of foramina at the posterior portion of the bone and a round tooth apex. The fossil 870 material, however, differs in its larger size (Fig. 7Q). Taking this difference as well as the 871 similarities into account, we tentatively consider the Mynsualmas record as cf. Alsophylax sp. 872 This fossil probably represents a larger Alsophylax species than those registered in the Weste 873 Siberia. 874 875 Suborder Lacertilia Owen, 1842 sensu Estes, Queiroz & Gauthier, 1988 876 Family Lacertidae Fitzinger, 1826 877 Genus Lacerta Linnaeus, 1758 878 879 Remarks. The generic assignment of fossil lacertid remains is extremely difficult. This group is 880 anatomically generalised (Lacera sensu lato) and shows very few characteristic features (e.g. 881 bone and teeth morphology) for detailed taxonomic assignments (Böhme, 2010; Böhme & 882 Vasilyan, 2014). 883 884 Lacerta s.l. sp. 1. 885 886 (Fig. 8A) 867 material revealed that the posterior fragment of a right maxilla shows morphology similar to 868 Alsophylax sp. from the Cherlak locality in the presence of a lingual longitudinal groove, the 869 absence of foramina at the posterior portion of the bone and a round tooth apex. The fossil 870 material, however, differs in its larger size (Fig. 7Q). Taking this difference as well as the 871 similarities into account, we tentatively consider the Mynsualmas record as cf. Alsophylax sp. 872 This fossil probably represents a larger Alsophylax species than those registered in the Western 873 Siberia. 874 875 Suborder Lacertilia Owen, 1842 sensu Estes, Queiroz & Gauthier, 1988 876 Family Lacertidae Fitzinger, 1826 877 Genus Lacerta Linnaeus, 1758 878 879 Remarks. The generic assignment of fossil lacertid remains is extremely difficult. This group is 880 anatomically generalised (Lacera sensu lato) and shows very few characteristic features (e.g. PeerJ reviewing PDF \| (2016:08:12606:4:0:NEW 23 Jan 2017) Manuscript to be reviewed 887 Material. Baikadam, GIN 650/2001-RE07 – -RE09, 2 (3?) left dentaries, GIN 650/2001-RE10, 1 888 postsacral vertebra; Pavlodar 1A, GIN 640/5001-RE01 – -RE15, -RE41 – -RE4217 left dentaries, 889 GIN 640/5001-RE16 – -RE25, 10 right dentaries. 887 Material. Baikadam, GIN 650/2001-RE07 – -RE09, 2 (3?) left dentaries, GIN 650/2001-RE10, 1 888 postsacral vertebra; Pavlodar 1A, GIN 640/5001-RE01 – -RE15, -RE41 – -RE4217 left dentaries, 889 GIN 640/5001-RE16 – -RE25, 10 right dentaries. 890 Description. The bones bear pleurodont bicuspid teeth. The most completely preserved 891 dentary possesses at least 20 teeth. The pars dentalis is tall , with its height corresponding to 892 two-thirds of the tooth length. The Meckelian groove is open ventrolingually. It starts from the 893 ventral side of the symphysis and posteriorly increases in height. The lamina horizontalis is 894 slightly curved. The anterior portion of the lamina horizontalis is high and broad, reaching its 895 maximal height in its middle section which corresponds to the tenth tooth position. Behind this 896 point, the lamina horizontalis articulates ventrally with the dorsal margin of the splenial and 897 gradually narrows posteriorly. The articulation surface is lingually exposed. The crista dentalis, 898 sensu Roček (1984), is not higher but is longer than the ventral margin of the lamina 890 Description. The bones bear pleurodont bicuspid teeth. The most completely preserved 891 dentary possesses at least 20 teeth. The pars dentalis is tall , with its height corresponding to 892 two-thirds of the tooth length. The Meckelian groove is open ventrolingually. It starts from the 893 ventral side of the symphysis and posteriorly increases in height. The lamina horizontalis is 894 slightly curved. The anterior portion of the lamina horizontalis is high and broad, reaching its 895 maximal height in its middle section which corresponds to the tenth tooth position. Behind this 896 point, the lamina horizontalis articulates ventrally with the dorsal margin of the splenial and 897 gradually narrows posteriorly. The articulation surface is lingually exposed. The crista dentalis, 898 sensu Roček (1984), is not higher but is longer than the ventral margin of the lamina 899 horizontalis. The ventral margin of the crista dentalis, in its posterior half, bears an articulation 900 surface with the ventral margin of the coronoid. A lingually exposed articulation surface of the 901 splenial is located at the posterior portion of the ventral surface of the lamina horizontalis. Manuscript to be reviewed Up 902 to eight small foramina are present in labial view (Fig. 8A). 903 Comments. See in Lacerta s.l. sp. 2. PeerJ reviewing PDF \| (2016:08:12606:4:0:NEW 23 Jan 2017) PeerJ reviewing PDF \| (2016:08:12606:4:0:NEW 23 Jan 2017) Manuscript to be reviewed 908 Material. Pavlodar 1A, GIN 640/5001-RE27 – -RE33, 7 left dentaries, GIN 640/5001-RE34 – - 909 RE39, 6 right dentaries; Cherlak, GIN 1110/2001-RE51, 1 right dentary. 910 Description. The dentaries possess 19 bicuspid teeth. The pars dentalis is high with its height 911 corresponding to two-thirds of the teeth length. The lamina horizontalis is curved and 912 maintains almost the same height along its entire length. The lamina horizontalis decreases 913 slightly in height only at the 9th-10th tooth positions, where the splenial articulates with the 914 lamina horizontalis. The articulation facet is lingually exposed only in its most posterior portion. 915 The crista dentalis is short but is longer than the ventral margin of the lamina horizontalis. The 916 Meckelian groove is low and ventrolingually open. Up to seven small foramina are present in 917 labial view (Fig. 8B). 908 Material. Pavlodar 1A, GIN 640/5001-RE27 – -RE33, 7 left dentaries, GIN 640/5001-RE34 – - 909 RE39, 6 right dentaries; Cherlak, GIN 1110/2001-RE51, 1 right dentary. 908 Material. Pavlodar 1A, GIN 640/5001-RE27 – -RE33, 7 left dentaries, GIN 640/5001-RE34 – - 909 RE39, 6 right dentaries; Cherlak, GIN 1110/2001-RE51, 1 right dentary. 909 RE39, 6 right dentaries; Cherlak, GIN 1110/2001 RE51, 1 right dentary. 910 Description. The dentaries possess 19 bicuspid teeth. The pars dentalis is high with its height 911 corresponding to two-thirds of the teeth length. The lamina horizontalis is curved and 912 maintains almost the same height along its entire length. The lamina horizontalis decreases 913 slightly in height only at the 9th-10th tooth positions, where the splenial articulates with the 914 lamina horizontalis. The articulation facet is lingually exposed only in its most posterior portion. 915 The crista dentalis is short but is longer than the ventral margin of the lamina horizontalis. The 916 Meckelian groove is low and ventrolingually open. Up to seven small foramina are present in 917 labial view (Fig. 8B). 918 Comments. Lacerta s.l. sp. 2 differs from Lacerta s.l. sp. 1 in having: (1) a more curved lamina 919 horizontalis that maintains nearly the same height along its length; (2) a higher and broader 920 anterior portion of the lamina horizontalis; (3) a shorter cirsta dentalis; and (4) a lower 921 Meckelian groove. 918 Comments. Lacerta s.l. sp. 2 differs from Lacerta s.l. sp. Manuscript to be reviewed 1 in having: (1) a more curved lamina 919 horizontalis that maintains nearly the same height along its length; (2) a higher and broader 920 anterior portion of the lamina horizontalis; (3) a shorter cirsta dentalis; and (4) a lower 921 Meckelian groove. 924 Material. Malyi Kalkaman 2, GIN 1107/2001-RE01, 1 vertebra; Olkhovka 1A, GIN 1111/1001- 925 RE01 and –RE02, 1 anterior and 1 posterior trunk vertebrae; Cherlak, GIN 1110/2001-RE06, - 926 RE52 – -RE57, 7 trunk vertebrae, GIN 1110/2001-RE47, -RE48, 2 left maxillae, GIN 1110/2001- 927 RE49, 1 right maxilla, GIN 1110/2001-RE50, 1 left dentary; Pavlodar 1A, GIN 640/5001-RE40, 1 928 premaxilla, GIN 640/5001-RE26, numerous fragments of dentaries and maxillae, GIN 640/5001- 929 RE43, 77 vertebrae; Pavlodar 1B, GIN 640/6001-RE01, -RE02, 2 left dentaries, GIN 640/6001- PeerJ reviewing PDF \| (2016:08:12606:4:0:NEW 23 Jan 2017) Manuscript to be reviewed Manuscript to be reviewed 930 RE03, -RE04, 2 right dentaries; Olkhovka 1B, GIN 1111/2001-RE01, 1 right dentary; Pavlodar 3A, 931 GIN 1108/3001-RE01, 1 right maxilla; Beteke 2, GIN 945/6001-RE01, 1 left dentary; Beteke 4, 932 GIN 945/8001-RE01, 1 left dentary. 930 RE03, -RE04, 2 right dentaries; Olkhovka 1B, GIN 1111/2001-RE01, 1 right dentary; Pavlodar 3A, 931 GIN 1108/3001-RE01, 1 right maxilla; Beteke 2, GIN 945/6001-RE01, 1 left dentary; Beteke 4, 932 GIN 945/8001-RE01, 1 left dentary. 933 Description and comments. The preserved maxillaries and dentaries possess pleurodont 934 bicuspid teeth. The Meckelian groove is lingually open. The labial surfaces of the maxillaries 935 show no ornamentation. In labial view, the foramina for mandibular division of the fifth cranial 936 (trigeminal) nerve are observable. They are situated along a longitudinal line, parallel to the 937 ventral margin of the bone. The opening of the superior alveolar canal is large. In lingual view, a 938 shallow but broad groove is present at the anterior portion of the frontal process. The large 939 foramen of the fifth cranial (trigeminal) nerve opens at the ventral surface of the lamina 940 horizontalis. A single premaxilla from Pavlodar 1A, GIN 640/5001-RE40 has a tapering nasal 941 process with a row of seven pleurodont and monocuspid teeth. 933 Description and comments. The preserved maxillaries and dentaries possess pleurodont 934 bicuspid teeth. The Meckelian groove is lingually open. The labial surfaces of the maxillaries 935 show no ornamentation. In labial view, the foramina for mandibular division of the fifth cranial 936 (trigeminal) nerve are observable. They are situated along a longitudinal line, parallel to the 937 ventral margin of the bone. The opening of the superior alveolar canal is large. In lingual view, a 938 shallow but broad groove is present at the anterior portion of the frontal process. The large 939 foramen of the fifth cranial (trigeminal) nerve opens at the ventral surface of the lamina 940 horizontalis. A single premaxilla from Pavlodar 1A, GIN 640/5001-RE40 has a tapering nasal 941 process with a row of seven pleurodont and monocuspid teeth. 942 The bone material is extremely fragmentary, and any comparison between different localities 943 was impossible. The fossil remains (maxillae and premaxilla) from Pavlodar 1A do not show any 944 taxonomical differences, so we were not able to group them neither to Lacerta s.l. sp. 1 nor 945 Lacerta s.l. sp. 2. Manuscript to be reviewed Besides the jaw material, vertebrae from the trunk region are available in the 946 Maly Kalkaman 2, Olkhovka 1A and Cherlak localities. It was not possible to identify all of 947 remains below the family level. 942 The bone material is extremely fragmentary, and any comparison between different localities 943 was impossible. The fossil remains (maxillae and premaxilla) from Pavlodar 1A do not show any 944 taxonomical differences, so we were not able to group them neither to Lacerta s.l. sp. 1 nor 945 Lacerta s.l. sp. 2. Besides the jaw material, vertebrae from the trunk region are available in the 946 Maly Kalkaman 2, Olkhovka 1A and Cherlak localities. It was not possible to identify all of 947 remains below the family level. 949 Genus Eremias Fitzinger, 1843 951 PeerJ reviewing PDF \| (2016:08:12606:4:0:NEW 23 Jan 2017) Manuscript to be reviewed Manuscript to be reviewed 953 Material. Pavlodar 2B, GIN 1108/2001-RE01, -RE02, 1 frontal and 1 trunk vertebra. 954 Description. The preserved frontal has a sandglass shape and the most anterior and posterior 955 portions are broken. The bone is slightly curved in lateral view. The posterior portion of the 956 dorsal surface is rough. The crista cranii are round and slightly elevated at the narrowest 957 portion of the bone. Anteriorly, these crista cranii increase in height and build the lateral walls 958 of the cranial vault. The anteroventral surface of the bone has two drop-shaped grooves. The 959 posteroventral surface is plain and slightly lower than the anterocentral surface. The prefrontal 960 facets are developed but do not show any lateral extension. The bone margin that connects 961 both facets is concave. In dorsal view, the nasal facets that are situated at the anterolateral 962 corners, are narrow, deep, and elongated (Fig. 8A, 8D). 953 Material. Pavlodar 2B, GIN 1108/2001-RE01, -RE02, 1 frontal and 1 trunk vertebra. 954 Description. The preserved frontal has a sandglass shape and the most anterior and posterior 955 portions are broken. The bone is slightly curved in lateral view. The posterior portion of the 956 dorsal surface is rough. The crista cranii are round and slightly elevated at the narrowest 957 portion of the bone. Anteriorly, these crista cranii increase in height and build the lateral walls 958 of the cranial vault. The anteroventral surface of the bone has two drop-shaped grooves. The 959 posteroventral surface is plain and slightly lower than the anterocentral surface. The prefrontal 960 facets are developed but do not show any lateral extension. The bone margin that connects 961 both facets is concave. In dorsal view, the nasal facets that are situated at the anterolateral 962 corners, are narrow, deep, and elongated (Fig. 8A, 8D). 954 Description. The preserved frontal has a sandglass shape and the most anterior and posterior 955 portions are broken. The bone is slightly curved in lateral view. The posterior portion of the 956 dorsal surface is rough. The crista cranii are round and slightly elevated at the narrowest 957 portion of the bone. Anteriorly, these crista cranii increase in height and build the lateral walls 958 of the cranial vault. The anteroventral surface of the bone has two drop-shaped grooves. 2 We follow taxonomy suggested by Fritz, Schmidt & Ernst, 2011 recognizing Emydoidea as a distinct genus from Emys. The 959 posteroventral surface is plain and slightly lower than the anterocentral surface. The prefrontal 960 facets are developed but do not show any lateral extension. The bone margin that connects 961 both facets is concave. In dorsal view, the nasal facets that are situated at the anterolateral 962 corners, are narrow, deep, and elongated (Fig. 8A, 8D). 963 In lateral view, a single preserved trunk vertebra has a rectangular shape. The neural arch is 964 moderately convex. A narrow and deep groove is present at the transition of the neural arch 965 and prezygapophysis. The neural spine is reduced and posteriorly builds a rounded process, 966 projecting over the posterior margin of the arch. The centrum is compressed anteroposteriorly 967 and possesses two shallow subcentral grooves, with a subcentral foramina in each one. The 968 condyle is small, round and is situated in the middle part of the posterior margin of the 969 centrum. 970 Comments. Among the Eurasian lacertids, fused dorsally sculptured frontals are known in 971 Acanthodactylus, Eremias, Ophisops (Evans, 2008). Our own observations of Recent species of 972 these genera (Eremias strauchi, Eremias pleskei, Eremias arguta, Eremias multicellata, Ophisops 973 elegans, Acanthodactylus erythrurus) allowed for the assignment of the frontals to the genus 970 Comments. Among the Eurasian lacertids, fused dorsally sculptured frontals are known in 971 Acanthodactylus, Eremias, Ophisops (Evans, 2008). Our own observations of Recent species of 972 these genera (Eremias strauchi, Eremias pleskei, Eremias arguta, Eremias multicellata, Ophisops 973 elegans, Acanthodactylus erythrurus) allowed for the assignment of the frontals to the genus 970 Comments. Among the Eurasian lacertids, fused dorsally sculptured frontals are known in 971 Acanthodactylus, Eremias, Ophisops (Evans, 2008). Our own observations of Recent species of 972 these genera (Eremias strauchi, Eremias pleskei, Eremias arguta, Eremias multicellata, Ophisops 973 elegans, Acanthodactylus erythrurus) allowed for the assignment of the frontals to the genus PeerJ reviewing PDF \| (2016:08:12606:4:0:NEW 23 Jan 2017) Manuscript to be reviewed 974 Eremias and to separate them from: (1) Oph 975 grooves at the anteroventral bone surface a 976 facet; and (2) Acanthodactylus by a flat post 977 outline in lateral view. The preserved single 978 found in Eremias (Rage, 1976). 979 980 Order Testudines Linnaeus, 1758 981 Suborder Cryptodira Cope, 1868 982 Family Emydidae (Rafinesque, 1815) 983 Genus Eymdoidea Gray, 18702 984 Emydoidea sp. 985 986 (Fig. 8E-8G) 987 Material. Novaya Stanitsa 1A, GIN 948/2001 988 GIN 948/2001-RE02, 1 left femur. 989 Description and comments. The caudal part 990 mm , is preserved (Fig. 8E) and probably bel 991 of the carapace, approximately 300 mm. In 992 straight and curves anteriorly only near the 993 the inguinal buttress. The bone is comparat 2 We follow taxonomy suggested by Fritz, Schmidt & Emys. 974 Eremias and to separate them from: (1) Ophisops by a robust frontal, more pronounced 975 grooves at the anteroventral bone surface and a lack of the lateral extension of the prefrontal 976 facet; and (2) Acanthodactylus by a flat posteroventral surface of the bone and a less curved 977 outline in lateral view. The preserved single vertebra strongly resembles the morphology that is 978 found in Eremias (Rage, 1976). 979 980 Order Testudines Linnaeus, 1758 981 Suborder Cryptodira Cope, 1868 982 Family Emydidae (Rafinesque, 1815) 983 Genus Eymdoidea Gray, 18702 984 Emydoidea sp. 985 986 (Fig. 8E-8G) 987 Material. Novaya Stanitsa 1A, GIN 948/2001-RE01, 1 posteriorly incomplete right hypoplastron, 988 GIN 948/2001-RE02, 1 left femur. 989 Description and comments. The caudal part of the left hypoplastron, which has a width of 54.3 990 mm , is preserved (Fig. 8E) and probably belongs to a middle-sized individual with a total length 991 of the carapace, approximately 300 mm. In ventral view, the femoral/abdominal sulcus is nearly 992 straight and curves anteriorly only near the lateral edge of the bone, terminating at the base of 993 the inguinal buttress. 1007 Testudines indet. 1008 Material. Malyi Kalkaman2, GIN 1107/2001-RE02, shell fragment; Shet-Irgyz 1, GIN 1106/1001- 1009 RE01, 1 neuralia; Petropavlovsk 1, GIN 952/1001-RE01, several fragments of carapax; Borki 1B, 1010 GIN 1115/2001-RE01, 1 fragment of carapax. 1008 Material. Malyi Kalkaman2, GIN 1107/2001-RE02, shell fragment; Shet-Irgyz 1, GIN 1106/1001- 1009 RE01, 1 neuralia; Petropavlovsk 1, GIN 952/1001-RE01, several fragments of carapax; Borki 1B, 1010 GIN 1115/2001-RE01, 1 fragment of carapax. 1008 Material. Malyi Kalkaman2, GIN 1107/2001-RE02, shell fragment; Shet-Irgyz 1, GIN 1106/1001- 1009 RE01, 1 neuralia; Petropavlovsk 1, GIN 952/1001-RE01, several fragments of carapax; Borki 1B, 1010 GIN 1115/2001-RE01, 1 fragment of carapax. 1011 Comments. The preserved remains were not sufficently informative for any other taxonomic 1012 interpretation. 1011 Comments. The preserved remains were not sufficently informative for any other taxonomic 1012 interpretation. Manuscript to be reviewed The bone is comparatively thin, medially from the bridge (4 mm) to 2 W f ll t t d b F it S h idt & E t 2011 i i E d id di ti t f 974 Eremias and to separate them from: (1) Ophisops by a robust frontal, more pronounced 975 grooves at the anteroventral bone surface and a lack of the lateral extension of the prefrontal 976 facet; and (2) Acanthodactylus by a flat posteroventral surface of the bone and a less curved 977 outline in lateral view. The preserved single vertebra strongly resembles the morphology that is 978 found in Eremias (Rage, 1976). 989 Description and comments. The caudal part of the left hypoplastron, which has a width of 54.3 990 mm , is preserved (Fig. 8E) and probably belongs to a middle-sized individual with a total length 991 of the carapace, approximately 300 mm. In ventral view, the femoral/abdominal sulcus is nearly 992 straight and curves anteriorly only near the lateral edge of the bone, terminating at the base of 993 the inguinal buttress. The bone is comparatively thin, medially from the bridge (4 mm) to PeerJ reviewing PDF \| (2016:08:12606:4:0:NEW 23 Jan 2017) Manuscript to be reviewed 994 behind the bridge (7.2 mm). The lateral edge of the bone projects slightly posterolaterally. The 995 outline of the femoral/abdominal sulcus and the profile of the lateral edge are similar to those 996 of the emydid genus Emydoidea (both fossil and Recent specimens) ((Chkhikvadze, 1983), figs. 997 26 and 27, p. 138; (Holman, 1995)). 994 behind the bridge (7.2 mm). The lateral edge of the bone projects slightly posterolaterally. The 995 outline of the femoral/abdominal sulcus and the profile of the lateral edge are similar to those 996 of the emydid genus Emydoidea (both fossil and Recent specimens) ((Chkhikvadze, 1983), figs. 997 26 and 27, p. 138; (Holman, 1995)). 998 An almost complete left femur is available from the same locality where the hypoplastron 999 fragment was found. The bone is slender and bent (Fig. 8F-8G), and is 50.6 mm in length. This 1000 bone could have belonged to an individual of about 300 mm of the carapace length. The femur 1001 lacks its proximal portion (i.e. femoral head, major and minor trochanters). In ventral view, the 1002 fossa is delimited by the trochanters and is observable below the femoral head. The dimension 1003 of the bone is characteristic of aquatic testudinoids. Taking this latter character into account, as 1004 well as the comparable reconstructed total body-sizes of both elements (hypoplastron and 1005 femur) (ca. 300 mm), we consider the remains to belong to the genus Emydoidea. Manuscript to be reviewed 1016 In general, and in contrast with the well-studied European fossil record, very is known about 1017 the Neogene herpetofauna from Asia. This record bias is owing to: (1) the less explored and less 1018 extensively studied Neogene deposits on the Asian continent; and (2) the entirely lack of study 1019 on Recent amphibians and reptiles, in spite of the intense investigations around small mammals 1020 by many scholars. The Western Siberian localities provide an exceptional opportunity to fill 1021 these gaps in information and to explore both the unknown diversity of the Asian 1022 herpetofaunal assemblages and the palaeobiogeographic affinities of the Western Siberian 1023 Neogene herpetofauna with the European faunas. Unfortunately, the yielded fossil material 1024 from this study and from previous investigations has thus far not been rich in amphibian and 1025 reptile remains. On average, only four taxa are available from each studied locality. Our 1026 faunistic, palaeogeographic and palaeoclimatic interpretations are, hence, very tentative and 1027 should be taken within this context. The unbiased comparison and analysis of our data are also 1028 hindered by the scarce record of the Asian Neogene fossil fauna. For the comparison with the 1029 European record, we used already published data on amphibian and reptile groups (families, 1030 genus, species, etc.) which have been summarised in the fosFARbase database (Böhme & Ilg, 1031 2003). These data are given in the Table S5. In the ʻEuropeʼ record, we consider all known fossil 1032 records from Western, Central, and Eastern Europe as well as from Anatolia (Fig. 9). By 1033 analysing the Neogene amphibian and reptile records from Europe and Asia, we were able to 1034 provide useful data that are applicable for fossil calibration of molecular clocks in the 1035 phylogenetic trees. 1016 In general, and in contrast with the well-studied European fossil record, very is known about 1017 the Neogene herpetofauna from Asia. This record bias is owing to: (1) the less explored and less 1018 extensively studied Neogene deposits on the Asian continent; and (2) the entirely lack of study 1019 on Recent amphibians and reptiles, in spite of the intense investigations around small mammals 1020 by many scholars. Manuscript to be reviewed The Western Siberian localities provide an exceptional opportunity to fill 1021 these gaps in information and to explore both the unknown diversity of the Asian 1033 analysing the Neogene amphibian and reptile records from Europe and Asia, we were able to 1034 provide useful data that are applicable for fossil calibration of molecular clocks in the 1035 phylogenetic trees. PeerJ reviewing PDF \| (2016:08:12606:4:0:NEW 23 Jan 2017) Manuscript to be reviewed 1038 The Asiatic salamanders (Salamandrella sp.) have the most abundant and frequent record 1039 among the studied Western Siberian localities. These organisms appeared in these areas in the 1040 middle Late Miocene (in the Selety 1A locality) and are present until the early Early Pleistocene. 1041 Although the herpetofaunal assemblages of the older localities are rich and represented by 1042 numerous taxa, they do not contain any hynobiid remains, demonstrating that there is no 1043 sampling bias in their record and that such specimen are not present in earlier localities. 1044 Recently, the oldest record of the genus, Salamandrella sp. has been described from the late (?) 1045 Early Miocene of Eastern Siberia (Lake Baikal) (Syromyatnikova, 2014), and a new species of 1046 Salamandrella is indicated to be present in the Late Miocene locality Ertemte 2, China (Vasilyan 1047 et al., 2013). Furthermore, the fossil Asiatic salamander, Ranodon cf. sibiricus was recovered 1048 from the Early Pleistocene of Southern Kazakhstan (Averianov & Tjutkova, 1995), and a 1049 Salamandrella sp. was reported from a few Middle Pleistocene age localities in European Russia 1050 (Ratnikov, 2010). 1051 In Central Europe, hynobiids (genus Parahynobius) appeared at the earliest Late Miocene and is 1052 present in the record until the Middle Pleistocene (Venczel, 1999a, 1999b; Venczel & Hír, 2013). 1053 According to our unpublished data, the hynobiids are also present in three Ukrainian localities – 1054 Grytsiv (11.1 Ma) (Kirscher et al., 2016), earliest Late Miocene; Cherevichnoe lower level, 1055 middle Late Miocene; and Kotlovina lower level, late Pliocene. The Ukrainian occurrences 1056 coincide with both the Central European and Western Siberian records of hynobiids, which at 1057 that time most probably characterised by favourable conditions for hynobiid distribution. 1058 Considering their oldest records, the origin of Hynobiidae was most probably in Eastern Asia in 1051 In Central Europe, hynobiids (genus Parahynobius) appeared at the earliest Late Miocene and is 1052 present in the record until the Middle Pleistocene (Venczel, 1999a, 1999b; Venczel & Hír, 2013). 1053 According to our unpublished data, the hynobiids are also present in three Ukrainian localities – 1054 Grytsiv (11.1 Ma) (Kirscher et al., 2016), earliest Late Miocene; Cherevichnoe lower level, 1055 middle Late Miocene; and Kotlovina lower level, late Pliocene. Manuscript to be reviewed The Ukrainian occurrences 1056 coincide with both the Central European and Western Siberian records of hynobiids, which at 1057 that time most probably characterised by favourable conditions for hynobiid distribution. d h ld d h f b d b bl 1058 Considering their oldest records, the origin of Hynobiidae was most probably in Eastern Asia in PeerJ reviewing PDF \| (2016:08:12606:4:0:NEW 23 Jan 2017) Manuscript to be reviewed 1059 the early Miocene. We will present a detailed study on the Cenozoic record of fossil Hynobiidae 1060 including the Western Siberian material in a forthcoming paper. 1061 1062 Cryptobranchidae 1063 The cryptobranchid remains are known from two localities in the town of Pavlodar and from 1064 three localities in the Zaisan Basin. The stratigraphic positions of the Pavlodar localities are not 1065 clear. The only record of giant salamander that we were able to study is stored at the 1066 Palaeontological Institute of Moscow, Russia. The collection label provides the following 1067 information: ʻcollected by Gaiduchenko, in 1970, from the Gusiniy Perelet locality, at the 1068 contact of the Aral clays with overlying sands, about 200-300 m south far from the ‘Gusiniy 1069 Perelet’ [=Pavlodar 1A] locality ʼ. The only explanation of the stratigraphic allocation of the 1070 giant salamander remains is that they originated from the basal horizon of the Pavlodar Svita, 1071 overlaying the ʻAral claysʼ (or= limnic clays of the Kalkaman Svita). Gaiduchenko (1984) and 1072 Gaiduchenko & Chkhikvadze (1985) mention a giant salamander (Cryptobranchidae indet.) from 1073 a locality named Detskaya Zheleznaja Doroga (engl. Children Railway) (Fig. 2, Table S1, Data S3), 1074 a sand pit located 10 km south-east from the ‘Gusiniy Perelet’ [=Pavlodar 1A] locality. The age 1075 of this fossiliferous horizon may fall near the Miocene-Pliocene boundary, an assumption that is 1076 mostly based on geology, age and accompanying fauna (see Data S2). This record from the 1077 Detskaya Zheleznaja Doroga presents the most northern (52.3° N) occurrence of the giant 1078 salamanders in the Northern Hemisphere known so far. Unfortunately, this material was not 1079 available for our study 1059 the early Miocene. We will present a detailed study on the Cenozoic record of fossil Hynobiidae 1060 including the Western Siberian material in a forthcoming paper. 1061 1062 Cryptobranchidae 1063 The cryptobranchid remains are known from two localities in the town of Pavlodar and from 1064 three localities in the Zaisan Basin. The stratigraphic positions of the Pavlodar localities are not 1065 clear. The only record of giant salamander that we were able to study is stored at the 1066 Palaeontological Institute of Moscow, Russia. Manuscript to be reviewed 1080 Giant salamander remains have also been reported from three Burdigalian localities – Tri 1081 Bogatyrya, Vympel, Poltinik of the Zaisan Basin (Fig. 1, Table S1) (Chkhikvadze, 1984; Böhme, 1082 Vasilyan & Winklhofer, 2012). The remains were assigned to the species Andrias karelcapeki by 1083 Chkhikvadze (1984). The taxonomic validity of the species still requires revision, which is 1084 necessary for any further interpretations. 1085 1086 Proteidae 1087 The oldest record of the genus is described as being from the Late Oligocene and was found in 1088 the Aral Formation in the Akespe locality, on the north coast of the Aral Sea, Kazakhstan (cf. 1089 Mioproteus,) (Malakhov, 2003; Bendukidze, Bruijn & Van den Hoek Ostende, Lars W., 2009). 1090 Here we add to the record a new, more recent Miocene (Aquitanian) Asian occurrence from the 1091 Ayakoz locality, Kazakhstan (Fig. 3D, Table S1). In the Middle Miocene, representatives of this 1092 genus occur in several localities in southern Russia and northern Kazakhstan (Table S1). 1093 According to our assessment, proteids survived until latest Miocene/earliest Pliocene (locality 1094 Petropavlovsk 1/2). The oldest stratigraphic record of Mioproteus (Mioproteus caucasicus) in 1080 Giant salamander remains have also been reported from three Burdigalian localities – Tri 1081 Bogatyrya, Vympel, Poltinik of the Zaisan Basin (Fig. 1, Table S1) (Chkhikvadze, 1984; Böhme, 1082 Vasilyan & Winklhofer, 2012). The remains were assigned to the species Andrias karelcapeki by 1083 Chkhikvadze (1984). The taxonomic validity of the species still requires revision, which is 1084 necessary for any further interpretations. 1085 1086 Proteidae 1087 The oldest record of the genus is described as being from the Late Oligocene and was found in 1088 the Aral Formation in the Akespe locality, on the north coast of the Aral Sea, Kazakhstan (cf. 1089 Mioproteus,) (Malakhov, 2003; Bendukidze, Bruijn & Van den Hoek Ostende, Lars W., 2009). 1090 Here we add to the record a new, more recent Miocene (Aquitanian) Asian occurrence from the 1091 Ayakoz locality, Kazakhstan (Fig. 3D, Table S1). In the Middle Miocene, representatives of this 1092 genus occur in several localities in southern Russia and northern Kazakhstan (Table S1). 1093 According to our assessment, proteids survived until latest Miocene/earliest Pliocene (locality 1094 Petropavlovsk 1/2). Manuscript to be reviewed The collection label provides the following 1067 information: ʻcollected by Gaiduchenko, in 1970, from the Gusiniy Perelet locality, at the 1068 contact of the Aral clays with overlying sands, about 200-300 m south far from the ‘Gusiniy 1069 Perelet’ [=Pavlodar 1A] locality ʼ. The only explanation of the stratigraphic allocation of the 1070 giant salamander remains is that they originated from the basal horizon of the Pavlodar Svita, 1071 overlaying the ʻAral claysʼ (or= limnic clays of the Kalkaman Svita). Gaiduchenko (1984) and 1072 Gaiduchenko & Chkhikvadze (1985) mention a giant salamander (Cryptobranchidae indet.) from 1073 a locality named Detskaya Zheleznaja Doroga (engl. Children Railway) (Fig. 2, Table S1, Data S3), 1074 a sand pit located 10 km south-east from the ‘Gusiniy Perelet’ [=Pavlodar 1A] locality. The age 1075 of this fossiliferous horizon may fall near the Miocene-Pliocene boundary, an assumption that is 1076 mostly based on geology, age and accompanying fauna (see Data S2). This record from the 1077 Detskaya Zheleznaja Doroga presents the most northern (52.3° N) occurrence of the giant 1078 salamanders in the Northern Hemisphere known so far. Unfortunately, this material was not 1079 available for our study PeerJ reviewing PDF \| (2016:08:12606:4:0:NEW 23 Jan 2017) PeerJ reviewing PDF \| (2016:08:12606:4:0:NEW 23 Jan 2017) Manuscript to be reviewed 1102 Mioproteus wezei from the Pliocene of Poland and from the Early Pleistocene of Moldavia 1103 (Malakhov, 2003). Later, Roček (2005) considered M. wezei as a junior synonym of M. 1104 caucasicus, although as already mentioned by Malakhov (2003), an adequate amount of 1105 material including cranial and postcranial elements is necessary to solve the taxonomic 1106 problems of the genus. Malakhov (2003) also suggested an Asiatic origin for the ʻMioproteus 1107 caucasicus complexʼ and their later distribution into Europe. In summary, the oldest Late 1108 Oligocene record of Mioproteus (Mioproteus sp.) from Akespe, Kazakhstan and other localities 1109 of younger ages suggest: (1) a probable Asian origin of the genus; (2) the genus was 1110 continuously present in Central Asia/Western Siberia until the early Pliocene; and (3) 1111 Mioproteus migrated into Europe in the Early Miocene. 1112 1113 Salamandridae 1114 As has already been established, Chelotriton is a basket taxon (Böhme, 2008) that needs further 1115 taxonomic study. It is one of the fossil amphibians that has an abundant and wide distribution 1116 in the late Paleogene and Neogene localities of Europe. In Asia, the genus was previously 1117 known only from the late Middle Miocene locality Malyi Kalkaman 1 (Tleuberdina, 1993). Our 1118 study showed that this genus was present at least since the Aquitanian age (the Aykoz locality 1119 in Kazakhstan, Early Miocene) (Table S1), making their Asiatic record older than previously 1120 assumed. 1121 Two localities (Ayakoz and Baikadam) from Western Siberia revealed aff. Tylototriton. The 1122 vertebrae showed significant similarities with the Recent East Asiatic genus Tylototriton. In 1123 Böhme & Ilg (2003) and Böhme (2010), the genus Tylototriton (cf. Tylototriton sp. and 1102 Mioproteus wezei from the Pliocene of Poland and from the Early Pleistocene of Moldavia 1103 (Malakhov, 2003). Later, Roček (2005) considered M. wezei as a junior synonym of M. 1104 caucasicus, although as already mentioned by Malakhov (2003), an adequate amount of 1105 material including cranial and postcranial elements is necessary to solve the taxonomic 1106 problems of the genus. Malakhov (2003) also suggested an Asiatic origin for the ʻMioproteus 1107 caucasicus complexʼ and their later distribution into Europe. Manuscript to be reviewed The oldest stratigraphic record of Mioproteus (Mioproteus caucasicus) in 1095 Europe is described from the mid Aquitanian (early Early Miocene about 20.5-22 Ma) at two 1096 localities Ulm-Uniklinik and Ulm Westtangente of the North Alpine Foreland Basin (Heizmann et 1097 al., 1989). The fossil proteids are known in Europe until the Pleistocene Epoch (Böhme & Ilg, 1098 2003).Due to the lack of complete fossil skeletons and unclear taxonomic assignments of the 1099 fossil records, Malakhov (2003) preferred to refer all known specimens of Mioproteus to the 1100 ʻMioproteus caucasicus complexʼ, including Mioproteus from Ashut, Kazakhstan, 1080 Giant salamander remains have also been reported from three Burdigalian localities – Tri 1081 Bogatyrya, Vympel, Poltinik of the Zaisan Basin (Fig. 1, Table S1) (Chkhikvadze, 1984; Böhme, 1082 Vasilyan & Winklhofer, 2012). The remains were assigned to the species Andrias karelcapeki by 1083 Chkhikvadze (1984). The taxonomic validity of the species still requires revision, which is 1084 necessary for any further interpretations. 1093 According to our assessment, proteids survived until latest Miocene/earliest Pliocene (locality 1094 Petropavlovsk 1/2). The oldest stratigraphic record of Mioproteus (Mioproteus caucasicus) in 1095 Europe is described from the mid Aquitanian (early Early Miocene about 20.5-22 Ma) at two 1096 localities Ulm-Uniklinik and Ulm Westtangente of the North Alpine Foreland Basin (Heizmann et 1097 al., 1989). The fossil proteids are known in Europe until the Pleistocene Epoch (Böhme & Ilg, 1098 2003).Due to the lack of complete fossil skeletons and unclear taxonomic assignments of the 1099 fossil records, Malakhov (2003) preferred to refer all known specimens of Mioproteus to the 1100 ʻMioproteus caucasicus complexʼ, including Mioproteus from Ashut, Kazakhstan, 1101 Mioproteuscaucasicus from type locality, as well as from the Late Miocene of Czech Republic, 1101 Mioproteuscaucasicus from type locality, as well as from the Late Miocene of Czech Republic, PeerJ reviewing PDF \| (2016:08:12606:4:0:NEW 23 Jan 2017) Manuscript to be reviewed 1113 Salamandridae 1114 As has already been established, Chelotriton is a basket taxon (Böhme, 2008) that needs further 1115 taxonomic study. It is one of the fossil amphibians that has an abundant and wide distribution 1116 in the late Paleogene and Neogene localities of Europe. In Asia, the genus was previously 1117 known only from the late Middle Miocene locality Malyi Kalkaman 1 (Tleuberdina, 1993). Our 1118 study showed that this genus was present at least since the Aquitanian age (the Aykoz locality 1119 in Kazakhstan, Early Miocene) (Table S1), making their Asiatic record older than previously 1120 assumed. 1121 Two localities (Ayakoz and Baikadam) from Western Siberia revealed aff. Tylototriton. The 1122 vertebrae showed significant similarities with the Recent East Asiatic genus Tylototriton. In 1123 Böhme & Ilg (2003) and Böhme (2010), the genus Tylototriton (cf. Tylototriton sp. and Manuscript to be reviewed In summary, the oldest Late 1108 Oligocene record of Mioproteus (Mioproteus sp.) from Akespe, Kazakhstan and other localities 1109 of younger ages suggest: (1) a probable Asian origin of the genus; (2) the genus was 1110 continuously present in Central Asia/Western Siberia until the early Pliocene; and (3) 1111 Mioproteus migrated into Europe in the Early Miocene. PeerJ reviewing PDF \| (2016:08:12606:4:0:NEW 23 Jan 2017) PeerJ reviewing PDF \| (2016:08:12606:4:0:NEW 23 Jan 2017) Manuscript to be reviewed 1124 Tylototriton sp. nov.) has been reported from several Early Oligocene localities in southern 1125 Germany. Two Siberian records represent the first fossil occurrence of the genus in Asia, which 1126 appeared more recently in the fossil record than in the European occurrence. These Western 1127 Siberian specimens and the European specimens can be clearly separated from each other by 1128 the morphology of the trunk vertebrae. The Siberian salamanders probably represent new 1129 forms, strongly related to the East Asian terrestrial salamander, Tylototriton. The sympatric 1130 occurrence of two fossil terrestrial salamander genera Chelotriton and Tylototriton was 1131 documented for the first time from the Aquitanian age locality Ayakoz. 1132 1133 Palaeobatrachidae 1134 The palaeobratrachids are considered a European family, with probable occurrence in North 1135 America at the terminal Cretaceous (Wuttke et al., 2012). Records of the palaeobratrachids are 1136 known from the Paleogene Western and Central Europe. It should be taken into account, 1137 however, that records from the Paleogene of Turkey, as well as from the Paleogene and Early to 1138 Middle Miocene of Eastern Europe, are very scarcely known. In the Miocene, palaeobatrachids 1139 appear to have expanded their distribution to Eastern Europe and also reached Anatolia, where 1140 they existed from the latest Oligocene and remained during the entire Early Miocene. During 1141 the Middle Miocene, palaeobatrachids were present in Europe, from Germany to Ukraine 1142 (Wuttke et al., 2012). The palaeobatrachid record in Europe is characterised by a four-million- 1143 year-long (ca. 5.6-9.78 Ma) gap in the Late Miocene (Fig. 9). During this gap, no palaeobatrachid 1144 is known from Western to Eastern Europe even in localities rich in diverse herpteofaunal 1145 assemblages (e.g. Staniantsi, Bulgaria; Morskaya 2, Russia, Böhme & Ilg, 2003) and where 1124 Tylototriton sp. nov.) has been reported from several Early Oligocene localities in southern 1125 Germany. Two Siberian records represent the first fossil occurrence of the genus in Asia, which 1126 appeared more recently in the fossil record than in the European occurrence. These Western 1127 Siberian specimens and the European specimens can be clearly separated from each other by 1128 the morphology of the trunk vertebrae. The Siberian salamanders probably represent new 1129 forms, strongly related to the East Asian terrestrial salamander, Tylototriton. Manuscript to be reviewed The sympatric 1130 occurrence of two fossil terrestrial salamander genera Chelotriton and Tylototriton was 1131 documented for the first time from the Aquitanian age locality Ayakoz. PeerJ reviewing PDF \| (2016:08:12606:4:0:NEW 23 Jan 2017) Manuscript to be reviewed 1146 characterised by favourable environmental conditions for their distribution. After this gap, 1147 palaeobatrachids occur near the Mio-Pliocene transition in studied localities from Italy (Ciabot 1148 Cagna), (Cavallo et al., 1993) and Hungary (Osztramos 1C) (Venczel, 2001)). They seems to have 1149 disappeared from Western (Tegelen locality in Holland, Villa et al., 2016) and Central Europe 1150 (Betfia IX/B locality in Romania, Venczel, 2000) after the Early Pleistocene and remained 1151 exclusively in Eastern Europe until the middle Pleistocene (Poland – European Russia) (Wuttke 1152 et al., 2012). The palaeobatrachids appear to have never reached the east of the Ural 1153 Mountains. Their most eastern distribution is recorded in the Late Pleistocene locality of 1154 Apastovo, in Russia, which is about 600 km west from the Ural Mountains (Wuttke et al., 2012). 1155 The Western Siberian record does not only represent the first and only out-of-Europe 1156 occurrence of the family, but, surprisingly, falls within the Late Miocene palaeobatrachid gap of 1157 the European record. It is possible that palaeobatrachids occupied Western Eurasia again at the 1158 Mio-Pliocene boundary, from the east. 1159 1160 Bombinatoridae 1161 The primitive family of aquatic toads Bombinatoridae includes two recent genera: Bombina and 1162 Barbourula. The family is known since the Maastrichtian, Late Cretaceous in Romania, genus 1163 Hatzegobatrachus (Venczel et al., 2016) and the Early Eocene in India, genus Eobarbourula 1164 (Folie et al., 2013). The Recent distribution of Bombina is confined to continental Europe and 1165 East Asia, representing the western and eastern genetic clades of the genus respectively. In 1166 Europe, two species Bombina bombina and Bombina variegata are known. B. bombina has the 1167 widest distribution and is found in Central to Eastern Europe, whereas B. variegata occurs in 1146 characterised by favourable environmental conditions for their distribution. After this gap, 1147 palaeobatrachids occur near the Mio-Pliocene transition in studied localities from Italy (Ciabot 1148 Cagna), (Cavallo et al., 1993) and Hungary (Osztramos 1C) (Venczel, 2001)). They seems to have 1149 disappeared from Western (Tegelen locality in Holland, Villa et al., 2016) and Central Europe 1150 (Betfia IX/B locality in Romania, Venczel, 2000) after the Early Pleistocene and remained 1151 exclusively in Eastern Europe until the middle Pleistocene (Poland – European Russia) (Wuttke 1152 et al., 2012). The palaeobatrachids appear to have never reached the east of the Ural 1153 Mountains. PeerJ reviewing PDF \| (2016:08:12606:4:0:NEW 23 Jan 2017) Manuscript to be reviewed Their most eastern distribution is recorded in the Late Pleistocene locality of 1154 Apastovo, in Russia, which is about 600 km west from the Ural Mountains (Wuttke et al., 2012). 1155 The Western Siberian record does not only represent the first and only out-of-Europe 1156 occurrence of the family, but, surprisingly, falls within the Late Miocene palaeobatrachid gap of 1157 the European record. It is possible that palaeobatrachids occupied Western Eurasia again at the 1158 Mio-Pliocene boundary, from the east. 1146 characterised by favourable environmental conditions for their distribution. After this gap, 1147 palaeobatrachids occur near the Mio-Pliocene transition in studied localities from Italy (Ciabot 1148 Cagna), (Cavallo et al., 1993) and Hungary (Osztramos 1C) (Venczel, 2001)). They seems to have 1149 disappeared from Western (Tegelen locality in Holland, Villa et al., 2016) and Central Europe 1150 (Betfia IX/B locality in Romania, Venczel, 2000) after the Early Pleistocene and remained 1151 exclusively in Eastern Europe until the middle Pleistocene (Poland – European Russia) (Wuttke 1152 et al., 2012). The palaeobatrachids appear to have never reached the east of the Ural 1153 Mountains. Their most eastern distribution is recorded in the Late Pleistocene locality of 1154 Apastovo, in Russia, which is about 600 km west from the Ural Mountains (Wuttke et al., 2012). 1155 The Western Siberian record does not only represent the first and only out-of-Europe 1156 occurrence of the family, but, surprisingly, falls within the Late Miocene palaeobatrachid gap of 1157 the European record. It is possible that palaeobatrachids occupied Western Eurasia again at the 1158 Mio-Pliocene boundary, from the east. 1160 Bombinatoridae 1161 The primitive family of aquatic toads Bombinatoridae includes two recent genera: Bombina and 1162 Barbourula. The family is known since the Maastrichtian, Late Cretaceous in Romania, genus 1163 Hatzegobatrachus (Venczel et al., 2016) and the Early Eocene in India, genus Eobarbourula 1164 (Folie et al., 2013). The Recent distribution of Bombina is confined to continental Europe and 1165 East Asia, representing the western and eastern genetic clades of the genus respectively. In 1166 Europe, two species Bombina bombina and Bombina variegata are known. B. bombina has the 1167 widest distribution and is found in Central to Eastern Europe, whereas B. variegata occurs in PeerJ reviewing PDF \| (2016:08:12606:4:0:NEW 23 Jan 2017) Manuscript to be reviewed 1168 Central Europe and in the south-eastern and western parts of Eastern Europe (Pabijan et al., 1169 2013). The fossil record of the fire-bellied toad Bombina is patchy and limited to the Neogene 1170 of continental Europe. According to Sanchíz & Schleich (1986), the oldest fossil occurrences of 1171 the genus (Bombina sp.) are known from two localities in Germany: Weißenburg 6 (early 1172 Aquitanian) and Stubersheim 3 (early Burdigalian) (Sanchiz & Schleich, 1986; Böhme & Ilg, 1173 2003). The personal observations of one of the co-authors of this study (MB) did not confirm 1174 the Weißenburg 6 record of Bombina. Therefore, in the present study, we consider 1175 Stubersheim 3 to be the earliest occurrence of the genus. 1176 Bombinatorids later appeared in Central Europe in the mid Middle Miocene (Bombina sp., 1177 Opole 2, Poland) (Młynarski et al., 1982). Later, fire-bellied toads are present in three localities, 1178 representing the middle Tortonian age, including also the first fossil occurrences of the Recent 1179 European species – Bombina sp. from Rudabánya in Hungary (9.9-10.30 Ma) (Roček, 2005), B. 1180 cf. bombina from Kohfidisch in Austria (8.55-8.95 Ma) (Tempfer, 2005), and B. cf. variegata 1181 from Suchomasty in Czech Republic (8.8-9.2 Ma) (Hodrová, 1987). During the Pliocene, 1182 bombinatorids are represented mainly by the species B. bombina in six localities within Central 1183 Europe (Böhme & Ilg, 2003). The Pleistocene record is the richest in bombinatorid specimens 1184 with record from over 15 localities ranging from Central to Eastern Europe, and in which both 1185 Recent European species, B. variegata and B. bombina, are documented (Böhme & Ilg, 2003) 1186 (Fig. 9, Table S5). 1187 In Western Siberia, bombinatorids are known from three localities: Malyi Kalkaman 2, Selety 1188 1A, and Cherlak. The oldest known record dated back to the late Serravalian (Middle Miocene). 1189 The oldest Messinian Selety 1A locality provided the fossil form of the Recent B. bombina (B. cf. 1187 In Western Siberia, bombinatorids are known from three localities: Malyi Kalkaman 2, Selety 1188 1A, and Cherlak. The oldest known record dated back to the late Serravalian (Middle Miocene). 1189 The oldest Messinian Selety 1A locality provided the fossil form of the Recent B. bombina (B. cf. Manuscript to be reviewed Later in the Middle Miocene they expanded into Weste 1196 the Ural Mountains. The Western Siberian fossil Bombina can be 1197 separated from Bombina orientalis, a member of the East Asian c 1198 recent distribution as well as the fossil record into account, a spl 1199 Bombina clades seems most probable in Asia during the Paleoge 1200 1201 Pelobatidae 1202 The family of European spadefoot toads Pelobatidae includes on 1203 species distributed in north-western Africa, Europe, in small area 1204 Mountains in Russia and in the north of Kazakhstan (Kuzmin, 199 1205 affinities and records are known from the Late Cretaceous in No 1206 pelobatids in Europe dates back to the Early Eocene, as indicated 1207 Eopelobates (Middle Eocene – late Pliocene), as well as by the fo 1208 Pelobates (Middle Oligocene – Recent) (Roček et al., 2014). The A 1209 very limited and includes forms from the Eo-Oligocene of Kazakh 1210 Eocene of India (Folie et al., 2013). Recently, Roček et al. (2014) e 1211 (Oligocene, Mongolia) (Gubin, 1995) from the family Pelobatidae 1190 bombina) (Fig. 9). The last record of the genus dates back to the early Messinian (Late 1191 Miocene). It is interesting to note that the Western Siberian record of the genus does not 1192 coincide with their European occurrences, i.e. they are present during those periods when 1193 Bombina is absent in Europe. According to our analysis, it is clear that the ancestor of the ʻB. 1194 bombina – B. variegataʼ clade was present in Europe from, at least, the later part of the Early 1195 Miocene. Later in the Middle Miocene they expanded into Western Asia, reaching the east from 1196 the Ural Mountains. The Western Siberian fossil Bombina can be clearly osteologically 1197 separated from Bombina orientalis, a member of the East Asian clade of the genus. Taking their 1198 recent distribution as well as the fossil record into account, a split of the European and Asian 1199 Bombina clades seems most probable in Asia during the Paleogene. 1190 bombina) (Fig. 9). The last record of the genus dates back to the early Messinian (Late 1191 Miocene). It is interesting to note that the Western Siberian record of the genus does not 1192 coincide with their European occurrences, i.e. they are present during those periods when 1193 Bombina is absent in Europe. PeerJ reviewing PDF \| (2016:08:12606:4:0:NEW 23 Jan 2017) PeerJ reviewing PDF \| (2016:08:12606:4:0:NEW 23 Jan 2017) Manuscript to be reviewed 1190 bombina) (Fig. 9). The last record of the genus dates back to the early Messinian (Late 1191 Miocene). It is interesting to note that the Western Siberian record of the genus does not 1192 coincide with their European occurrences, i.e. they are present during those periods when 1193 Bombina is absent in Europe. According to our analysis, it is clear that the ancestor of the ʻB. 1194 bombina – B. variegataʼ clade was present in Europe from, at least, the later part of the Early 1195 Miocene. Later in the Middle Miocene they expanded into Western Asia, reaching the east from 1196 the Ural Mountains. The Western Siberian fossil Bombina can be clearly osteologically 1197 separated from Bombina orientalis, a member of the East Asian clade of the genus. Taking their 1198 recent distribution as well as the fossil record into account, a split of the European and Asian 1199 Bombina clades seems most probable in Asia during the Paleogene. 1200 1201 Pelobatidae 1202 The family of European spadefoot toads Pelobatidae includes only one extant genus with four 1203 species distributed in north-western Africa, Europe, in small areas that are east of the Ural 1204 Mountains in Russia and in the north of Kazakhstan (Kuzmin, 1995). The family has Laurasian 1205 affinities and records are known from the Late Cretaceous in North America. The presence of 1206 pelobatids in Europe dates back to the Early Eocene, as indicated by the fossil genus 1207 Eopelobates (Middle Eocene – late Pliocene), as well as by the fossil forms of the Recent genus 1208 Pelobates (Middle Oligocene – Recent) (Roček et al., 2014). The Asian record of Pelobatidae is 1209 very limited and includes forms from the Eo-Oligocene of Kazakhstan (Chkhikvadze, 1985) and 1210 Eocene of India (Folie et al., 2013). Recently, Roček et al. (2014) excluded the genus Uldzinia 1211 (Oligocene, Mongolia) (Gubin, 1995) from the family Pelobatidae. The Kazakhstan fossil record Manuscript to be review 1190 bombina) (Fig. 9). The last record of the genus dates back to the 1191 Miocene). It is interesting to note that the Western Siberian reco 1192 coincide with their European occurrences, i.e. they are present d 1193 Bombina is absent in Europe. According to our analysis, it is clear 1194 bombina – B. variegataʼ clade was present in Europe from, at lea 1195 Miocene. 4 localities: main level of Plesh, Tuzkabak, Cherepakhovoe Pole [Tayzhuzgen Section], Raskop [Aksyr Section], Tyubiteika, sopki ʻRybnayaʼ and Kontrolnaya [Juvan-Kara Section] 3 localities: Zertsalo [Sunduk Section], lager Biryukova [Kiin-Kerish Section], lower faunistic level of Plesh [Kusto- Kyzylkain Section], probably also Tabtym [Sarykamysh Section] Manuscript to be reviewed 1212 of the family (Chkhikvadze, 1985, 1998) includes numerous remains of Pelobatidae indet. from: 1213 (1) the localities of the Zaisan Basin from the Upper Aksyr Svita3, early Priabonian; rare finds in 1214 the Kusto Svita and basal horizon of Buran Svita4, late Priabonian and earliest Rupelian; 1215 abundant occurrences in the Buran Svita5, early Rupelian and (2) large-sized spadefoot toads 1216 from the Kyzyl-Kak locality of the Turgay Basin, Central Kazakhstan, Late Oligocene 1217 (Chkhikvadze, 1998). Revision of this rich pelobatid record from the Zaisan Basin was not 1218 possible due to the lack of descriptions and illustrations of the material as well as the difficulty 1219 in accessing the specimens. Nevertheless, taking the Paleogene fossil records into account, we 1220 inferred that the spadefoot toads may have dispersed from Europe to Western Asia during the 1221 Late Eocene to Early Oligocene. It cannot be ascertained if the Pelobates sp. from the Selety 1A 1222 (early Messinian, Miocene) is a European or Asian migrant. 1223 1224 Hylidae 1225 The family of tree frogs, Hylidae, has a wide distribution in Eurasia and is represented by the 1226 monophyletic genus Hyla. The most recent phylogenetic study of the genus Hyla by Li et al. 1227 (2015) recognised two closely related clades in Eurasia, namely the West Palaearctic arborea- 1228 group and East Palaearctic chinensis-group, as well as a small East Palaearctic japonica-group 1229 that is related to the North American clade of Hyla. The revision of the Western Eurasian Hyla 3 localities: Zertsalo [Sunduk Section], lager Biryukova [Kiin-Kerish Section], lower faunistic level of Plesh [Kusto- Kyzylkain Section], probably also Tabtym [Sarykamysh Section] 4 localities: main level of Plesh, Tuzkabak, Cherepakhovoe Pole [Tayzhuzgen Section], Raskop [Aksyr Section], Tyubiteika, sopki ʻRybnayaʼ and Kontrolnaya [Juvan-Kara Section] 5 localities: Maylibay Tologay [Tayzhuzgen Section] Podorozhnik [Jaman-Kara Section] Manuscript to be review 1212 of the family (Chkhikvadze, 1985, 1998) includes numerous remains of Pelobatidae indet. from: 1213 (1) the localities of the Zaisan Basin from the Upper Aksyr Svita3, early Priabonian; rare finds in 1214 the Kusto Svita and basal horizon of Buran Svita4, late Priabonian and earliest Rupelian; 1215 abundant occurrences in the Buran Svita5, early Rupelian and (2) large-sized spadefoot toads 1216 from the Kyzyl-Kak locality of the Turgay Basin, Central Kazakhstan, Late Oligocene 1217 (Chkhikvadze, 1998). 5 localities: Maylibay, Tologay [Tayzhuzgen Section], Podorozhnik [Jaman-Kara Section] Manuscript to be reviewed According to our analysis, it is clear that the ancestor of the ʻB. 1194 bombina – B. variegataʼ clade was present in Europe from, at least, the later part of the Early 1195 Miocene. Later in the Middle Miocene they expanded into Western Asia, reaching the east from 1196 the Ural Mountains. The Western Siberian fossil Bombina can be clearly osteologically 1197 separated from Bombina orientalis, a member of the East Asian clade of the genus. Taking their 1198 recent distribution as well as the fossil record into account, a split of the European and Asian 1199 Bombina clades seems most probable in Asia during the Paleogene. 1202 The family of European spadefoot toads Pelobatidae includes only one extant genus with four 1203 species distributed in north-western Africa, Europe, in small areas that are east of the Ural 1204 Mountains in Russia and in the north of Kazakhstan (Kuzmin, 1995). The family has Laurasian 1205 affinities and records are known from the Late Cretaceous in North America. The presence of 1206 pelobatids in Europe dates back to the Early Eocene, as indicated by the fossil genus 1207 Eopelobates (Middle Eocene – late Pliocene), as well as by the fossil forms of the Recent genus 1208 Pelobates (Middle Oligocene – Recent) (Roček et al., 2014). The Asian record of Pelobatidae is 1209 very limited and includes forms from the Eo-Oligocene of Kazakhstan (Chkhikvadze, 1985) and 1210 Eocene of India (Folie et al., 2013). Recently, Roček et al. (2014) excluded the genus Uldzinia 1211 (Oligocene, Mongolia) (Gubin, 1995) from the family Pelobatidae. The Kazakhstan fossil record PeerJ reviewing PDF \| (2016:08:12606:4:0:NEW 23 Jan 2017) 6 The divergence dates of split events were estimated by a relaxed molecular clock approach, based on the mitochondrial data set, where the calibration with fossil record is missing. Manuscript to be reviewed Revision of this rich pelobatid record from the Zaisan Basin was not 1218 possible due to the lack of descriptions and illustrations of the material as well as the difficulty 1219 in accessing the specimens. Nevertheless, taking the Paleogene fossil records into account, we 1220 inferred that the spadefoot toads may have dispersed from Europe to Western Asia during the 1221 Late Eocene to Early Oligocene. It cannot be ascertained if the Pelobates sp. from the Selety 1A 1222 (early Messinian, Miocene) is a European or Asian migrant. 1212 of the family (Chkhikvadze, 1985, 1998) includes numerous remains of Pelobatidae indet. from: 1213 (1) the localities of the Zaisan Basin from the Upper Aksyr Svita3, early Priabonian; rare finds in 1214 the Kusto Svita and basal horizon of Buran Svita4, late Priabonian and earliest Rupelian; 1215 abundant occurrences in the Buran Svita5, early Rupelian and (2) large-sized spadefoot toads 1216 from the Kyzyl-Kak locality of the Turgay Basin, Central Kazakhstan, Late Oligocene 1217 (Chkhikvadze, 1998). Revision of this rich pelobatid record from the Zaisan Basin was not 1218 possible due to the lack of descriptions and illustrations of the material as well as the difficulty 1219 in accessing the specimens. Nevertheless, taking the Paleogene fossil records into account, we 1220 inferred that the spadefoot toads may have dispersed from Europe to Western Asia during the 1221 Late Eocene to Early Oligocene. It cannot be ascertained if the Pelobates sp. from the Selety 1A 1222 (early Messinian, Miocene) is a European or Asian migrant. 4 localities: main level of Plesh, Tuzkabak, Cherepakhovoe Pole [Tayzhuzgen Section], Raskop [Aksyr Section], Tyubiteika, sopki ʻRybnayaʼ and Kontrolnaya [Juvan-Kara Section] PeerJ reviewing PDF \| (2016:08:12606:4:0:NEW 23 Jan 2017) mitochondrial data set, where the calibration with fossil record is missing. Manuscript to be reviewed 1230 phylogeny, based on molecular genetic studies, revealed a high diversity in the area containing 1231 about eight (?nine) (Li et al., 2015) or ten (Gvoždík et al., 2010) species. Among them are two 1232 clades: (1) H. savignyi in the east (Levant and the area of Turkey, Iran, Armenia, Georgia) and (2) 1233 H. arborea (Western, Central Europe and Balkan) + H. orientalis (South-eastern Europe, 1234 Georgia, Armenia, Iran), which have wide distributions in the east and west respectively (Stöck 1235 et al., 2008a; Gvoždík et al., 2010). 1230 phylogeny, based on molecular genetic studies, revealed a high diversity in the area containing 1231 about eight (?nine) (Li et al., 2015) or ten (Gvoždík et al., 2010) species. Among them are two 1232 clades: (1) H. savignyi in the east (Levant and the area of Turkey, Iran, Armenia, Georgia) and (2) 1233 H. arborea (Western, Central Europe and Balkan) + H. orientalis (South-eastern Europe, 1230 phylogeny, based on molecular genetic studies, revealed a high diversity in the area containing 1231 about eight (?nine) (Li et al., 2015) or ten (Gvoždík et al., 2010) species. Among them are two 1232 clades: (1) H. savignyi in the east (Levant and the area of Turkey, Iran, Armenia, Georgia) and (2) 1233 H. arborea (Western, Central Europe and Balkan) + H. orientalis (South-eastern Europe, 1234 Georgia, Armenia, Iran), which have wide distributions in the east and west respectively (Stöck 1235 et al., 2008a; Gvoždík et al., 2010). 1236 The oldest European record of the genus is known from the Oberdorf O4 locality, late Early 1237 Miocene, Austria (Sanchíz, 1998b). After an interruption/gap of approximately three million 1238 years, records of the genus continued in the late Langhian with the first fossil appearance of the 1239 Recent species H. arborea (H. cf. arborea, Mátraszőlős 2, Hungary) (Venczel, 2004). The record 1240 is almost consistent in the entire Neogene and Quaternary periods of Europe (Fig. 9). There is 1241 quite an abundant record of the genus with the oldest and first occurrences of H. savignyi (H. 1242 cf. savignyi) derived from five localities in Western Siberia, dating back to the late Late Miocene 1243 and early Early Pliocene. Apart from the distribution in Siberia, H. savignyi also may occur in 1244 southern Russia, in the Middle Pleistocene (Ratnikov, 2002)(see ʻComparison and Discussionʼ in 1245 Hyla gr. H. Manuscript to be reviewed savignyi), representing the youngest fossil record of the species. 1246 Based on the fossil record of the tree frogs, we concluded that two large Western Eurasian 1247 clades split in Europe during the Middle Miocene. Our data indicated older ages for the first 1248 fossil occurrences of these clades than has been previously estimated from molecular data in 1249 two recent studies (Gvoždík et al., 2010; Li et al., 2015). Gvoždík et al. (2010)6 suggested that 6 The divergence dates of split events were estimated by a relaxed molecular clock approach, based on the mitochondrial data set, where the calibration with fossil record is missing. PeerJ reviewing PDF \| (2016:08:12606:4:0:NEW 23 Jan 2017) Manuscript to be reviewed 1250 the split of H. orientalis/arborea and H. savignyi occurred 11.1 Ma (early Late Miocene, early 1251 Tortonian), which is approximately three million years younger than the first fossil occurrence 1252 of H. cf. arborea (Table S5). Whereas, without calibrating the molecular clock using the oldest 1253 European fossil Hyla (Hyla sp., Oberdorf O4 locality in Austria), Li et al. (2015) estimated this 1254 split to have occurred at 12-20 Ma, during a time interval in which the oldest fossil tree frogs 1255 related to the Recent H. arborea occurred. In both of the cases, the interpretation of the 1256 molecular phylogeny of the group can be improved by calibrating the phylogenetic tree with 1257 the fossil record introduced in this study. 1257 the fossil record introduced in this study. 1258 Considering our data and the results presented by Li et al. (2015), we suggest the following 1259 distribution pattern for the West Palaearctic H. arborea-group: (1) the group entered Eurasia 1260 from the east via Beringia from North America, during the Paleogene; and (2) the ancestors of 1261 the group reached Europe during the early Miocene via the Tugai Strait between Europe and 1262 Asia (the Turgai Strait) and diversified, apparently, in Western Siberian. The Late Miocene and 1263 Early Pliocene records represent the most eastern expansion of the European genera, when the 1264 climatic conditions were still favourable for their distribution; it is conceivable for us that the H. 1265 savignyi may have potentially so far not found fossil occurrences in the Miocene of Eastern 1266 Europe and/or from the Caucasus in the south. 1267 1268 Bufonidae 1269 Two groups of toads were found in the studied localities in Western Siberian; namely the 1270 common (Bufo bufo) and the green toads (Bufotes cf. viridis) (Fig. 7F-7K, Table S1). The toads of 1258 Considering our data and the results presented by Li et al. (2015), we suggest the following 1259 distribution pattern for the West Palaearctic H. arborea-group: (1) the group entered Eurasia 1260 from the east via Beringia from North America, during the Paleogene; and (2) the ancestors of 1261 the group reached Europe during the early Miocene via the Tugai Strait between Europe and 1262 Asia (the Turgai Strait) and diversified, apparently, in Western Siberian. Manuscript to be reviewed The Late Miocene and 1263 Early Pliocene records represent the most eastern expansion of the European genera, when the 1264 climatic conditions were still favourable for their distribution; it is conceivable for us that the H. 1265 savignyi may have potentially so far not found fossil occurrences in the Miocene of Eastern 1266 Europe and/or from the Caucasus in the south. 1258 Considering our data and the results presented by Li et al. (2015), we suggest the following 1259 distribution pattern for the West Palaearctic H. arborea-group: (1) the group entered Eurasia 1260 from the east via Beringia from North America, during the Paleogene; and (2) the ancestors of 1261 the group reached Europe during the early Miocene via the Tugai Strait between Europe and 1262 Asia (the Turgai Strait) and diversified, apparently, in Western Siberian. The Late Miocene and 1263 Early Pliocene records represent the most eastern expansion of the European genera, when the 1264 climatic conditions were still favourable for their distribution; it is conceivable for us that the H. 1265 savignyi may have potentially so far not found fossil occurrences in the Miocene of Eastern 1266 Europe and/or from the Caucasus in the south. PeerJ reviewing PDF \| (2016:08:12606:4:0:NEW 23 Jan 2017) PeerJ reviewing PDF \| (2016:08:12606:4:0:NEW 23 Jan 2017) Manuscript to be reviewed 1271 both groups are, with their occurrences are the most abundant among frogs remains found at 1272 the fossil localities. 1271 both groups are, with their occurrences are the most abundant among frogs remains found at 1272 the fossil localities. 1271 both groups are, with their occurrences are the most abundant among frogs remains found at 1272 the fossil localities. g p , g g 1272 the fossil localities. 1273 Common toads. Bufo bufo is the Recent species with the widest distribution (i.e. Central, 1274 Southern, Eastern Europe and Western and Eastern Asia) of all members of the common toads 1275 Bufo bufo species group. This group includes three other species with limited distribution, 1276 namely: Bufo spinosus (northern Africa, Western Europe), Bufo eichwaldi (south coast of the 1277 Caspian Sea), and Bufo verrucosissimus (east of the Black Sea) (Arntzen et al., 2013). These 1278 species are known also as the western group of the genus. Their nearby Eastern Asian relatives 1279 - the eastern group, include the Bufo gargarizans species group. The Western Siberian fossil 1280 record of the Bufo bufo species group is restricted to the late Late Miocene to the early Early 1281 Pliocene, which in comparison to the European record, is very poorly represented. The oldest 1282 toad remains that are assigned to the Bufo bufo species group are from the Middle Miocene of 1283 Slovakia: B. bufo from the Devinská Nová Ves – Zapfe’s fissure locality, 13.7-14 Ma (Hodrova, 1284 1980; Böhme, 2003) and B. cf. bufo from the Devinská Nová Ves – Bonanza locality, 13.5-13.7 1285 Ma (Hodrová, 1988). Then, since 9.2 Ma during the Late Miocene (Suchomasty locality in the 1286 Czech Republic) (Hodrová, 1987), B. bufo representatives are present in Central Europe and 1287 extend their distribution across Europe. At ca. 4.7 Ma, remains of the common toad, exhibiting 1288 characters of the Recent B. spinosus, appeared in Spain, in the Celadas 6 locality (Böhme & Ilg, 1289 2003). The oldest fossil remains referred to B. verrucosissimus were recovered from the Late 1290 Pliocene (3.0-3.8 Ma) in the Apastovo locality in Russia (Ratnikov, 2001). The Western Siberian 1291 record suggests at least a late Miocene dispersal of B. bufo to the east, reaching the present 1292 distribution area of the species. Considering the genomic data of Recuero et al. (2012), these PeerJ reviewing PDF \| (2016:08:12606:4:0:NEW 23 Jan 2017) Manuscript to be reviewed 1293 ʻmigrantsʼ should represent the common ancestor of the B. bufo + B. verrucosissimus clade, 1294 expanding to the east into Asia and to the south into Eastern Europe. This bufonids most 1295 probably remained, permanently, in these areas, until present times. The lack of their 1296 representation in the fossil record in the Late Pliocene and Quaternary sites can be explained 1297 by sampling bias. Although B. bufo and B. verrucosissimus do not occur sympatrically nowadays, 1298 specimens of both these species have been found together in two Middle Pleistocene localities 1299 (Koziy Ovrag and Yablonovets from Russia; see more in Table S5). 1293 ʻmigrantsʼ should represent the common ancestor of the B. bufo + B. verrucosissimus clade, 1294 expanding to the east into Asia and to the south into Eastern Europe. This bufonids most 1295 probably remained, permanently, in these areas, until present times. The lack of their 1296 representation in the fossil record in the Late Pliocene and Quaternary sites can be explained 1297 by sampling bias. Although B. bufo and B. verrucosissimus do not occur sympatrically nowadays, 1298 specimens of both these species have been found together in two Middle Pleistocene localities 1299 (Koziy Ovrag and Yablonovets from Russia; see more in Table S5). 1300 Two recent molecular studies (Garcia-Porta et al., 2012; Recuero et al., 2012; pp. 71-86) 1313 Green toads. The range of the widely distributed Bufotes viridis species group (or Bufotes viridis 1314 sensu lato) extends across Central Europe to Central Asia, as well as the entire northern Africa 1313 Green toads. The range of the widely distributed Bufotes viridis species group (or Bufotes viridis 1314 sensu lato) extends across Central Europe to Central Asia, as well as the entire northern Africa 1313 Green toads. The range of the widely distributed Bufotes viridis species group (or Bufotes viridis 1314 sensu lato) extends across Central Europe to Central Asia, as well as the entire northern Africa PeerJ reviewing PDF \| (2016:08:12606:4:0:NEW 23 Jan 2017) Manuscript to be reviewed 1315 and Mediterranean area, including numerous islands. The species complex is highly diverse and 1316 includes over ten recognised species, e.g. Bufotes balearicus (southern Mediterranean and 1317 Apennine Peninsula, Corsica, Sardinia, Balearic Islands), Bufotes boulengeri (northern Africa), 1318 Bufotes siculus (Sicily), B. viridis (Central and Eastern Europe), Bufotes variabilis (Balkans, 1319 Anatolia, Caucasus) etc., found in a diverse range of environments (Stöck et al., 2006; Stöck et 1320 al., 2008b). Among them, however, no valuable osteological characters has been established for 1321 taxonomic identification (Blain, Gibert & Ferràndez-Cañadell, 2010). Hence, no precise specific 1322 assignment of any fossil material is possible. Blain, Gibert & Ferràndez-Cañadell (2010) recently 1323 showedthat the green toads were also present in the Iberian Peninsula in the Early Pleistocene, 1324 1.1-1.3 Ma, and suggested that they became extinct due to climatic changes and/or 1325 competition. 1315 and Mediterranean area, including numerous islands. The species complex is highly diverse and 1316 includes over ten recognised species, e.g. Bufotes balearicus (southern Mediterranean and 1317 Apennine Peninsula, Corsica, Sardinia, Balearic Islands), Bufotes boulengeri (northern Africa), 1318 Bufotes siculus (Sicily), B. viridis (Central and Eastern Europe), Bufotes variabilis (Balkans, 1319 Anatolia, Caucasus) etc., found in a diverse range of environments (Stöck et al., 2006; Stöck et 1320 al., 2008b). Among them, however, no valuable osteological characters has been established for 1321 taxonomic identification (Blain, Gibert & Ferràndez-Cañadell, 2010). Hence, no precise specific 1322 assignment of any fossil material is possible. Blain, Gibert & Ferràndez-Cañadell (2010) recently 1323 showedthat the green toads were also present in the Iberian Peninsula in the Early Pleistocene, 1324 1.1-1.3 Ma, and suggested that they became extinct due to climatic changes and/or 1325 competition. 1326 In the studied Western Siberian localities, fossil remains that are related to B. viridis are the 1327 most frequently occurring element in the Western Siberian herpetofauna. This species is almost 1328 permanently present from the Middle Miocene to Early Pleistocene. Specimens are found in 1329 the late Middle Miocene localities, and although there are gaps in the record, remains are 1330 present in the late Late Miocene to Early Pleistocene localities (Table S1). In the youngest 1331 localities (Olkhovka 1A, Lebiazhie 1A, Lebiazhie 1B), they are found as a sole taxon. Manuscript to be reviewed Further 1332 fossils assigned to the family Bufonidae (Bufonidae indet.) were already reported from the 1333 Kentyubek locality in the Turgay Basin, from the Middle Miocene (Bendukidze & Chkhikvadze, 1334 1976), and from two localities in the Zaisan Basin: the Zmei Gorynych locality in Akzhar Svita, 1335 from the Early Miocene (Chkhikvadze, 1985) and from the early Rupelian age fossil sites (see 1336 section ʻPelobatidae’) of the Buran Svita (Chkhikvadze, 1998). Malakhov (2005) described the 1326 In the studied Western Siberian localities, fossil remains that are related to B. viridis are the 1327 most frequently occurring element in the Western Siberian herpetofauna. This species is almost 1328 permanently present from the Middle Miocene to Early Pleistocene. Specimens are found in 1329 the late Middle Miocene localities, and although there are gaps in the record, remains are 1330 present in the late Late Miocene to Early Pleistocene localities (Table S1). In the youngest 1331 localities (Olkhovka 1A, Lebiazhie 1A, Lebiazhie 1B), they are found as a sole taxon. Further 1332 fossils assigned to the family Bufonidae (Bufonidae indet.) were already reported from the 1333 Kentyubek locality in the Turgay Basin, from the Middle Miocene (Bendukidze & Chkhikvadze, 1334 1976), and from two localities in the Zaisan Basin: the Zmei Gorynych locality in Akzhar Svita, 1335 from the Early Miocene (Chkhikvadze, 1985) and from the early Rupelian age fossil sites (see 1336 section ʻPelobatidae’) of the Buran Svita (Chkhikvadze, 1998). Malakhov (2005) described the PeerJ reviewing PDF \| (2016:08:12606:4:0:NEW 23 Jan 2017) PeerJ reviewing PDF \| (2016:08:12606:4:0:NEW 23 Jan 2017) Manuscript to be reviewed 1359 the early Burdigalian into Europe during the middle Burdigalian. Apparently, the European 1360 Neogene record should not necessarily represent one ʻlineageʼ or one dispersal event of the B. 1361 viridis group from Asia. Several migration events most probably took place during the Miocene. 1362 The descendants of these events were replaced later by the ancestors of the Recent species B. 1363 viridis, B. variabilis, etc. as indicated by the genetic data at the Mio-Pliocene transition (Stöck et 1364 al., 2006). Prospective further studies could include: (1) the verification of dispersal events in 1365 the European fossil record, with help of an abundant and species-rich fossil material from 1366 stratigraphically well-dated localities; (2) the exploration the Miocene record of Anatolian and 1367 South-eastern Europe, as well as the Paleogene record of Asia; and (3) a challenging project of 1368 establishing the osteological characters that are important for the systematic identification of 1369 the members of the B. viridis species group. 1370 1371 Ranidae 1372 The family of true frogs, Ranidae, are present in the Western Siberian record by both green 1373 (Pelophylax sp.) and brown (Rana sp.) frogs. The green frogs appear more frequently in the 1374 record than the brown frogs. Both frog genera are common amphibians in the Recent 1375 herpetofauna of the area. Besides this record, further true frog finds (e.g. Ranidae indet.) are 1376 reported from the early Rupelian age fossil sites (see the list of the locality section 1377 ‘Pelobatidae’) of the Buran Svita, Zaisan Basin. We were not able to revise their taxonomic 1378 validity due to lack of figured fossils and the inaccessibility of the material. 1379 Green frogs. The genus Rana includes 21 Recent species of aquatic frogs having a wide 1380 distribution ranging from northern Africa, Europe to Eastern Asia. Two genetically distinct 1359 the early Burdigalian into Europe during the middle Burdigalian. Apparently, the European 1360 Neogene record should not necessarily represent one ʻlineageʼ or one dispersal event of the B. 1361 viridis group from Asia. Several migration events most probably took place during the Miocene. 1362 The descendants of these events were replaced later by the ancestors of the Recent species B. 1363 viridis, B. variabilis, etc. as indicated by the genetic data at the Mio-Pliocene transition (Stöck et 1364 al., 2006). Manuscript to be reviewed 1337 stratigraphically oldest green toad fossil, B. aff. viridis, from the early Early Miocene (20.4-22.5 1338 Ma, Aquitanian) locality of Ayakoz in North-eastern Kazakhstan (Fig. 1, Table S1). B. aff. viridis 1339 from the Ayakoz locality is older than the B. aff. viridis from the Early Miocene Keseköy locality 1340 (18-20 Ma) in north-western Turkey (Claessens, Leon P. A. M., 1997). All the occurrences of the 1341 oldest European fossils of green toads are from the Early Miocene: Vieux-Collonges locality in 1342 France (14-17 Ma), (Bailon & Hossini, 1990); Petersbuch 2 and 7 (17.5-18 Ma) localities in 1343 Germany (Böhme & Ilg, 2003); and probably the Córcoles locality (17-18 Ma) in Spain (Sanchíz, 1344 1998a). Once the green toads entered Europe, they became a regular element of the European 1345 Neogene and Quaternary herpetofaunal assemblages (Fig. 9). Besides B. aff. viridis, the 1346 European record of green toads includes another species, Bufotes priscus, from four localities of 1347 the latest Early Miocene to the earliest Late Miocene age (see Table S5). Taking into account 1348 the B. viridis Neogene records and the bufonid records from the Eurasian Paleogene, we 1349 suggest that the group arrived in the Old World in the Paleocene (Rage, 2003), entered Central 1350 Asia in the Early Oligocene and diversified there. Although we were not able to study the 1351 Paleogene bufonid record from Kazakhstan, taking into consideration the palaeogeography of 1352 common and green frogs, the assignment of the Early Oligocene Kazakhstan record to the green 1353 toad seems most probable. Apparently, the Early Oligocene forms were ancestral to the B. 1354 viridis lineage, which evolved in Central Asia in the Early Miocene. This assumption is also 1355 supported by molecular data suggesting that: (1) the green toad clade underwent 1356 diversification in Asia during the Oligocene/early Miocene; and (2) a high genomic and specific 1357 diversity is found within the Central Asian green toads (Stöck et al., 2006). Present in the 1358 Central Asian fossil record from the Early Miocene; they consequently dispersed via Anatolia in PeerJ reviewing PDF \| (2016:08:12606:4:0:NEW 23 Jan 2017) Manuscript to be reviewed Prospective further studies could include: (1) the verification of dispersal events in 1365 the European fossil record, with help of an abundant and species-rich fossil material from 1366 stratigraphically well-dated localities; (2) the exploration the Miocene record of Anatolian and 1367 South-eastern Europe, as well as the Paleogene record of Asia; and (3) a challenging project of 1368 establishing the osteological characters that are important for the systematic identification of 1369 the members of the B. viridis species group. 1359 the early Burdigalian into Europe during the middle Burdigalian. Apparently, the European 1360 Neogene record should not necessarily represent one ʻlineageʼ or one dispersal event of the B. 1361 viridis group from Asia. Several migration events most probably took place during the Miocene. 1362 The descendants of these events were replaced later by the ancestors of the Recent species B. 1363 viridis, B. variabilis, etc. as indicated by the genetic data at the Mio-Pliocene transition (Stöck et 1364 al., 2006). Prospective further studies could include: (1) the verification of dispersal events in 1365 the European fossil record, with help of an abundant and species-rich fossil material from 1366 stratigraphically well-dated localities; (2) the exploration the Miocene record of Anatolian and 1367 South-eastern Europe, as well as the Paleogene record of Asia; and (3) a challenging project of 1368 establishing the osteological characters that are important for the systematic identification of 1369 the members of the B. viridis species group. 1379 Green frogs. The genus Rana includes 21 Recent species of aquatic frogs having a wide 1380 distribution ranging from northern Africa, Europe to Eastern Asia. Two genetically distinct PeerJ reviewing PDF \| (2016:08:12606:4:0:NEW 23 Jan 2017) Manuscript to be reviewed 1381 clades, i.e. Western Palaearctic and the Far East, are recognised within the green frogs genus 1382 Pelophylax (Lymberakis et al., 2007). The oldest green frog record from the Western Siberian 1383 (Pelophylax sp.) is dated back to the late Middle Miocene, coinciding stratigraphically with the 1384 Eastern Siberian record of the group (Middle Miocene, ca. 13 Ma, Tagay Section, Baikal Lake, 1385 Russia) (Daxner-Höck et al., 2013). Records of this group are present in the studied localities 1386 until the late Early Pliocene with long (during the Late Miocene) and short (during the Early 1387 Pliocene) gaps in the fossil record. Due to the fragmentary preservation of the studied bones as 1388 well as the lack of other informative elements of the skeleton (e.g. frontoparietals), any 1389 assignment to the Recent green frog species was impossible. Considering the present 1390 distribution of the two green frog clades, an affiliation of the Western Siberian fossil record to 1391 the Western Palaearctic clade is most probable. 1381 clades, i.e. Western Palaearctic and the Far East, are recognised within the green frogs genus 1382 Pelophylax (Lymberakis et al., 2007). The oldest green frog record from the Western Siberian 1383 (Pelophylax sp.) is dated back to the late Middle Miocene, coinciding stratigraphically with the 1384 Eastern Siberian record of the group (Middle Miocene, ca. 13 Ma, Tagay Section, Baikal Lake, 1385 Russia) (Daxner-Höck et al., 2013). Records of this group are present in the studied localities 1386 until the late Early Pliocene with long (during the Late Miocene) and short (during the Early 1387 Pliocene) gaps in the fossil record. Due to the fragmentary preservation of the studied bones as 1388 well as the lack of other informative elements of the skeleton (e.g. frontoparietals), any 1389 assignment to the Recent green frog species was impossible. Considering the present 1390 distribution of the two green frog clades, an affiliation of the Western Siberian fossil record to 1391 the Western Palaearctic clade is most probable. 1392 Despite the being only a few green frog records described in this study, these records still 1393 significantly expand the previously scarce and poorly known fossil history of the genus. 1393 significantly expand the previously scarce and poorly known fossil history of the genus. Manuscript to be reviewed 1394 Moreover, both of the Middle Miocene records from Western and Eastern Siberia represent the 1395 oldest records of the green frogs in the Asian continent. Although an Asiatic origin of the green 1396 frogs has been already assumed by several authors e.g. Sanchíz, Schleich & Esteban (1993), 1397 Lymberakis et al. (2007), the earliest frog remains have been assigned to the Pelophylax 1398 ridibundus species group, which occurred in Europe in the early Oligocene (Möhren 13 locality, 1399 Germany) (Sanchíz, Schleich & Esteban, 1993). Its affiliation to a living species is impossible. In 1400 Europe, the fossil record of Pelophylax is continuous and is maintained through the Oligocene 1401 and entire Neogene (Table S5). Nevertheless, a well-documented Paleogene record of the 1402 group is not available from Asia and, therefore, any interpretations would not be confident. The 1393 significantly expand the previously scarce and poorly known fossil history of the genus. 1394 Moreover, both of the Middle Miocene records from Western and Eastern Siberia represent the 1395 oldest records of the green frogs in the Asian continent. Although an Asiatic origin of the green 1396 frogs has been already assumed by several authors e.g. Sanchíz, Schleich & Esteban (1993), 1397 Lymberakis et al. (2007), the earliest frog remains have been assigned to the Pelophylax 1398 ridibundus species group, which occurred in Europe in the early Oligocene (Möhren 13 locality, 1399 Germany) (Sanchíz, Schleich & Esteban, 1993). Its affiliation to a living species is impossible. In 1400 Europe, the fossil record of Pelophylax is continuous and is maintained through the Oligocene 1401 and entire Neogene (Table S5). Nevertheless, a well-documented Paleogene record of the 1402 group is not available from Asia and, therefore, any interpretations would not be confident. The PeerJ reviewing PDF \| (2016:08:12606:4:0:NEW 23 Jan 2017) Manuscript to be reviewed 1403 only possible scenario, taking into account both the fossil record and genomic data, is that the 1404 Western Palaearctic green frogs split from their Far East sister clade during the Eocene; they 1405 diversified in the territory of Europe and/or Western Asia during the Oligocene; they dispersed 1406 back to the East in the middle Miocene; and eventually reached the territory of the Western 1407 Siberia. 1403 only possible scenario, taking into account both the fossil record and genomic data, is that the 1404 Western Palaearctic green frogs split from their Far East sister clade during the Eocene; they 1405 diversified in the territory of Europe and/or Western Asia during the Oligocene; they dispersed 1406 back to the East in the middle Miocene; and eventually reached the territory of the Western 1407 Siberia. 1408 Brown frogs. The genus Rana (subgenus Rana sensu Veith, Kosuch & Vences, (2003) is 1409 comprised of more than 15 species that are distributed throughout Eurasia. Similar to green 1410 frogs, there are two known lineages from the brown frog species, namely: the Western and the 1411 Eastern Palaearctic lineages (Veith, Kosuch & Vences, 2003). Based on the osteological 1412 characters, the studied Western Siberian brown frog remains show a relation to the Western 1413 Asiatic lineage of the genus Rana, more precisely to the Rana temporaria species group (sensu 1414 Veith, Kosuch & Vences, 2003). Among the late Paleogene and Early Miocene fossil frogs 1415 (Böhme & Ilg, 2003), in which the generic identification is unclear (Rana vel Pelophylax), only 1416 the frog remains from the Early Miocene in Dietrichsberg, Germany (Böhme, 2001) have 1417 definitely been assigned to the brown frog R. cf. temporaria, representing the oldest known 1418 record of the group so far. As already suggested by Böhme (2001), brown frogs migrated from 1419 their possible centre of origin in Western or Central Asia to Europe during the second half of 1420 the Early Miocene. This hypothesis is confirmed by the brown frog fossils from the Ayakoz 1421 locality in Kazakhstan, which dates back to the Aquitanian age and are stratigraphically older 1422 than the Dietrichsberg fossil frogs. The present-day biogeography and diversity of brown frogs, 1423 the presence of a distinct Eastern Palaearctic lineage in Eastern Asia as well as the Asian 1424 distribution of many European species provide further support for an Asiatic origin. Most likely, 1408 Brown frogs. Manuscript to be reviewed 1425 the dispersal route of the brown frogs was similar to that of the green toad (Bufotes cf. viridis) 1426 whereby dispersal into Europe occured via Anatolia, during the Early Miocene. 1427 It is interesting to note that the earliest brown frog from the studied Western Siberian localities 1428 (Malyi Kalkaman 2) shows osteological similarities with the Recent species R. temporaria, 1429 representing herewith the oldest fossil record of the species in the east. 1430 Previous molecular studies (Veith, Kosuch & Vences, 2003; Lymberakis et al., 2007), on both 1431 green and brown frogs, aimed to reconstruct their phylogenetic relationships, suggest models 1432 of biogeographic history as well as suggest when the splits between different genera, clades, 1433 species, etc. occurred. Such studies have provided contradictory results also for this group, e.g. 1434 the split of Rana and Pelophylax was at 9.32 Ma (Veith, Kosuch & Vences, 2003), whereas 1435 Lymberakis et al. (2007) estimated the split of the Western Palaearctic and Far East lineages of 1436 Pelophylax to have occurred significantly earlier, i.e. 15 Ma before. Here neither geologic 1437 events nor the fossil records have been used consistently for the calibration of the molecular 1438 clock. Thus, the recalibrating of the timing for the splits with the new fossil finds provides a 1439 more reliable basis for phylogenetic reconstructions. 1440 For the better understanding of relationships between these groups, as well as to reveal more 1425 the dispersal route of the brown frogs was similar to that of the green toad (Bufotes cf. viridis) 1426 whereby dispersal into Europe occured via Anatolia, during the Early Miocene. 1425 the dispersal route of the brown frogs was similar to that of the green toad (Bufotes cf. viridis) 1426 whereby dispersal into Europe occured via Anatolia, during the Early Miocene. 1427 It is interesting to note that the earliest brown frog from the studied Western Siberian localities 1428 (Malyi Kalkaman 2) shows osteological similarities with the Recent species R. temporaria, 1429 representing herewith the oldest fossil record of the species in the east. Manuscript to be reviewed The genus Rana (subgenus Rana sensu Veith, Kosuch & Vences, (2003) is 1409 comprised of more than 15 species that are distributed throughout Eurasia. Similar to green 1410 frogs, there are two known lineages from the brown frog species, namely: the Western and the 1411 Eastern Palaearctic lineages (Veith, Kosuch & Vences, 2003). Based on the osteological 1412 characters, the studied Western Siberian brown frog remains show a relation to the Western 1413 Asiatic lineage of the genus Rana, more precisely to the Rana temporaria species group (sensu 1414 Veith, Kosuch & Vences, 2003). Among the late Paleogene and Early Miocene fossil frogs 1415 (Böhme & Ilg, 2003), in which the generic identification is unclear (Rana vel Pelophylax), only 1416 the frog remains from the Early Miocene in Dietrichsberg, Germany (Böhme, 2001) have 1417 definitely been assigned to the brown frog R. cf. temporaria, representing the oldest known 1418 record of the group so far. As already suggested by Böhme (2001), brown frogs migrated from 1419 their possible centre of origin in Western or Central Asia to Europe during the second half of 1420 the Early Miocene. This hypothesis is confirmed by the brown frog fossils from the Ayakoz 1421 locality in Kazakhstan, which dates back to the Aquitanian age and are stratigraphically older 1422 than the Dietrichsberg fossil frogs. The present-day biogeography and diversity of brown frogs, 1423 the presence of a distinct Eastern Palaearctic lineage in Eastern Asia as well as the Asian 1424 distribution of many European species provide further support for an Asiatic origin. Most likely, PeerJ reviewing PDF \| (2016:08:12606:4:0:NEW 23 Jan 2017) Manuscript to be reviewed 1440 For the better understanding of relationships between these groups, as well as to reveal more 1441 around the origin and palaeobiogeographic history of them, it would be interesting to review 1442 the specimens of the Paleocene frogs (Ranidae indet.) from the early Rupelian fossil sites (see 1443 section ʻPelobatidae’) of the Buran Svita in the Zaisan Basin (Chkhikvadze, 1998). The 1444 incorporation of such a review, however, was not possible in the present study, due to the lack 1445 of figures of the fossils and the inaccessibility of the material. 1446 1440 For the better understanding of relationships between these groups, as well as to reveal more 1441 around the origin and palaeobiogeographic history of them, it would be interesting to review 1442 the specimens of the Paleocene frogs (Ranidae indet.) from the early Rupelian fossil sites (see 1443 section ʻPelobatidae’) of the Buran Svita in the Zaisan Basin (Chkhikvadze, 1998). The 1444 incorporation of such a review, however, was not possible in the present study, due to the lack 1445 of figures of the fossils and the inaccessibility of the material. 1446 PeerJ reviewing PDF \| (2016:08:12606:4:0:NEW 23 Jan 2017) Manuscript to be reviewed 1447 Gekkonidae 1448 The family Gekkonidae is represented in the Western Siberian fossil record by the straight- 1449 fingered or even-fingered geckos, genus Alsophylax. They occur only in the Cherlak locality, 1450 dated back to the terminal Miocene, ca. 5.9 Ma. Alsophylax sp. is the most abundant element in 1451 the herpetofaunal assemblage of the Cherlak locality, with approximately 70% of the 1452 identifiable bone material belonging to this taxon. The genus Alsophylax is mainly distributed in 1453 Central Asia, partly occurring also in Mongolia and China. These geckos prefer habitats in arid 1454 and warm landscapes (Ananjeva et al., 2006). The appearance of these dry and warm adapted 1455 geckos in Western Siberia, which is four degrees north of their present occurrence, suggests a 1456 shift of the arid environment from the south to the north at the end of the Late Miocene (see 1457 below). It is interesting to note that out of the seven gecko genera, e.g. Eublephareus, 1458 Mediadactylus, Terratoscincus (Ananjeva et al., 2006) inhabiting Central Asia, only Alsophylax, 1459 which has the most northern distribution, occurs in the fossil record. Apparently, this genus is 1460 ecologically more adaptable in comparison to other genera, not only in the present, but 1461 probably also in the past. 1462 1463 Lacertidae 1464 Lacertid remains are the most frequent fossil bones among those of lizards occurring in 1465 Western Siberian localities. They are very rare in the Middle Miocene faunas, but occur more 1466 frequently in the Late Miocene, Pliocene, and Pleistocene localities. In the middle Late Miocene 1467 locality Pavlodar 1A (ca. 7.25 Ma), two taxa (Lacerta s.l. sp. 1 and sp. 2) occur sympatrically. 1468 Eremias sp. appears in the Western Siberian record in the Pliocene. This genus is widely 1447 Gekkonidae 1448 The family Gekkonidae is represented in the Western Siberian fossil record by the straight- 1449 fingered or even-fingered geckos, genus Alsophylax. They occur only in the Cherlak locality, 1450 dated back to the terminal Miocene, ca. 5.9 Ma. Alsophylax sp. is the most abundant element in 1451 the herpetofaunal assemblage of the Cherlak locality, with approximately 70% of the 1452 identifiable bone material belonging to this taxon. The genus Alsophylax is mainly distributed in 1453 Central Asia, partly occurring also in Mongolia and China. These geckos prefer habitats in arid 1454 and warm landscapes (Ananjeva et al., 2006). Manuscript to be reviewed The appearance of these dry and warm adapted 1455 geckos in Western Siberia, which is four degrees north of their present occurrence, suggests a 1456 shift of the arid environment from the south to the north at the end of the Late Miocene (see 1457 below). It is interesting to note that out of the seven gecko genera, e.g. Eublephareus, 1458 M di d l T i (A j l 2006) i h bi i C l A i l Al h l 1447 Gekkonidae 1448 The family Gekkonidae is represented in the Western Siberian fossil record by the straight- 1449 fingered or even-fingered geckos, genus Alsophylax. They occur only in the Cherlak locality, 1450 dated back to the terminal Miocene, ca. 5.9 Ma. Alsophylax sp. is the most abundant element in 1451 the herpetofaunal assemblage of the Cherlak locality, with approximately 70% of the 1452 identifiable bone material belonging to this taxon. The genus Alsophylax is mainly distributed in 1453 Central Asia, partly occurring also in Mongolia and China. These geckos prefer habitats in arid 1454 and warm landscapes (Ananjeva et al., 2006). The appearance of these dry and warm adapted 1455 geckos in Western Siberia, which is four degrees north of their present occurrence, suggests a 1456 shift of the arid environment from the south to the north at the end of the Late Miocene (see 1457 below). It is interesting to note that out of the seven gecko genera, e.g. Eublephareus, 1453 Central Asia, partly occurring also in Mongolia and China. These geckos prefer habitats in arid 1454 and warm landscapes (Ananjeva et al., 2006). The appearance of these dry and warm adapted 1455 geckos in Western Siberia, which is four degrees north of their present occurrence, suggests a 1456 shift of the arid environment from the south to the north at the end of the Late Miocene (see 1457 below). It is interesting to note that out of the seven gecko genera, e.g. Eublephareus, 1458 Mediadactylus, Terratoscincus (Ananjeva et al., 2006) inhabiting Central Asia, only Alsophylax, 1459 which has the most northern distribution, occurs in the fossil record. Apparently, this genus is 1460 ecologically more adaptable in comparison to other genera, not only in the present, but 1461 probably also in the past. 1462 1463 Lacertidae 1464 Lacertid remains are the most frequent fossil bones among those of lizards occurring in 1465 Western Siberian localities. Manuscript to be reviewed They are very rare in the Middle Miocene faunas, but occur more 1466 frequently in the Late Miocene, Pliocene, and Pleistocene localities. In the middle Late Miocene 1467 locality Pavlodar 1A (ca. 7.25 Ma), two taxa (Lacerta s.l. sp. 1 and sp. 2) occur sympatrically. 1468 Eremias sp. appears in the Western Siberian record in the Pliocene. This genus is widely 1458 Mediadactylus, Terratoscincus (Ananjeva et al., 2006) inhabiting Central Asia, only Alsophylax, 1459 which has the most northern distribution, occurs in the fossil record. Apparently, this genus is 1460 ecologically more adaptable in comparison to other genera, not only in the present, but 1461 probably also in the past. PeerJ reviewing PDF \| (2016:08:12606:4:0:NEW 23 Jan 2017) PeerJ reviewing PDF \| (2016:08:12606:4:0:NEW 23 Jan 2017) Manuscript to be reviewed 1469 distributed in the Central Asian steppes, inhabiting dry and warm habitats (Ananjeva et al., 1470 2006). 1471 1472 Emydidae 1473 Emydoidea sp. is the only turtle identified from the studied fossil sites. The present-day 1474 distribution of the monotypic genus Emydoidea is restricted to the water bodies of the north- 1475 eastern territory of the USA. In Eurasia, fossil forms of this aquatic genus appear in the fossil 1476 record in Central Kazakhstan since the Middle Miocene (Emydoidea tasbaka, the Kentyubek 1477 locality in the Turgay Basin) (Chkhikvadze, 1989). Fossil forms have also been reported in 1478 Eastern Europe from the Late Miocene (Emydoidea tarashchuki, Krivoy Rog locality in Ukraine 1479 and Pantishara (8.7-9.2 Ma) in Georgia) (Chkhikvadze, 1980);(Chkhikvadze, 2003). The Siberian 1480 record indicates their occurrence in Asia also during the Late Miocene, which, interestingly, is 1481 located much further north than their Middle Miocene record from Kazakhstan. According to 1482 Chkhikvadze (2003), representatives may have also been present in Eastern Europe during the 1483 Pliocene. We avoid interpreting palaeobiogeography, stratigraphic distribution, etc. of this 1484 genus, since the available published material (e.g. Chkhikvadze, 1983, 1989), together with 1485 other extinct testudinoid taxa from Kazakhstan and Eastern Europe, is insufficiently described 1486 and poorly illustrated, requiring thorough revision. Nevertheless, we used the available 1487 published data on both freshwater turtles and terrestrial tortoises to attempt to interpret the 1488 record at the family level (Table 2). The turtle records from three well-explored regions in the 1489 studying area, i.e. Zaisan Basin, Turgay Basins and Western Siberia, are summarised in the Table 1490 2 . Throughout the entire Early Miocene in the Zaisan Basin, the turtle fauna is dominated by 1469 distributed in the Central Asian steppes, inhabiting dry and warm habitats (Ananjeva et al., 1470 2006). PeerJ reviewing PDF \| (2016:08:12606:4:0:NEW 23 Jan 2017) Manuscript to be reviewed 1491 aquatic forms, i.e. out of eight taxa only two are tortoises (Protestudo spp.). The aquatic forms 1492 remained dominant in the Zaisan Basin during the Middle Miocene, the terrestrial family 1493 Testudinidae completely replaced the aquatic turtles (Emydidae, Triochynidae) in the end of the 1494 Middle Miocene and became the only family present in the younger deposits of the Late 1495 Miocene. Similar to the Zaisan Basin, the aquatic forms represent the Middle Miocene turtle 1496 fauna in two adjacent regions, the Turgay Basin in the west and Western Siberia in the north. 1497 Subsequently, in the beginning of the early Late Miocene, a testudinid appears in Western 1498 Siberia and is replaced by an emydid towards the end of the late Late Miocene and a chelydrid 1499 at the Mio-Pliocene transition. The absence of tortoises since the end of the Late Miocene in 1500 Western Siberia and the Plio-Pleistocene in the Zaisan Basin can be explained by a less 1501 favourable, probably colder (MAT <15˚C, cold month temperature CMT <8˚C) climate. Since the 1502 late Late Miocene, the emydid and chelydrid aquatic turtles are the only chelonids in Western 1503 Siberia. The presence of these chelonids not only indicate a humid environment with standing 1504 water-bodies but most probably also a cooler climate (for emydids: MAT>8˚C, CM>-1.4 ˚C), 1505 since, in general, aquatic turtles can tolerate much colder conditions than tortoises, in that an 1506 aquatic environment acts as thermal buffer, consequently enabling aquatic turtles to populate 1507 higher poleward latitudes. 1508 1509 Palaeobiogeographic considerations 1510 By comparing the spatial and temporal patterns between European and Asian fossil records, 1511 including the first and last fossil occurrences, combined with an analysis of the available 1510 By comparing the spatial and temporal patterns between European and Asian fossil records, 1511 including the first and last fossil occurrences, combined with an analysis of the available PeerJ reviewing PDF \| (2016:08:12606:4:0:NEW 23 Jan 2017) Manuscript to be reviewed 1512 genomic data of the recent relatives of the fossil groups present in the studied material, certain 1513 palaeogeographic distribution patterns can be revealed along with new interpretations. 1514 Our analysis suggests a Western Asiatic origin for Hynobiidae, Proteidae, aff. Tylototriton, 1515 Bufotes viridis species group and brown frogs, Rana. The green toads and brown frogs 1516 dispersed coincidentally in the earliest Miocene wherein, and at least for the Bufotes viridis 1517 group, Anatolia was involved. Anatolia also played an important role in the distribution of the 1518 Bufo bufo species group; however, any age estimation of the event is not available. A 1519 salamander, showing affinities to the clade of the Recent East Asian genera Tylototriton + 1520 Echinotriton, is present in Western Siberia, most probably representing the forms similar to that 1521 of the early Oligocene (aff. Tylototriton) in Europe, a sister group of the recent clade. In order to 1522 resolve the affiliations of these fossils, further Paleogene material from both the Asia and 1523 European continents are necessary. 1512 genomic data of the recent relatives of the fossil groups present in the studied material, certain 1513 palaeogeographic distribution patterns can be revealed along with new interpretations. 1512 genomic data of the recent relatives of the fossil groups present in the studied material, certain 1513 palaeogeographic distribution patterns can be revealed along with new interpretations. 1514 Our analysis suggests a Western Asiatic origin for Hynobiidae, Proteidae, aff. Tylototriton, 1515 Bufotes viridis species group and brown frogs, Rana. The green toads and brown frogs 1516 dispersed coincidentally in the earliest Miocene wherein, and at least for the Bufotes viridis 1517 group, Anatolia was involved. Anatolia also played an important role in the distribution of the 1518 Bufo bufo species group; however, any age estimation of the event is not available. A 1519 salamander, showing affinities to the clade of the Recent East Asian genera Tylototriton + 1520 Echinotriton, is present in Western Siberia, most probably representing the forms similar to that 1521 of the early Oligocene (aff. Tylototriton) in Europe, a sister group of the recent clade. In order to 1522 resolve the affiliations of these fossils, further Paleogene material from both the Asia and 1523 European continents are necessary. Manuscript to be reviewed 1524 An eastward dispersal from Europe into Western Asia can be observed over a period ranging 1525 from the Middle to Late Miocene, based on the current data available from both European and 1526 Asiatic records, for at least seven amphibian groups (family Palaeobatrachidae, genera 1527 Chelotriton, Pelobates, Bombina (i.e. Bombina (cf.) bombina), Hyla (i.e. Hyla cf. savignyi), 1528 Pelophylax ?, Bufo bufo species group). Besides the amphibians, some Western Siberian 1529 reptiles, such as the glass lizards and snakes from the Middle Miocene, show European 1530 affinities, resembling the Central European faunas (Vasilyan, Böhme & Klembara, 2016). 1531 The amphibian genera Bombina, Hyla, Bufo, Rana, and Pelophylax resemble a comparable 1532 palaeobiogeographic pattern: the molecular genetic data showed the presence of two clearly 1533 separable western and eastern clades (species groups) in each of these genera. In all cases, it 1524 An eastward dispersal from Europe into Western Asia can be observed over a period ranging 1525 from the Middle to Late Miocene, based on the current data available from both European and 1526 Asiatic records, for at least seven amphibian groups (family Palaeobatrachidae, genera 1527 Chelotriton, Pelobates, Bombina (i.e. Bombina (cf.) bombina), Hyla (i.e. Hyla cf. savignyi), 1528 Pelophylax ?, Bufo bufo species group). Besides the amphibians, some Western Siberian 1529 reptiles, such as the glass lizards and snakes from the Middle Miocene, show European 1530 affinities, resembling the Central European faunas (Vasilyan, Böhme & Klembara, 2016). 1531 The amphibian genera Bombina, Hyla, Bufo, Rana, and Pelophylax resemble a comparable 1532 palaeobiogeographic pattern: the molecular genetic data showed the presence of two clearly 1533 separable western and eastern clades (species groups) in each of these genera. In all cases, it PeerJ reviewing PDF \| (2016:08:12606:4:0:NEW 23 Jan 2017) Manuscript to be reviewed 1534 was possible to morphologically attribute the Western Siberian fossil amphibians to the 1535 western clades or species of the clades. It is interesting to note that even though the first fossil 1536 occurrences of these genera have different stratigraphic ages, they are found exclusively in 1537 Europe (see Fig. 9, Table S5). To explain this common pattern, we hypothesise that the western 1538 and eastern clades had already split in the Paleogene, most probably in the western or central 1539 parts of Asia, and subsequently dispersed into Europe. 1540 The Western Siberian fossil Mioproteus, Chelotriton, Bombina, Paleobatidae, Hyla, Bufo bufo 1541 and Rana temporaria represent the most eastern records of those groups found in the Eurasian 1542 fossil record. In comparison to their present-day geography, the Western Eurasian species of 1543 the genera Bombina and Hyla, respectively, show wider distribution ranges during the Middle 1544 to Late Miocene, and Late Miocene to Early Pliocene. The palaeogeographic affinity of the 1545 earliest Messinian pelobatid (locality Selety 1A) is still unclear. Considering the geographic 1546 location of the fossil site, its relation to the Recent genus Pelobates seems most possible. 1547 In Chkhikvadze (1985), two lizards Varanus sp. and Agamidae indet. have been reported from 1548 three Miocene localities of the Zaisan Basin. Although the taxonomic assignment of the remains 1549 could not be verified in this study, we adopt the identifications for biogeographic and 1550 palaeoenvironmental interpretations. These lizards are currently widely distributed in Central 1551 Asia. Varanus, being a thermophilous reptile species, is restricted to the southern part of the 1552 region. Its presence in the early Late Miocene of the Zaisan Basin aids in characterisation of the 1553 climate of the Sarybulak Svita, in the beginning of the late Miocene, i.e. a probable mean 1554 annual temperature of not less than 14.8 °C (Böhme, 2003). 1540 The Western Siberian fossil Mioproteus, Chelotriton, Bombina, Paleobatidae, Hyla, Bufo bufo 1541 and Rana temporaria represent the most eastern records of those groups found in the Eurasian 1542 fossil record. In comparison to their present-day geography, the Western Eurasian species of 1543 the genera Bombina and Hyla, respectively, show wider distribution ranges during the Middle 1544 to Late Miocene, and Late Miocene to Early Pliocene. The palaeogeographic affinity of the 1545 earliest Messinian pelobatid (locality Selety 1A) is still unclear. Manuscript to be reviewed Considering the geographic 1546 location of the fossil site, its relation to the Recent genus Pelobates seems most possible. PeerJ reviewing PDF \| (2016:08:12606:4:0:NEW 23 Jan 2017) Manuscript to be reviewed 1567 The palaeobiogeographic analysis of the recent amphibian faunas of Western Asia (Savage; 1568 Garcia-Porta et al., 2012) hypothesised a progressive aridification of Central Asia linked with 1569 the global cooling trends during the Miocene, forcing amphibians to shift their distribution to 1570 the south. 1571 1572 Palaeoclimatic implications 1573 The Neogene climate evolution of Western Siberia has been previously reconstructed based on 1574 palynofloras, showing a progressive change in environmental conditions, i.e. in the climate and 1575 vegetation, during the Miocene (Arkhipov et al., 2005). Between the Early to Late Miocene, a 1576 warm and humid climate was replaced by a warm temperate climate in the Middle Miocene PeerJ reviewing PDF \| (2016:08:12606:4:0:NEW 23 Jan 2017) 1567 The palaeobiogeographic analysis of the recent amphibian faunas of Western Asia (Savage; 1568 Garcia-Porta et al., 2012) hypothesised a progressive aridification of Central Asia linked with 1569 the global cooling trends during the Miocene, forcing amphibians to shift their distribution to 1570 the south. PeerJ reviewing PDF \| (2016:08:12606:4:0:NEW 23 Jan 2017) Manuscript to be reviewed 1555 In summary, Western Siberia (Central Asia) can be hypothesised as a centre of evolution and 1556 dispersal for several temperate Neogene herpetofaunal taxa, e.g. the genera Salamandrella and 1557 Mioproteus, the green toad Bufotes viridis species group and brown frog Rana. The Neogene 1558 herpetofauna of Western Siberia and the adjacent areas has significant similarities with the 1559 European amphibian and reptile assemblages. The Western Palaearctic herpetofauna gradually 1560 entered the Siberian territory from Europe, between the Middle Miocene to Early Pliocene, 1561 strongly shaping the herpetofauna of Western Siberia and partially retaining the faunal 1562 elements of an Asiatic origin (e.g. Hynobiidae, Proteidae, and Alsophylax). The faunal diversity 1563 of the fossil record collapses significantly after the Early Pliocene. Only a few amphibians and 1564 reptiles, e.g. Salamandrella, Bufotes, Lacerta, and Vipera are present in the Pliocene fossil 1565 record, being able to survive in the increasingly less favourable environments to form the main 1566 part of the present-day Western Siberian herpetofauna. 1555 In summary, Western Siberia (Central Asia) can be hypothesised as a centre of evolution and 1556 dispersal for several temperate Neogene herpetofaunal taxa, e.g. the genera Salamandrella and 1557 Mioproteus, the green toad Bufotes viridis species group and brown frog Rana. The Neogene 1558 herpetofauna of Western Siberia and the adjacent areas has significant similarities with the 1559 European amphibian and reptile assemblages. The Western Palaearctic herpetofauna gradually 1560 entered the Siberian territory from Europe, between the Middle Miocene to Early Pliocene, 1561 strongly shaping the herpetofauna of Western Siberia and partially retaining the faunal 1562 elements of an Asiatic origin (e.g. Hynobiidae, Proteidae, and Alsophylax). The faunal diversity 1563 of the fossil record collapses significantly after the Early Pliocene. Only a few amphibians and 1564 reptiles, e.g. Salamandrella, Bufotes, Lacerta, and Vipera are present in the Pliocene fossil 1565 record, being able to survive in the increasingly less favourable environments to form the main 1566 part of the present-day Western Siberian herpetofauna. 1563 of the fossil record collapses significantly after the Early Pliocene. Only a few amphibians and 1564 reptiles, e.g. Salamandrella, Bufotes, Lacerta, and Vipera are present in the Pliocene fossil 1565 record, being able to survive in the increasingly less favourable environments to form the main 1566 part of the present-day Western Siberian herpetofauna. 1572 Palaeoclimatic implications 1573 The Neogene climate evolution of Western Siberia has been previously reconstructed based on 1574 palynofloras, showing a progressive change in environmental conditions, i.e. in the climate and 1575 vegetation, during the Miocene (Arkhipov et al., 2005). Between the Early to Late Miocene, a 1576 warm and humid climate was replaced by a warm temperate climate in the Middle Miocene 1573 The Neogene climate evolution of Western Siberia has been previously reconstructed based on 1574 palynofloras, showing a progressive change in environmental conditions, i.e. in the climate and 1575 vegetation, during the Miocene (Arkhipov et al., 2005). Between the Early to Late Miocene, a 1576 warm and humid climate was replaced by a warm temperate climate in the Middle Miocene PeerJ reviewing PDF \| (2016:08:12606:4:0:NEW 23 Jan 2017) Manuscript to be reviewed 1577 and a boreal-warm temperate climate in the Late Miocene. Towards the end of the Miocene, a 1578 drastic climatic shift took place resulting in semiarid and arid conditions. The Pliocene climate is 1579 predominated by frequent changes between semiarid forest-steppe/steppe and arid desert 1580 environments, however, from the Late Pliocene the environment changes into subarctic 1581 (Arkhipov et al., 2005; fig. 46, p. 76). 1582 At a lower temporal resolution, the testudinoid fossil records from the Zaisan Basin, the Turgay 1583 Basin, and Western Siberia confirm a general trend towards aridity in the Neogene (Data S4). 1584 Based on the environmental requirement (aquatic or terrestrial) of the testudinoids from the 1585 Zaisan Basin, we infer that the climate changed from humid to dry. We further infer that the 1586 Early and Middle Miocene was mostly humid (dominance of aquatic families), whereas the 1587 presence of exclusively terrestrial forms (tortoises) from the latest Middle Miocene to Late 1588 Miocene indicates dry and open habitats in the Zaisan Basin. Unfortunately, it is impossible to 1589 make any quantification of the palaeoprecipitation values based on these limited taxa and well- 1590 documented herpetofaunal assemblages are necessary from these deposits for further 1591 environmental reconstructions. 1592 To establish a better palaeoclimatic understanding, we estimated palaeoprecipitation values for 1593 12 data points (Table S4) These localities provided six and more amphibian and reptile taxa 1577 and a boreal-warm temperate climate in the Late Miocene. Towards the end of the Miocene, a 1578 drastic climatic shift took place resulting in semiarid and arid conditions. The Pliocene climate is 1579 predominated by frequent changes between semiarid forest-steppe/steppe and arid desert 1580 environments, however, from the Late Pliocene the environment changes into subarctic 1581 (Arkhipov et al., 2005; fig. 46, p. 76). 1577 and a boreal-warm temperate climate in the Late Miocene. Towards the end of the Miocene, a 1578 drastic climatic shift took place resulting in semiarid and arid conditions. The Pliocene climate is 1579 predominated by frequent changes between semiarid forest-steppe/steppe and arid desert 1580 environments, however, from the Late Pliocene the environment changes into subarctic 1581 (Arkhipov et al., 2005; fig. 46, p. 76). 1582 At a lower temporal resolution, the testudinoid fossil records from the Zaisan Basin, the Turgay 1583 Basin, and Western Siberia confirm a general trend towards aridity in the Neogene (Data S4). PeerJ reviewing PDF \| (2016:08:12606:4:0:NEW 23 Jan 2017) Manuscript to be reviewed Apart from the fluctuating humidity factor, in 1601 general, the MAP was significantly above the present day values (reaching 550% of the present- 1602 day values) (Fig. 10). Only two localities are characterised by drier climates, the late Serravallian 1603 (ca. 12.1 Ma) and the late Messinian (5.9 Ma), exhibiting either present-day or below present- 1604 day levels. Manuscript to be reviewed 1584 Based on the environmental requirement (aquatic or terrestrial) of the testudinoids from the 1585 Zaisan Basin, we infer that the climate changed from humid to dry. We further infer that the 1586 Early and Middle Miocene was mostly humid (dominance of aquatic families), whereas the 1587 presence of exclusively terrestrial forms (tortoises) from the latest Middle Miocene to Late 1588 Miocene indicates dry and open habitats in the Zaisan Basin. Unfortunately, it is impossible to 1589 make any quantification of the palaeoprecipitation values based on these limited taxa and well- 1590 documented herpetofaunal assemblages are necessary from these deposits for further 1591 environmental reconstructions. 1592 To establish a better palaeoclimatic understanding, we estimated palaeoprecipitation values for 1593 12 data points (Table S4). These localities provided six and more amphibian and reptile taxa, 1594 applicable for the bioclimatic analysis (Böhme et al., 2006). Even so, our data do not be enable 1595 accurate reconstruction of the climate development over the Middle Miocene to earliest 1596 Pleistocene in Western Siberia. The climate development can, therefore, only be reconstructed 1597 and discussed for several short intervals. Nevertheless, our estimations rather show a dynamic 1598 climate development in the Neogene of Western Siberia, with larger precipitation amplitudes, 1592 To establish a better palaeoclimatic understanding, we estimated palaeoprecipitation values for 1593 12 data points (Table S4). These localities provided six and more amphibian and reptile taxa, 1594 applicable for the bioclimatic analysis (Böhme et al., 2006). Even so, our data do not be enable 1595 accurate reconstruction of the climate development over the Middle Miocene to earliest 1596 Pleistocene in Western Siberia. The climate development can, therefore, only be reconstructed 1597 and discussed for several short intervals. Nevertheless, our estimations rather show a dynamic 1598 climate development in the Neogene of Western Siberia, with larger precipitation amplitudes, PeerJ reviewing PDF \| (2016:08:12606:4:0:NEW 23 Jan 2017) 1606 Reliability of precipitation estimates 1607 The accuracy of precipitation estimates, based on bioclimatic analysis of herpetofauna, 1607 The accuracy of precipitation estimates, based on bioclimatic analysis of herpetofauna, 1608 depends primarily on the taxon counts and the assumption of low (stochastic) taphonomic bias 1609 (Böhme et al., 2006). In Western Siberia, some of the documented localities were rich in aquatic 1610 herpetofauna, e.g. composed by freshwater turtles, giant salamanders, proteids, etc. , but small 1611 terrestrial forms (e.g. lizards and anguids) were absent, indicating a possible non-stochastic 1612 taphonomic bias (i.e. exclusion of elements of certain habitats). These localities will result in a 1613 bias in humidity estimates toward the wet end. Examples of such localities include Kentyubek 1614 and Novaya Stanitsa 1A, where the numeric results well exceed the MAP of 1600 mm, the 1615 upper limit to which the eco-physiologic index – humidity relation is calibrated (see details in 1616 Böhme et al., 2006). In these cases, we restrict our estimates to a limit of 1500 mm. 1618 For the Aquitanian age Ayakoz locality, we estimated a MAP value of 945 mm, representing 1619 more than three times higher rainfall in comparison to the recent times. Using the palynologic 1620 data, Arkhipov et al. (2005) estimated a humid climate with MAP 800 mm for the Abrosimov Manuscript to be reviewed 1599 ranging from 158 mm to over 1500 mm per year (Table S1, Fig. 10), than previously estimated 1600 using palynological data (Arkhipov et al., 2005). Apart from the fluctuating humidity factor, in 1601 general, the MAP was significantly above the present day values (reaching 550% of the present- 1602 day values) (Fig. 10). Only two localities are characterised by drier climates, the late Serravallian 1603 (ca. 12.1 Ma) and the late Messinian (5.9 Ma), exhibiting either present-day or below present- 1604 day levels. 1605 1606 Reliability of precipitation estimates 1607 The accuracy of precipitation estimates, based on bioclimatic analysis of herpetofauna, 1608 depends primarily on the taxon counts and the assumption of low (stochastic) taphonomic bias 1609 (Böhme et al., 2006). In Western Siberia, some of the documented localities were rich in aquatic 1610 herpetofauna, e.g. composed by freshwater turtles, giant salamanders, proteids, etc. , but small 1611 terrestrial forms (e.g. lizards and anguids) were absent, indicating a possible non-stochastic 1612 taphonomic bias (i.e. exclusion of elements of certain habitats). These localities will result in a 1613 bias in humidity estimates toward the wet end. Examples of such localities include Kentyubek 1614 and Novaya Stanitsa 1A, where the numeric results well exceed the MAP of 1600 mm, the 1615 upper limit to which the eco-physiologic index – humidity relation is calibrated (see details in 1616 Böhme et al., 2006). In these cases, we restrict our estimates to a limit of 1500 mm. 1617 Aquitanian 1618 For the Aquitanian age Ayakoz locality, we estimated a MAP value of 945 mm, representing 1619 more than three times higher rainfall in comparison to the recent times. Using the palynologic 1599 ranging from 158 mm to over 1500 mm per year (Table S1, Fig. 10), than previously estimated 1600 using palynological data (Arkhipov et al., 2005). Apart from the fluctuating humidity factor, in 1601 general, the MAP was significantly above the present day values (reaching 550% of the present- 1602 day values) (Fig. 10). Only two localities are characterised by drier climates, the late Serravallian 1603 (ca. 12.1 Ma) and the late Messinian (5.9 Ma), exhibiting either present-day or below present- 1604 day levels. 1599 ranging from 158 mm to over 1500 mm per year (Table S1, Fig. 10), than previously estimated 1600 using palynological data (Arkhipov et al., 2005). Manuscript to be reviewed 1626 Akzhar Svita 1627 Towards the end of the late Early Miocene (Burdigalian), an elevated humidity in Western 1628 Siberia can be suggested based on the presence of the giant salamander in three localities of 1629 the Zaisan Basin (Tri Bogatyrya, Vympel, and Poltinik). As already suggested, their occurrence 1630 indicates a high rainfall for those time periods (MAP > 900 mm), as well as an increased basinal 1631 relief enabling the distribution and reproduction of this group in the lowland settings (Böhme, 1632 Vasilyan & Winklhofer, 2012). This period of the Akzhar Svita also corresponds to the folding 1633 and uplift of the Altai Mountains (Zykin, 2012; p. 394), from which the establishment of the 1634 higher basinal relief was possible. PeerJ reviewing PDF \| (2016:08:12606:4:0:NEW 23 Jan 2017) PeerJ reviewing PDF \| (2016:08:12606:4:0:NEW 23 Jan 2017) PeerJ reviewing PDF \| (2016:08:12606:4:0:NEW 23 Jan 2017) Manuscript to be reviewed 1621 Svita (Aquitanian age) in Western Siberia. Besides this study and based on the data of fossil 1622 macroflora, Bruch & Zhilin (2007) estimated similar values of precipitation (935 to 1232 mm) 1623 for about 30 Aquitanian age localities, distributed from Western to Eastern Kazakhstan. Our 1624 reconstruction, therefore, appears to fit well within the historical precipitation estimates of the 1625 region. 1626 Akzhar Svita 1627 Towards the end of the late Early Miocene (Burdigalian), an elevated humidity in Western 1628 Siberia can be suggested based on the presence of the giant salamander in three localities of 1629 the Zaisan Basin (Tri Bogatyrya, Vympel, and Poltinik). As already suggested, their occurrence 1630 indicates a high rainfall for those time periods (MAP > 900 mm), as well as an increased basinal 1631 relief enabling the distribution and reproduction of this group in the lowland settings (Böhme, 1632 Vasilyan & Winklhofer, 2012). This period of the Akzhar Svita also corresponds to the folding 1633 and uplift of the Altai Mountains (Zykin, 2012; p. 394), from which the establishment of the 1634 higher basinal relief was possible. 1621 Svita (Aquitanian age) in Western Siberia. Besides this study and based on the data of fossil 1622 macroflora, Bruch & Zhilin (2007) estimated similar values of precipitation (935 to 1232 mm) 1623 for about 30 Aquitanian age localities, distributed from Western to Eastern Kazakhstan. Our 1624 reconstruction, therefore, appears to fit well within the historical precipitation estimates of the 1625 region. 1626 Akzhar Svita 1627 Towards the end of the late Early Miocene (Burdigalian), an elevated humidity in Western 1628 Siberia can be suggested based on the presence of the giant salamander in three localities of 1629 the Zaisan Basin (Tri Bogatyrya, Vympel, and Poltinik). As already suggested, their occurrence 1630 indicates a high rainfall for those time periods (MAP > 900 mm), as well as an increased basinal 1631 relief enabling the distribution and reproduction of this group in the lowland settings (Böhme, 1632 Vasilyan & Winklhofer, 2012). This period of the Akzhar Svita also corresponds to the folding 1633 and uplift of the Altai Mountains (Zykin, 2012; p. 394), from which the establishment of the 1634 higher basinal relief was possible. 1648 Rytov Svita 1649 The Cherlak locality (5.9 Ma, Rytov Suite) is characterised by a rather dry climate (MAP 255 1650 mm), with a similar humidity level to that of the present-day (Fig. 10). Our data for a warm and 1651 dry climate are confirmed by the presence of: (1) gekkonid Alsophylax; (2) mollusc fauna 1652 containing thermophilous species; (3) the small mammal fauna, represented mainly by pikas, 1653 hamsters and jerboas, characteristic for open and dry habitats (Zykin, 2012); and (4) ostriches 1654 (Struthiolithus sp.) and camels (Paracamelus sp.) in this svita (Shpanskiy, 2008). Arkhipov et al. 1655 (2005) summarised the available palynological and vegetation data of the svita and reported 1656 the presence of a poor (due to the oxidation) spectra containing xerophyte plants (Asteraceae, 1657 Chenopodiacea), characterising desert and steppe environments. Interestingly, his results 1658 proposed a northward shift of dry steppe and desert environments by 4° (to the latitude of 1659 56°), which concurs with our data, as is indicated by the presence of the steppe-dwelling 1660 gekkonid Alsophylax sp. (see the section ‘Gekkonidae’). 1649 The Cherlak locality (5.9 Ma, Rytov Suite) is characterised by a rather dry climate (MAP 255 1650 mm), with a similar humidity level to that of the present-day (Fig. 10). Our data for a warm and 1651 dry climate are confirmed by the presence of: (1) gekkonid Alsophylax; (2) mollusc fauna 1652 containing thermophilous species; (3) the small mammal fauna, represented mainly by pikas, 1653 hamsters and jerboas, characteristic for open and dry habitats (Zykin, 2012); and (4) ostriches 1654 (Struthiolithus sp.) and camels (Paracamelus sp.) in this svita (Shpanskiy, 2008). Arkhipov et al. 1655 (2005) summarised the available palynological and vegetation data of the svita and reported 1656 the presence of a poor (due to the oxidation) spectra containing xerophyte plants (Asteraceae, 1657 Chenopodiacea), characterising desert and steppe environments. Interestingly, his results 1658 proposed a northward shift of dry steppe and desert environments by 4° (to the latitude of 1659 56°), which concurs with our data, as is indicated by the presence of the steppe-dwelling 1660 gekkonid Alsophylax sp. (see the section ‘Gekkonidae’). 1635 Late Serravallian 1636 In contrast to the already known climate development suggested by Arkhipov et al. (2005), our 1637 data suggest that there were strong humidity fluctuations during the late Middle Miocene (late 1638 Serravallian), with MAP values ranging between 282, 884 and 1108 mm (Fig. 10). The only 1639 botanical data of this time (Bescheul macroflora) point to a warm-temperate and humid (MAP 1640 ~700 mm) climate (Arkhipov et al., 2005), which best compares to our Malyi Kalkaman 2 results 1641 (MAP 884 mm). PeerJ reviewing PDF \| (2016:08:12606:4:0:NEW 23 Jan 2017) Manuscript to be reviewed 1643 Although the herpetofaunal assemblage for the early Messinian locality Novaya Stanitsa 1A is 1644 incomplete, a very high MAP value of at least 1500 mm can be estimated. The value indicates a 1645 significantly higher humidity than of Tortonian-Messinian boundary and late Messinian (see 1646 below). Our data are contrary to the palynologic results, which gave lower estimates (400-450 1647 mm; Arkhipov et al. 2005). 1643 Although the herpetofaunal assemblage for the early Messinian locality Novaya Stanitsa 1A is 1644 incomplete, a very high MAP value of at least 1500 mm can be estimated. The value indicates a 1645 significantly higher humidity than of Tortonian-Messinian boundary and late Messinian (see 1646 below). Our data are contrary to the palynologic results, which gave lower estimates (400-450 1647 mm; Arkhipov et al. 2005). 1661 Miocene-Pliocene transition (Detskaya Zhelznaja Daroga) 1662 Even though the precise taxonomic identification of the Western Siberian and Zaisan 1663 cryptobranchids, is unclear at the generic or species level, their occurrence indicates a high 1664 rainfall > 900 mm MAP (Böhme, Vasilyan & Winklhofer, 2012) during the Burdigalian age in the PeerJ reviewing PDF \| (2016:08:12606:4:0:NEW 23 Jan 2017) PeerJ reviewing PDF \| (2016:08:12606:4:0:NEW 23 Jan 2017) PeerJ reviewing PDF \| (2016:08:12606:4:0:NEW 23 Jan 2017) Manuscript to be reviewed 1665 Zaisan Basin and the Miocene-Pliocene transition in Western Siberia. Besides the presence of 1666 Cryptobranchidae indet. from the locality Detskaya Zheleznaja Doroga, the co-occurrence of the 1667 aquatic chelonids Chelydropsis kuznetsovi and probable Sakya sp. (Gaiduchenko, 1984; 1668 Gaiduchenko & Chkhikvadze, 1985) confirms the presence of a high degree of precipitation at 1669 the Miocene-Pliocene boundary in Western Siberia. 1665 Zaisan Basin and the Miocene-Pliocene transition in Western Siberia. Besides the presence of 1666 Cryptobranchidae indet. from the locality Detskaya Zheleznaja Doroga, the co-occurrence of the 1667 aquatic chelonids Chelydropsis kuznetsovi and probable Sakya sp. (Gaiduchenko, 1984; 1668 Gaiduchenko & Chkhikvadze, 1985) confirms the presence of a high degree of precipitation at 1669 the Miocene-Pliocene boundary in Western Siberia. 1670 Earliest Pliocene (Olkhovka 1A-1C) ( ) 1671 Our earliest Pliocene humidity data are estimated based on the fauna from the localities 1672 Olkhovka 1A, 1B and 1C, for which no correlation data is available for regional svitas (see 1673 ‘Geology and Stratigraphy section’). Nevertheless, the results still indicate significant 1674 precipitation (MAP 575 mm), well above the present-day values for this region. These findings 1675 correspond well with the similar aged Speranovskaya palynoflora (Volkova, 1984), which 1676 indicates the presence of warm forests and forest-steppes with MAP estimates between 500- 1677 550 mm (Arkhipov et al., 2005). 1678 1679 Conclusions 1680 In summary, over 50 salamander, frog, lizard, snake and turtle taxa have been assigned to 1681 specimens from more than 40 Western Siberian localities that range in age from the Middle 1682 Miocene to the Pleistocene (Table S1). The late Middle Miocene localities have the most diverse 1683 faunas including all the main groups of the herpetofauna. According to our analysis, the fossil 1684 fauna contains taxa showing an Asian (Eastern Palaearctic) origin, such as Hynobiidae, 1685 Proteidae, Bufotes viridis species group and Rana, Varanus, and Agamidae. The main part of the 1686 herpetofaunal assemblage, including Palaeobatrachidae, Paleobatidae, the genera Chelotriton, PeerJ reviewing PDF \| (2016:08:12606:4:0:NEW 23 Jan 2017) Manuscript to be reviewed 1687 Bombina (i.e. Bombina (cf.) bombina), Hyla (i.e. Hyla (cf.) savignyi), Pelophylax ?, Bufo bufo, 1688 Ophisaurus sp. (Vasilyan, Böhme & Klembara, 2016), has European (Western Palaearctic) 1689 affinities. The Western Siberian records of Mioproteus, Chelotriton, Bombina, Paleobatidae, 1690 Hyla, Bufo bufo, and Rana temporaria represent the most eastern occurrences of these groups 1691 in Eurasia. The earliest Miocene dispersal of the green toad, Bufotes viridis species group into 1692 Europe from Asia via Anatolia, can be inferred. We suggest the same distribution pattern for 1693 brown frogs, Rana, too. In this scope, it will be important to perform future detailed studies on 1694 the Neogene record of the amphibian and reptile faunas in Anatolia and analyse them in a 1695 palaeobiogeographic context. 1687 Bombina (i.e. Bombina (cf.) bombina), Hyla (i.e. Hyla (cf.) savignyi), Pelophylax ?, Bufo bufo, 1688 Ophisaurus sp. (Vasilyan, Böhme & Klembara, 2016), has European (Western Palaearctic) 1689 affinities. The Western Siberian records of Mioproteus, Chelotriton, Bombina, Paleobatidae, 1690 Hyla, Bufo bufo, and Rana temporaria represent the most eastern occurrences of these groups 1691 in Eurasia. The earliest Miocene dispersal of the green toad, Bufotes viridis species group into 1692 Europe from Asia via Anatolia, can be inferred. We suggest the same distribution pattern for 1693 brown frogs, Rana, too. In this scope, it will be important to perform future detailed studies on 1694 the Neogene record of the amphibian and reptile faunas in Anatolia and analyse them in a 1695 palaeobiogeographic context. 1696 According to our study, the precipitation development in Western Siberia shows high- 1697 amplitude changes during the studied intervals. Aside from the certain time periods, i.e. late 1698 Seravalian and late Messinian, the palaeorainfall in Western Siberia was estimated to be 1699 significantly higher than the present-day values. The best results on precipitation estimates that 1700 we were able to reconstruct, with reliable age constrain, were for the period from 6.6 to ~4.5 1701 Ma. These results indicate a humid climate during the early Messinian; a dry climate during the 1702 late Messinian; a very humid climate during the Miocene-Pliocene transition and a humid 1703 climate during the earliest Pliocene (Data S4, Fig. 10). The decreasing tendency of the 1704 herpetofaunal diversity towards the end of the Neogene and Quaternary could be attributed to 1705 the progressive global cooling and forced ice-sheet development in the Northern Hemisphere. Manuscript to be reviewed 1706 1696 According to our study, the precipitation development in Western Siberia shows high- 1697 amplitude changes during the studied intervals. Aside from the certain time periods, i.e. late 1698 Seravalian and late Messinian, the palaeorainfall in Western Siberia was estimated to be 1699 significantly higher than the present-day values. The best results on precipitation estimates that 1700 we were able to reconstruct, with reliable age constrain, were for the period from 6.6 to ~4.5 1701 Ma. These results indicate a humid climate during the early Messinian; a dry climate during the 1702 late Messinian; a very humid climate during the Miocene-Pliocene transition and a humid 1703 climate during the earliest Pliocene (Data S4, Fig. 10). The decreasing tendency of the 1704 herpetofaunal diversity towards the end of the Neogene and Quaternary could be attributed to 1705 the progressive global cooling and forced ice-sheet development in the Northern Hemisphere. 1706 PeerJ reviewing PDF \| (2016:08:12606:4:0:NEW 23 Jan 2017) 1718 1719 REFERENCES 1720 Amphibiaweb. 2016. <http://amphibiaweb.org>: Information on amphibian biology and 1721 conservation (accessed 1 July 2016). 1722 Ananjeva NB, Orlov NL, Khalikov RG, Darevsky I, Ryaboc SA, Barabanov AV. 2006. The Reptiles 1723 of Northern Eurasia: Taxonomic Diversity, Distribution, Conservation Status. Sofia: Pensoft. 1724 Anderson J. 1871. Description of a new genus of newts from western Yunan. Proceedings of the 1725 Zoological Society of London 1871:423–425. 1726 Arkhipov SA, Volkova VS, Zolnikov ID, Zykina VS, Krukover AA, Kul'kova LA. 2005. West 1727 Siberia. Geological Society of America Special Papers 382:67–88. 1728 Arntzen JW, Recuero E, Canestrelli D, Martínez-Solano I. 2013. How complex is the Bufo bufo 1729 species group? Molecular Phylogenetics and Evolution 69 (3):1203–1208. 1730 Audouin JV. 1827. Explication sommaire des planches de Reptiles (supplément) … offrant un 1731 exposé des caracteres naturelles des genres, avec la distinction des especes: Description de 1732 l'Égypte, ou Recueil des Observations et des Recherches qui ont été faites en Égypte pendant 1733 l'Expedition de l'Armée Française, publie par les Ordres de sa Majesté l'Empereur Napoléon le 1734 Grand. Histoire Naturelle 1 (4):161–184. 1735 Averianov AO, Tjutkova L. 1995. Ranodon cf. sibiricus (Amphibia, Caudata) from the Upper 1736 Pliocene of Southern Kazakhstan: The first fossil record of the family Hynobiidae. 1737 Paläontologische Zeitschrift 69 (1):257–264. Manuscript to be reviewed 1708 We sincerely thank B. Sanchiz (Madrid), Z. Roček (Prague), J. Prieto (Munich), M. Rabi 1709 (Tübingen) M. Delfino (Turin), and V. Ratnikov (Voronezh) for their constructive discussions and 1710 comments. We are grateful to V. Chkhikvadze (Tbilisi) for providing material from the localities: 1711 Pavlodar 1A, Ayakoz, Petropavlovsk 1/2, Malyi Kalkaman 1; to Dr. L. Maul (Weimar) for 1712 providing details of the ages of the Quaternary localities where palaeobatrachid frogs occur; to 1713 A. Fatz (Tübingen) for making figure and table images; to I. Stepanyan (Yerevan) for literature 1714 help; and A. Ilg (Düsseldorf) for providing support with the database ʻfosFARbaseʼ. Finally, we 1715 would like to express our gratitude to G. Piñeiro (Montevideo) for editorial comments and 1716 guiding, and both reviewers M. Venczel (Oradea) and Sh. Meiri (Tel Aviv) for their criticism and 1717 constructive reviews. 1708 We sincerely thank B. Sanchiz (Madrid), Z. Roček (Prague), J. Prieto (Munich), M. Rabi 1709 (Tübingen) M. Delfino (Turin), and V. Ratnikov (Voronezh) for their constructive discussions and 1710 comments. We are grateful to V. Chkhikvadze (Tbilisi) for providing material from the localities: 1711 Pavlodar 1A, Ayakoz, Petropavlovsk 1/2, Malyi Kalkaman 1; to Dr. L. Maul (Weimar) for 1712 providing details of the ages of the Quaternary localities where palaeobatrachid frogs occur; to 1713 A. Fatz (Tübingen) for making figure and table images; to I. Stepanyan (Yerevan) for literature 1714 help; and A. Ilg (Düsseldorf) for providing support with the database ʻfosFARbaseʼ. Finally, we 1715 would like to express our gratitude to G. Piñeiro (Montevideo) for editorial comments and 1716 guiding, and both reviewers M. Venczel (Oradea) and Sh. Meiri (Tel Aviv) for their criticism and 1717 constructive reviews. 1718 Manuscript to be reviewed 1738 Bailon S. 1999. Différenciation ostéologique des Anoures (Amphibia, Anura) de France. Antibes: 1739 Centre de Recherches Archéologiques du CNRS. 1740 Bailon S, Hossini S. 1990. Les plus anciens Bufonidae (Amphibia, Anura) d'Europe: les espèces du 1741 Miocène français. Annales de Paléontologie 76:121–132. 1742 Barry JC, Morgan ME, Flynn LJ, Pilbeam D, Behrensmeyer AK, Raza SM, Khan IA, Badgley C, 1743 Hicks J, Kelley J. 2002. Faunal and environmental change in the late Miocene Siwaliks of 1744 northern Pakistan. Paleobiology 28 (sp3):1–71. 1745 Batsch, A. J. G. K. 1796. Umriß der gesammten Naturgeschichte: ein Auszug aus den frühern 1746 Handbüchern des Verfassers für seine Vorfesungen. Jena, Leipzig: Christian Ernst Gabler. 1747 Bendukidze OG, Bruijn H de, Van den Hoek Ostende, Lars W. 2009. A revision of Late Oligocene 1748 associations of small mammals from the Aral Formation (Kazakhstan) in the National Museum 1749 of Georgia, Tbilissi. Palaeodiversity 2:343–377. 1750 Bendukidze OG, Chkhikvadze VM. 1976. Preliminary results of study on fossil amphibians, 1751 reptiles and birds from Turgay and Ustyurt. Bulletin of the Moscow Society of Naturalists, 1752 Geological Series 51 (5):156. 1753 Blain H-A. 2009. Contribution de la paléoherpétofaune (Amphibia & Squamata) à la connaissance 1754 de l'évolution du climat et du paysage du Pliocène supèrieur au Pléistocène moyen d'Espagne. 1755 Treballs del Museu de Geologia de Barcelona 16:39–170. 1756 Blain H-A, Bailon S, Agustí J. 2007. Anurans and squamate reptiles from the latest early 1757 Pleistocene of Almenara-Casablanca-3 (Castellón, East of Spain). Systematic, climatic and 1758 environmental considerations. Geodiversitas 29 (2):269–295. 1756 Blain H-A, Bailon S, Agustí J. 2007. Anurans and squamate reptiles from the latest early 1757 Pleistocene of Almenara-Casablanca-3 (Castellón, East of Spain). Systematic, climatic and 1758 environmental considerations. Geodiversitas 29 (2):269–295. 1759 Blain H-A, Gibert L, Ferràndez-Cañadell C. 2010. First report of a green toad (Bufo viridis sensu 1760 lato) in the Early Pleistocene of Spain: Palaeobiogeographical and palaeoecological implications. 1761 Comptes Rendus Palevol 9 (8):487–497. 1759 Blain H-A, Gibert L, Ferràndez-Cañadell C. 2010. First report of a green toad (Bufo viridis sensu 1760 lato) in the Early Pleistocene of Spain: Palaeobiogeographical and palaeoecological implications. 1761 Comptes Rendus Palevol 9 (8):487–497. 1762 Blain H-A, Villa P. 2006. Amphibians and squamate reptiles from the early Upper Pleistocene of 1763 Bois Roche Cave (Charente, southwestern France). Acta zoologica cracoviensia 49A (1-2):1–32. 1762 Blain H-A, Villa P. 2006. 1719 REFERENCES 1720 Amphibiaweb. 2016. <http://amphibiaweb.org>: Information on amphibian biology and 1721 conservation (accessed 1 July 2016). 1722 Ananjeva NB, Orlov NL, Khalikov RG, Darevsky I, Ryaboc SA, Barabanov AV. 2006. The Reptiles 1723 of Northern Eurasia: Taxonomic Diversity, Distribution, Conservation Status. Sofia: Pensoft. 1724 Anderson J. 1871. Description of a new genus of newts from western Yunan. Proceedings of the 1722 Ananjeva NB, Orlov NL, Khalikov RG, Darevsky I, Ryaboc SA, Barabanov AV. 2006. The Reptiles 1723 of Northern Eurasia: Taxonomic Diversity, Distribution, Conservation Status. Sofia: Pensoft. 1723 of Northern Eurasia: Taxonomic Diversity, Distribution, Conservation Status. Sofia: Pensoft. 1724 Anderson J. 1871. Description of a new genus of newts from western Yunan. Proceedings of the 1725 Zoological Society of London 1871:423–425. 1726 Arkhipov SA, Volkova VS, Zolnikov ID, Zykina VS, Krukover AA, Kul'kova LA. 2005. West 1727 Siberia. Geological Society of America Special Papers 382:67–88. 1726 Arkhipov SA, Volkova VS, Zolnikov ID, Zykina VS, Krukover AA, Kul'kova LA. 2005. West 1727 Siberia. Geological Society of America Special Papers 382:67–88. 1728 Arntzen JW, Recuero E, Canestrelli D, Martínez-Solano I. 2013. How complex is the Bufo bufo 1729 species group? Molecular Phylogenetics and Evolution 69 (3):1203–1208. 1728 Arntzen JW, Recuero E, Canestrelli D, Martínez-Solano I. 2013. How complex is the Bufo bufo 1729 species group? Molecular Phylogenetics and Evolution 69 (3):1203–1208. 1730 Audouin JV. 1827. Explication sommaire des planches de Reptiles (supplément) … offrant un 1731 exposé des caracteres naturelles des genres, avec la distinction des especes: Description de 1732 l'Égypte, ou Recueil des Observations et des Recherches qui ont été faites en Égypte pendant 1733 l'Expedition de l'Armée Française, publie par les Ordres de sa Majesté l'Empereur Napoléon le 1734 Grand. Histoire Naturelle 1 (4):161–184. 1735 Averianov AO, Tjutkova L. 1995. Ranodon cf. sibiricus (Amphibia, Caudata) from the Upper 1736 Pliocene of Southern Kazakhstan: The first fossil record of the family Hynobiidae. 1737 Paläontologische Zeitschrift 69 (1):257–264. PeerJ reviewing PDF \| (2016:08:12606:4:0:NEW 23 Jan 2017) Manuscript to be reviewed 1780 Böhme M, Ilg A. 2003. fosFARbase. Available at www.wahre-staerke.com (accessed 1 December 1781 2015). 1782 Böhme M, Ilg A, Ossig A, Küchenhoff H. 2006. New method to estimate paleoprecipitation using 1783 fossil amphibians and reptiles and the middle and late Miocene precipitation gradients in 1784 Europe. Geology 34 (6):425–428. 1785 Böhme M, Vasilyan D. 2014. Ectothermic vertebrates from the late Middle Miocene of Gratkorn 1786 (Austria, Styria). Palaeobiodiversity and Palaeoenvironments 94 (1):21–40. 1787 Böhme M, Vasilyan D, Winklhofer M. 2012. Habitat tracking, range dynamics and palaeoclimatic 1788 significance of Eurasian giant salamanders (Cryptobranchidae) — indications for elevated 1789 Central Asian humidity during Cenozoic global warm periods. Palaeogeography, 1790 Palaeoclimatology, Palaeoecology 342–343:64–72. 1791 Bonaparte CL. 1850. Conspectus systematum. Mastozoölogiae. Ornithologiae. Herpetologiae et 1792 Amphibiologiae. Ichthyologiae. Lugduni Batavorum: E. J. Brill. 1793 Borisov BA. 1963. Stratigraphy of upper Cretaceus and Paleogene-Neogene of Zaisan basin. 1794 Transactions of Pansoviet scientific-research geological institute New Series, 94:11–75. 1795 Borkin LJ. 1999. Distribution of Amphibians in North Africa, Europe, Western Asia, and the Former 1796 Soviet Union. In: Duellman WE, ed. Patterns of Distribution of Amphibians. Baltimore and 1797 London: The Johns Hopkins University Press, 329–420. 1798 Bruch AA, Zhilin SG. 2007. Early miocene climate of Central Eurasia — Evidence from Aquitanian 1799 floras of Kazakhstan. Palaeogeography, Palaeoclimatology, Palaeoecology 248 (1-2):32–48. 1800 Chkhikvadze VM. 1980. Systematic position of the Neogene freshwater turtle of Moldavia, Ukraine 1801 and some Central European Countries. Bulletin of the Academy of Sciences of Georgian SSR 99 1802 (3):721–724. 1803 Chkhikvadze VM. 1983. The fossil turtles of Caucasus and Northern Black Sea region. Tbilisi: 1804 Metsniereba. 1805 Chkhikvadze VM. 1984. Survey of the fossil urodelan and anuran amphibians from the USSR. 1806 Izvestia Akademii Nauk Gruzinska SSR, Seria Biologitcheskaya 10 (1):5–13. 1807 Chkhikvadze VM. 1985. Preliminary results of studies on tertiary amphibians and squamate 1808 reptiles of the Zaisan Basin. In: Darevsky I, ed.: Nauka, 234–235. 1809 Chkhikvadze VM. 1989. Neogene turtles of USSR. Tbilisi: Metsniereba. 1810 Chkhikvadze VM. 1998. Eocene - Oligocene herpetofauna of USSR. In: Krasheninnikov VA, 1811 Akhmetiev MA, eds. Late Eocene - Early Oligocene geological and biotic events. On the territory of 1812 the former Soviet Union, II. Moscow: Geos, 61–66. 1810 Chkhikvadze VM. 1998. Eocene - Oligocene herpetofauna of USSR. In: Krasheninnikov VA, 1811 Akhmetiev MA, eds. Late Eocene - Early Oligocene geological and biotic events. On the territory of 1812 the former Soviet Union, II. Manuscript to be reviewed Amphibians and squamate reptiles from the early Upper Pleistocene of 1763 Bois Roche Cave (Charente, southwestern France). Acta zoologica cracoviensia 49A (1-2):1–32. 1764 Böhme G. 1977. Zur Bestimmung quartärer Anuren Europas an Hand von Skelettelementen. 1765 Wissenschaftliche Zeitschrift der Humboldt-Universität zu Berlin, Mathematisch- 1766 Naturwissenschaftliche Reihe 26 (3):283–299. 1764 Böhme G. 1977. Zur Bestimmung quartärer Anuren Europas an Hand von Skelettelementen. 1765 Wissenschaftliche Zeitschrift der Humboldt-Universität zu Berlin, Mathematisch- 1766 Naturwissenschaftliche Reihe 26 (3):283–299. 1767 Böhme M. 1998. Archeotriton basalticus (v. Mayer, 1859) (Urodela, Salamandridae) aus dem 1768 Unteroligozän von Hammerunterwiesenthal (Freistaat Sachsen). Abhandlungen des staatlichen 1769 Museums für Mineralogie und Geologie zu Dresden 43/44:265–280. 1767 Böhme M. 1998. Archeotriton basalticus (v. Mayer, 1859) (Urodela, Salamandridae) aus dem 1768 Unteroligozän von Hammerunterwiesenthal (Freistaat Sachsen). Abhandlungen des staatlichen 1769 Museums für Mineralogie und Geologie zu Dresden 43/44:265–280. 1770 Böhme M. 2001. The oldest representative of a brown frog (Ranidae) from the Early Miocene of 1771 Germany. Acta Palaeontologica Polonica 46 (1):119–124. 1770 Böhme M. 2001. The oldest representative of a brown frog (Ranidae) from the Early Miocene of 1771 Germany. Acta Palaeontologica Polonica 46 (1):119–124. 1772 Böhme M. 2003. The Miocene Climatic Optimum: evidence from ectothermic vertebrates of Central 1773 Europe. Palaeogeography, Palaeoclimatology, Palaeoecology 195 (3–4):389–401. 1772 Böhme M. 2003. The Miocene Climatic Optimum: evidence from ectothermic vertebrates of Central 1773 Europe. Palaeogeography, Palaeoclimatology, Palaeoecology 195 (3–4):389–401. 1774 Böhme M. 2008. Ectothermic vertebrates (Teleostei, Allocaudata, Urodela, Anura, Testudines, 1775 Choristodera, Crocodylia, Squamata) from the Upper Oligocene of Oberleichtersbach (Northern 1776 Bavaria, Germany). Courier Forschungsinstitut Senckenberg 260:161–183. 1774 Böhme M. 2008. Ectothermic vertebrates (Teleostei, Allocaudata, Urodela, Anura, Testudines, 1775 Choristodera, Crocodylia, Squamata) from the Upper Oligocene of Oberleichtersbach (Northern 1776 Bavaria, Germany). Courier Forschungsinstitut Senckenberg 260:161–183. 1777 Böhme M. 2010. Ectothermic vertebrates (Actinopterygii, Allocaudata, Urodela, Anura, Crocodylia, 1778 Squamata) from the Miocene of Sandelzhausen (Germany, Bavaria) and their implications for 1779 environment reconstruction and palaeoclimate. Paläontologische Zeitschrift 84 (1):3–41. 1777 Böhme M. 2010. Ectothermic vertebrates (Actinopterygii, Allocaudata, Urodela, Anura, Crocodylia, 1778 Squamata) from the Miocene of Sandelzhausen (Germany, Bavaria) and their implications for 1779 environment reconstruction and palaeoclimate. Paläontologische Zeitschrift 84 (1):3–41. PeerJ reviewing PDF \| (2016:08:12606:4:0:NEW 23 Jan 2017) Manuscript to be reviewed 1822 Cope ED. 1865. Sketch of the primary groups of Batrachia Salientia. Natural history review 5:97– 1823 120. 1824 Cope ED. 1868. On the origin of genera. Proceedings of the Academy of Natural Sciences of 1825 Philadelphia 20:242–300. 1826 Cuvier G. 1817. Le Règne Animal distribué d’après son organisation pour servir de base à l’histoire 1827 naturelle des animaux et d’introduction à l’anatomie comparée. Les reptiles, les poissons, les 1828 mollusques et les annélides 2:1–532. 1829 Danilov IG, Cherepanov GO, Vitek NS. 2013. Chelonological studies of l. I. Khosatzky with his 1830 annotated bibliography on turtles. Proceedings of the Zoological Institute RAS 317 (4):382–425. 1831 Daxner-Höck G, Badamgarav D, Erbaeva MA, Göhlich UB. 2013. Miocene Mammal 1832 Biostratigraphiy of Central Mongolia (Valley of Lakes): New Results. In: Wang X, J FL, Fortelius 1833 M, eds. Fossil mammals of Asia: Neogene biostratigraphy and chronology. New York: Columbia 1834 University Press, 477–507. 1835 Daza JD, Alifanov VR, Bauer AM. 2012. A redescription and phylogenetic reinterpretation of the 1836 fossil lizard Hoburogekko suchanovi Alifanov, 1989 (Squamata, Gekkota), from the Early 1837 Cretaceous of Mongolia. Journal of Vertebrate Paleontology 32 (6):1303–1312. 1838 Daza JD, Aurich J, Bauer AM. 2011. Anatomy of an enigma: an osteological investigation of the 1839 Namibian festive gecko (Narudasia festiva: Gekkonidae: Gekkota). Acta Zoologica 93 (4):465– 1840 486. 1841 Daza JD, Bauer AM. 2010. The Circumorbital Bones of the Gekkota (Reptilia: Squamata). The 1842 Anatomical Record: Advances in Integrative Anatomy and Evolutionary Biology 293 (3):402–413 1843 Delfino M, Bailon S, Pitruzzella G. 2011. The Late Pliocene amphibians and reptiles from “Capo 1844 Mannu D1 Local Fauna” (Mandriola, Sardinia, Italy). Geodiversitas 33 (2):357–382. 1845 Dybowski B. 1870. Beitrag zur Kenntniss der Wassermolche Sibiriens. Verhandlungen des 1846 Zoologisch-Botanischen Vereins in Wien 20:237–242. 1847 Edwards JL. 1976. Spinal nerves and their bearing on salamander phylogeny. Journal of 1848 Morphology 148 (3):305–327. 1849 Estes R. 1969. Die Fauna der miozänen Spaltenfüllung von Neudorf an der March (ČSSR) Reptilia 1850 (Lacertilia). Österreichische Akademie der Wissenschaften, Mathematisch-Naturwissenschaftliche 1851 Klasse Abteilung I, Sitzungsberichte 178:77–82. 1852 Estes R, Darevsky I. 1977. Fossil amphibians from the Miocene of North Caucasus, USSR. Journal of 1853 Palaeontological Society of India 20:164–169. 1854 Estes R, Queiroz K de, Gauthier J. 1988. Phylogenetic relationships within Squamata. In: Estes R, 1855 Pregill G, eds. Phylogenetic relationships of the lizard families. Stanford, California: Stanford 1856 University Press, 119–281. 1857 Evans SE. 2008. Manuscript to be reviewed Moscow: Geos, 61–66. 1813 Chkhikvadze VM. 2003. Perspectives of paleontological studies of Late Neogene vertebrate 1814 localities in the valley of River Iori, Eastern Georgia and Western Azerbaijan. In: Abdaladze O, 1815 ed. Transactions of scientific works: Conservation arid and semiarid ecosystems in Transcaucasia. 1816 Tbilisi: NAGRES, 63–67. 1817 Claessens, Leon P. A. M. 1997. On the herpteofauna of some Neogene Eastern Mediterranean 1818 localities and the occurence of Palaeobatrachus and Bufo (Amphibia, Anura) in the Lower 1819 Miocene of Turkey. Journal of Vertebrate Paleontology 17 (Suppl. 3):39. 1817 Claessens, Leon P. A. M. 1997. On the herpteofauna of some Neogene Eastern Mediterranean 1818 localities and the occurence of Palaeobatrachus and Bufo (Amphibia, Anura) in the Lower 1819 Miocene of Turkey. Journal of Vertebrate Paleontology 17 (Suppl. 3):39. 1820 Cope ED. 1859. On the primary division of the Salamandirdae, with description of two new species. 1821 Proceedings of the Academy of Natural Sciences of Philadelphia 11:122–128. 1820 Cope ED. 1859. On the primary division of the Salamandirdae, with description of two new species. 1821 Proceedings of the Academy of Natural Sciences of Philadelphia 11:122–128. PeerJ reviewing PDF \| (2016:08:12606:4:0:NEW 23 Jan 2017) Manuscript to be reviewed 1865 Fitzinger L. 1826. Neue Classification der Reptilien nach ihrer natürlichen Verwandtschaft. Wien: J. 1866 G. Heubner. 1865 Fitzinger L. 1826. Neue Classification der Reptilien nach ihrer natürlichen Verwandtschaft. Wien: J. 1866 G. Heubner. Fitzinger L. 1843. Systema Reptilium. Fasciculus Primus. Wien: Braumüller et Seidel. 1868 Folie A, Rana RS, Rose KD, Sahni A, Kumar K, Singh L, Smith T. 2013. Early Eocene frogs from 1869 Vastan Lignite Mine, Gujarat, India. Acta Palaeontologica Polonica 58 (3):511–524. 1870 Fritz U, Schmidt C, Ernst CH. 2011. Competing generic concepts for Blanding’s, Pacific and 1871 European pond turtles (Emydoidea, Actinemys and Emys)—Which is best. Zootaxa 2971:41–53. 1872 Gaiduchenko LL. 1984. On stratigraphy of Neogene sediments of the southmost Westsiberian 1873 Plain. In: Volkova VS, Kul'kova LA, eds. Environment and ligfe at the boundaries of Cenozoic 1874 epochs in SIberia and Far East. Novosibirsk: Nauka, 172–184. 1875 Gaiduchenko LL, Chkhikvadze VM. 1985. First record of chelydrid turtle from the Neogene 1876 sediments of Pavlodarian Priirtyshya. Geologiya i geofizika (1):116–118. 1877 Garcia-Porta J, Litvinchuk SN, Crochet PA, Romano A, Geniez PH, Lo-Valvo M, Lymberakis P, 1878 Carranza S. 2012. Molecular phylogenetics and historical biogeography of the west-palearctic 1879 common toads (Bufo bufo species complex). Molecular Phylogenetics and Evolution 63 (1):113– 1880 130. 1881 Gnibidenko ZN, Volkova VS, Kuz’mina OB, Dolya ZA, Khazina IV, Levicheva AV. 2011. 1882 Stratigraphic, paleomagnetic, and palynological data on the Paleogene–Neogene continental 1883 sediments of southwestern West Siberia. Russian Geology and Geophysics 52 (4):466–473. 1882 Stratigraphic, paleomagnetic, and palynological data on the Paleogene–Neogene continental 1883 sediments of southwestern West Siberia. Russian Geology and Geophysics 52 (4):466–473. 1884 Gnibitenko ZN. 2006. Cenozoic paleomagnetism of the West Siberian Plate. Novosibirsk: Geo. 1883 sediments of southwestern West Siberia. Russian Geology and Geophysics 52 (4):466–473. 1884 Gnibitenko ZN. 2006. Cenozoic paleomagnetism of the West Siberian Plate. Novosibirsk: Geo. 1884 Gnibitenko ZN. 2006. Cenozoic paleomagnetism of the West Siberian Plate. Novosibirsk: Geo. 1885 Goldfuss GA. 1820. Handbuch der Zoologie. Nürnberg: J. L. Schrag. 1886 Gray JE. 1825. A synopsis of the genera of reptiles and Amphibia, with a description of some new 1887 species. Annals of Philosophy, London 10:193–217. 1888 Gray JE. 1870. Supplement to the Catalogue of shield reptiles in the collection of the British Museum. 1889 Part 1. Testudinata (Tortoises). London: Taylor & Francis. 1890 Gubin YM. 1995. Тhе First Find оf Pelobatids (Аnurа) in the Paleogene of Моngоliа. 1891 Paleontologicheskiy zhurnal (4):73–76. Manuscript to be reviewed The skull of lizards and tuatara. In: Gans C, ed. The skull of Lepidosauria, Volume 1858 20, Morphology H. Ithaca: Society for the Study of Amphibians and Reptiles, 1–347. 1859 Fejfar O, Heinrich W-D, Pevzner MA, Vangengeim EA. 1997. Late Cenozoic sequences of 1860 mammalian sites in Eurasia: an updated correlation. Palaeogeography, Palaeoclimatology, 1861 Palaeoecology 133 (3–4):259–288. 1862 Fischer von Waldheim G. 1813. Zoognosia. Tabulis Synopticis Illustrata, in Usum Prælectionum 1863 Academiæ Imperialis Medico-Chirurgicæ Mosquensis Edita. Moscow: Typis Nicolai Sergeidis 1864 Vsevolozsky. PeerJ reviewing PDF \| (2016:08:12606:4:0:NEW 23 Jan 2017) Manuscript to be reviewed 1908 Iskakova K. 1969. Fossil amphibians from Priirtyshie. Proceedings of Academy of Sciences of 1909 Kazakhian SSR, Series Biological (1):48–53. 1910 Ivanov M. 2008. Early Miocene Amphibians (Caudata, Salientia) from the Mokrá-Western Quarry 1911 (Czech Republic) with comments on the evolution of Early Miocene amphibian assemblages in 1912 Central Europe. Geobios 41 (4):465–492. 1913 Kirscher U, Prieto J, Bachtadse V, Abdul Aziz H, Doppler G, Hagmaier M, Böhme M. 2016. A 1914 biochronologic tie-point for the base of the Tortonian stage in European terrestrial settings: 1915 Magnetostratigraphy of the topmost Upper Freshwater Molasse sediments of the North Alpine 1916 Foreland Basin in Bavaria (Germany). Newsletters on Stratigraphy 49 (3):445–467. 1917 Kordikova EG. 1994. Review of fossil Trionychid localities in the Soviet Union. Courier 1918 Forschungsinstitut Senckenberg 173:341–358. 1919 Kuzmin SL. 1995. Die Amphibien Russlands und angrenzender Gebiete. Magdeburg: Westarp 1920 Wissenschaften. 1921 Laurenti JN. 1768. Specimen medicum, exhibens synopsin reptilium emendatum cum experimentis 1922 circa venena et antidota Reptilium Austriacorum. Viennae: Typ. Joan. Thom. nob. de Trattnern. 1923 Li J-T, Wang J-S, Nian H-H, Litvinchuk SN, Wang J, Li Y, Rao D-Q, Klaus S. 2015. Amphibians 1924 crossing the Bering Land Bridge: Evidence from holarctic treefrogs (Hyla, Hylidae, Anura). 1925 Molecular Phylogenetics and Evolution 87:80–90. 1926 Linnaeus C. 1758. Systema naturae per regna tria naturae, secundum classes, ordines, genera, 1927 species, cum characteribus, differentiis, synonymis, locis. L. Salvi: Stockholm. 1928 Linnaeus C. 1761. Fauna Svecica sisten Animalia Sveciae Regni Mammalia, Aves, Amphibia, Pisces, 1929 Insecta, Vermes. Distributa per Classes & Ordines, Genera & Species, cum Differentiis Specierum, 1930 Synonymis Auctorum, Nominibus Incolarum, Locis Natalium Descriptionibus Insectorum. Editio 1931 altera. Stockholm: Laurentius Salvus. 1932 Lucas SG, Bray ES, Emry RJ, Hirsch KF. 2012. Dinour eggshell and the Cretaceous-Paleogene 1933 boundary in the Zaysan Basin, Eastern Kazakstan. Journal of Stratigraphy 36 (2):417–435. 1934 Lychev GF. 1990. Use of relative higth of the tooth crown of castorids as a indicator of geological 1935 time. In: Tleuberdina PA, Kojamkulova BS, Rajushkina GS, eds. Vertebrate fauna and flora of the 1936 Mesozoic and Cenozoic of Kazakhstan, vol. 11. Alma-Ata: Nauka, 54–60. 1937 Lymberakis P, Poulakakis N, Manthalou G, Tsigenopoulos CS, Magoulas A, Mylonas M. 2007. 1938 Mitochondrial phylogeography of Rana (Pelophylax) populations in the Eastern Mediterranean 1939 region. Molecular Phylogenetics and Evolution 44 (1):115–125. 1940 Malakhov DV. 2003. The earliest known record of Mioproteus (Caudata; Proteidae) from the 1941 Middle Miocene of Central Kazakhstan. Manuscript to be reviewed 1892 Gvoždík V, Moravec J, Klütsch C, Kotlík P. 2010. Phylogeography of the Middle Eastern tree frogs 1893 (Hyla, Hylidae, Amphibia) as inferred from nuclear and mitochondrial DNA variation, with a 1894 description of a new species. Molecular Phylogenetics and Evolution 55 (3):1146–1166. 1895 Heizmann EPJ, Bloos G, Böttcher R, Werner J, Ziegler R. 1989. Ulm-Westtangente und Ulm- 1896 Uniklinik: Zwei neue Wirbeltier-Faunen aus der Unteren Süßwasser-Molasse (Untermiozän) 1897 von Ulm (Baden-Württemberg). Stuttgarter Beiträge Naturkunde Serie B 153:1–14. 1898 Hodrova M. 1980. A toad from the Middle Miocene at Devínska Nová Ves near Bratislava. Věstník 1899 Ústředního ústavu geologichého 55 (5):311–316. Č 1900 Hodrová M. 1987. Amphibians from the Miocene sediments of the Bohemian Karst. Časopis pro 1901 mineralogii a geologii 32 (4):345–356. 1902 Hodrová M. 1988. Miocene frog fauna from the locality Devínska Nová Ves - Bonanza. Věstník 1903 Ústředního ústavu geologichého 63 (5):305–310. 1904 Hoffstetter R, Gasc J-P. 1969. Vertebrae and rips of modern reptiles. In: Gans C, ed. Morphology A, 1905 Biology of the Reptilia. London, New York: Academic Press, 201–310. 1906 Holman AJ. 1995. A New Species of Emydoidea (Reptilia: Testudines) from the Late Barstovian 1907 (Medial Miocene) of Cherry County, Nebraska. Journal of Herpetology 29 (4):548–553. PeerJ reviewing PDF \| (2016:08:12606:4:0:NEW 23 Jan 2017) Manuscript to be reviewed 1951 Młynarski M, Szyndlar Z, Estes R, Sanchíz B. 1982. Lower vertebrate fauna from the Miocene of 1952 Opole (Poland). Estudios geológicos 38:103–119. 1951 Młynarski M, Szyndlar Z, Estes R, Sanchíz B. 1982. Lower vertebrate fauna from the Miocene of 1952 Opole (Poland). Estudios geológicos 38:103–119. 1953 Müller MJ, Hennings D. 2000. The Global Climate Data Atlas on CD-Rom. Fle 1953 Müller MJ, Hennings D. 2000. The Global Climate Data Atlas on CD Rom. Flensburg and Köln. 1954 Nikitina N, Ananjeva NB. 2009. Characteristics of dentition in gekkonid lizards of the genus 1955 Teratoscincus and other Gekkota (Sauria, Reptilia). Biology Bulletin 36 (2):193–198. 1954 Nikitina N, Ananjeva NB. 2009. Characteristics of dentition in gekkonid lizards of the genus 1955 Teratoscincus and other Gekkota (Sauria, Reptilia). Biology Bulletin 36 (2):193–198. 1956 Nikitina NG. 2009. Pecularities of skull morphology and skin of geckos (Reptilia: Sauria: Gekkota) 1957 and their phylogenetic relevance, Zoological Institute of RAS. 1956 Nikitina NG. 2009. Pecularities of skull morphology and skin of geckos (Reptilia: Sauria: Gekkota) 1957 and their phylogenetic relevance, Zoological Institute of RAS. 1958 Nilsson S. 1842. Skandinavisk herpetologi eller beskrifning öfver de sköldpaddor, ödlor, ormar och 1959 grodor, som förekomma i Sverige och Norrige, jemte deras lefnadssätt, födoämnen, nytta och 1960 skada m.m. Lund: Tryckt uti Borlingska Boktryckeriet. 1958 Nilsson S. 1842. Skandinavisk herpetologi eller beskrifning öfver de sköldpaddor, ödlor, ormar och 1959 grodor, som förekomma i Sverige och Norrige, jemte deras lefnadssätt, födoämnen, nytta och 1960 skada m.m. Lund: Tryckt uti Borlingska Boktryckeriet. 1961 Nokariya H. 1983. Comparative osteology of Japanese frogs and toads for paleontological studies 1962 (I): Bufo, Hyla, Microhyla and Bombina. Bulletin of Natural Science Museum, Series C 9 (1):23–40. 1961 Nokariya H. 1983. Comparative osteology of Japanese frogs and toads for paleontological studies 1962 (I): Bufo, Hyla, Microhyla and Bombina. Bulletin of Natural Science Museum, Series C 9 (1):23–40. 1963 Oken L 1816 Lehrbuch der Naturgeschichte Leipzig Jena: Reclam ( ) f , y , y f , ( ) 1963 Oken L. 1816. Lehrbuch der Naturgeschichte. Leipzig, Jena: Reclam. 1963 Oken L. 1816. Lehrbuch der Naturgeschichte. Leipzig, Jena: Reclam. 1964 Oppel M. 1811. Die Ordnungen, Familien und Gattungen der Reptilien als Prodrom einer 1965 Naturgeschichte derselben. München: Joseph Lindauer. 1964 Oppel M. 1811. Die Ordnungen, Familien und Gattungen der Reptilien als Prodrom einer 1965 Naturgeschichte derselben. München: Joseph Lindauer. Manuscript to be reviewed Biota 4 (1-2):67–72. 1942 Malakhov DV. 2004. Toads (Anura, Bufonidae) from the Middle Miocene in the Turgay Depression 1943 (Central Kazakhstan). Biota 5 (1-2):41–46. 1944 Malakhov DV. 2005. The early Miocene herpetofauna of Ayakoz (Eastern Kazakhstan). Biota 6 (1- 1945 2):29–35. 1946 Malakhov DV. 2009. Fossil amphibians and reptiles from Cenozoic of Kazakhstan: state of art and 1947 new materials. Transactions of the Institute of Zoology MES RK 50:25–34. 1948 Miklas PM. 2002. Die Amphibienfauna (Amphibia: Caudata, Anura) der obermiozänen Fundstelle 1949 Götzendorf an der Leitha (südliches Wiener Becken, Niederösterreich). Annalen des 1950 Naturhistorischen Museums in Wien 103A:161–211. PeerJ reviewing PDF \| (2016:08:12606:4:0:NEW 23 Jan 2017) Manuscript to be reviewed 1966 Pabijan M, Wandycz A, Hofman S, Węcek K, Piwczyński M, Szymura JM. 2013. Complete 1967 mitochondrial genomes resolve phylogenetic relationships within Bombina (Anura: 1968 Bombinatoridae). Molecular Phylogenetics and Evolution 69 (1):63–74. 1969 Pomel A. 1853. Catalogue méthodique et descriptif des vertébrés fossiles découverts dans le bassin 1970 hydrographique supérieur de la Loire. Paris: J. B. Baillières. 1971 Rafinesque CS. 1815. Analyse de Nature, ou Tableau de l’Universe et des Corps Organisés. Palermo: 1972 Jean Barravecchia. 1973 Rage J-C. 1976. Les Squamates du Miocène de Béni Mellal, Maroc. Géologie méditerranéenne 3 1974 (2):57–70. 1975 Rage J-C. 2003. Oldest Bufonidae (Amphibia, Anura) from the Old World: a bufonid from the 1976 Paleocene of France. Journal of Vertebrate Paleontology 23 (2):462–463. 1977 Rage J-C, Hossini S. 2000. Les Amphibiens du Miocène moyen de Sansan. Mémoires du Muséum 1978 national d'histoire naturelle 183:177–217. 1979 Ratnikov VY. 1997. Tailless amphibians and lanscape settings of the Late Cenozoic in Western 1980 Transbaikalia. Geologiya i geofizika 39 (9):1458–1464. 1981 Ratnikov VY. 2001. Pliocene anurans of East-European platform. Russian Journal of Herpetology 8 1982 (3):171–178. 1983 Ratnikov VY. 2002. New find of amphibians and reptiles in type localities of Muchkapian, upper 1984 Don Basin. Bulletin of Voronezh State University: Geology (1):73–79. 1985 Ratnikov VY. 2010. A review of tailed Amphibian remains from Late Cenozoic sediments of the 1986 East European plain. Russian Journal of Herpetology 17 (1):59–66. 1987 Ratnikov VY, Litvinchuk SN. 2009. Atlant vertebrae of tailed amphibians of Russia and adjacent 1988 countries. Russian Journal of Herpetology 19 (1):57–68. 1989 Ravkin YS, Bogomolova IN, Chesnokova SV. 2010. Amphibian and reptile biogeographic regions 1990 of Northern Eurasia, mapped separately. Contemporary Problems of Ecology 3 (5):562–571. 1991 Ravkin YS, Yudkin VA, Tsybulin SM, Kuranova VN, Borisovich OB, Bulakhova NA, Patrakov SV, 1992 Shamgunova RR. 2008. Spatial-typological structure and mapping of reptile population of 1993 West Siberia. Contemporary Problems of Ecology 1 (2):214–220. PeerJ reviewing PDF \| (2016:08:12606:4:0:NEW 23 Jan 2017) Manuscript to be reviewed 1994 Recuero E, Canestrelli D, Vörös J, Szabó K, Poyarkov NA, Arntzen JW, Crnobrnja-Isailovic J, 1995 Kidov AA, Cogălniceanu D, Caputo FP, Nascetti G, Martínez-Solano I. 2012. Multilocus 1996 species tree analyses resolve the radiation of the widespread Bufo bufo species group (Anura, 1997 Bufonidae). Molecular Phylogenetics and Evolution 62 (1):71–86. 1998 Reichenbacher B, Böhme M, Heissig K, Prieto J, Kossler A. 2004. New approach to assess 1999 biostratigraphy, palaeoecology and past climate in the South German Molasse Basin during the 2000 Early Miocene (Ottnangian, Karpatian). Courier Forschungsinstitut Senckenberg 249:71–89. 2001 Reichenbacher B, Krijgsman W, Lataster Y, Pippèrr M, Baak CC, Chang L, Kälin D, Jost J, 2002 Doppler G, Jung D, Prieto J, Abdul Aziz H, Böhme M, Garnish J, Kirscher U, Bachtadse V. 2003 2013. A new magnetostratigraphic framework for the Lower Miocene (Burdigalian/Ottnangian, 2004 Karpatian) in the North Alpine Foreland Basin. Swiss Journal of Geosciences 106 (2):309–334. 2005 Roček Z. 1984. Lizards (Reptili: Sauria) from the Lower Miocene locality Dolnice (Bohemia, 2006 Czechoslovakia). Řada matematických a přirodních vĕd 94 (1):4–69. 2007 Roček Z. 2005. Late Miocene Amphibia from Rudabánya. Palaeontographia Italica 90:11–29. 2008 Roček Z, Wuttke M, Gardner J, Singh Bhullar B-A. 2014. The Euro-American genus Eopelobates, 2009 and a re-definition of the family Pelobatidae (Amphibia, Anura). Palaeobiodiversity and 2010 Palaeoenvironments 94 (4):529–567. 2011 Sanchiz B, Schleich H. 1986. Erstnachweis der Gattung Bombina (Amphibia: Anura) im 2012 Untermiozän Deutschlands. Mitteilungen der Bayerischen Staatssammlung für Paläontologie und 2013 historische Geologie 26:41–44. 2014 Sanchíz B. 1998a. Salientia. München: Verlag Dr. Friedrich Pfeil. 2015 Sanchíz B. 1998b. Vertebrates from the Early Miocene lignite deposits of the opencast mine 2016 Oberdorf (Western Styrian Basin, Austria). Annalen des Naturhistorischen Museums in Wien 2017 99A:13–29. 2018 Sanchíz B, Młynarski M. 1979. Remarks on the Fossil Anurans from the Polish Neogene. Acta 2019 zoologica cracoviensia 24 (3):153–174. 2020 Sanchíz B, Schleich H-H. 1986. Erstnachweis der Gattung Bombina (Amphibia: Anura) im 2021 Untermiozän Deutschlands. Mitteilungen der Bayerischen Staatssammlung für Paläontologie und 2022 historische Geologie 26:41–44. 2020 Sanchíz B, Schleich H-H. 1986. Erstnachweis der Gattung Bombina (Amphibia: Anura) im 2021 Untermiozän Deutschlands. Mitteilungen der Bayerischen Staatssammlung für Paläontologie und 2022 historische Geologie 26:41–44. 2023 Sanchíz B, Schleich H-H, Esteban M. 1993. Water frogs (Ranidae) from the Oligocene of Germany. 2024 Journal of Herpetology 27 (4):486–489. 2023 Sanchíz B, Schleich H-H, Esteban M. 1993. Water frogs (Ranidae) from the Oligocene of Germany. Manuscript to be reviewed 2037 Palearctic green toads (Bufo viridis subgroup) with insights in their genomic plasticity. 2038 Molecular Phylogenetics and Evolution 41 (3):663–689. 2039 Stöck M, Sicilia A, Belfiore N, Buckley D, Lo Brutto S, Lo Valvo M, Arculeo M. 2008b. Post- 2040 Messinian evolutionary relationships across the Sicilian channel: Mitochondrial and nuclear 2041 markers link a new green toad from Sicily to African relatives. BMC Evolutionary Biology 8 (1). 2042 Syromyatnikova EV. 2014. The first record of Salamandrella (Caudata: Hynobiidae) from the 2043 Neogene of Russia. Russian Journal of Herpteology 21 (3):217–220. 2044 Tempfer PM. 2005. The herpetofauna (Amphibia: Caudata, Anura; Reptilia: Scleroglossa) of the 2045 Upper Miocene locality Kohfidisch, Burgenland, Austria, Wien, Universität. 2046 Tleuberdina PA, ed. 1993. Faunistic and floristic complexes of Mesozoic and Cenozoic of 2047 Kazakhstan. Almaty: Baspager. 2048 Tleuberdina PA, Kozhamkulova GS, Kondratenko BS. 1989. Catalogue of the Cenozoic mammals 2049 from Kazakhstan. Alma-Ata: Nauka Kazakhian SSR. 2050 Tleuberdina PA, Volkova VS, Lushczaeva TT, Lychev GF, Pita OM, Tjutkova LA, Chkhikvadze 2051 VM. 1993. Vertebrate fauna of Kalkaman (Pavlodar Priirtyshe). In: Tleuberdina PA, ed. Faunistic 2052 and floristic complexes of Mesozoic and Cenozoic of Kazakhstan, vol. 12. Almaty: Baspager, 132– 2053 157. 2054 Tschudi JJv. 1838. Classification der Batrachier, Mit Berucksichtigung der fossilen Thiere dieser 2055 Abtheilung der Reptilien. Neuchâtel: Petitpierre. 2056 Vangengeim EA, Pevzner MA, Tesakov AS. 2005. Ruscinian and Lower Villafranchian: age of 2057 boundaries and position in magnetochronological scale. Stratigraphy and Geological Correlation 2058 13 (5):530–546. 2059 Vasilyan D, Böhme M, Chkhikvadze VM, Semenov YA, Joyce WG. 2013. A new giant salamander 2060 (Urodela, Pancryptobrancha) from the Miocene of Eastern Europe (Grytsiv, Ukraine). Journal of 2061 Vertebrate Paleontology 33 (2):301–318. 2062 Vasilyan D, Böhme M, Klembara J. 2016. First record of fossil Ophisaurus (Anguimorpha, 2063 Anguidae, Anguinae) in Asia (Russia and Kazakhstan). Journal of Vertebrate Paleontology:1–6. 2064 Veith M, Kosuch J, Vences M. 2003. Climatic oscillations triggered post-Messinian speciation of 2065 Western Palearctic brown frogs (Amphibia, Ranidae). Molecular Phylogenetics and Evolution 26 2066 (2):310–327. 2067 Venczel M. 1999a. Fossil land salamanders (Caudata, Hynobiidae) from the Carpathian basin: 2068 relation between extinct and extant genera. Acta Palaeontologica Romaniae 2:489–492. 2069 Venczel M. 1999b. Land salamanders of the family Hynobiidae from the Neogene and Quaternary 2070 of Europe. Amphibia-Reptilia 20:401–412. 2071 Venczel M. 2000. Amphibians from the Lower Pleistocene Betfia 9 locality (Bihor country, 2072 Romania). Satu Mare - Studii şi comunicări, seria stiintele naturale 1:28–37. Manuscript to be reviewed 2024 Journal of Herpetology 27 (4):486–489. 2025 Savage JM. The geographic dstribution of frogs: patterns and predictions. In: Val (Ed.) 1973 – 2026 Evolutionary biology of the anurans, 351–445. 2025 Savage JM. The geographic dstribution of frogs: patterns and predictions. In: Val (Ed.) 1973 – 2026 Evolutionary biology of the anurans, 351–445. 2027 Scopoli GA. 1777. Introductio ad historiam naturalem, sistens genera lapidum, plantarum et 2028 animalium hactenus detecta, caracteribus essentialibus donata, in tribus divisa, subinde ad leges 2029 naturae. Prague: Apud Wolfgangum Gerle. 2027 Scopoli GA. 1777. Introductio ad historiam naturalem, sistens genera lapidum, plantarum et 2028 animalium hactenus detecta, caracteribus essentialibus donata, in tribus divisa, subinde ad leges 2029 naturae. Prague: Apud Wolfgangum Gerle. 2030 Shpanskiy AV. 2008. Pecularities of the development of the hipparion fauna of the Pavlodar 2031 Priirtyshe. In: Tleuberdina PA, Erzhanov NT, Zykin VS, eds.: Pavlodarian State University, 92–96. 2030 Shpanskiy AV. 2008. Pecularities of the development of the hipparion fauna of the Pavlodar 2031 Priirtyshe. In: Tleuberdina PA, Erzhanov NT, Zykin VS, eds.: Pavlodarian State University, 92–96. 2032 Stöck M, Dubey S, Klütsch C, Litvinchuk SN, Scheidt U, Perrin N. 2008a. Mitochondrial and 2033 nuclear phylogeny of circum-Mediterranean tree frogs from the Hyla arborea group. Molecular 2034 Phylogenetics and Evolution 49 (3):1019–1024. 2032 Stöck M, Dubey S, Klütsch C, Litvinchuk SN, Scheidt U, Perrin N. 2008a. Mitochondrial and 2033 nuclear phylogeny of circum-Mediterranean tree frogs from the Hyla arborea group. Molecular 2034 Phylogenetics and Evolution 49 (3):1019–1024. ck M, Moritz C, Hickerson M, Frynta D, Dujsebayeva T, Eremchenko V, Macey JR, 2035 Stöck M, Moritz C, Hickerson M, Frynta D, Dujsebayeva T, Eremchenko V, Macey JR, 2036 Papenfuss TJ, Wake DB. 2006. Evolution of mitochondrial relationships and biogeography of 2035 Stöck M, Moritz C, Hickerson M, Frynta D, Dujsebayeva T, Eremchenko V, Macey JR, 2036 Papenfuss TJ, Wake DB. 2006. Evolution of mitochondrial relationships and biogeography of 2035 Stöck M, Moritz C, Hickerson M, Frynta D, Dujsebayeva T, Eremchenko V, Macey JR, 2036 Papenfuss TJ, Wake DB. 2006. Evolution of mitochondrial relationships and biogeography of PeerJ reviewing PDF \| (2016:08:12606:4:0:NEW 23 Jan 2017) Manuscript to be reviewed 2073 Venczel M. 2001. Anurans and squamates from the Lower Pliocene (MN 14) Osztramos 1 locality 2074 (Northern Hungary). Palaeontologica Hungarica 19:79–90. 2075 Venczel M. 2004. Middle Miocene anurans from the Carpathian Basin. Palaeontographica Abt. A 2076 271:151–174. 2077 Venczel M. 2008. A new salamandrid amphibian from the Middle Miocene of Hungary and its 2078 phylogenetic relationships. Journal of Systematic Palaeontology 6 (1):41–59. PeerJ reviewing PDF \| (2016:08:12606:4:0:NEW 23 Jan 2017) Manuscript to be reviewed 2079 Venczel M, Codrea V, Fărcaş C. 2012. A new palaeobatrachid frog from the early Oligocene of 2080 Suceag, Romania. Journal of Systematic Palaeontology 11 (2):179–189. 2081 Venczel M, Gardner JD, Codrea VA, Csiki-Sava Z, Vasile Ş, Solomon AA. 2016. New insights into 2082 Europe’s most diverse Late Cretaceous anuran assemblage from the Maastrichtian of western 2083 Romania. Palaeobiodiversity and Palaeoenvironments 96 (1):61–95. 2084 Venczel M, Hír J. 2013. Amphibians and Squamates from the Miocene of Felsötárkány Basin, N- 2085 Hungary. Palaeontographica Abteilung A 300 (1-6):117–147. 2086 Vergnaud-Grazzini C, Młynarski M. 1969. Position systématic du genre Pliobatrachus Fejérváry 2087 1917. Comptes Rendus hebdomadaires des Séances de l'Académie des Sciences, Série D: Sciences 2088 naturelles 268:2399–2402. 2089 Villa A, Roček Z, Tschopp E, Van den Hoek Ostende, Lars W., Delfino M. 2016. Palaeobatrachus 2090 eurydices, sp. nov. (Amphibia, Anura), the last western European palaeobatrachid. Journal of 2091 Vertebrate Paleontology:e1211664. 2092 Volkova VS. 1984. Changes in the palinofloras of Siberia in the late Cenozoic. In: Volkova VS, 2093 Kul'kova LA, eds. Environment and ligfe at the boundaries of Cenozoic epochs in SIberia and Far 2094 East. Novosibirsk: Nauka, 54–69. 2095 Wagler JG. 1830. Natürliches System der Amphibien, mit vorangehender Classification der 2096 Säugethiere und Vögel . Ein Beitrag zur vergleichenden Zoologie. München, Stuttgart, Tübingen: 2097 J.G. Cotta. 2098 Wuttke M, Přikryl T, Ratnikov VY, Dvořák Z, Roček Z. 2012. Generic diversity and distributional 2099 dynamics of the Palaeobatrachidae (Amphibia: Anura). Palaeobiodiversity and 2100 Palaeoenvironments 92 (3):367–395. 2101 Zykin VS. 1979. Stratigraphy and unionids of the Pliocene of southern part of Western Siberian plain. 2102 Novosibirsk: Nauka. 2103 Zykin VS. 2012. Stratigraphy and evolution of environments and climate during Late Cenozoic in the 2104 Southern West Siberia. Novosibirsk: Geo. 2105 Zykin VS, Zazhigin VS. 2004. A new biostratigraphic level of the Pliocene in Western Siberia and 2106 the age of the Lower-Middle Miocene stratotype of the Beshcheul Horizon. Doklady Earth 2107 Sciences 398 (7):904–907. 2108 Zykin VS, Zazhigin VS. 2008. On the Neogene stratigraphy of Pavlodarian Priirtyshya. In: 2109 Tleuberdina PA, Erzhanov NT, Zykin VS, eds.: Pavlodarian State University, 15–21. 2110 Zykin VS, Zykina VS, Zazhigin VS. 2007. Issues in separating and correlating Pliocene and 2111 Quaternary sediments of Southerwestern Siberia. Archeology, Ethnology & Anthropology of 2112 Eurasia 30 (2):24–40. Manuscript to be reviewed Figure 1(on next page) Figure 1(on next page) Figure 1(on next page) Map of Eurasia (A) showing location of the Western Siberian studied fossil sites (B) (1- 38, 58; black–thin outlined circles) as well as localities known from the literature (39-57; white – thick outlined circles). 1, Baikadam; 2, Malyi Kalkaman 2; 3, Malyi Kalkaman 1; 4, Shet-Irgyz 1; 5, Petropavlovsk 1; 6, Znamenka; 7, Pavlodar 1A; 8, Selety 1A; 9, Kedey; 10, Novaya Stanitsa 1A; 11, Borki 1A; 12, Lezhanka 2A; 13, Cherlak; 14, Pavlodar 1B; 15, Lezhanka 2B; 16, Olkhovka 1A; 17, Olkhovka 1B; 18, Olkhovka 1C; 19, Iskakovka 2A; 20, Isakovka 1A; 21, Peshniovo 3; 22, Isakovka 1B; 23, Kamyshlovo; 24, Beteke 1B; 25, Pavlodar 2B; 26, Pavlodar 3A; 27, Lezhanka 1; 28, Andreievka-Speransko; 29, Andreievka 1; 30, Livenka; 31, Beteke 1C; 32, Lebiazhie 1A; 33, Lebiazhie 1B; 34, Podpusk 1; 35, Beteke 2; 36, Beteke 4; 37, Kamen-na-Obi; 38, Razdole; 39, Akespe; 40, Ayakoz; 41, Golubye Peski; 42, Zmei Gorynych; 43, Vympel; 44, Poltinik; 45, Zaezd; 46, Tri Bogatyrja; 47, Kaymanovaja cherepakha; 48, Ryzhaya II; 49, Kentyubek; 50, Ashut; 51, Point “Y”; 52, Sarybulak Svita; 53, Kalmakpai Svita; 54, Karabastuz; 55, Kalmakpai; 56, Petropavlovsk 1/2; 57, Detskaya zheleznaja doroga; 58, Shet-Irgyz 2. Map data © 2016 Google and Map data © OpenStreetMap contributors, CC BY-SA. 1, Baikadam; 2, Malyi Kalkaman 2; 3, Malyi Kalkaman 1; 4, Shet-Irgyz 1; 5, Petropavlovsk 1; 6, Znamenka; 7, Pavlodar 1A; 8, Selety 1A; 9, Kedey; 10, Novaya Stanitsa 1A; 11, Borki 1A; 12, Lezhanka 2A; 13, Cherlak; 14, Pavlodar 1B; 15, Lezhanka 2B; 16, Olkhovka 1A; 17, Olkhovka 1B; 18, Olkhovka 1C; 19, Iskakovka 2A; 20, Isakovka 1A; 21, Peshniovo 3; 22, Isakovka 1B; 23, Kamyshlovo; 24, Beteke 1B; 25, Pavlodar 2B; 26, Pavlodar 3A; 27, Lezhanka 1; 28, Andreievka-Speransko; 29, Andreievka 1; 30, Livenka; 31, Beteke 1C; 32, Lebiazhie 1A; 33, Lebiazhie 1B; 34, Podpusk 1; 35, Beteke 2; 36, Beteke 4; 37, Kamen-na-Obi; 38, Razdole; 39, Akespe; 40, Ayakoz; 41, Golubye Peski; 42, Zmei Gorynych; 43, Vympel; 44, Poltinik; 45, Zaezd; 46, Tri Bogatyrja; 47, Kaymanovaja cherepakha; 48, Ryzhaya II; 49, Kentyubek; 50, Ashut; 51, Point “Y”; 52, Sarybulak Svita; 53, Kalmakpai Svita; 54, Karabastuz; 55, Kalmakpai; 56, Petropavlovsk 1/2; 57, Detskaya zheleznaja doroga; 58, Shet-Irgyz 2. Map data © 2016 Google and Map data © OpenStreetMap contributors, CC BY-SA. PeerJ reviewing PDF \| (2016:08:12606:4:0:NEW 23 Jan 2017) PeerJ reviewing PDF \| (2016:08:12606:4:0:NEW 23 Jan 2017) Manuscript to be reviewed PeerJ reviewing PDF \| (2016:08:12606:4:0:NEW 23 Jan 2017) PeerJ reviewing PDF \| (2016:08:12606:4:0:NEW 23 Jan 2017) PeerJ reviewing PDF \| (2016:08:12606:4:0:NEW 23 Jan 2017) Manuscript to be reviewed Figure 2(on next page) Figure 2(on next page) Compiled stratigraphy of the Middle Miocene – Early Pleistocene studied localities in Western Siberia grouped in the svitas and relative to their geographic positions. The localities without assignment into a certain svita are given in coloured frames according to age. In the right column the small mammalian biochronologic data (species or lineages) used for age estimations are given, the numbers accompanying the species refer the locality numbers. The arrows left from the small mammal taxa indicates its/their first appearance. Abbreviations: klm – Kalkaman, ish – Ishim, pv – Pavlodar, kd – Kedey, nst – Novaya Stanitsa, rt – Rytov, is – Isakov, psh – Peshnev, krt – Krutogor, bt – Betekey, liv – Levetin, irt – Irtysh, kar – Karagash. PeerJ reviewing PDF \| (2016:08:12606:4:0:NEW 23 Jan 2017) age chrones epoch MN Cherlak (r), Pavlodar 1B Novaya Stanitsa 1A (n) Borki 1A Lezhanka 2A Lezhanka 2B Beteke 1C (n) Lebiazhie 1A (r) Lebiazhie 1B (r) Podpusk 1 (r) Beteka 4 (r) Beteke 2 (n) Kedey Selety 1A West East lineages of small mammals ish nst rt kd 17 16 15 14 13 12 10 9 11 Ma C3n.4n C3r C3An.1n C3An.1r C3An.2n C3Ar C3Bn C3Br.2r C3Br.3r C4n.1n C4n.2n C4An C4Ar.1n C4Ar.2n C4r.1n C5n.1n C2An.3n C2r.2r C3n.3r C3n.3n C3n.2r C3n.2n C3n.1r C3n.1n C2Ar C2An.2r C2An.2n C2An.1r C2An.1n C2r.1r C1r.1n C1r.3r C1r.2r C2n C1r.1r Petropavlovsk 1 Sh t I 1 Sh t I 2 kar Razdole (r) 1 2 3 4 5 6 7 9 10 Miocene Late Late Early Early Tortonian Messinian Zanclean Piacen Gelas Calb Pleistocene Pliocene ~ ~ ~ ~ Kamen-na-Obi (n) irt sel kc irt 5 4 58 5 5 pv Pavlodar 1A (r) 7 8 8 10 8 9 10 11 12 13 13 14 14 8 8 9 10 10 13 13, 14 15 Olkhovka 1A-1C 20-22 Iskakova 1A, 2A (r) is 16, 17 16 Peshniovo 3 (n) psh 18 18 18 19 23 29, 30 31 24,25 28 24 Iskakova 1B (n) krt 19 20-22, 23-25 28, 30 Beteke 1B Pavlodar 2B Pavlodar 3A Kamyshlovo bt bt 23 25 26 24 31 34 32-34 35 36 37 38 32 33 37 38 35 Andreievka - Speransko Andreievka 1 Livenka (r) Lezhanka 1 liv 27 29 30 28 36 ? gracilis vinogradovi vinogradovi afanasievi sibiricus sibiricus antiquus cf. Manuscript to be reviewed Manuscript to be reviewed Figure 2(on next page) davakosi pauli irthyshensis irtyshensis seletyensis ultimus ultimus Lophocricetus Rhinocerodon Allophaiomys pliocaenicus seletyensis latidens betekensis latidens Sibirosminthus sp. 1 sp. 2 insuliferus Promimomys Prosomys gracilis gracilis polonicus pliocaenicus pliocaenicus ? gracilis Mimomys Ischymomys quadriradicatus minuscilus H t i th d h i Manuscript to be reviewed West lineages of small mammals lineages of small mammals Manuscript to be reviewed Beteka 4 (r) Beteke 2 (n) 37 36 32-34 35 pliocaenicus pliocaenicus Mimomys 31 polonicus Beteke 1C (n) sel 31 Pliocene 28, 30 ultimus Lophocricetus Andreievka - Speransko Livenka (r) 0 8 a - Speransko Livenka (r) Andreievka 1 Lezhanka 1 24 betekensis Sibirosminthus Beteke 1B 24 Iskakova 1B (n) 19 Peshniovo 3 (n) Iskakova 1A, 2A (r) is 6, 17 Cherlak (r), Pavlodar 1B Messinian nst Novaya Stanitsa 1A (n) Lezhanka 2A Borki 1A Kedey Selety 1A 1 1 2 3 5 4, 58 5 8 8 8 8 9 Herterosminthus aff. orientalis Microtocricetus molassicus vinogradovi vinogradovi pauli seletyensis seletyensis Ischymomys quadriradicatus minuscilus Heterosminthus mugodzharicus pv Pavlodar 1A (r) 5 Ischymomys quadriradicatus Petropavlovsk 1 Shet Irgyz 1, Shet Irgyz 2 Microtocricetus molassicus Malyi Kalkaman 2 Malyi Kalkaman 1 Salamander remains from Western Siberian localities. Salamander remains from Western Siberian localities. A-C, Cryptobranchidae indet. from the loc. ? Gusiny Perelet, unnr. PIN specimens; A, fragmentary right dentary, natural cross-section; B, the same dentary, in lingual view; C, a jaw fragment, lingual view; D-G, Salamandrella sp., Lezhanka 2A, GIN 1130/1001-AM01, trunk vertebra; H-S, Mioproteus sp.; H-L, loc. Ayakoz, trunk vertebra, GNM unnr. specimen; M-O, trunk vertebra, Borki 1A, GIN 1115/1001-AM01; P-Q, right premaxilla, Malyi Kalkaman 2, GIN 1107/2001-AM01; R-S, left premaxilla loc. Grytsiv (Ukraine), unnr. MNMHK specimen; T-X, trunk vertebrae of aff. Chelotriton sp., loc. Ayakoz, GNM unnr. specimen; Y, Chelotriton sp. from Malyi Kalkaman 2, GNM unnr. specimen; D, H, M, P, R, T, Y, dorsal view; E, I, N, Q, S, U, ventral view; F, J, O, V, lateral view; G, K, W, anterior view; L, X, posterior view. Scale bars: A-C = 5 mm; D-G = 0.5 mm; H-Y = 1 mm. Manuscript to be reviewed PeerJ reviewing PDF \| (2016:08:12606:4:0:NEW 23 Jan 2017) PeerJ reviewing PDF \| (2016:08:12606:4:0:NEW 23 Jan 2017) Manuscript to be reviewed Manuscript to be reviewed PeerJ reviewing PDF \| (2016:08:12606:4:0:NEW 23 Jan 2017) PeerJ reviewing PDF \| (2016:08:12606:4:0:NEW 23 Jan 2017) 016:08:12606:4:0:NEW 23 Jan 2017) PeerJ reviewing PDF \| (2016:08:12606:4:0:NEW 23 Jan 2017) Figure 4 Trunk vertebrae of fossil aff. Tylototriton (A-K), and recent Tylototriton, Echinotriton and Cynops (L-AE). A-E, aff. Tylototriton sp., locality Ayakoz, GNM unnr. specimen; F-K, GIN 950/2001-AM14 and GIN 950/2001-AM01, loc. Baikadam; L-P, Tylototriton verrucosus, GPIT unnr. specimen; Q-U, Tylototriton shanjing, GPIT unnr. specimens; V-Z, Echinotriton andersoni, GPIT unnr. specimen; AA-AE, Cynops pyrrhogaster, GPIT unnr. specimen; A, F, G, L, Q, V, AA, lateral view; B, H, M, R, W, AB, dorsal view; C, I, N, S, X, AC, ventral view; D, J, O, T, Y, AD, anterior view; E, K, P, U, Z, AE, posterior view. Scale bars = 2 mm. Trunk vertebrae of fossil aff. Tylototriton (A-K), and recent Tylototriton, Echinotriton and Cynops (L-AE). A-E, aff. Tylototriton sp., locality Ayakoz, GNM unnr. specimen; F-K, GIN 950/2001-AM14 and GIN 950/2001-AM01, loc. Baikadam; L-P, Tylototriton verrucosus, GPIT unnr. specimen; Q-U, Tylototriton shanjing, GPIT unnr. specimens; V-Z, Echinotriton andersoni, GPIT unnr. specimen; AA-AE, Cynops pyrrhogaster, GPIT unnr. specimen; A, F, G, L, Q, V, AA, lateral view; B, H, M, R, W, AB, dorsal view; C, I, N, S, X, AC, ventral view; D, J, O, T, Y, AD, anterior view; E, K, P, U, Z, AE, posterior view. Scale bars = 2 mm. PeerJ reviewing PDF \| (2016:08:12606:4:0:NEW 23 Jan 2017) Manuscript to be reviewed Figure 5 Palaeobatrichid sphenethmoids. Manuscript to be reviewed PeerJ reviewing PDF \| (2016:08:12606:4:0:NEW 23 Jan 2017) PeerJ reviewing PDF \| (2016:08:12606:4:0:NEW 23 Jan 2017) Palaeobatrichid sphenethmoids. A-D, Palaeobatrachidae indet., Novaya Stanitsa 1A, GIN 948/2001-AM12; E-H, Palaeobatrachus sp. from Grytsiv (Ukraine), unnr. NMNHK specimen; A, E, ventral view; B, F,, dorsal view; C, G, anterior view; D, H, lateral view. Abbreviations: ao, antrum olfactorium; alo, antrum pro lobo olfactorio; is, incisura semielliptical; ff, frontoparietal facet; lp, lateral processes; ls, lamina supraorbitalis; nf, nasal facet; onf, orbitonasal foramina; olf, olfactory foramina; pf, parasphenoid facet. Scale equals 1 mm PeerJ reviewing PDF \| (2016:08:12606:4:0:NEW 23 Jan 2017) Manuscript to be reviewed Manuscript to be reviewed Figure 6 Fossil frogs from Western Siberia. Manuscript to be reviewed Fossil frogs from Western Siberia. Fossil frogs from Western Siberia. A-L, P-Q, X, AA, AB, A, AF, Ilia; A-C Bombina cf. bombina, Selety 1A, GIN 951/1001-AM06; D-E, Bombina sp., Cherlak, GIN 1110/2001-AM13; G-I, Pelobates sp., Selety 1A, GIN 951/1001-AM07; J-L, Hyla gr. H. savignyi, Lezhanka 2A, GIN 1130/1001-AM29; P, Q, Bufo bufo, Olkhovka 1B, GIN 11 11/2001-AM03; X, Bufotes cf. viridis, Pavlodar 1A, GIN 640/5001- AM01; AA, AB, Pelophylax sp., Lezhanka 1, GIN 1129/1001-AM05; AE-AF, Rana arvalis, Malyi Kalkaman 1, GIN 1107/1001-AM10; A, D, G, J, P, U, AA, AE, in lateral view; B, E, H, K, Q, AB, in proximal view; C, F, I, L, in medial view; M-O, R-T, Y, Z, AC, AD, AG, AH, Scapulae of frogs; M-O, Hyla gr. H. savignyi from Lezhanka 2A, GIN 1130/1001-AM33; R-T, Bufo bufo, Olkhovka 1C, GIN 1111/3001-AM01; Y-Z, Bufotes cf. viridis, Pavlodar 1A, GIN 640/5001- AM63; AC-AD, Pelophylax sp., Lezhanka 1, GIN 1129/1001-AM07; AG-AH, Rana temporaria, Malyi Kalkaman 1, GIN 1107/1001-AM01; M, R, Y, AC, AG, dorsal view; N, S, Z, AD, AH, ventral view; O, T, posterior view; U, V, trunk vertebra of Bufo bufo, Olkhovka 1C, GIN 1111/3001-AM02; U, anterior view; V, lateral view; W, urostyle of Bufo bufo, Olkhovka 1C, GIN 1111/3001-AM03, dorsal view. The arrows show the position of the angular fossa. Scale bars: A-Q, AA-AD, AG, AH = 1 mm, R-Z, AE, AF = 2 mm. PeerJ reviewing PDF \| (2016:08:12606:4:0:NEW 23 Jan 2017) PeerJ reviewing PDF \| (2016:08:12606:4:0:NEW 23 Jan 2017) Manuscript to be reviewed PeerJ reviewing PDF \| (2016:08:12606:4:0:NEW 23 Jan 2017) PeerJ reviewing PDF \| (2016:08:12606:4:0:NEW 23 Jan 2017) Manuscript to be reviewed Manuscript to be reviewed PeerJ reviewing PDF \| (2016:08:12606:4:0:NEW 23 Jan 2017) Alsophylax sp. from the localities Cherlak (A-P) and Mynsualmas-MSA 3 (Q). Alsophylax sp. from the localities Cherlak (A-P) and Mynsualmas-MSA 3 (Q). A-E, two left dentaries; A-D, left dentary, GIN 1110/2001-RE11; A, mirrored labial view; B-D, mirrored lingual view; C, symphyseal region in lingual view; D, the same region in ventral view, both display the symphyseal groove; E, posterior fragment of left dentary, GIN 1110/2001-RE12, lingual view; F-M, five maxillae; F-G, left maxilla, GIN 1110/2001-RE26, lingual view; H, right maxilla, GIN 1110/2001-RE39, lingual view; I-J, right, GIN 1110/2001- RE40 and K-L, left maxillae GIN 1110/2001-RE27; I, K, lingual view; J, L, labial view; M, left maxilla, GIN 1110/2001-RE28, labial view; N-P, cervical vertebra, GIN 1110/2001-RE44; N, anterior view; O, left lateral view; P, posterior view; Q, right dentary, unnr. GPIT specimen, lingual view. Abbreviations: dl, dental lamina; ds, dental shelf; fcpr, facial process of maxilla; fMx5, foramina for mandibular division of the fifth cranial (trigeminal) nerve; hfr, haemal foramen; hl, horizontal lamella; lf, lacrimal facet; lg, longitudinal groove; lh, lamina horzontalis; mc, Meckelian canal; na, neural arch; nc, neural canal; nf, nasal facet; pfc, palatine facet; ph, paries horizontalis; prz, prezygapophysis; psz, postzygapophysis; pv, paries verticalis; pxp, premaxillary process; pyp, pterygapophysis; sac, opening of superior alveolar canal; sg, symphyseal groove; sf, splenial facet; tpr, transverse process. Figure 7 Alsophylax sp. from the localities Cherlak (A-P) and Mynsualmas-MSA 3 (Q). Manuscript to be reviewed Figure 8 Lizard and turtle remains from the Western Siberian localities. PeerJ reviewing PDF \| (2016:08:12606:4:0:NEW 23 Jan 2017) PeerJ reviewing PDF \| (2016:08:12606:4:0:NEW 23 Jan 2017) Manuscript to be reviewed PeerJ reviewing PDF \| (2016:08:12606:4:0:NEW 23 Jan 2017) PeerJ reviewing PDF \| (2016:08:12606:4:0:NEW 23 Jan 2017) PeerJ reviewing PDF \| (2016:08:12606:4:0:NEW 23 Jan 2017) PeerJ reviewing PDF \| (2016:08:12606:4:0:NEW 23 Jan 2017) J Manuscript to be reviewed Lizard and turtle remains from the Western Siberian localities. A, Lacerta s.l. sp. 1, left dentary, Pavlodar 1A, GIN 640/5001-RE01, lingual view; B, Lacerta s.l. sp. 2, right dentary, Pavlodar 1A, GIN 640/5001-RE34, lingual view; C-D, Eremias sp., frontal, Pavlodar 2B, GIN 1108/2001-RE01; C, dorsal view; D, ventral views; E, Emydoidea sp., fragment of right hypoplastron, GIN 948/2001-RE01, ventral view; F-G, Emydoidea sp., left femur, GIN 948/2001-RE02; F cranial view; G, ventral view. Scale bars: A, C, D = 2 mm; B = 1 mm; E-G = 1 cm. PeerJ reviewing PDF \| (2016:08:12606:4:0:NEW 23 Jan 2017) PeerJ reviewing PDF \| (2016:08:12606:4:0:NEW 23 Jan 2017) PeerJ reviewing PDF \| (2016:08:12606:4:0:NEW 23 Jan 2017) PeerJ reviewing PDF \| (2016:08:12606:4:0:NEW 23 Jan 2017) Manuscript to be reviewed Figure 9(on next page) Figure 9(on next page) The European (Suppl. 3) and Western Siberian (present study) Neogene fossil record of the studied amphibian groups. Detailed list of the localities see the Additional file 3 and for the family Cryptobranchidae – Böhme et al. (2012; table 1). The occurrences of each group in Europe and Western Siberia are given in the same colour. The Paleogene records of the groups are indicated by arrows. Abbreviations: Hyn, Hynobiidae; Cry, Cryptobranchidae; Prot, Proteidae; Chel, Chelotriton; Tylt, Tylototriton; Bomb, Bombina; red balk, Bombina (cf.) variegata; black balk, Bombina (cf.) bombina; Palbr, Palaeobatrachidae; Pelb, Pelobatidae; Hyla, Hyla; white balk, Hyla (cf.) arborea; Bbuf, Bufo bufo (group); Bvir, Bufotes (cf.) viridis / group of Bufotes viridis; Rana, Rana (cf.) temporaria; Pelx, Pelophylax. PeerJ reviewing PDF \| (2016:08:12606:4:0:NEW 23 Jan 2017) epoch M I O C E N E OLIGOCENE LATE EARLY Bufonidae indet. MIDDLE LATE LATE EARLY EARLY MIDDLE early PLEISTOCENE PLIOCENE 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 Cry Prot Chel Tylt Bomb Palbr Pelb Bbuf Bvir Bvir Rana Pelx Rana Pelx Hyla Hyn Cry Prot Chel Tylt Bomb Palbr Pelb Bbuf Hyla 30.3-33.8 Ma Bombina (cf.) variegata Bombina (cf.) bombina Hyla (cf.) arborea Hyla (cf.) savignyi ? ? PeerJ reviewing PDF \| (2016:08:12606:4:0:NEW 23 Jan 2017) Manuscript to be reviewed Manuscript to be reviewed Manuscript to be reviewed Table 1(on next page) Table 1(on next page) Figure 10(on next page) Figure 10(on next page) Palaeoprecipiation development of Western Siberia including the Zaisan Basin. A, curve displaying the development of the absolute values of mean annual precipitation (MAP); B, the ratio of MAP to recent precipitation value (MAP/ MAPrecent100% ), dashed black line (100 %) indicates the recent precipitation values. Localities: 1, Ayakoz; 2, Vympel; 3, Poltinik; 4, Tri Bogatyrja; 5, Kentyubek; 6, Malyi Kalkaman 2; 7, Malyi Kalkaman 1; 8, Baikadam; 9, Novaya Stanitsa 1A; 10, Cherlak; 11, Detskaya zheleznaya doroga; 12, Olkhovka 1B. PeerJ reviewing PDF \| (2016:08:12606:4:0:NEW 23 Jan 2017) 20 19 18 17 21 22 23 sent day value Middle Miocene Early Miocene Late Miocene Pliocene age error MAP error 6 6 8 8 2-4 2-4 1 1 5 5 12 12 10 11 9 9 9 9 7 7 10 11 5 6 7 12 13 Miocene Middle Late Early Pliocene ~ Serravalian Aquitanian Burdigalian Messinian Zanclean ~ ~ Early ~ ~ PeerJ reviewing PDF \| (2016:08:12606:4:0:NEW 23 Jan 2017) Manuscript to be reviewed Manuscript to be reviewed B MAP/MAPrecent % MAP/MAPrecent % present day value 0 500 1000 1500 A 0 MAP (mm) 0 500 1000 1500 A 0 A MAP (mm) MAP (mm) MAP (mm) Middle Miocene Middle Miocene Early Miocene Late Miocene Pliocene g MAP error Early Miocene 0 Burdigalian Messinian Aquitanian Early Ma 7 12 PeerJ reviewing PDF \| (2016:08:12606:4:0:NEW 23 Jan 201 23 Table 1(on next page) Recent herpetofauna of south-western part of Siberia (Ob and Irtysh River drainages) according to different authors. Recent herpetofauna of south-western part of Siberia (Ob and Irtysh River drainages) according to different authors. References: 1 - Kuzmin (1995) ; 2 - Borkin (1999) ; 3 - Ravkin, Bogomolova & Chesnokova (2010) ; 4 - Amphibiaweb (2016) ; 5 - Ananjeva et al. (2006). Abbreviation: is – insular occurrence. PeerJ reviewing PDF \| (2016:08:12606:4:0:NEW 23 Jan 2017) Manuscript to be reviewed 1 References Taxa 1 2 3 4 5 Salamandrella keyserlingii + + + + Caudata Lissotriton vulgaris is + - + Rana arvalis + + + + Rana amurensis + + + + Rana temporaria + - + + Pelophylax ridibundus is - + is Bufotes viridis is + + is Bufotes variabilis - - - + Bufo bufo + + + + Anura Bufo gargarizans - ? - is Lacerta agilis + Lacertoidea Zootoca vivipara + Elaphe dione + Natrix natrix + Vipera berus + Vipera renardi + Serpentes Gloydius halys + Manuscript to be reviewed PeerJ reviewing PDF \| (2016:08:12606:4:0:NEW 23 Jan 2017) PeerJ reviewing PDF \| (2016:08:12606:4:0:NEW 23 Jan 2017) Table 2(on next page) Table 2(on next page) Neogene testudinoid fauna of Western Siberia and the Zaisan and Turgay Basins. According to Chkhikvadze (1989), as well as the superscriptions before the taxa indicate the references: * Kordikova (1994) ; Danilov, Cherepanov & Vitek (2013) , * Kentyubek fauna (Suppl. 3), - Chkhikvadze (1989) and + Tleuberdina et al. (1993) , K Kuznetsov (1982) , O our results. The aquatic families are indicated with blue and terrestrial families with dark yellow colour. Abbreviations: Ty – Trionychidae, Cy – Chelydridae, Pl – Platysternidae, Ts – Testudinidae, Ey – Emydidae, Gey – Geoemydidae. PeerJ reviewing PDF \| (2016:08:12606:4:0:NEW 23 Jan 2017) Manuscript to be reviewed 1 2 1 2 Zaisan Basin Turgay Basin Western Siberia Stage Suite Turtle “Stage”- taxa taxa taxa Pliocene Chelydropsis kuznetsovi(Cy), ?Sakya sp. (Ey) Karabulak Protestudo illiberalis (Ts) OEmydoidea sp. (Ey) Late Kalmakpai Protestudo kegenica (Ts) KProtestudo karabastusica (Ts) up. Protestudo darewskii (Ts) Sarybulak low. + Chrysemy sp. (Ey), Ocadia sp. (Gey) up. Pelodiscus jakhimovitchae (Ty) Baicalemys moschifera (Ey) Middle Zaisan low. ()Baicalemys sp. (Ey) *Chrysemys sp. (Ey), ?Ocadia sp. (Gey), Emydoidea tasbaka (Ey), Kazakhemys zaisanensis (Pl), ?Chelydropsis sp. (Cy) up. Protestudo sp. (Ts) - Chelydropsis poena (Cy) Pelodiscus sp. (Ty) - Kazakhemys zaisanensis (Pl) Baicalemys jegalloi (Ey) midd. Ocadia iliensis (Gey) -Protestudo alba (Ts) ocene Early Akzhar low. Emydidae gen. indet. (Ey) Zaisan Basin Turgay Basin Western Siberia Stage Suite Turtle “Stage”- taxa taxa taxa Pliocene Chelydropsis kuznetsovi(Cy), ?Sakya sp. (Ey) Karabulak Protestudo illiberalis (Ts) OEmydoidea sp. (Ey) Late Kalmakpai Protestudo kegenica (Ts) KProtestudo karabastusica (Ts) up. Protestudo darewskii (Ts) Sarybulak low. + Chrysemy sp. (Ey), Ocadia sp. (Gey) up. Pelodiscus jakhimovitchae (Ty) Baicalemys moschifera (Ey) Middle Zaisan low. ()Baicalemys sp. (Ey) **Chrysemys sp. (Ey), ?Ocadia sp. (Gey), Emydoidea tasbaka (Ey), Kazakhemys zaisanensis (Pl), ?Chelydropsis sp. (Cy) up. Protestudo sp. (Ts) - Chelydropsis poena (Cy) Pelodiscus sp. (Ty) - Kazakhemys zaisanensis (Pl) Baicalemys jegalloi (Ey) midd. Ocadia iliensis (Gey) *-Protestudo alba (Ts) Miocene Early Akzhar low. Emydidae gen. indet. (Ey) PeerJ reviewing PDF \| (2016:08:12606:4:0:NEW 23 Jan 2017)
https://openalex.org/W4361232897	https://zenodo.org/records/7785338/files/B-108.pdf	English	null	TASK-BASED LANGUAGE LEARNING	Zenodo (CERN European Organization for Nuclear Research)	2,023	cc-by	1,456	TASK-BASED LANGUAGE LEARNING Abstract. Task-based learning is an approach to language learning where learners are given interactive tasks to complete. In order to do this, they need to communicate. Once the task is complete, then the teacher discusses the language used. The learners plan an itinerary for a guest who is coming to stay with their teacher. Keywords: authentic language, adopted task-based language, concept, information-gap activity, reasoning-gap activity, opinion-gap activity. Introduction: Task-based language teaching (TBLT), also known as task-based instruction (TBI), focuses on the use of authentic language to complete meaningful tasks in the target language. Such tasks can include visiting a doctor, conducting an interview, or calling customer service for help. Assessment is primarily based on task outcome (the appropriate completion of real-world tasks) rather than on accuracy of prescribed language forms. This makes TBLT especially popular for developing target language fluency and student confidence. As such, TBLT can be considered a branch of communicative language teaching (CLT). Task-based language learning has its origins in communicative language teaching, and is a subcategory of it. Educators adopted task-based language learning for a variety of reasons. Some moved to a task-based syllabus in an attempt to develop learner capacity to express meaning, while others wanted to make language in the classroom truly communicative, rather than the pseudo- communication that results from classroom activities with no direct connection to real-life situations. Others, like Prabhu in the Bangalore Project, thought that tasks were a way of tapping into learners' natural mechanisms for second-language acquisition, and weren't concerned with real-life communication. TBLT was popularized by N. S. Prabhu while working in Bangalore, India, according to Jeremy Harmer. Prabhu noticed that his students could learn language just as easily with a non- linguistic problem as when they were concentrating on linguistic questions. Major scholars who have done research in this area include Teresa P. Pica, Martin East, and Michael Long. A concept, earlier known as the "communicative activity" in 1970s and 80's was later replaced by the term task has since been defined differently by different scholars. Willis (1996) has defined a task as a goal based activity involving the use of the learners' existing language resources, that leads to the outcome. Examples include playing games, and solving problems and puzzles etc. SCIENCE AND INNOVATION INTERNATIONAL SCIENTIFIC JOURNAL VOLUME 2 ISSUE 3 MARCH 2023 UIF-2022: 8.2 \| ISSN: 2181-3337 \| SCIENTISTS.UZ SCIENCE AND INNOVATION INTERNATIONAL SCIENTIFIC JOURNAL VOLUME 2 ISSUE 3 MARCH 2023 UIF-2022: 8.2 \| ISSN: 2181-3337 \| SCIENTISTS.UZ SCIENCE AND INNOVATION INTERNATIONAL SCIENTIFIC JOURNAL VOLUME 2 ISSUE 3 MARCH 2023 UIF-2022: 8.2 \| ISSN: 2181-3337 \| SCIENTISTS.UZ TASK-BASED LANGUAGE LEARNING Bozorova Madina Makhmudovna Samarkand state architecture and civil engineering university https://doi.org/10.5281/zenodo.7785338 TASK-BASED LANGUAGE LEARNING Ellis (2003) defines a task as a work plan that involves a pragmatic processing of language, using the learners' existing language resources and attention to meaning, and resulting in the completion of an outcome which can be assessed for its communicative function. David Nunan (2004) draws upon the definitions given by other experts, of two types of tasks: target tasks and pedagogical tasks. Targets tasks refer to doing something outside the classroom and in the real world; whereas pedagogical tasks refer to the tasks students perform inside the classroom and in response to target language input or processing. Nunan concludes that target tasks may be non- linguistic. He defines pedagogical task as a classroom activity that involves a student to understand and produce the target language while focusing on conveying the meaning and not being too 494 SCIENCE AND INNOVATION INTERNATIONAL SCIENTIFIC JOURNAL VOLUME 2 ISSUE 3 MARCH 2023 concerned with form. On the other hand, Long (1985) defines a task as things people do in everyday life. concerned with form. On the other hand, Long (1985) defines a task as things people do in everyday life. According to Rod Ellis, a task has four main characteristics: 1. A task involves a primary focus on (pragmatic) meaning. 1. A task involves a primary focus on (pragmatic) meaning. 2. A task has some kind of ‘gap’. (Prabhu identified the three main types as information gap, reasoning gap, and opinion gap.) 2. A task has some kind of ‘gap’. (Prabhu identified the three main types as information gap, reasoning gap, and opinion gap.) 3. The participants choose the linguistic resources needed to complete the task. 4. A task has a clearly defined, non-linguistic outcome. 4. A task has a clearly defined, non-linguistic outcome. The core of the lesson or project is, as the name suggests, the task. Teachers and curriculum developers should bear in mind that any attention to form, i.e., grammar or vocabulary, increases the likelihood that learners may be distracted from the task itself and become preoccupied with detecting and correcting errors and/or looking up language in dictionaries and grammar references. Although there may be several effective frameworks for creating a task-based learning lesson, here is a basic outline: Pre-task: In the pre-task, the teacher will present what will be expected from the students in the task phase. SCIENCE AND INNOVATION INTERNATIONAL SCIENTIFIC JOURNAL VOLUME 2 ISSUE 3 MARCH 2023 UIF-2022: 8.2 \| ISSN: 2181-3337 \| SCIENTISTS.UZ quickest) for a given purpose and within given constraints. The activity necessarily involves comprehending and conveying information, as in an information-gap activity, but the information to be conveyed is not identical with that initially comprehended. There is a piece of reasoning which connects the two. Opinion-gap activity, which involves identifying and articulating a personal preference, feeling, or attitude in response to a given situation. One example is story completion; another is taking part in the discussion of a social issue. The activity may involve using factual information and formulating arguments to justify one's opinion, but there is no objective procedure for demonstrating outcomes as right or wrong, and no reason to expect the same outcome from different individuals or on different occasions. Task-based learning benefits students because it is more student-centered, allows for more meaningful communication, and often provides for practical extra-linguistic skill building. As the tasks are likely to be familiar to the students (e.g.: visiting the doctor), students are more likely to be engaged, which may further motivate them in their language learning. TASK-BASED LANGUAGE LEARNING Additionally, in the "weak" form of TBLT, the teacher may prime the students with key vocabulary or grammatical constructs, although this can mean that the activity is, in effect, more similar to the more traditional present-practice-produce (PPP) paradigm. In "strong" task-based learning lessons, learners are responsible for selecting the appropriate language for any given context themselves. The instructors may also present a model of the task by either doing it themselves or by presenting picture, audio, or video demonstrating the task. Task: During the task phase, the students perform the task, typically in small groups, although this depends on the type of activity. Unless the teacher plays a particular role in the task, the teacher's role is typically limited to one of an observer or counselor—thereby making it a more student-centered methodology. Review: If learners have created tangible linguistic products, e.g. text, montage, presentation, audio or video recording, learners can review each other's work and offer constructive feedback. If a task is set to extend over longer periods of time, e.g. weeks, and includes iterative cycles of constructive activity followed by review, TBLT can be seen as analogous to Project-based learning. According to N. S. Prabhu, there are three main categories of task: information-gap, reasoning-gap, and opinion-gap. Information-gap activity, which involves a transfer of given information from one person to another – or from one form to another, or from one place to another – generally calling for the decoding or encoding of information from or into language. One example is pair work in which each member of the pair has a part of the total information (for example an incomplete picture) and attempts to convey it verbally to the other. Another example is completing a tabular representation with information available in a given piece of text. The activity often involves selection of relevant information as well, and learners may have to meet criteria of completeness and correctness in making the transfer. Reasoning-gap activity, which involves deriving some new information from given information through processes of inference, deduction, practical reasoning, or a perception of relationships or patterns. One example is working out a teacher's timetable on the basis of given class timetables. Another is deciding what course of action is best (for example cheapest or 495 REFERENCES 1. Maxkamova G.T. Konsepsiya formirovaniya mejkulturnoy kompetensii studentov angliyskogo yazika. -Tashkent: fan, 2010. -208 s. 2. Maxkamova G.T., Alimova Sh.S., Ziyayev A.I. Innovative pedagogical technologies in the English language teaching. -T.:2017. 3. Macmillian Dictionary/ 6 th edition/ UK.2016. 4. https://www.teachingenglish.org.uk/article/task-based-speaking 5. https://www.teachingenglish.org.uk/blogs/jane-willis/grammar-based-teaching-task-based- teaching-making-shift
https://openalex.org/W2314681778	http://pdf.blucher.com.br/chemicalengineeringproceedings/cobeq2014/1872-17090-177068.pdf	es		Uso del gas pobre en motores de combustión interna para la generación de electricidad en Cuba.	null	2,015	cc-by	2,541	XX CONGRESSO BRASILEIRO DE ENGENHARIA QUÍMICA Título: Uso del gas pobre en motores de combustión interna para la generación de electricidad en Cuba. Autores: R. Lesme1, L. Garcia1, A. A. Recio1, L. O. Oliva1, J. A. Pajarín1, Yuniel Iglesias2. 1 Universidade de Oriente, Centro de Estudos de Eficiencia Energética, Cuba. 2 Empresa Forestal Integral Gran Piedra Baconao, Cuba. Email: lesme1958@yahoo.es 1. INTRODUCCIÓN. La biomasa ha sido presentada como una alternativa promisoria de energía, no sólo por su potencial mundial, sino porque es una fuente de CO2 neutral, su termo conversión genera bajas emisiones de SO2 y NOX y es la única fuente de energía renovable capaz de sustituir a los combustibles fósiles en todos sus mercados de energía. Entre las fuentes de energía a partir de la biomasa los residuos de los aserraderos son particularmente atractivos económicamente, ellos están concentrados en un determinado depósito y sus costo de transportación es relativamente bajo. Se estima que durante el proceso de elaboración de la madera se obtienen 0.50 m3 de residuos/m3 de madera en bolos (Lesme et al, 2009) El gas pobre, resultado de la gasificación de biomasa con aire tiene un elevado contenido de sustancias volátiles, una alta estabilidad al encendido y puede ser utilizado en motores de combustión interna, el esquema tecnológico se presenta en la figura 1. Figura 1. Esquema general de la tecnología. El tipo de reactor depende fundamentalmente del rango de potencia eléctrica, de la cantidad y características de la biomasa y de la calidad del gas producto. Para la generación de electricidad a pequeña (< 1MW) los más utilizados son los gasificadores downdraft, ellos son capaces de garantizar los requerimientos de calidad del gas a suministrar al motor. 1 Área temática: Engenharia Ambiental e Tecnologias Limpas 1 2. CARACTERÍSTICAS DEL GAS. INFLUENCIA EN EL TRABAJO DEL MOTOR. Las características del gas que más influyen en el trabajo del motor son: Composición, densidad energética de la mezcla aire-gas pobre, número de metano y velocidad de la llama. 2.1. Composición del gas pobre. En un gasificador, la biomasa puede ser convertida en un gas compuesto por H2, CO, CO2, H2O, CH4, alquitrán y particulados. La composición exacta depende de las características de la biomasa, de las condiciones de operación y el tipo de reactor. (Martinez et al, 2011) 2.2. Densidad energética Tomando en cuenta que los motores a gas trabajan aproximadamente en condiciones estequiométricas de combustión, la densidad energética de la mezcla puede ser determinada por la siguiente expresión: (Sridhar et al 2001). HVm      1  2.38Y   2.38Y   9.53Y  10768 YH 2  12696YCO   35866 YCH4 H2 CO 1 CH4 Donde: HVm. Densidad energética de la mezcla aire-gas pobre, kJ/Nm3 YH2, YCO, YCH4. Concentraciones de hidrógeno, monóxido de carbono y metano en el gas. 2.3. Número de metano. El número de metano es un indicador de los combustibles gaseosos que mide su tendencia a la detonación, mientras mayor sea su valor, menor la tendencia. Existe una escala denominada número de metano (NM) que utiliza como referencias el metano puro (NM = 100) y el hidrógeno (NM = 0) que valora la aplicabilidad del gas en los motores, considerándose un valor entre 75-80 como satisfactorio (Przybyla G et al, 2010) 2.4. Velocidad de la llama y ángulo de avance. La velocidad de la llama es un parámetro que define las características de oxidación de un combustible, su valor dependen de su composición química, de la cantidad de aire empleada para su combustión, así como, las condiciones de presión y temperatura de la mezcla. Es posible el uso de un combustible en motores de combustión interna cuando la velocidad de la llama está por encima de 8 cm/s en condiciones estequiométricas de combustión. El ángulo de avance de la ignición está íntimamente ligado a la velocidad de la llama, el mismo deberá ser atrasado cuando la mezcla presente una elevada velocidad, con el fin de alcanzar mayores potencias y eficiencias. (J et al, 2006) En la tabla 1 se presentan la densidad energética, el número de metano y velocidad de la llama del gas pobre, como puede observarse de acuerdo a los valores del número de metano y velocidad de la llama su empleo en motores resulta satisfactorio. 2 Área temática: Engenharia Ambiental e Tecnologias Limpas 2 Tabla 1. Densidad energética, número de metano y velocidad de la llama del gas natural y el gas pobre. (Muñoz, 2000) Indicadores Gas Natural Gas Pobre Densidad Energética de la mezcla, (MJ/Nm3) 3.32 2.20 Número de metano 105 80 Velocidad de la llama, cm/s 35 50 3. ESTADO DEL ARTE DE LOS MOTORES DE COMBUSTIÓN INTERNA ALIMENTADOS CON GAS POBRE. Las aplicaciones comerciales de la tecnología se localizan en un rango de potencia menor de 1 MW. Los motores utilizados pueden ser de ignición por chispa y Diesel. En los motores de ignición por chispa el gas pobre puede sustituir el 100% del combustible. La principal modificación a realizar consiste en el atraso del tiempo de ignición a fin de lograr que la presión máxima de combustión se alcance entre 10 y 12 grados DPMS y con ello mejores desempeños del motor (Sridhar et al, 2001) Los motores Diesel operan a mayores relaciones de compresión, su operación con gas pobre debe ser dual (sustitución de hasta el 85% del Diesel). La ventaja radica en su flexibilidad, en caso de un mal funcionamiento del gasificador es posible un cambio inmediato de combustible, operando el motor totalmente con Diesel. Cuando se quiere utilizar un motor Diesel con gas pobre (100%) hay que implementar un mezclador y un sistema de ignición por chispa y modificarse el ángulo de avance de la ignición (Ramadhas et al 2008) En los últimos años varios trabajos se han reportado sobre la evaluación del desempeño de los motores accionados con gas pobre, en la tabla 2 se ofrecen algunos resultados. Tabla 2- Motores de combustión interna trabajando con gas pobre. Motor rpm RC Gas. Mod. (%) Pot PI Pp ηT ηG (kW) (PMS) (%) (%) (%) Ref. Diesel 1500 11.5:1 100 SIG 12-16(1) 35 nd 28-32 21-24(2) Shashhika ntha,et al,1993 Diesel 1500 17:1 100 SIG 2.3(3) 10 20 19.05(4) nd Ramachan dra, 1993 Diesel nd nd 100 SIG 15-20(1) nd 20 28(4) 25(2) Henriskse n et al, 2006 Otto 1800 12:1 100 MGN 40.7(4) 30 37.38 nd nd Muraro, 2006 Diesel 1500 18.5:1 81 No 11.44(3) nd 21 14.7(5) 11.69(2) Bhattachat ya et al, 2001 Diesel 1500 17:1 100 SIG 17.5(1) 6 16.7(6) nd 16.6(7) Shridha 3 Área temática: Engenharia Ambiental e Tecnologias Limpas 3 Gr, 2001 Diesel 1500 17.5:1 65 No 60 SIG 4(3) 27 20 27 nd 24 nd Ahrefeldt et al, 2000 Diesel 1800 12.5:1 100 SIG 5.3(1) 15 47 nd nd Viera R et al, 2011 Gás 1500 12:1 100 No 75 28 14 20.7 - Raman P et al, 2013 (1) Potencia eléctrica, nd. No declarado, (2) Biomasa-electricidad, (3) Potencia al freno, (4) En el eje, (5) Motor-generador, (6) Potencia mecánica, (7) Biomasa-salida en el eje, RC. Relación de Compresión, Mod. Modificaciones en el motor, SIG. Sistema de ignición, PI. Punto de ignición, APMS. Antes del punto muerto superior, η T. Eficiencia Térmica, ηG. Eficiencia Global. 4. EVALUACIÓN DE UN MOTOR ASHOK LEYLAND TRABAJANDO CON GAS POBRE. MATERIALES Y MÉTODOS. 4.1. Equipamiento. El gas es producido en un reactor downdraft (Figura 2), modelo COMBO-80, (Owner’s Manual, 2009) Figura 2- Gasificador downdraft El gas se somete a un proceso de limpieza en un sistema compuesto por un filtro grueso, un filtro fino activo y otro pasivo, y un filtro de seguridad (Figura 3.) 4 Área temática: Engenharia Ambiental e Tecnologias Limpas 4 Figura 3- Sistema de limpieza seca. La unidad generadora está formada por un motor y un generador eléctrico. El generador, directamente acoplado al motor, es trifásico, gira a 1800 rpm de 220 V y 60 Hz. (Figura 4) Figura 4. Unidad generadora de electricidad. El motor Diesel modelo Leyland fue modificado para trabajar con gas pobre mediante la implementación de un mezclador tipo Venturi y un sistema de encendido por chispa eléctrica, sus características técnicas aparecen en la “Tabla 3” (Leyland P76, 1973) Tabla 3- Características técnicas del motor. Modelo Revoluciones No. de Cilindros Cilindrada Diámetro del pistón Carrera Potencia Nominal Relación de compresión Leyland 1800 6 2623 cm3 76,2 mm 95.7 mm 90.2 kW (Diesel) 9:1 5 Área temática: Engenharia Ambiental e Tecnologias Limpas 5 4.2. Técnica Experimental. El motor se evaluó para tres condiciones de trabajo, en cada caso se midió la carga eléctrica y el flujo de gas suministrado al motor. La carga eléctrica a través de un Watímetro situado en el panel de control y el flujo de gases con un tubo Venturi (Mgases) situado a la entrada del motor. La composición del gas y el consumo de biomasa del gasificador (Mbio) se obtuvieron mediante un modelo de equilibrio termodinámico no estequiométrico del gasificador. Los resultados de las pruebas y el modelo se recogen en la ¨Tabla 4” ( Trabajo de Grado, 2013) Tabla 4. Mediciones experimentales y resultados del modelo termoquímico. No 1 2 3 No 1 2 3 No 1 2 3 Potencia Eléctrica (Pele) = 5 kW Composición del gas, (%v) Mgases (kg/h) CO CO2 CH4 N2 H2 21 13 3 46 16 56.32 Potencia Eléctrica (Pele) = 20 kW Composición del gas, (%v) Mgases (kg/h) CO CO2 CH4 N2 H2 20 12 2 46 18 7.7.33 Potencia Eléctrica (Pele) = 40 kW Composición del gas, (%v) Mgases (kg/h) CO CO2 CH4 N2 H2 19 10 2 50 18 155.03 Mbio (kg/h) 29.48 Mbio (kg/h) 30.28 Mbio (kg/h) 60.00 5. RESULTADOS. 5.1. Eficiencia térmica del motor. Evalúa el proceso de transformación de energía de la mezcla combustible en electricidad. T  Potencia Eléctrica HVM xM mezcla 2 Donde: Mmezcla. Flujo volumétrico de la mezcla, Nm3/s El flujo volumétrico de la mezcla se determinó por la expresión (3) (Trabajo de Grado, 2013) M mezcla  M gases  M aire 3 La masa de aire (Maire) se determinó través de la expresión general de la combustión del gas pobre, teniendo en cuenta su composición y los resultados de balance molares de cada uno de los elementos que lo conforman. (Trabajo de grado, 2013) 6 Área temática: Engenharia Ambiental e Tecnologias Limpas 6 5.2. Eficiencia global de la instalación. Evalúa el proceso de transformación de la energía de la biomasa en electricidad (Ec. 4) G  Potencia Eléctrica M bio xPCI bio 4 Donde PCIbio. Poder calórico bajo de la biomasa, kJ/kg Los resultados de los cálculos se reportan en la “Tabla 5” Tabla 5. Eficiencia del motor y la instalación. Pele (kW) 5 20 40 HVm ( MJ/Nm3) 2.05 2.45 2.40 Mmezcla ( Nm3/s) 0.024 0.033 0.065 ηT ( %) 10.2 24.0 25.7 ηG ( %) 3.6 14.0 14.2 Los resultados muestran una eficiencia térmica del motor entre 10,2 y 25,7 % manteniéndose de forma estable en un amplio rango de carga (20-40 kW) 6. CONCLUSIONES 1. El gas pobre presenta propiedades combustibles inferiores a la del gas natural (densidad energética y número de metano), no obstante puede ser utilizado en motores de combustión interna de forma satisfactoria, esperándose una reducción de la potencia del motor con respecto a los combustibles tradicionales proporcional a la disminución de su densidad energética. En el caso del motor Leyland P76, la reducción fue de 42% con respecto al Diesel. 2. La mayor parte de los motores utilizados para trabajar con gas pobre son Diesel a los cuales se le implementa un mezclador, un sistema de ignición por chispa y se modifica el ángulo de avance. 3. La evaluación del desempeño de los motores de combustión interna accionados con gas pobre para la generación de electricidad a pequeña escala, muestran que ellos trabajan con una eficiencia térmica en un rango de 20-32% y una eficiencia global entre 12-25%. En el caso del motor Leyland P76 que trabaja con gas pobre producto de la gasificación de residuos de la industria forestal posee una eficiencia térmica de 25.7 % y una eficiencia global de 14.2 %. 7. BIBLIOGRAFÍA. 1. Ahrenfeldt, J., Henriksen, U., Schramm, J. Experimental on Wood Gas Engines. The Technical University of Denmark. Department of Energy Engineering, 40 p. Año 2000 7 Área temática: Engenharia Ambiental e Tecnologias Limpas 7 2. Bhattachatya, S. C., Hla, S. S., Pham, H. L. “A Study on a Multi- Stage Hybrid Gasifier-Engine System”, Biomass and Bioenergy, vol. 21, pp. 445-460. Año 2001 3. Henriksen, U., Ahrenfeldt, J., Jensen, T, Gobel, B, Bentzen, J, Hindsgaul, C, Sorensen, L.H. “The Design, Construction and Operation of a 75 kW Two-Stage Gasifier”, Energy, vol. 31, pp. 1542–1553. Año 2006 4. Lesme Jaén R, Oliva Ruiz L. Potencial energético de los residuos de la industria de la madera en el aserrío “El Brujo” de la Empresa Gran Piedra Baconao. The 8th latinamerican congress on electricity generation and transmission - CLAGTEE 2009 5. Leyland P76, 1973 especificaciones técnicas 257884 6. Martínez D. J, Viera Andrade R, Silva Lora E, Lesme Jaén R. Experimental study on biomass gasification in a double air stage downdraft reactor. Revista: Biomass and Bioenergy XXX. Pag.1-17. Año 2011. 7. Muñoz M., Moreno, F., Morea-Roy, J., Ruiz J., Arauzo, J. “Low Heating Value Gas on Spark Ignition Engines”, Biomass and Bioenergy, vol. 18, pp. 431-439. Año 2000. 8. Muraro, W. Avaliação do Funcionamento de Motor ICE com Gás de Baixo Poder Calorífico Proveniente da Gaseificação de Casca de Arroz, Dissertação de Mestrado, Universidade Estadual de Campinas, UNICAMP, 86p. Año 2006 9. Owner’s Manual. Ankur Scientific Energy Technologies Pvt. Ltd. ISO 9001 / ISO 14001 / BS OHSAS 18001 Certified Company. Año 2009. 10. Ramadhas A, Jayaraj, Muraleedharan C. Dual fuel mode operation in diesel engines using renewable fuels: Rubber seed oil and coir-pith producer gas. Renewable Energy No. 33 PP. 2077-2083. Año 2008. 11. Ramachandra, A. “Performance Studies on a Wood Gas Run IC Engine”, Recent Advances in Biomass Gasification and Combustion. Proceedings of Fourth National Meet on Biomass Gasification and Combustion, pp. 213-218. Año 1993 12. Raman P, N.K. Ram N, Gupta R. A dual fired downdraft gasifier system to produce cleaner gas for power generation: Design, development and performance analysis. Energy 54 (2013) 302-314. 13. Shashikantha, B, Khairnar P, Kamat, Parikh P. Development and Performance Analysis of a 15 kWe Producer Gas Operated SI Engine. Recent Advances in Biomass Gasification and Combustion. Proceedings of Fourth National Meet on Biomass Gasification and Combustion,. pp. 219-231. Año 1993 14. Sridhar G, Paul J, Mukunda H. Biomass derived producer gas as a reciprocating engine fuel—an experimental analysis. Biomass and Bioenergy No. 21 Año (2001) PP. 61–72. 15. Trabajo de Grado. Evaluación teórico experimental de un gasificador downdraft modelo COMBO-80. Centro de Estudios de Eficiencia Energética. Universidad de Oriente. Santiago de Cuba. Año 2013. 16. Trabajo de Grado. Evaluación Termodinámica del Motor Ashok Leyland LTD alimentado con gas pobre. Centro de Estudios de Eficiencia energética. Universidad de Oriente. Santiago de Cuba. Año 2013. 17. Vieira R, Silva E, Lesme R. Análise da operação de um gaseificador co-corrente de duplo estágio acoplado a um motor de combustão interna de ignição por centelha. Congreso Iberoamericano de Ingeniería Mecánica. CIBIM 2011. Porto Potugal. Año 2011 8 Área temática: Engenharia Ambiental e Tecnologias Limpas 8
https://openalex.org/W2976684639	https://peerj.com/articles/7762.pdf	English	null	Simulated manned Mars exploration: effects of dietary and diurnal cycle variations on the gut microbiome of crew members in a controlled ecological life support system	PeerJ	2,019	cc-by	9,007	How to cite this article Dong H-S, Chen P, Yu Y-B, Zang P, Wei Z. 2019. Simulated manned Mars exploration: effects of dietary and diurnal cycle variations on the gut microbiome of crew members in a controlled ecological life support system. PeerJ 7:e7762 DOI 10.7717/peerj.7762 Subjects Biodiversity, Bioinformatics, Ecology, Genomics, Microbiology Keywords Manned Mars exploration, 16S gene sequencing, Controlled Ecological Life Support System Environment, Alpha diversity, Core microbiome, Longitudinal variation, Biomarker screening, Crewmembers, 25-hydroxyvitamin D Simulated manned Mars exploration: effects of dietary and diurnal cycle variations on the gut microbiome of crew members in a controlled ecological life support system 2 State Key Lab of Space Medicine Fundamentals and Application, Key Laboratory of Space Nutrition and Food Engineering, China Astronaut Research and Training Center, Beijing, China 3 SPACEnter Space Science and Technology Institute, Shenzhen, China INTRODUCTION Long-term manned spaceﬂight and the development of extraterrestrial planet settlement measures are of great interest in aerospace technology. An effective approach to realize this goal is the establishment of a controlled ecological life support system (CELSS) that provides all the essential living requirements, such as food, oxygen, and water through continuous regeneration (Nelson, Dempster & Allen, 2008). The CELSS is also termed as a bio-regenerative life support system, biological life system, or third-generation life support system, and is a closed micro-ecological circulation environment that is artiﬁcially constructed based on the characteristics of the space environment (Guo et al., 2008). In this system, plants provide food and oxygen for heterotrophic organisms (humans and animals) through photosynthesis, and convert carbon dioxide and other waste from heterotrophic organisms into useful products, constituting the carbon and oxygen cycles of the system (MacElroy & Bredt, 1984). Plants participate in the system’s water puriﬁcation cycle through root absorption and leaf transpiration. Microbes degrade, mineralize, and regenerate inedible parts of plants, occupant excrement, and domestic wastewater; providing water and nutrients for plants and food for the inhabitants. This completes the food recycling process and creates a closed-circuit ecosystem that is continually renewed by plants, animals, microorganisms, humans, and the necessary organic and inorganic environment (Degermendzhi & Tikhomirov, 2014). In current short or medium-term manned space activities, aerospace foods are periodically transported from the ground to space stations through a cargo spaceship, which is technically and economically feasible. However, for the future establishment of long-term manned space missions, including Lunar or Mars bases, the existing supply mode of food systems will shift from on-earth supplies to biologically regenerated food systems in situ (Meinen et al., 2018). CELSS is an important environmental model for long-term manned space station missions and human activities at extra-territorial bases. Studies have shown that the frequency of health problems increase in a closed environment, including physical and psychological issues (Stein, 2001). For manned space ﬂight studies, the major drawbacks include limited spacecraft space, which makes large scale studies on numerous individual subjects unfeasible (Lang et al., 2017). Repeated measurements of individuals during longitude research, particularly for twin samples, can serve as an efﬁcient study method (Garrett-Bakelman et al., 2019). ABSTRACT Background: Changes in gut microbiome are closely related to dietary and environment variations, and diurnal circle interventions impact on human metabolism and the microbiome. Changes in human gut microbiome and serum biochemical parameters during long-term isolation in a controlled ecological life support system (CELSS) are of great signiﬁcance for maintaining the health of crewmembers. The Green Star 180 project performed an integrated study involving a four-person, 180-day duration assessment in a CELSS, during which variations in gut microbiome and the concentration of serum 25-hydroxyvitamin D, a-tocopherol, retinol and folic acid from the crewmembers were determined. Results: Energy intake and body mass index decreased during the experiment. A trade-off between Firmicutes and Bacteroidetes during the study period was observed. Dynamic variations in the two dominant genus Bacteroides and Prevotella indicated a variation of enterotypes. Both the evenness and richness of the fecal microbiome decreased during the isolation in the CELSS. Transition of diurnal circle from Earth to Mars increased the abundance of Fusobacteria phylum and decreased alpha diversity of the fecal microbiome. The levels of serum 25-hydroxyvitamin D in the CELSS were signiﬁcantly lower than those outside the CELSS. Submitted 8 March 2019 Accepted 26 August 2019 Published 26 September 2019 Corresponding author Hai-Sheng Dong, donghs16@mails.tsinghua.edu.cn Academic editor Joseph Gillespie Additional Information and Declarations can be found on page 13 DOI 10.7717/peerj.7762 Copyright 2019 Dong et al. Distributed under C ti C CC BY 4 0 Conclusions: The unique isolation process in the CELSS led to a loss of alpha diversity and a transition of enterotypes between Bacteroides and Prevotella. Attention should therefore be paid to the transition of the diurnal circle and its effects on the gut microbiome during manned Mars explorations. In particular, serum 25-hydroxyvitamin D levels require monitoring under artiﬁcial light environments and during long-term space ﬂight. Large-scale studies are required to further consolidate our ﬁndings. Subjects Biodiversity, Bioinformatics, Ecology, Genomics, Microbiology Keywords Manned Mars exploration, 16S gene sequencing, Controlled Ecological Life Support System Environment, Alpha diversity, Core microbiome, Longitudinal variation, Biomarker screening, Crewmembers, 25-hydroxyvitamin D Subjects Biodiversity, Bioinformatics, Ecology, Genomics, Microbiology Keywords Manned Mars exploration, 16S gene sequencing, Controlled Ecological Life Support System Environment, Alpha diversity, Core microbiome, Longitudinal variation, Biomarker screening, Crewmembers, 25-hydroxyvitamin D Distributed under Creative Commons CC-BY 4.0 Dong et al. (2019), PeerJ, DOI 10.7717/peerj.7762 INTRODUCTION Studies investigating the temporal dynamics of the gut microbiome in six individuals sharing a conﬁned environment during a 520-day ground-based space simulation indicated the importance of maintaining a health-promoting, mutualistic microbiome environment (Mardanov et al., 2013; Turroni et al., 2017). Gut microorganisms are the largest micro-ecological systems in the human body d h d f h ( kb l ) Th l Gut microorganisms are the largest micro-ecological systems in the human body encoding ≥100 times the coding genes of humans (Eckburg et al., 2005). There are at least 160 dominant microorganisms in each individual, more than 90% of which belong to Bacteroidetes and Firmicutes (Shafquat et al., 2014). In recent years, the continuous development and application of new technologies such as high-throughput sequencing and metagenomics have revealed new perspectives for understanding human life activities Dong et al. (2019), PeerJ, DOI 10.7717/peerj.7762 2/17 2/17 (Yibin, 2014; Yu et al., 2017). Gut microorganisms profoundly affect human health by participating in nutrient metabolism and immune system development, and have emerged as new targets for precision medical interventions (Botero et al., 2016; Saraf et al., 2017). Changes in the internal and external environment of the human body can affect the species, structure, and function of the gut microbiome (He et al., 2015; Lin et al., 2015). Through constant adaptation and regulation, the ﬂora, host, and environment together maintain physiological balance and health (Botero et al., 2016; He et al., 2015). The lower gastrointestinal tract of humans hosts trillions of diverse microbial communities, each with unique and extensive metabolic capacities, producing a variety of nutrients including B vitamins, and vitamin K (LeBlanc et al., 2013). The gut microbiome also contribute to human metabolism and digestion which have profound effects on the pathogenesis of metabolic syndrome (Ramakrishna, 2013). The low levels of natural light in a manned spacecraft has negative effects on serum vitamin D levels, whilst changes in the intestinal microbiome inﬂuence vitamin D metabolism (Bora et al., 2018). Furthermore, evidence suggests that consistent long-term physical and psychological stress damage in an isolated weightlessness environment leads to decreased serum vitamin E and vitamin A levels, and a loss of antioxidants (Smith et al., 2005). Meanwhile, variations in normal human daily behaviors (wakefulness) and rest (sleep) rhythmicity or diurnal rhythmicity have been reported to impact metabolism, the microbiome and circadian clocks (Nobs, Tuganbaev & Elinav, 2019; Reinke & Asher, 2019). INTRODUCTION Since the dynamic changes of the microbiome inﬂuences serum biochemistry for crewmembers living in CELSS, studies have focused on their responses during simulated Mars exploration missions for a 180-day period. In this study, we hypothesized that during the 180 days in the CELSS, dietary changes, the isolated environment and scheduled diurnal cycles may lead to ﬂuctuations in the gut microbiome and speciﬁc serum vitamins. We performed 16S gene sequencing technology to analyze longitudinal changes in the human gut microbiome during 180-day long-term residence in CELSS. Blood samples were collected for the analysis of serum 25-hydroxyvitamin D, a-tocopherol, retinol and folate. The experiments were designed to simulate changes in the intestinal microbiome structure during the long-term residency of astronauts in the controlled environment of extraterrestrial planets. Our ﬁndings provide a basis for promoting the intestinal and nutritional health of astronauts during long-term space ﬂight missions. Dong et al. (2019), PeerJ, DOI 10.7717/peerj.7762 MATERIALS AND METHODS To simulate the manned Mars exploration mission, both Earth (24 h) and Martian diurnal cycles (MDCs) (24 h 40 min) were employed. The Earth diurnal cycle (EDC) was adopted after entering the CELSS from day 1 to day 72 and day 110 to day 180, whilst days 73 to 109 used the MDC through time information control, scheduled light exposure (150 Lx white light LED) and the adjustment of waking and rest rhythmicity. Fecal samples collected in sterile sample collection tubes are shown in Fig. 1. The prescribed time points were as follows: 30 days before entering the CELSS, 15 days before entering the CELSS, on days 2d, 15d, 30d, 45d, 60d, 75d, 90d, 105d, 120d, 135d, 150d, 165d and 175d after entering the CELSS. All the fecal samples were collected in the morning into clean stool collection tubes at scheduled time points, sealed and transferred to a −80 C freezers prior to analysis. affecting the gut microbiome, and were compatible in terms of dietary habits. No antibiotics or probiotics was used within 3 months of the trial. Various activities were scheduled for the crewmembers such as system monitoring and recording, biochemistry experimental projects, crop cultivation and harvesting, solid waste incineration, equipment operating and maintenance and physical training. Crewmembers were familiar with the trial process and mastered each professional technique. The CELSS consisting of eight modules: crew module I, crew module II, the CELSS module, the resource module, plant module I, plant module II, plant module III, and plant module IV, in which speciﬁc crops were planted are shown in Table S1. The living environment in the CELSS was suitable for human physiological requirements with a temperature 23–25 C, humidity 46–77% R.H., an atmospheric pressure of 81.3–104.3 kPa, O2 18–26% and CO2 ≤0.8%. The foodstuffs provided were served as prepackaged food outside the CELSS and freshly prepared foods supplied by the self-producing plants (including ﬁve crops, 16 vegetable plants, and two fruit plants) grown in the CELSS. Recipes for the four crewmembers during their 180-day residence in the CELSS are shown in Table S2. The shelf life of the space foods, ingredients and seasoning provided covered the task duration. The dietary design for the crewmembers was in accordance with the customs of the Chinese population including three meals a day, namely breakfast, lunch and supper. p p g y y pp Complementary foods were provided as snacks or refreshments. Dong et al. (2019), PeerJ, DOI 10.7717/peerj.7762 MATERIALS AND METHODS The main compositions of the diet included carbohydrates, protein, fat, and dietary ﬁber which were measured through the calculation of nutrient composition in the edible portion of speciﬁc food items through reference to the authoritative handbook: China Food Composition recommended by the Chinese Nutrition Society (Yang & Pan, 2009). The speciﬁc calculation strategy was in accordance with recently published references (Hao et al., 2018; Kirchhoff, 2002). To simulate the manned Mars exploration mission, both Earth (24 h) and Martian diurnal cycles (MDCs) (24 h 40 min) were employed. The Earth diurnal cycle (EDC) was adopted after entering the CELSS from day 1 to day 72 and day 110 to day 180, whilst days 73 to 109 used the MDC through time information control, scheduled light exposure (150 Lx white light LED) and the adjustment of waking and rest rhythmicity. Fecal samples collected in sterile sample collection tubes are shown in Fig. 1. The prescribed time points were as follows: 30 days before entering the CELSS, 15 days before entering the CELSS, on days 2d, 15d, 30d, 45d, 60d, 75d, 90d, 105d, 120d, 135d, 150d, 165d and 175d after entering the CELSS. All the fecal samples were collected in the morning into clean stool collection tubes at scheduled time points, sealed and transferred to a −80 C freezers prior to analysis. MATERIALS AND METHODS MATERIALS AND METHODS Four volunteers (three males and one female, aged 26–36 years) that passed rigorous health examinations and had comparable body mass indexes similar to those of astronauts (BMI = 18.6–24.6 kg/m2) were screened as crewmembers for the Green Star 180 project (GS-180). The Ethics Committee of the Astronaut Research and Training Center of China approved the project (ethics approval number: Szsisc EA201601). Crewmembers were healthy with no digestive system disease, alcohol abuse, diabetes or other problems Dong et al. (2019), PeerJ, DOI 10.7717/peerj.7762 affecting the gut microbiome, and were compatible in terms of dietary habits. No antibiotics or probiotics was used within 3 months of the trial. Various activities were scheduled for the crewmembers such as system monitoring and recording, biochemistry experimental projects, crop cultivation and harvesting, solid waste incineration, equipment operating and maintenance and physical training. Crewmembers were familiar with the trial process and mastered each professional technique. The CELSS consisting of eight modules: crew module I, crew module II, the CELSS module, the resource module, plant module I, plant module II, plant module III, and plant module IV, in which speciﬁc crops were planted are shown in Table S1. The living environment in the CELSS was suitable for human physiological requirements with a temperature 23–25 C, humidity 46–77% R.H., an atmospheric pressure of 81.3–104.3 kPa, O2 18–26% and CO2 ≤0.8%. The foodstuffs provided were served as prepackaged food outside the CELSS and freshly prepared foods supplied by the self-producing plants (including ﬁve crops, 16 vegetable plants, and two fruit plants) grown in the CELSS. Recipes for the four crewmembers during their 180-day residence in the CELSS are shown in Table S2. The shelf life of the space foods, ingredients and seasoning provided covered the task duration. The dietary design for the crewmembers was in accordance with the customs of the Chinese population including three meals a day, namely breakfast, lunch and supper. Complementary foods were provided as snacks or refreshments. The main compositions of the diet included carbohydrates, protein, fat, and dietary ﬁber which were measured through the calculation of nutrient composition in the edible portion of speciﬁc food items through reference to the authoritative handbook: China Food Composition recommended by the Chinese Nutrition Society (Yang & Pan, 2009). The speciﬁc calculation strategy was in accordance with recently published references (Hao et al., 2018; Kirchhoff, 2002). Microbial DNA extraction, Illumina MiSeq sequencing and data analysis After thawing at 4 C, DNA was extracted from the fecal samples using the cetyl trimethylammonium bromide method (Hamilton et al., 2013; Wu et al., 2016) to remove impurities including proteins and polysaccharides. Total DNA was extracted from 150 mg of fecal samples using commercially available DNA stool MiniKits (Qiagen, Hilden, Germany) and stored at −80 C until analysis. Forward PCR Dong et al. (2019), PeerJ, DOI 10.7717/peerj.7762 Figure 1 Scheduled fecal and blood sampling time points. MDC, Mars diurnal circle; EDC, Earth diurnal circle. FS1–FS15 refers to fecal sampling time points. BS1–BS9 refers to blood sampling time points. Full-size  DOI: 10.7717/peerj.7762/ﬁg-1 Figure 1 Scheduled fecal and blood sampling time points. MDC, Mars diurnal circle; EDC, Earth diurnal circle. FS1–FS15 refers to fecal sampling time points. BS1–BS9 refers to blood sampling time points. Full-size  DOI: 10.7717/peerj.7762/ﬁg-1 primers: 338F: 5′ACTCCTACGGGAGGCAGCAG3′; and reverse primers were 806R: 5′GGACTACHVGGGTWTCTAAT3′ from the V3–V4 conserved region of 16S rDNA. PCR reactions to the target regions were performed using the 96 well AB9902 PCR system in 50 mL reactions: 25 mL KOD FX Neo Buf, one mL KOD FX Neo, 1.5 mL of forward and reverse primers (10 mM), 10 mL dNTP (two mM), ﬁve nL (40–60 ng) template DNA, up to 50 mL in ddH2O. Reaction sequence conditions were as follows: initial denaturation at 95 C for 5 min, 15 cycles of denaturation at 95 C for 1 min, annealing at 50 C for 1 min, and elongation at 72 C for 1 min. Final elongation at 72 C for 7 min. PCR products were sequenced using paired-end sequencing platforms on Illumina Hiseq 2500. Raw sequence data obtained were stored in the NCBI online public database with accession number PRJNA557020. Raw tag acquisition: sequence reads were merged using FLASH v1.2 software with a minimum overlap length of 10 bp and a maximum mismatch ratio of 0.2 to obtain Raw Tags. Clean tag acquisition: clean tags were obtained by ﬁltering merged raw tags with Trimmomatic v0.33 software (threshold: 50 bp). Effective tag acquisition: chimera sequences were removed using UCHIME v2.4 software. Operational taxonomy units (OTU) analysis: OTU refer to clusters of (uncultivated or unknown) microbial organisms at different taxonomic levels grouped by sequence similarities in speciﬁc marker genes. OTU categorization was performed at thresholds ≥97% using USEARCH software version 10.0. Dong et al. (2019), PeerJ, DOI 10.7717/peerj.7762 Microbial DNA extraction, Illumina MiSeq sequencing and data analysis Taxonomic annotation of the OTUs was accomplished by matching all OTU tags to the SILVA (128) database (http://www.arb-silva.de). Dong et al. (2019), PeerJ, DOI 10.7717/peerj.7762 5/17 Determination of serum 25-hydroxyvitamin D, α-tocopherol, retinol and folate All crewmembers were asked to collect blood samples according to prescribed time points: 30 days before entry, on days 2d, 30d, 60d, 90d, 120d, 150d, and 180d, and 30 days after exiting the CELSS. All blood samples were collected in the morning using vacuum blood collection tubes, and centrifuged at 4,000×g for 15 min (4 C). Supernatants were transferred to a −80 C freezer for further analysis. The determination of serum 25-hydroxyvitamin D, a-tocopherol and retinol was performed using conventional High Performance Liquid Chromatography (HPLC) quantitative analysis methods (Rock et al., 1999). Serum folate determinations were performed through competitive chemiluminescent enzyme immunoassay methods (Cumurcu, Sahin & Aydin, 2006; Kullak-Ublick et al., 2016). Dong et al. (2019), PeerJ, DOI 10.7717/peerj.7762 Statistical analysis Data analysis and visual explorations were performed using the online statistical analysis platform Calypso Version 8.84 (Zakrzewski et al., 2016) and GraphPad Prism 5.0. Alpha diversity was performed by straight observational diversity metrics as evenness and richness. Multivariate data visualization and repeated measurement ANOVA assessments were used to measure the accumulation factor of the environment on crewmembers with p-value cutoff 0.05. Biomarker screening Repeated ANOVA analysis was used to explore fecal microbiome at the phylum and genus level (p < 0.05). Results are shown in Table 1. The level of Fusobacteria, Paraprevotella, Phascolarctobacterium, Pseudobutyrivibrio, Bacteroides and Collinsella when the EDC schedule was performed were signiﬁcantly lower than those of the MDC schedule. The levels of Bacteroidetes, Firmicutes, Prevotella, Alistipes, Megamonas and Sutterella when the EDC schedule was performed were signiﬁcantly higher than those of when the MDC schedule was performed. Diversity analysis Alpha diversity of microbiomes reﬂects the general richness and evenness of the species communities in the habitats. As shown in Fig. 3 and Fig. S5, compared to the fecal microbiome outside the system, both evenness and richness showed signiﬁcant decreases after CELSS admission (p < 0.01). Meanwhile, the richness of the fecal microbiome decreased when the crewmembers performed MDC schedules, which recovered to their original status when the EDC schedule was resumed. Longitudinal variations of gut microbiome S1 refer to data from fecal samples for crewmember 0# 30 days before entering the CELSS, 15 days before entering the CELSS, and 2d, 15d, 30d, 45d, 60d, 75d, 90d, 105d, 120d, 135d, 150d, 165d and 175d after entering the CELSS. MDC, Mars diurnal cycle; EDC, Earth diurnal cycle; OC, period when crewmembers were out of the CELSS (before entry); IC, period when the crewmembers were in the CELSS. Full-size  DOI: 10.7717/peerj.7762/ﬁg-2 Figure 2 Relative abundance variation proﬁling of fecal microbiota. (A) At phylum level. (B) At genus level. Sequenced C0#S01 to C0#S15 in Fig. S1 refer to data from fecal samples for crewmember 0# 30 days before entering the CELSS, 15 days before entering the CELSS, and 2d, 15d, 30d, 45d, 60d, 75d, 90d, 105d, 120d, 135d, 150d, 165d and 175d after entering the CELSS. MDC, Mars diurnal cycle; EDC, Earth diurnal cycle; OC, period when crewmembers were out of the CELSS (before entry); IC, period when the crewmembers were in the CELSS. Full-size  DOI: 10.7717/peerj.7762/ﬁg-2 Figure 2 Relative abundance variation proﬁling of fecal microbiota. (A) At phylum level. (B) At genus level. Sequenced C0#S01 to C0#S15 in Fig. S1 refer to data from fecal samples for crewmember 0# 30 days before entering the CELSS, 15 days before entering the CELSS, and 2d, 15d, 30d, 45d, 60d, 75d, 90d, 105d, 120d, 135d, 150d, 165d and 175d after entering the CELSS. MDC, Mars diurnal cycle; EDC, Earth diurnal cycle; OC, period when crewmembers were out of the CELSS (before entry); IC, period when the crewmembers were in the CELSS. Full-size  DOI: 10.7717/peerj.7762/ﬁg-2 Longitudinal variations of gut microbiome A total of 54 fecal samples passed the quality control and were identiﬁed by next-generation sequencing of the V3–V4 hypervariable region of the 16S rRNA gene, with a total of 3,345,159 high-quality sequence reads produced (mean per sample, 61,947; range, 57,791–65,818). Data missing at certain sampling time points was attributed to that certain raw data reads obtained from the HiSeq platform did not pass the quality control. Reads were clustered into ﬁltered OTUs at 97% sequence similarity. OTU annotations were performed using SILVA, and 357 OTUs were identiﬁed according to previously described normalization strategies (Singh et al., 2012; Wang et al., 2009; Wong, Chaston & Douglas, 2013). Relative abundance proﬁling at the phylum and genus level for each crewmember was visualized as shown in Fig. 2, and relative abundance proﬁling at the order, family and species levels are shown in Fig. S1. Core microbiomes for each crewmember at the phylum and genus level (top 10) are shown in Fig. S2. The results showed dominant microbiomes at the phylum level in fecal samples from the four crewmembers were Firmicutes, Bacteroidetes, Proteobacteria, Actinobacteria and Fusobacteria. Dynamic tradeoff of the two most abundant fecal microbiomes Firmicutes and Bacteroidetes were observed during the process (shown in Fig. S3). Dominant fecal microbiomes at the genus level in fecal samples from the four crewmembers were Bacteroides, Prevotella, Incertae_Sedis, Faecalibacterium, Pseudobutyrivibrio, Alistipes, Dialister and Roseburia. We also observed a dynamic tradeoff of Bacteroides and Prevotella during the study period (Fig. S4). Dong et al. (2019), PeerJ, DOI 10.7717/peerj.7762 6/17 6/17 Figure 2 Relative abundance variation proﬁling of fecal microbiota. (A) At phylum level. (B) At genus level. Sequenced C0#S01 to C0#S15 in Fig. S1 refer to data from fecal samples for crewmember 0# 30 days before entering the CELSS, 15 days before entering the CELSS, and 2d, 15d, 30d, 45d, 60d, 75d, 90d, 105d, 120d, 135d, 150d, 165d and 175d after entering the CELSS. MDC, Mars diurnal cycle; EDC, Earth diurnal cycle; OC, period when crewmembers were out of the CELSS (before entry); IC, period when the crewmembers were in the CELSS. Full-size  DOI: 10.7717/peerj.7762/ﬁg-2 Figure 2 Relative abundance variation proﬁling of fecal microbiota. (A) At phylum level. (B) At genus level. Sequenced C0#S01 to C0#S15 in Fig. Dong et al. (2019), PeerJ, DOI 10.7717/peerj.7762 Dietary variations As shown in Fig. 4, decrease of the BMI values of the four crewmembers was accompanied by the reduced energy intake. During the 180-day mission, the controlled eco-surveillance Dong et al. (2019), PeerJ, DOI 10.7717/peerj.7762 Figure 3 Proﬁling of alpha diversity for fecal microbiome. (A) Richness comparison between EDC and MDC. (B) Richness comparison between IC and OC. (C) Evenness comparison between IC and OC. (D) Evenness comparison between EDC and MDC. MDC, Mars diurnal cycle; EDC, Earth diurnal cycle; OC, period when crewmembers were out of the CELSS (before entry); IC, period when the crewmembers were in the CELSS. Full-size  DOI: 10.7717/peerj.7762/ﬁg-3 Figure 3 Proﬁling of alpha diversity for fecal microbiome. (A) Richness comparison between EDC and MDC. (B) Richness comparison between IC and OC. (C) Evenness comparison between IC and OC. (D) Evenness comparison between EDC and MDC. MDC, Mars diurnal cycle; EDC, Earth diurnal cycle; OC, period when crewmembers were out of the CELSS (before entry); IC, period when the crewmembers were in the CELSS. Full-size  DOI: 10.7717/peerj.7762/ﬁg-3 system did not achieve 100% of the required food supply, which was supplemented with pre-packaged food (detailed recipe is shown in a Supplemental File). Due to the higher consumption of prepackaged food compared to fresh food at the beginning of the study, dietary ﬁber was lower during these stages. The proportions of the three major energy nutrients during dietary intake are shown in Fig. 4. Energy was mainly provided by carbohydrates (48.9–65.5%) which increased over the study period. The energy provided by protein was 14.9–22.4%, which also increased across the duration of the study. The energy provided by fat was 13.6–35.4%, which decreased as the study progressed. Variation of the serum 25-hydroxyvitamin D, α-tocopherol, retinol and Repeated ANOVA analysis was used to explore the temporal dynamics of the four serum vitamins (Fig. 5; Fig. S6). Fluctuations were observed for serum a-tocopherol, retinol and folate but the differences were not signiﬁcant. Serum 25-hydroxyvitamin D in 3/4 of the study participants signiﬁcantly decreased during CELSS residence, but increased on day 30 after CELSS departure. The levels of serum 25-hydroxyvitamin D in the CELSS were thus lower than those outside the CELSS. DISCUSSION Enterotypes and diversity variation in the CELSS system The CELSS included a reliable functional system, including basic environmental regulation, regenerative life support, critical life support measurements and control systems, which could accurately control temperature, humidity, atmospheric pressure and air composition of the enclosed environment. Changes in temperature and atmospheric Dong et al. (2019), PeerJ, DOI 10.7717/peerj.7762 (Continued) Enterotypes and diversity variation in the CELSS system Dong et al. (2019), PeerJ, DOI 10.7717/peerj.7762 Table 1 Fecal microbiota at phylum and genus level with signiﬁcant differences by repeated measurement ANOVA analysis. Taxa p-value Comparison between treatments Phylum level IC vs OC Firmicutes 0.0082 < Fusobacteria 0.0099 > Bacteroidetes 0.0170 > Genus level IC vs OC Faecalibacterium 0.0001 > Alistipes 0.0001 < Lachnospira 0.0004 < Prevotella 0.0010 > Ruminococcus 0.0017 < Incertae_Sedis 0.0045 < Bacteroides 0.0088 < Pseudobutyrivibrio 0.0300 > Butyricimonas 0.0310 < Parabacteroides 0.0380 < Subdoligranulum 0.0390 > Dialister 0.0480 < Phylum level EDC vs MDC Fusobacteria 0.0003 < Bacteroidetes 0.0084 > Firmicutes 0.0170 > Genus level EDC vs MDC Paraprevotella 0.0000 < Phascolarctobacterium 0.0002 < Pseudobutyrivibrio 0.0028 < Prevotella 0.0034 > Bacteroides 0.0061 < Alistipes 0.0081 > Megamonas 0.0120 > Collinsella 0.0190 < Sutterella 0.0340 > Phylum level Crewmembers Fusobacteria 0.0011 C03 > C02 > C04 >> C03 Bacteroidetes 0.0001 C02 > C01 > C03 > C04 Firmicutes 0.0003 C04 > C03 > C01 > C02 Proteobacteria 0.0360 C01 > C03 > C04 > C02 Genus level Crewmembers Bacteroides 0.0005 C02 > C01 > C03 > C04 Prevotella 0.0006 C03 > C04 > C01 > C02 Pseudobutyrivibrio 0.0035 C03 > C01 > C04 > C02 Megamonas 0.0070 C04 > C03 > C02 > C01 (Continued) Table 1 Fecal microbiota at phylum and genus level with signiﬁcant differences by repeated measurement ANOVA analysis. (Continued) Dong et al. (2019), PeerJ, DOI 10.7717/peerj.7762 9/17 pressure in the CELSS were minor and identical for all crewmembers (Dai et al., 2017). Studies on host and gut microbiome interactions are complicated by compound environmental factors, such as dietary habits, the host genotype, and occupations, that can lead to drastic variations in gut microbiome. The gut microbiome of the Asian population can be classiﬁed into three enterotypes namely: Bacteroides, Prevotella, and Enterobacteriaceae (Arumugam et al., 2011; Liang et al., 2017). In this study, dynamic Table 1 (continued). Enterotypes and diversity variation in the CELSS system Taxa p-value Comparison between treatments Incertae_Sedis 0.0086 C04 > C03 > C01 > C02 Faecalibacterium 0.0180 C04 > C03 > C01 > C02 Dialister 0.0180 C01 > C02 > C04 > C03 Alistipes 0.0210 C02 > C01 > C03 > C04 Subdoligranulum 0.0380 C04 > C03 > C02 > C01 Roseburia 0.0420 C02 > C01 > C04 > C03 Note: MDC, Mars diurnal cycle; EDC, Earth diurnal cycle; OC, period when crewmembers were out of the CELSS (before entry); IC, period when the crewmembers were in the CELSS. C01, C02, C03 and C04 refers to crewmember 01, 02, 03 and 04. Figure 4 Dietary and BMI variations. (A) Energy intake. (B) Energy supply proportion. (C) Average dietary ﬁber intake. (D) BMI. C01, C02, C03 and C04 refers to crewmember 01, 02, 03 and 04. Full-size  DOI: 10.7717/peerj.7762/ﬁg-4 Table 1 (continued). Taxa p-value Comparison between treatments Incertae_Sedis 0.0086 C04 > C03 > C01 > C02 Faecalibacterium 0.0180 C04 > C03 > C01 > C02 Dialister 0.0180 C01 > C02 > C04 > C03 Alistipes 0.0210 C02 > C01 > C03 > C04 Subdoligranulum 0.0380 C04 > C03 > C02 > C01 Roseburia 0.0420 C02 > C01 > C04 > C03 Note: MDC, Mars diurnal cycle; EDC, Earth diurnal cycle; OC, period when crewmembers were out of the CELSS (before entry); IC, period when the crewmembers were in the CELSS. C01, C02, C03 and C04 refers to crewmember 01, 02, 03 and 04. pressure in the CELSS were minor and identical for all crewmembers (Dai et al., 2017). Studies on host and gut microbiome interactions are complicated by compound environmental factors, such as dietary habits, the host genotype, and occupations, that can lead to drastic variations in gut microbiome. The gut microbiome of the Asian population can be classiﬁed into three enterotypes namely: Bacteroides, Prevotella, and Enterobacteriaceae (Arumugam et al., 2011; Liang et al., 2017). In this study, dynamic variations of Bacteroides-Prevotella-Bacteroides-Prevotella were observed (Fig. S4), which y); , p , , , , and 04. Figure 4 Dietary and BMI variations. (A) Energy intake. (B) Energy supply proportion. (C) Average dietary ﬁber intake. (D) BMI. C01, C02, C03 and C04 refers to crewmember 01, 02, 03 and 04. Full-size  DOI: 10.7717/peerj.7762/ﬁg-4 Figure 4 Dietary and BMI variations. (A) Energy intake. (B) Energy supply proportion. (C) Average dietary ﬁber intake. (D) BMI. Dong et al. (2019), PeerJ, DOI 10.7717/peerj.7762 Enterotypes and diversity variation in the CELSS system S01–S09 refers to blood sampling time points: 30 days before entry (S01), on days 2d (S02), 30d (S03), 60d (S04), 90d (S05), 120d (S06), 150d (S07), and 180d (S08), and 30 days after exiting the CELSS (S09). Full-size  DOI: 10.7717/peerj.7762/ﬁg-5 may have resulted from a cumulative effect of both the diet and environment (Liang et al., 2017). The variations in diversity showed that changes in the fecal microbiome in the isolated environment were closely related to scheduled diurnal cycles. The host can inﬂuence the circadian activity of the intestinal microbiome through biological clock mechanisms (intestinal epithelial peristalsis, secretion) and dietary behaviors. Diurnal alterations of the intestinal microbiome in turn inﬂuence the rhythmic behavior of the host. A loss of circadian rhythm increases the susceptibility to viral and bacterial infections, leading to metabolic disorders as the intestinal microbiomes help to resist the invasion of pathogens through ecological competition (Thaiss et al., 2015). The recoverable properties of gut microbiome diversity observed in this study were in accordance with the MARS 500 trial (Turroni et al., 2017). may have resulted from a cumulative effect of both the diet and environment (Liang et al., 2017). The variations in diversity showed that changes in the fecal microbiome in the isolated environment were closely related to scheduled diurnal cycles. The host can inﬂuence the circadian activity of the intestinal microbiome through biological clock mechanisms (intestinal epithelial peristalsis, secretion) and dietary behaviors. Diurnal alterations of the intestinal microbiome in turn inﬂuence the rhythmic behavior of the host. A loss of circadian rhythm increases the susceptibility to viral and bacterial infections, leading to metabolic disorders as the intestinal microbiomes help to resist the invasion of pathogens through ecological competition (Thaiss et al., 2015). The recoverable properties of gut microbiome diversity observed in this study were in accordance with the MARS 500 trial (Turroni et al., 2017). may have resulted from a cumulative effect of both the diet and environment (Liang et al., Enterotypes and diversity variation in the CELSS system C01, C02, C03 and C04 refers to crewmember 01, 02, 03 and 04. Full-size  DOI: 10.7717/peerj.7762/ﬁg-4 pressure in the CELSS were minor and identical for all crewmembers (Dai et al., 2017). Studies on host and gut microbiome interactions are complicated by compound environmental factors, such as dietary habits, the host genotype, and occupations, that can lead to drastic variations in gut microbiome. The gut microbiome of the Asian population can be classiﬁed into three enterotypes namely: Bacteroides, Prevotella, and Enterobacteriaceae (Arumugam et al., 2011; Liang et al., 2017). In this study, dynamic variations of Bacteroides-Prevotella-Bacteroides-Prevotella were observed (Fig. S4), which Dong et al. (2019), PeerJ, DOI 10.7717/peerj.7762 Figure 5 Proﬁling of the serum 25-hydroxyvitamin D concentrations. (A) Variation during the experiment. (B) Comparison between IC and OC. S01–S09 refers to blood sampling time points: 30 days before entry (S01), on days 2d (S02), 30d (S03), 60d (S04), 90d (S05), 120d (S06), 150d (S07), and 180d (S08), and 30 days after exiting the CELSS (S09). Full-size  DOI: 10.7717/peerj.7762/ﬁg-5 may have resulted from a cumulative effect of both the diet and environment (Liang et al., 2017). The variations in diversity showed that changes in the fecal microbiome in the isolated environment were closely related to scheduled diurnal cycles. The host can inﬂuence the circadian activity of the intestinal microbiome through biological clock mechanisms (intestinal epithelial peristalsis, secretion) and dietary behaviors. Diurnal alterations of the intestinal microbiome in turn inﬂuence the rhythmic behavior of the host. A loss of circadian rhythm increases the susceptibility to viral and bacterial infections, leading to metabolic disorders as the intestinal microbiomes help to resist the invasion of pathogens through ecological competition (Thaiss et al., 2015). The recoverable properties of gut microbiome diversity observed in this study were in accordance with the MARS 500 trial (Turroni et al., 2017). Figure 5 Proﬁling of the serum 25-hydroxyvitamin D concentrations. (A) Variation during the experiment. (B) Comparison between IC and OC. S01–S09 refers to blood sampling time points: 30 days before entry (S01), on days 2d (S02), 30d (S03), 60d (S04), 90d (S05), 120d (S06), 150d (S07), and 180d (S08), and 30 days after exiting the CELSS (S09). Full-size  DOI: 10.7717/peerj.7762/ﬁg-5 Figure 5 Proﬁling of the serum 25-hydroxyvitamin D concentrations. (A) Variation during the experiment. (B) Comparison between IC and OC. Dong et al. (2019), PeerJ, DOI 10.7717/peerj.7762 Dietary and serum 25-hydroxyvitamin D variations Dietary and serum 25-hydroxyvitamin D variations As shown in Fig. S4, decreases in BMI, energy intake and energy supply ratio by fat were common amongst the four crewmembers, which was in accordance with the dynamic variation of the Firmicutes: Bacteroidetes ratio (Fig. S7). The lower body weight and Firmicutes:Bacteroidetes (F:B) ratio during later experimental stages was related to energy restriction and psychological pressure (Bailey et al., 2011; Xu et al., 2016). Vitamin D3 is a fat-soluble vitamin partly derived from the dietary supply. It is mainly found in animal foods such as marine ﬁsh, animal liver, egg yolk and lean meat. Vitamin D3 can also be synthesized in a UV dependent manner (Ann, 2006). As the experiment was performed in an artiﬁcially enclosed environment, natural light was limited leading to decreased levels of vitamin D3 synthesis. After leaving the CELSS, the occupants were exposed to natural light and their diet was more diversiﬁed, boosting the body’s ability to synthesize vitamin D3 (Pasco et al., 2001). The levels of serum 25-hydroxyvitamin D of the four crewmembers increased at S5 compared to S4 which may have been due to the rise in inﬂammation inducing Fusobacteria (Bora et al., 2018). Consistent with these ﬁndings, Vitamin D3 levels of crewmembers have been reported to decrease during space-ﬂight missions (Lane et al., 2013; Smith et al., 2005). Vitamin D3 supplements are therefore required to overcome this deﬁciency. Dong et al. (2019), PeerJ, DOI 10.7717/peerj.7762 Fusobacteria responding to the diurnal cycle variations Repeated measurements and ANOVA analysis showed that the levels of Fusobacteria were signiﬁcantly higher when MDC was performed (S8, S9 and S10) compared to EDC, indicating that the diurnal cycle caused gut microbiome variations at the phylum level. Fusobacteria are non-spore-forming, obligate anaerobic and gram-negative bacteria that frequently colonize the human oral cavity (Fan et al., 2018). It has been reported that boneC3G7 at the family level (phylum Fusobacteria) respond directly to diurnal circle variations, whilst the interactions between oral and intestinal microbiome may be enhanced (Derrien et al., 2010). Higher relative abundances of Fusobacteria were found in bronchial aspirates from patients with severe asthma (Millares et al., 2017). Furthermore, signiﬁcantly increased levels were observed in the inﬂammatory bowel disease group (Santoru et al., 2018). Thus, further attention should be paid to explore the effects of increased Fusobacteria abundance when MDC is performed. Improper dietary control inﬂuences the gut microbiome leading to a loss of body weight (Maslowski & Mackay, 2011), and the psychological stress caused by limited space constraints and long-term isolation, also challenges the gut microbiome and health (Marchesi et al., 2016). During the Mars 500 trial, the structure of the bacterial ﬂora also changed throughout the trial period. Dong et al. (2019), PeerJ, DOI 10.7717/peerj.7762 Speciﬁcally, the relative abundance of the anti-inﬂammatory bacterium Faecalibacterium signiﬁcantly changed in the latter half of the year, reaching its lowest point after 1 year of entry (Turroni et al., 2017). Food provisions were identical for all crewmembers, but repeated ANOVA measurements of speciﬁc fecal microbiome at the phylum and genus level of the four crewmembers showed signiﬁcant differences (Table 1), indicating that the individuals responded differently to the same dietary patterns. These individual responses may be due to differences in their original inherent gut microbiome and physiological state (Orland et al., 2019). CONCLUSIONS A controlled ecological system environment provides an important model for future long-term manned space ﬂight activities and life support for extraterrestrial planetary bases. During studies in the CELSS, diet and diurnal cycles underwent major changes, leading to variations of the gut microbiome. The dynamic changes of Firmicutes and Bacteroidetes were contrasting during the transitional period of the diurnal cycles. The gut microbiome of the crewmembers responded differently to identical dietary and environmental patterns both at the phylum and genus level. Both the unique environment of the CELSS, and MDC treatment caused signiﬁcant differences of bacteria at the phylum and genus levels. Decreased BMI was correlated to decreased energy intake, which was in accordance with the increased F:B ratio during the later experimental stages. A consistent decrease in serum 25-hydroxyvitamin D was observed, most likely due to sunlight deprivation. The factors affecting the human gut microbiome diversity included the environment, diet, psychological stress and physical activity. Dynamic variations of the enterotypes for the four crewmembers as Bacteroides-Prevotella-Bacteroides-Prevotella Dong et al. (2019), PeerJ, DOI 10.7717/peerj.7762 during the isolated residence was a unique phenomenon under compound factors. Therefore, for long-term residence under a controlled ecological health system environment, it is necessary to study multi-factorial adjustment measures to regulate the gut microbiome, maintaining its diversity. As this study was limited by its small sample size, meaningful large-scale studies with close coordination of menu and environmental parameters should now be performed to verify our ﬁndings. Competing Interests Competing Interests The authors declare that they have no competing interests. Author Contributions Hai-Sheng Dong conceived and designed the experiments, prepared ﬁgures and/or tables, authored or reviewed drafts of the paper. Hai-Sheng Dong conceived and designed the experiments, prepared ﬁgures and/or tables, authored or reviewed drafts of the paper. Pu Chen performed the experiments. Yan-Bo Yu analyzed the data. Peng Zang analyzed the data. Zhao Wei approved the ﬁnal draft. Funding This work was funded by the National Natural Science Foundation of China (NO. 21435002, 21621003) and the Shenzhen Science and Technology Plan Project (JCYJ20170818100846805, JCYJ20180305163647462). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript. Grant Disclosures The following grant information was disclosed by the authors: National Natural Science Foundation of China: 21435002, 21621003. Shenzhen Science and Technology Plan: JCYJ20170818100846805, JCYJ20180305163647462. The following grant information was disclosed by the authors: National Natural Science Foundation of China: 21435002, 21621003. Shenzhen Science and Technology Plan: JCYJ20170818100846805, JCYJ20180305163647462. ACKNOWLEDGEMENTS We are grateful to the Astronaut Center of China for the logistic support and the Green Star 180 project facility. We thank the four crewmembers and all researchers involved in the project. We are also grateful to Professor Jin-Ming Lin at Tsinghua University for kind advice, data analysis and statistics. Dong et al. (2019), PeerJ, DOI 10.7717/peerj.7762 Data Availability The following information was supplied regarding data availability: The following information was supplied regarding data availability: The raw sequence data reads is available at NCBI SRA: PRJNA557020. The diet data is available as a Supplemental File. Human Ethics The following information was supplied relating to ethical approvals (i.e., approving body and any reference numbers): The ethics committee of Astronaut Research and Training Center of China approved all scientiﬁc investigations carried out in the context of the project (ethics approval number: Szsisc EA201601). 13/17 Dong et al. (2019), PeerJ, DOI 10.7717/peerj.7762 Supplemental Information Supplemental information for this article can be found online at http://dx.doi.org/10.7717/ peerj.7762#supplemental-information. Supplemental information for this article can be found online at http://dx.doi.org/10.7717/ peerj.7762#supplemental-information. Dong et al. (2019), PeerJ, DOI 10.7717/peerj.7762 REFERENCES (2019), PeerJ, DOI 10.7717/peerj.7762 14/17 Callahan CM, Chen S, Choi AMK, Chlipala GE, Contrepois K, Covington M, Crucian BE, De Vivo I, Dinges DF, Ebert DJ, Feinberg JI, Gandara JA, George KA, Goutsias J, Grills GS, Hargens AR, Heer M, Hillary RP, Hoofnagle AN, Hook VYH, Jenkinson G, Jiang P, Keshavarzian A, Laurie SS, Lee-McMullen B, Lumpkins SB, MacKay M, Maienschein-Cline MG, Melnick AM, Moore TM, Nakahira K, Patel HH, Pietrzyk R, Rao V, Saito R, Salins DN, Schilling JM, Sears DD, Sheridan CK, Stenger MB, Tryggvadottir R, Urban AE, Vaisar T, Van Espen B, Zhang J, Ziegler MG, Zwart SR, Charles JB, Kundrot CE, Scott GBI, Bailey SM, Basner M, Feinberg AP, Lee SMC, Mason CE, Mignot E, Rana BK, Smith SM, Snyder MP, Turek FW. 2019. The NASA twins study: a multidimensional analysis of a year-long human spaceﬂight. Science 364:eaau8650. Callahan CM, Chen S, Choi AMK, Chlipala GE, Contrepois K, Covington M, Crucian BE, De Vivo I, Dinges DF, Ebert DJ, Feinberg JI, Gandara JA, George KA, Goutsias J, Grills GS, Hargens AR, Heer M, Hillary RP, Hoofnagle AN, Hook VYH, Jenkinson G, Jiang P, Keshavarzian A, Laurie SS, Lee-McMullen B, Lumpkins SB, MacKay M, Maienschein-Cline MG, Melnick AM, Moore TM, Nakahira K, Patel HH, Pietrzyk R, Rao V, Maienschein-Cline MG, Melnick AM, Moore TM, Nakahira K, Patel HH, Pietrzyk R, Rao V, Saito R, Salins DN, Schilling JM, Sears DD, Sheridan CK, Stenger MB, Tryggvadottir R, Urban AE, Vaisar T, Van Espen B, Zhang J, Ziegler MG, Zwart SR, Charles JB, Kundrot CE, Scott GBI, Bailey SM, Basner M, Feinberg AP, Lee SMC, Mason CE, Mignot E, Rana BK, Smith SM, Snyder MP, Turek FW. 2019. The NASA twins study: a multidimensional analysis of a year-long human spaceﬂight. Science 364:eaau8650. Guo S, Tang Y, Zhu J, Wang X, Yin Y, Feng H, Ai W, Liu X, Qin L. 2008. Development of a CELSS experimental facility. Advances in Space Research 41(5):725–729 DOI 10.1016/j.asr.2007.09.029. Hamilton MJ, Weingarden AR, Unno T, Khoruts A, Sadowsky MJ. 2013. High-throughput DNA sequence analysis reveals stable engraftment of gut microbiota following transplantation of previously frozen fecal bacteria. Gut Microbes 4(2):125–135 DOI 10.4161/gmic.23571. Hao Z, Li L, Fu Y, Liu H. 2018. The inﬂuence of bioregenerative life-support system dietary structure and lifestyle on the gut microbiota: a 105-day ground-based space simulation in Lunar Palace 1. Environmental Microbiology 20(10):3643–3656 DOI 10.1111/1462-2920.14358. REFERENCES Ann P. 2006. Vitamin D deﬁciency: a global perspective. Arquivos Brasileiros De Endocrinologia E Metabologia 50:153–164. Arumugam M, Raes J, Pelletier E, Le PD, Yamada T, Mende DR, Fernandes GR, Tap J, Bruls T, Batto JM. 2011. Enterotypes of the human gut microbiome. Nature 473(7346):174–180 DOI 10.1038/nature09944. Bailey MT, Dowd SE, Galley JD, Hufnagle AR, Allen RG, Lyte M. 2011. Exposure to a social stressor alters the structure of the intestinal microbiota: implications for stressor-induced immunomodulation. Brain, Behavior, and Immunity 25(3):397–407 DOI 10 1016/j bbi 2010 10 023 DOI 10.1016/j.bbi.2010.10.023. Bora SA, Kennett MJ, Smith PB, Patterson AD, Cantorna MT. 2018. The gut microbiota regulates endocrine Vitamin D metabolism through ﬁbroblast growth factor 23. Frontiers in Immunology 9:408. Botero LE, Delgado-Serrano L, Cepeda Hernandez ML, Del Portillo Obando P, Zambrano Eder MM. 2016. The human microbiota: the role of microbial communities in health and disease. Acta Biológica Colombiana 21:5–15. Cumurcu T, Sahin S, Aydin E. 2006. Serum homocysteine, vitamin B 12 and folic acid levels in different types of glaucoma. BMC Ophthalmology 6(1):1–5 DOI 10.1186/1471-2415-6-6. Dai K, Yu Q, Zhang Z, Wang Y, Wang X. 2017. Aromatic hydrocarbons in a controlled ecological life support system during a 4-person-180-day integrated experiment. Science of the Total Environment 610-611:905–911 DOI 10.1016/j.scitotenv.2017.08.164. Degermendzhi AG, Tikhomirov AA. 2014. Designing artiﬁcial closed land- and space-based ecosystems. Herald of the Russian Academy of Sciences 84(2):124–130 DOI 10.1134/S1019331614020026. Derrien M, Van Passel MW, Van De Bovenkamp JH, Schipper RG, De Vos WM, Dekker J. 2010. Mucin-bacterial interactions in the human oral cavity and digestive tract. Gut Microbes 1(4):254–268 DOI 10.4161/gmic.1.4.12778. Eckburg PB, Bik EM, Bernstein CN, Purdom E, Dethlefsen L, Sargent M, Gill SR, Nelson KE, Relman DA. 2005. Diversity of the human intestinal microbial ﬂora. Science 308(5728):1635–1638 DOI 10.1126/science.1110591. Fan X, Alekseyenko AV, Wu J, Peters BA, Jacobs EJ, Gapstur SM, Purdue MP, Abnet CC, Stolzenberg-Solomon R, Miller G, Ravel J, Hayes RB, Ahn J. 2018. Human oral microbiome and prospective risk for pancreatic cancer: a population-based nested case-control study. Gut 67(1):120–127 DOI 10.1136/gutjnl-2016-312580. Garrett-Bakelman FE, Darshi M, Green SJ, Gur RC, Lin L, Macias BR, McKenna MJ, Meydan C, Mishra T, Nasrini J, Piening BD, Rizzardi LF, Sharma K, Siamwala JH, Taylor L, Vitaterna MH, Afkarian M, Afshinnekoo E, Ahadi S, Ambati A, Arya M, Bezdan D, 14/17 Dong et al. Dong et al. (2019), PeerJ, DOI 10.7717/peerj.7762 REFERENCES He Y, Caporaso JG, Jiang X-T, Sheng H-F, Huse SM, Rideout JR, Edgar RC, Kopylova E, Walters WA, Knight R, Zhou H-W. 2015. Stability of operational taxonomic units: an important but neglected property for analyzing microbial diversity. Microbiome 3(1):20 DOI 10.1186/s40168-015-0081-x. Kirchhoff E. 2002. Online-publication of the German food composition table “Souci-Fachmann– Kraut“ on the internet. Journal of Food Composition and Analysis 15(4):465–472 DOI 10.1006/jfca.2002.1091. Kullak-Ublick GA, Gubler C, Spanaus K, Ismair MG, Da Silva TC, Jetter A. 2016. No major effects of vitamin D3 (1,25 dihydroxyvitamin D3) on absorption and pharmacokinetics of folic acid and fexofenadine in healthy volunteers. European Journal of Clinical Pharmacology 72(7):797–805 DOI 10.1007/s00228-016-2050-0. Lane HW, Bourland C, Barrett A, Heer M, Smith SM. 2013. The role of nutritional research in the success of human space ﬂight. Advances in Nutrition 4(5):521–523 DOI 10.3945/an.113.004101. Lang JM, Coil DA, Neches RY, Brown WE, Cavalier D, Severance M, Hampton-Marcell JT, Gilbert JA, Eisen JA. 2017. A microbial survey of the International Space Station (ISS). PeerJ 5(18):e4029 DOI 10.7717/peerj.4029. LeBlanc JG, Milani C, De Giori GS, Sesma F, Van Sinderen D, Ventura M. 2013. Bacteria as vitamin suppliers to their host: a gut microbiota perspective. Current Opinion in Biotechnology 24(2):160–168 DOI 10.1016/j.copbio.2012.08.005. Liang C, Tseng H-C, Chen H-M, Wang W-C, Chiu C-M, Chang J-Y, Lu K-Y, Weng S-L, Chang T-H, Chang C-H, Weng C-T, Wang H-M, Huang H-D. 2017. Diversity and enterotype in gut bacterial community of adults in Taiwan. BMC Genomics 18:932. Lin S-W, Freedman ND, Shi J, Gail MH, Vogtmann E, Yu G, Klepac-Ceraj V, Paster BJ, Dye BA, Wang G-Q, Wei W-Q, Fan J-H, Qiao Y-L, Dawsey SM, Abnet CC. 2015. Beta-diversity metrics of the upper digestive tract microbiome are associated with body mass index. Obesity 23(4):862–869 DOI 10.1002/oby.21020. Dong et al. (2019), PeerJ, DOI 10.7717/peerj.7762 15/17 MacElroy RD, Bredt J. 1984. Current concepts and future directions of CELSS. Advances in Space Research 4(12):221–229 DOI 10.1016/0273-1177(84)90566-0. Marchesi JR, Adams DH, Fava F, Hermes GDA, Hirschﬁeld GM, Hold G, Quraishi MN, Kinross J, Smidt H, Tuohy KM, Thomas LV, Zoetendal EG, Hart A. 2016. The gut microbiota and host health: a new clinical frontier. Gut 65(2):330–339 DOI 10.1136/gutjnl-2015-309990. Mardanov AV, Babykin MM, Beletsky AV, Grigoriev AI, Zinchenko VV, Kadnikov VV, Kirpichnikov MP, Mazur AM, Nedoluzhko AV, Novikova ND, Prokhortchouk EB, Ravin NV, Skryabin KG, Shestakov SV. 2013. REFERENCES Metagenomic analysis of the dynamic changes in the gut microbiome of the participants of the MARS-500 experiment, simulating long term space ﬂight. Acta Naturae 5(3):116–125 DOI 10.32607/20758251-2013-5-3-116-125. Maslowski KM, Mackay CR. 2011. Diet, gut microbiota and immune responses. Nature Immunology 12(1):5–9 DOI 10.1038/ni0111-5. Meinen E, Dueck T, Kempkes F, Stanghellini C. 2018. Growing fresh food on future space missions: environmental conditions and crop management. Scientia Horticulturae 235:270–278 DOI 10.1016/j.scienta.2018.03.002. Millares L, Bermudo G, Pérez-Brocal V, Domingo C, Garcia-Nuñez M, Pomares X, Moya A, Monsó E. 2017. The respiratory microbiome in bronchial mucosa and secretions from severe IgE-mediated asthma patients. BMC Microbiology 17(1):20 DOI 10.1186/s12866-017-0933-6. Nelson M, Dempster WF, Allen JP. 2008. Integration of lessons from recent research for “Earth to Mars” life support systems. Advances in Space Research 41(5):675–683 DOI 10.1016/j.asr.2007.02.075. Nobs SP, Tuganbaev T, Elinav E. 2019. Microbiome diurnal rhythmicity and its impact on host physiology and disease risk. EMBO Reports 20(4):e47129 DOI 10.15252/embr.201847129. Orland C, Emilson EJS, Basiliko N, Mykytczuk NCS, Gunn JM, Tanentzap AJ. 2019. Microbiome functioning depends on individual and interactive effects of the environment and community structure. ISME Journal 13(1):1–11 DOI 10.1038/s41396-018-0230-x. Pasco JA, Henry MJ, Nicholson GC, Sanders KM, Kotowicz MA. 2001. Vitamin D status of women in the Geelong Osteoporosis Study: association with diet and casual exposure to sunlight. Medical Journal of Australia 175:401–405. Ramakrishna BS. 2013. Role of the gut microbiota in human nutrition and metabolism. Journal of Gastroenterology and Hepatology 28:9–17 DOI 10.1111/jgh.12294. Reinke H, Asher G. 2019. Crosstalk between metabolism and circadian clocks. Nature Reviews Molecular Cell Biology 20(4):227–241 DOI 10.1038/s41580-018-0096-9. Rock CL, Thornquist MD, Kristal AR, Patterson RE, Cooper DA, Neuhouser ML, Neumark- Sztainer D, Cheskin LJ. 1999. Demographic, dietary and lifestyle factors differentially explain variability in serum carotenoids and fat-soluble vitamins: baseline results from the sentinel site of the Olestra Post-Marketing Surveillance Study. Journal of Nutrition 129(4):855–864 DOI 10.1093/jn/129.4.855. Santoru ML, Piras C, Murgia A, Palmas V, Camboni T, Liggi S, Ibba I, Lai MA, Orrù S, Blois S, Loizedda AL, Grifﬁn JL, Usai P, Caboni P, Atzori L, Manzin A. 2018. Cross sectional evaluation of the gut-microbiome metabolome axis in an Italian cohort of IBD patients. Scientiﬁc Reports 7(1):9523 DOI 10.1038/s41598-017-10034-5. Saraf MK, Piccolo BD, Bowlin AK, Mercer KE, LeRoith T, Chintapalli SV, Shankar K, Badger TM, Yeruva L. 2017. Formula diet driven microbiota shifts tryptophan metabolism from serotonin to tryptamine in neonatal porcine colon. Dong et al. (2019), PeerJ, DOI 10.7717/peerj.7762 REFERENCES Microbiome 5(1):77 DOI 10.1186/s40168-017-0297-z. 16/17 Dong et al. (2019), PeerJ, DOI 10.7717/peerj.7762 Shafquat A, Joice R, Simmons SL, Huttenhower C. 2014. Functional and phylogenetic assembly of microbial communities in the human microbiome. Trends in Microbiology 22(5):261–266 DOI 10.1016/j.tim.2014.01.011. Singh KM, Shah T, Deshpande S, Jakhesara SJ, Koringa PG, Rank DN, Joshi CG. 2012. High through put 16S rRNA gene-based pyrosequencing analysis of the fecal microbiota of high FCR and low FCR broiler growers. Molecular Biology Reports 39(12):10595–10602 DOI 10.1007/s11033-012-1947-7. Smith SM, Zwart SR, Gladys B, Rice BL, Davis-Street JE. 2005. The nutritional status of astronauts is altered after long-term space ﬂight aboard the International Space Station. Journal of Nutrition 135(3):437–443 DOI 10.1093/jn/135.3.437. Stein TP. 2001. Nutrition in the space station era. Nutrition Research Reviews 14(1):87–118 DOI 10.1079/095442201108729150. Thaiss CA, Zeevi D, Levy M, Segal E, Elinav E. 2015. A day in the life of the meta-organism: diurnal rhythms of the intestinal microbiome and its host. Gut Microbes 6(2):137–142 DOI 10.1080/19490976.2015.1016690. Turroni S, Rampelli S, Biagi E, Consolandi C, Severgnini M, Peano C, Quercia S, Soverini M, Carbonero FG, Bianconi G, Rettberg P, Canganella F, Brigidi P, Candela M. 2017. Temporal dynamics of the gut microbiota in people sharing a conﬁned environment, a 520-day ground-based space simulation, MARS500. Microbiome 5(1):39 DOI 10 1186/s40168-017-0256-8 Wang Y, Hoenig JD, Malin KJ, Qamar S, Petrof EO, Sun J, Antonopoulos DA, Chang EB, Claud EC. 2009. 16S rRNA gene-based analysis of fecal microbiota from preterm infants with and without necrotizing enterocolitis. ISME Journal 3(8):944–954 DOI 10.1038/ismej.2009.37. Wong AC, Chaston JM, Douglas AE. 2013. The inconstant gut microbiota of Drosophila species revealed by 16S rRNA gene analysis. ISME Journal 7(10):1922–1932 DOI 10.1038/ismej.2013.86. Wu X, Zhang H, Chen J, Shang S, Wei Q, Yan J, Tu X. 2016. Comparison of the fecal microbiota of dholes high-throughput Illumina sequencing of the V3–V4 region of the 16S rRNA gene. Applied Microbiology and Biotechnology 100(8):3577–3586 DOI 10.1007/s00253-015-7257-y. Xu J, Galley JD, Bailey MT, Thomas-Ahner JM, Clinton SK, Olivo-Marston SE. 2016. The impact of dietary energy intake early in life on the colonic microbiota of adult mice. Scientiﬁc Reports 6(1):19083 DOI 10.1038/srep19083. Yang YWG, Pan X. 2009. China food composition-book1. Beijing: Peking University Medical Press Yang YWG, Pan X. 2009. China food composition-book1. Beijing: Peking University Medical Press. Yibin G. 2014. Sequencing technology in gene diagnosis and its application. Hereditas 36:1121–1130. g f p j g g y Yibin G. 2014. Dong et al. (2019), PeerJ, DOI 10.7717/peerj.7762 REFERENCES Sequencing technology in gene diagnosis and its application. Hereditas 36:1121–1130. Yu M, Jia H, Zhou C, Yang Y, Zhao Y, Yang M, Zou Z. 2017. Variations in gut microbiota and fecal metabolic phenotype associated with depression by 16S rRNA gene sequencing and LC/ MS-based metabolomics. Journal of Pharmaceutical and Biomedical Analysis 138:231–239 DOI 10.1016/j.jpba.2017.02.008. Zakrzewski M, Proietti C, Ellis JJ, Hasan S, Brion M-J, Berger B, Krause L. 2016. Calypso: a user-friendly web-server for mining and visualizing microbiome-environment interactions. Bioinformatics 33:782–783. Dong et al. (2019), PeerJ, DOI 10.7717/peerj.7762 17/17
https://openalex.org/W3165700857	http://eprints.uad.ac.id/26857/1/Design%20of%20an%20Adaptive%20Super-Twisting%20Control%20for%20the%20Cart-Pole%20Inverted%20Pendulum%20System.pdf	English	null	Design of an Adaptive Super-Twisting Control for the Cart-Pole Inverted Pendulum System	JITEKI : Jurnal Ilmiah Teknik Elektro Komputer dan Informatika	2,021	cc-by-sa	9,395	Keywords: Inverted pendulum system; Stabilization control; Sliding mode control; Adaptive super-twisting algorithm; Open dynamic engine simulation; anyKode Marilou simulation This work is licensed under a Creative Commons Attribution-Share Alike 4.0 us e a Dept. of Electrical Engineering, Politeknik Negeri Banjarmasin, Jl. Brigjen H. Hasan Basry, Banjarmasin 70123 Indonesia. Email: yusie.rizal@poliban.ac.id Email: yusie.rizal@poliban.ac.id Jurnal Ilmiah Teknik Elektro Komputer dan Informatika (JITEKI) Vol. 7, No. 1, April 2021, pp. 161-174 ISSN: 2338-3070, DOI: 10.26555/jiteki.v7i1.20420 Jurnal Ilmiah Teknik Elektro Komputer dan Informatika (JITEKI) Vol. 7, No. 1, April 2021, pp. 161-174 ISSN: 2338-3070, DOI: 10.26555/jiteki.v7i1.20420 161 Journal homepage : http://journal.uad.ac.id/index.php/JITEKI ABSTRACT A cart-pole inverted pendulum system is one of the underactuated systems that has been used in many applications. This research aims to study the design and the effectiveness of the Adaptive Super-Twisting controller to stabilize the system by comparing it with other previous control methods. A stabilization control of the pendulum upright using the Adaptive Super- Twisting algorithm (ASTA), was investigated. The proposed controller was designed based on the decoupling algorithm method to solve the coupled control input in the system model. We then compared the proposed stabilizing controller with first-order sliding mode control (FOSMC) and Super-Twisting algorithm (STA) in Matlab/Simulink simulation and realistic computer simulation. We developed the computer simulation using anyKode Marilou software, which adopted Open-Dynamic Engine (ODE) as a physics engine. In Matlab/Simulink simulation, we considered three different scenarios: a nominal system, a system with uncertainty, and a disturbed system. Meanwhile, in a computer simulation, we only presented the comparison of different controllers' performances for the realized system. Both results showed that the three controllers could stabilize the pendulum upright with 0.1 rad initial angular position around the vertical axis. Under the same conditions, the ASTA and STA controllers had similar performances; they both have less chattering and faster convergence than the FOSMC approach. However, the FOSMC approach had the least energy delivered and smallest errors than the other two approaches. y Received March 24, 2021 Revised April 25, 2021 Accepted April 30, 2021 This work is licensed under a Creative Commons Attribution-Share Alike 4.0 Design of an Adaptive Super-Twisting Control for the Cart-Pole Inverted Pendulum System Yusie Rizal 1, Muhammad Wahyu 1, Imansyah Noor 1, Joni Riadi 1, Feriyadi 1, Ronny Mantala 2 1Dept. of Electrical Engineering, Politeknik Negeri Banjarmasin, Jl. Brigjen H. Hasan Basry, Banjarmasin 70123, Indonesia 2Dept. of Business Administration, Politeknik Negeri Banjarmasin, Jl. Brigjen H. Hasan Basry, Banjarmasin 70123, Indonesia Yusie Rizal Dept. of Electrical Engineering, Politeknik Negeri Banjarmasin, Jl. Brigjen H. Hasan Basry, Banjarmasin 70123, Indonesia. E il i i l@ lib id 1. INTRODUCTION p p y y g p g g Only conventional first-order and Super-Twisting second-order sliding mode control will be employed as a comparison to the presented controller. These comparisons will be shown in a realistic computer simulation and Matlab/Simulink. In Matlab/Simulink simulation, we provide detailed comparisons to study the control performances for different scenarios, namely, for the nominal system, uncertainty system, and disturbed system, while in a realistic computer simulation, we only present the comparison of controllers for virtual a cart-pole inverted pendulum system without uncertainty and disturbance. We consider this computer simulation because it offers realistic simulated physics [18] in which it adopts Open-Dynamic Engine (ODE). This engine is one of the physics engines in the market even there are other several physics engines such as Bullet physics, NVidea Physics, and DART [19]. The physics engine is the library for simulating rigid bodies in a real-world simulation. Commonly, this computer simulation is intended for robotic systems such as for simulating realistic multi-robot systems [20], modular robots [21], or virtual mechatronics systems. The contribution of this paper is not only the use of the ASTA controller for stabilizing the system plant but also to use the virtual cart-pole inverted pendulum as a testbed for implementing the designed controller. The virtual simulation of the system plant is stable and reliable enough for mimicking this type of underactuated system. In the realistic simulation presented in this paper, several embedded sensors and actuators are already provided in anykode Marilou, e.g., dc motor, encoder, accelerometer/gyroscope, etc. [22]. Thus, this computer simulation is good enough to imitate and simulate the realizing cart-pole inverted pendulum system. there is an interface program i.e., known as Marilou Open Device Access (MODA) to implement the designed controller in realistic simulation in anyKode Marilou. This interface program can be accessed by C/C++ programming language using Microsoft Visual C++ 2008 [22]. Since this work may become an alternative solution to solve the control problem of cart-pole inverted pendulum stabilization in a realistic simulation, then this simulation can also be applied as a low-cost mechatronics system for laboratory testbed in engineering classes. Students and teachers may find this virtual simulation very useful to study control systems and their implementation in robotics and mechatronics system. 1. INTRODUCTION In general, there are three types of control problems related to inverted pendulum systems: (i) swinging up the pendulum upright from the downward position [6][7][8], (ii) stabilization or balancing of the pendulum upright [7][9][10], and (iii) tracking control of the inverted pendulum [11][12][13][14]. Since this system is popular for several decades, many control methods have been proposed to solve these three problems. This paper investigated the second problem, where the stabilization control of the inverted pendulum upright from its initial position around the equilibrium point. p p g p q p The Adaptive Super-Twisting control is well-known as robust control and has been employed in many real-world applications. One of the advantages of using adaptive tuning law is the elimination of the requirement of the knowledge regarding upper bounds of external perturbation [1]. Previous researches related to the use of the Adaptive Super-Twisting control method for a cart-pole inverted pendulum system can be found in [1], [6], and [15]. In [1], they use adaptive-super twisting decoupled with the terminal sliding mode control scheme. The system is considered two separate subsystems, and a controller is proposed to steer each subsystem. In contrast to that, we use the coordinate transformation of the decoupling algorithm proposed by Olfati-Saber [16][17] to obtain its normal form. In other research, as presented in [15], they investigate the system with external perturbation and use global sliding mode surface for the system. This means the term of the sliding surface has an exponential term which is a different form of conventional terms for sliding surface. However, in this research, we use the conventional algebraic terms for sliding surface and study the parametric uncertainty (mass uncertainty) and external disturbance for the cart-pole inverted pendulum. Another approach is discussed in [6], they investigate output-feedback controller for a cart-pole inverted pendulum via coupled sliding-mode algorithm. Unlike in [6], we use a full state-feedback controller for such a system via decoupled algorithm to obtain regular form. Moreover, in this work, we improve the previous work as presented in [10] where the controller of LQR, first-order, and second-order sliding mode control was employed. The novelty of this research is the use of an Adaptive Super-Twisting Algorithm controller for a cart-pole inverted pendulum system by utilizing the decoupling algorithm method. 1. INTRODUCTION The inverted pendulum system is well known as the benchmark for problem-case studies in the control engineering field. Some researchers are attracted to this system because of its characteristics: an unstable, underactuated, and nonlinear system. Thus, some controllers' methods can be designed and tested for such a system plant. This underactuated system has significant applications in real-world engineering problems, for instance, in rockets, underwater robots, walking humanoid robots, overhead cranes, free-flying spacecraft, robotic manipulators, and many more [1]. One practical example of inverted pendulum applications can be found in [2]. Instead of using a complex dynamic model of a certain system, the inverted pendulum model can be employed as a replacement for a simplified model of a longitudinal dynamic of a unicycle robot. The conventional cart-pole inverted pendulum system consists of a cart with an unactuated pendulum attached to the body of the cart. The unactuated pendulum is maintained to stay around its unstable vertical position without falling by controlling the cart's movement. According to [1], an Adaptive Super-Twisting algorithm with decoupled terminal sliding mode control was used to control the inverted pendulum system. In other references, some researchers studied the stabilization control of inverted pendulum system based on output feedback control using Kalman Filter and LQR controller [3], and PID-like fuzzy logic controller [4]. A Email : jiteki@ee.uad.ac.id Journal homepage : http://journal.uad.ac.id/index.php/JITEKI Jurnal Ilmiah Teknik Elektro Komputer dan Informatika (JITEKI) Vol. 7, No. 1, April 2021, pp. 161-174 162 ISSN 2338-3070 different approach was proposed in [5], where the adaptive control logic was applied to solve the system problem. In general, there are three types of control problems related to inverted pendulum systems: (i) swinging up the pendulum upright from the downward position [6][7][8], (ii) stabilization or balancing of the pendulum upright [7][9][10], and (iii) tracking control of the inverted pendulum [11][12][13][14]. Since this system is popular for several decades, many control methods have been proposed to solve these three problems. This paper investigated the second problem, where the stabilization control of the inverted pendulum upright from its initial position around the equilibrium point. different approach was proposed in [5], where the adaptive control logic was applied to solve the system problem. y The system model of the cart-pole inverted pendulum is shown in Fig. 1. The dynamic model of this system is given in [25]: 1. INTRODUCTION Moreover, this virtual simulation can also be applied to the research proposed as given in [23] where the proposed teaching robot algorithm is tested in a virtual robot experimentation platform (V-REP). This similar idea can be found in [22][24] for a brain-computer interface to control the wheelchair that is virtually simulated in anyKode Marilou. 2.1. System Model y The system model of the cart-pole inverted pendulum is shown in Fig. 1. The dynamic model of this system is given in [25]: The system model of the cart-pole inverted pendulum is shown in Fig. 1. The dynamic model of this system is given in [25]: Design of an Adaptive Super-Twisting Control for the Cart-Pole Inverted Pendulum System (Yusie Rizal) ISSN 2338-3070 Jurnal Ilmiah Teknik Elektro Komputer dan Informatika (JITEKI) 163 Vol. 7, No. 1, April 2021, pp. 161-174 8-3070 Jurnal Ilmiah Teknik Elektro Komputer dan Informatika (JITEKI) 163 Vol. 7, No. 1, April 2021, pp. 161-174 163 (𝑚+ 𝑀)𝑦̈ + 𝑚𝑙(𝜃̈ 𝑐𝑜𝑠𝜃−𝜃̇ 2 𝑠𝑖𝑛𝜃) = 𝜏 𝑦̈ 𝑐𝑜𝑠𝜃+ 𝑙𝜃̈ + 𝑔𝑠𝑖𝑛𝜃= 0 (1) (1) where m and M are the mass of the pendulum and cart, respectively. The l is the length of the pendulum, τ is the force, y is the displacement, g is the gravitational force, and θ is the angle of a pendulum to the z-axis. The system (1) can be rewritten as follows: where m and M are the mass of the pendulum and cart, respectively. The l is the length of the pendulum, τ is the force, y is the displacement, g is the gravitational force, and θ is the angle of a pendulum to the z-axis. The system (1) can be rewritten as follows: 𝑦̈ = (𝑙 𝜃̇ 2 + 𝑔𝑐𝑜𝑠𝜃)𝑚𝑠𝑖𝑛𝜃 𝑀+ 𝑚𝑠𝑖𝑛2 𝜃 + 𝜏 𝑀+ 𝑚𝑠𝑖𝑛2 𝜃 𝜃̈ = − ((𝑚+ 𝑀)𝑔+ 𝑚𝑙𝜃̇ 2 𝑐𝑜𝑠𝜃) 𝑠𝑖𝑛𝜃 𝑀𝑙+ 𝑚𝑙𝑠𝑖𝑛2 𝜃 − 𝑐𝑜𝑠𝜃𝜏 𝑀𝑙+ 𝑚𝑙𝑠𝑖𝑛2 𝜃 (2) (2) It is important to note that the system in (2) is the correction for the dynamic model presented by Mahjoub [26][27], in particular, for the negative sign of the first term in the second row of the system (2). It is important to note that the system in (2) is the correction for the dynamic model presented by Mahjoub [26][27], in particular, for the negative sign of the first term in the second row of the system (2). is important to note that the system in (2) is the correction for the dynamic model presented by Mahjou 6][27], in particular, for the negative sign of the first term in the second row of the system (2). We let sin 𝜃≈𝜃, cos 𝜃≈1, 𝑢= 𝜏, and since 𝜃 is around equilibrium point then we assume that 𝑚𝑙sin2 𝜃≈0. 2.1. System Model We define 𝑥1 = 𝜃 𝑥2 = 𝜃̇ 𝑥3 = 𝑦 𝑥4 = 𝑦̇ (3) 𝑥1 = 𝜃 𝑥2 = 𝜃̇ 𝑥3 = 𝑦 𝑥4 = 𝑦̇ (3) 𝑥1 = 𝜃 𝑥2 = 𝜃̇ 𝑥3 = 𝑦 𝑥4 = 𝑦̇ (3) It follows that the state-space in (2) is: It follows that the state-space in (2) is: 𝑥̇1 = 𝑥2 𝑥̇2 = − (𝑚+ 𝑀) 𝑔 𝑥1 + 𝑚 𝑙 𝑥1 𝑥2 2 𝑀 𝑙 − 𝑢 𝑀 𝑙 𝑥̇3 = 𝑥4 𝑥̇4 = 𝑚 𝑙 𝑥1 𝑥2 2 + 𝑚 𝑔 𝑥1 𝑀 + 𝑢 𝑀 Fig. 1. Cart-pole inverted pendulum model. q m M l t z y 𝑥̇1 = 𝑥2 𝑥̇2 = − (𝑚+ 𝑀) 𝑔 𝑥1 + 𝑚 𝑙 𝑥1 𝑥2 2 𝑀 𝑙 − 𝑢 𝑀 𝑙 𝑥̇3 = 𝑥4 𝑥̇4 = 𝑚 𝑙 𝑥1 𝑥2 2 + 𝑚 𝑔 𝑥1 𝑀 + 𝑢 𝑀 Fig. 1. Cart-pole inverted pendulum model. q m M l t z y (4) Fig. 1. Cart-pole inverted pendulum model. Design of an Adaptive Super-Twisting Control for the Cart-Pole Inverted Pendulum System (Yusie Rizal) We let sin 𝜃≈𝜃, cos 𝜃≈1, 𝑢= 𝜏, and since 𝜃 is around equilibrium point then we assume th 𝑙sin2 𝜃≈0. We define 2.2. Decoupling Input 161-174 𝑞2 = 𝑦 𝑝2 = 𝑦̇ Thus, the equation (5) can be written as: Thus, the equation (5) can be written as: Thus, the equation (5) can be written as: 𝑞̇1 = 𝑝1 𝑝̇1 = 𝑓1(𝑝, 𝑞) + 𝑔1(𝑝, 𝑞)𝑢 𝑞̇2 = 𝑝2 𝑝̇2 = 𝑓2(𝑝, 𝑞) + 𝑔2(𝑝, 𝑞)𝑢 (6) 𝑞̇1 = 𝑝1 𝑝̇1 = 𝑓1(𝑝, 𝑞) + 𝑔1(𝑝, 𝑞)𝑢 𝑞̇2 = 𝑝2 𝑝̇2 = 𝑓2(𝑝, 𝑞) + 𝑔2(𝑝, 𝑞)𝑢 (6) (6) 𝑞̇2 = 𝑝2 𝑝̇ = 𝑓(𝑝𝑞) + 𝑔(𝑝𝑞)𝑢 (6) 𝑝2 𝑓2(𝑝, 𝑞) + 𝑔2(𝑝, 𝑞)𝑢 We use new variables 𝑧1 𝑧2 𝑧3 𝑧4 and by following global coordinate transformation in [28] we define 𝑝2 𝑓2(𝑝, 𝑞) 𝑔2(𝑝, 𝑞) We use new variables 𝑧1, 𝑧2, 𝑧3, 𝑧4 and by following global coordinate transformation in [28], we define We use new variables 𝑧1, 𝑧2, 𝑧3, 𝑧4 and by following global coordinate transformation in [28], we define 𝑧1 = 𝑞1 −∫ 𝑔1(𝑠) 𝑔2(𝑠) 𝑞2 0 𝑑𝑠= 𝑞1 + 𝑞2 𝑙 𝑧2 = 𝑝1 −𝑔1(𝑞2) 𝑔2(𝑞2) 𝑝2 = 𝑝1 + 𝑝2 𝑙 𝑧3 = 𝑞2 (7) 𝑧1 = 𝑞1 −∫ 𝑔1(𝑠) 𝑔2(𝑠) 𝑞2 0 𝑑𝑠= 𝑞1 + 𝑞2 𝑙 𝑧2 = 𝑝1 −𝑔1(𝑞2) 𝑔2(𝑞2) 𝑝2 = 𝑝1 + 𝑝2 𝑙 𝑧3 = 𝑞2 𝑧4 = 𝑝2 (7) 𝑧2 = 𝑝1 −𝑔1(𝑞2) 𝑔2(𝑞2) 𝑝2 = 𝑝1 + 𝑝2 𝑙 (7) (7) (7) 𝑧3 = 𝑞2 𝑧4 = 𝑝2 𝑧3 = 𝑞2 𝑧4 = 𝑝2 follows from (7) that 𝑧̇1 = 𝑧2 𝑧̇2 = −𝑔 𝑙𝑧1 + 𝑔 𝑙2 𝑧3 𝑧̇3 = 𝑧4 𝑧̇4 = 𝑚𝑔 𝑀𝑧1 −𝑚𝑔 𝑀𝑙𝑧3 + 𝑢 𝑀+ ∆ (8) 𝑧̇1 = 𝑧2 𝑧̇2 = −𝑔 𝑙𝑧1 + 𝑔 𝑙2 𝑧3 𝑧̇3 = 𝑧4 𝑧̇4 = 𝑚𝑔 𝑀𝑧1 −𝑚𝑔 𝑀𝑙𝑧3 + 𝑢 𝑀+ ∆ (8) (8) Thus, the control input that appears in (4) is removed as in (8). The term ∆ is assumed as disturbance so that it can be ignored in controller design. It is given by: Thus, the control input that appears in (4) is removed as in (8). The term ∆ is assumed as disturbance so that it can be ignored in controller design. It is given by: ∆ = 𝑚𝑙 𝑀(𝑧1 −𝑧3 𝑙) (𝑧2 −𝑧4 𝑙) 2 (9) (9) 2.2. Decoupling Input The control input in the system (4) appears in the second and fourth row. One way to remove the system input in the second row is by considering it as two subsystems as given in [26] and [1]. Another way to solve this problem is to employ the decoupling algorithm proposed by Olfati-Saber [16][17]. This algorithm also applies in different underactuated systems such as wheeled pendular-like systems (WPS) vehicle [28] and inertial wheel inverted pendulum [29]. To decouple these system inputs, we let 𝑞1 = 𝜃 𝑝1 = 𝜃̇ (5) 𝑞1 = 𝜃 𝑝1 = 𝜃̇ (5) 164 Jurnal Ilmiah Teknik Elektro Komputer dan Informatika (JITEKI) ISSN 2338-3070 Vol. 7, No. 1, April 2021, pp. 161-174 Jurnal Ilmiah Teknik Elektro Komputer dan Informatika (JITEKI) ISSN 2338-3070 Vol. 7, No. 1, April 2021, pp. 161-174 164 ISSN 2338-3070 p Vol. 7, No. 1, April 2021, pp. Design of an Adaptive Super-Twisting Control for the Cart-Pole Inverted Pendulum System (Yusie Rizal) 2.3. Controller Design In this paper, we assume that all states are available to be measured. We design the sliding surface, σ as following [30]: 𝜎= 𝑐1𝑧1 + 𝑐2𝑧2 + 𝑐3𝑧3 + 𝑧4 (10) (10) It follows from (10) that It follows from (10) that 𝜎̇ = 𝑐1𝑧̇1 + 𝑐2𝑧̇2 + 𝑐3𝑧̇3 + 𝑧̇4 (11) (11) 𝜎̇ = 𝑐1𝑧̇1 + 𝑐2𝑧̇2 + 𝑐3𝑧̇3 + 𝑧̇4 d by substituting (8) into (11) and ignore the disturbance term of Δ, then we have: tuting (8) into (11) and ignore the disturbance term of Δ, then we have: 𝜎̇ = 𝑐1𝑧1 + 𝑐2 (−𝑔 𝑙𝑧1 + 𝑔 𝑙2 𝑧3) + 𝑐3𝑧4 + 𝑚𝑔 𝑀𝑧1 −𝑚𝑔 𝑀𝑙𝑧3 + 𝑢 𝑀 (12) (12) The system (12) has relative degree 1, and thus, the Super-Twisting Algorithm can be applied. By ch has relative degree 1, and thus, the Super-Twisting Algorithm can be applied. By choosing The system (12) has relative degree 1, and thus, the Super-Twisting Algorithm can be applied 𝑢= 𝑀(𝑢𝑒𝑞+ 𝑢𝑠𝑡) (13) 𝑢= 𝑀(𝑢𝑒𝑞+ 𝑢𝑠𝑡) 𝑢= 𝑀(𝑢𝑒𝑞+ 𝑢𝑠𝑡) (13) where 𝑢𝑒𝑞= −(𝑚𝑔 𝑀−𝑐2𝑔 𝑙) 𝑧1 −𝑐1𝑧2 −(𝑔𝑐2 𝑙2 −𝑚𝑔 𝑀𝑙) 𝑧3 −𝑐3𝑧4 (14) (14) And 𝑢𝑠𝑡, the super-twisting algorithm [31][32] is given by: And 𝑢𝑠𝑡, the super-twisting algorithm [31][32] is given by: 𝑢𝑠𝑡= 𝑢1 + 𝑢2 (15) 𝑢𝑠𝑡= 𝑢1 + 𝑢2 (15) (15) 𝑢𝑠𝑡= 𝑢1 + 𝑢2 (15) 𝑢𝑠𝑡= 𝑢1 + 𝑢2 𝑢𝑠𝑡= 𝑢1 + 𝑢2 Design of an Adaptive Super-Twisting Control for the Cart-Pole Inverted Pendulum System (Yusie Rizal) ISSN 2338-3070 Jurnal Ilmiah Teknik Elektro Komputer dan Informatika (JITEKI) 165 Vol. 7, No. 1, April 2021, pp. 161-174 165 where 𝑢1 = −𝛼\|𝜎\|1/2𝑠𝑖𝑔𝑛(𝜎) 𝑢̇ 2 = {−𝑢𝑠𝑡 𝑖𝑓\|𝑢𝑠𝑡\| > 𝑢𝑚 −𝛽 𝑠𝑖𝑔𝑛(𝜎) 𝑖𝑓\|𝑢𝑠𝑡\| ≤𝑢𝑚 (16) 𝑢1 = −𝛼\|𝜎\|1/2𝑠𝑖𝑔𝑛(𝜎) 𝑢̇ 2 = {−𝑢𝑠𝑡 𝑖𝑓\|𝑢𝑠𝑡\| > 𝑢𝑚 𝛽𝑖 ( ) 𝑖𝑓\| \| ≤ (16) 𝑢1 = −𝛼\|𝜎\|1/2𝑠𝑖𝑔𝑛(𝜎) 𝑢̇ 2 = {−𝑢𝑠𝑡 𝑖𝑓\|𝑢𝑠𝑡\| > 𝑢𝑚 −𝛽 𝑠𝑖𝑔𝑛(𝜎) 𝑖𝑓\|𝑢𝑠𝑡\| ≤𝑢𝑚 (16) (16) 𝑢̇ 2 = {−𝑢𝑠𝑡 𝑖𝑓\|𝑢𝑠𝑡\| > 𝑢𝑚 −𝛽 𝑠𝑖𝑔𝑛(𝜎) 𝑖𝑓\|𝑢𝑠𝑡\| ≤𝑢𝑚 ( and 1 > 0 and and 𝑠𝑖𝑔𝑛(𝜎) = { 1 𝜎> 0 0 −1 𝜎= 0 𝜎< 0 (15) (15) with 𝑢𝑚 is the upper bound for 𝑢𝑠𝑡. The control gains 𝛼 and 𝛽 are adaptively changing by following algorithm [31][33]: with 𝑢𝑚 is the upper bound for 𝑢𝑠𝑡. 2.3. Controller Design The control gains 𝛼 and 𝛽 are adaptively changing by following algorithm [31][33]: 𝛼̇ = {𝜔1√𝛾1 2 𝑠𝑖𝑔𝑛(\|𝜎\| −𝜇0) 𝑖𝑓 𝛼> 𝛼𝑚 0 𝑖𝑓 𝛼≤𝛼𝑚 𝛽= 2𝜀𝛼 (17) (17) Where 𝜔1, 𝛾1, 𝜀, 𝜇0 and 𝛼𝑚 are nonzero and they need to be tuned to ensure stability. By substituting (13) into (12) using (14)–(15), it follows from (12), Where 𝜔1, 𝛾1, 𝜀, 𝜇0 and 𝛼𝑚 are nonzero and they need to be tuned to ensure stability. By substituting (13) into (12) using (14)–(15), it follows from (12), 𝜎̇ = 𝑢1 + 𝑢2 (18) (18) To stabilize the system, we urge to make the sliding surface 𝜎 and its derivative 𝜎̇ converge to zero, i.e., ize the system, we urge to make the sliding surface 𝜎 and its derivative 𝜎̇ converge to zero, i.e., To stabilize the system, we urge to make the sliding surface 𝜎 and its derivative 𝜎̇ converge to zero, i.e., (19) 𝜎= 𝜎̇ = 0 (19) 𝜎= 𝜎̇ = 0 𝜎= 𝜎̇ = 0 Hence, if 𝜎→0 in (10), this implies that 𝑧4, 𝑧3, 𝑧2, and 𝑧1 will also converge to zero. By following [34] to guarantee 𝜎→0 in (10) we choose 𝑐for 𝑖= 1 2 3 Let suppose 𝜎= 0 in (10) Hence, if 𝜎→0 in (10), this implies that 𝑧4, 𝑧3, 𝑧2, and 𝑧1 will also converge to zero. By following [34], to guarantee 𝜎→0 in (10), we choose 𝑐𝑖 for 𝑖= 1, 2, 3. Let suppose 𝜎= 0 in (10), it follows that Hence, if 𝜎→0 in (10), this implies that 𝑧4, 𝑧3, 𝑧2, and 𝑧1 will also converge to zero. By following [34], to guarantee 𝜎→0 in (10), we choose 𝑐𝑖 for 𝑖= 1, 2, 3. Let suppose 𝜎= 0 in (10), it follows that p 4 3 2 1 g By following [34], to guarantee 𝜎→0 in (10), we choose 𝑐𝑖 for 𝑖= 1, 2, 3. 2.3. Controller Design Let suppose 𝜎= 0 in (10), it follows that 𝑧4 = −𝑐1𝑧1 −𝑐2𝑧2 −𝑐3𝑧3 (20) (20) 𝑧4 = −𝑐1𝑧1 −𝑐2𝑧2 −𝑐3𝑧3 nd by denoting 𝜇1 = 𝑧4 and 𝜇2 = [𝑧1 𝑧2 𝑧3]𝑇, we may rewrite (20) as and by denoting 𝜇1 = 𝑧4 and 𝜇2 = [𝑧1 𝑧2 𝑧3]𝑇, we may rewrite (20) as and by denoting 𝜇1 = 𝑧4 and 𝜇2 = [𝑧1 𝑧2 𝑧3]𝑇, we may rewrite (20) as 𝜇2 = [𝑧1 𝑧2 𝑧3]𝑇, we may rewrite (20) as 𝜇1 = 𝑐𝜇2 (21) (21) 𝜇1 = 𝑐𝜇2 (21) 𝜇1 = 𝑐𝜇2 𝜇1 = 𝑐𝜇2 (21) [ ] here 𝑐= [−𝑐1 −𝑐2 −𝑐3]. By taking the first derivative of (10), we have: 𝜎̇ = 𝜇̇1 + 𝑐𝜇̇2 (22) 𝜎̇ = 𝜇̇1 + 𝑐𝜇̇2 𝜎̇ = 𝜇̇1 + 𝑐𝜇̇2 (22) (22) 𝜎̇ = 𝜇̇1 + 𝑐𝜇̇2 (22) and by relating to (9), it follows that 𝜇̇2 = [ 𝑧̇1 𝑧̇2 𝑧̇3 ] = [ 𝑧2 −𝑔 𝑙𝑧1 + 𝑔 𝑙2 𝑧3 𝑧4 ] (23) (23) Design of an Adaptive Super-Twisting Control for the Cart-Pole Inverted Pendulum System (Yusie Rizal) i.e., rewritten as: i.e., rewritten as: i.e., rewritten as: 𝜇̇2 = 𝐴1𝜇1 + 𝐴2𝜇2 (24) 𝜇̇2 = 𝐴1𝜇1 + 𝐴2𝜇2 (24) By substituting (21) into (24), we have 𝜇̇2 = (𝐴1𝑐+ 𝐴2)𝜇2 (25) (25) To guarantee 𝜇2 →0, we make (𝐴1𝑐+ 𝐴2) in (25) to be Hurwitz. Thus, if 𝜇2 →0, from (21) we also have 𝜇1 →0. Since 𝜇1 →0 and 𝜇2 →0 then 𝑧1 →0, 𝑧2 →0, 𝑧3 →0, 𝑧4 →0. This implies that 𝜎→0 is guaranteed. Thus, if this condition is guaranteed, then it is similar to say that by relating (6)-(8) with (10), we have, 𝑦→0, 𝑦̇ →0, 𝜃→0, and 𝜃̇ →0. 𝑦 𝑦 From the controller design in (13), we have the control law as follow: 𝑢= 𝑀[(𝑚 𝑀+ 𝑐2 𝑙) 𝑔𝜃+ (𝑚𝑔𝑙−𝑀(𝑙2𝑐1 + 𝑔𝑐2)) 𝑀𝑙2 𝜃̇ + (𝑀𝑐2 −𝑚𝑙) 𝑀𝑙2 𝑔𝑦− (𝑐1 + 𝑐3𝑙) 𝑙 𝑦̇] −𝛼\|𝜎\| 1 2 𝑠𝑖𝑔𝑛(𝜎) −∫𝛽 𝑠𝑖𝑔𝑛(𝜎)𝑑𝑡 (26) 𝑢= 𝑀[(𝑚 𝑀+ 𝑐2 𝑙) 𝑔𝜃+ (𝑚𝑔𝑙−𝑀(𝑙2𝑐1 + 𝑔𝑐2)) 𝑀𝑙2 𝜃̇ + (𝑀𝑐2 −𝑚𝑙) 𝑀𝑙2 𝑔𝑦− (𝑐1 + 𝑐3𝑙) 𝑙 𝑦̇] (26) Design of an Adaptive Super-Twisting Control for the Cart-Pole Inverted Pendulum System (Yusie Rizal) Design of an Adaptive Super-Twisting Control for the Cart-Pole Inverted Pendulum System (Yusie Rizal) Jurnal Ilmiah Teknik Elektro Komputer dan Informatika (JITEKI) ISSN 2338-3070 Vol. 7, No. 1, April 2021, pp. 161-174 Jurnal Ilmiah Teknik Elektro Komputer dan Informatika (JITEKI) ISSN 2338-3070 Vol. 7, No. 1, April 2021, pp. 161-174 166 ISSN 2338-3070 Where 𝑐1, 𝑐2, and 𝑐3 are nonzero. Furthermore, 𝛼 and 𝛽 are adaptively changing according to (17). Thus, 𝑐1, 𝑐2, 𝑐3, 𝜔1, 𝛾1, 𝜀, 𝜇0 and 𝛼𝑚 are needed to be tuned and chosen properly from experiments. Where 𝑐1, 𝑐2, and 𝑐3 are nonzero. Furthermore, 𝛼 and 𝛽 are adaptively changing according to (17). Thus, 𝑐1, 𝑐2, 𝑐3, 𝜔1, 𝛾1, 𝜀, 𝜇0 and 𝛼𝑚 are needed to be tuned and chosen properly from experiments. tability Analysis 2.4. Stability Analysis To prove the control stability, we use equations from (13)-(16) to substitute 𝑢 in (12). By calculating (12) for 𝜎̇, we have 𝜎̇ = 𝑢𝑠𝑡+ ∆(𝑧1, 𝑧2, 𝑧3, 𝑧4) (27) 𝜎̇ = 𝑢𝑠𝑡+ ∆(𝑧1, 𝑧2, 𝑧3, 𝑧4) (27) We assume that the disturbance ∆(𝑧1, 𝑧2, 𝑧3, 𝑧4) is differentiable and it satisfies the inequality \|∆̇(𝑧1, 𝑧2, 𝑧3, 𝑧4)\| < 𝛿 (28) (27) 𝜎 𝑢𝑠𝑡+ ∆(𝑧1, 𝑧2, 𝑧3, 𝑧4) We assume that the disturbance ∆(𝑧1, 𝑧2, 𝑧3, 𝑧4) is differentiable and it satisfies the inequality \|∆̇(𝑧1, 𝑧2, 𝑧3, 𝑧4)\| < 𝛿 \|∆̇(𝑧1, 𝑧2, 𝑧3, 𝑧4)\| < 𝛿 (28) (28) where 𝛿 is positive constant. We define the Lyapunov function candidate as where 𝛿 is positive constant. We define the Lyapunov function candidate as where 𝛿 is positive constant. We define the Lyapunov function candidate as 𝑉= 1 2 𝜎𝑇𝜎 (29) (29) By using 𝜎̇ in (27) and employing (15)-(16) for switching control 𝑢𝑠𝑡 and inequality in (28), we compute the first derivative of (29) as following 𝑉̇ 𝑇̇ 𝑉̇ = 𝜎𝑇𝜎̇ = 𝜎𝑇(𝑢𝑠𝑡+ ∆(𝑧1, 𝑧2, 𝑧3, 𝑧4)) = 𝜎𝑇(−𝛼√\|𝜎\| 𝑠𝑖𝑔𝑛(𝜎) −∫𝛽 𝑠𝑖𝑔𝑛(𝜎) 𝑑𝑡+ ∆(𝑧1, 𝑧2, 𝑧3, 𝑧4)) = −𝜎𝑇𝛼√\|𝜎\| 𝑠𝑖𝑔𝑛(𝜎) −𝜎𝑇∫𝛽 𝑠𝑖𝑔𝑛(𝜎) 𝑑𝑡+ 𝜎𝑇∆(𝑧1, 𝑧2, 𝑧3, 𝑧4) ≤−𝛼\|𝜎𝑇\|√\|𝜎\| −\|𝜎𝑇\| ∫𝛽 𝑑𝑡+ \|𝜎𝑇∆(𝑧1, 𝑧2, 𝑧3, 𝑧4)\| = −𝛼\|𝜎𝑇\|√\|𝜎\| −\|𝜎𝑇\| ∫𝛽 𝑑𝑡+ \|𝜎𝑇\| ∫∆̇(𝑧1, 𝑧2, 𝑧3, 𝑧4)𝑑𝑡 = −𝛼\|𝜎𝑇\|√\|𝜎\| −\|𝜎𝑇\| ∫𝛽 𝑑𝑡+ \|𝜎𝑇\| ∫𝛿 𝑑𝑡 = −𝛼\|𝜎𝑇\|√\|𝜎\| −\|𝜎𝑇\| ∫(𝛽−𝛿) 𝑑𝑡 ≤−𝛼\|𝜎𝑇\|√\|𝜎\| (30) (30) and since 𝑉̇ is negative semi-definite, then the controller law converges to zero asymptotically [35]. and since 𝑉̇ is negative semi-definite, then the controller law converges to zero asymptotically [35]. Design of an Adaptive Super-Twisting Control for the Cart-Pole Inverted Pendulum System (Yusie Rizal)  Choose the values for 𝑐1, 𝑐2, and 𝑐3 2.5. Implementation To verify the efficacy of the controller design, we conduct two simulations using Matlab/Simulink and ODE-based simulation. The implementation of Matlab/Simulink and ODE simulation can be depicted in the controller's structure as given in Fig. 2 (left), while the realistic simulation of anyKode Marilou is shown in Fig. 2 (right). All state variables of the system are available to measure. In the simulation, we choose the control gains properly, and if the control system is not stable, then the pendulum will fall immediately and vice-versa. To implement the controller laws in ODE simulation for the proposed controller (ASTA), we use the Visual C++ 2008 program to access the virtual inverted pendulum system plant in anyKode Marilou. The IMU (inertial measurement unit) sensor is applied to the realistic simulation to obtain angular position and velocity in the cart-pole inverted pendulum system. This sensor feedback is updated 100 times per second as the control sampling, Ts. The initial position for the pendulum was set up in anyKode Marilou and not from the C++ program. However, in this application program, we may detect this angular position sensor and use it as feedback for the controller. The following algorithm is given and is derived from the controller's structure of ASTA in Fig. 2 (left).  Choose the values for 𝑐1, 𝑐2, and 𝑐3  Choose values for 𝜔1, 𝛾1, 𝜀, 𝜇0, 𝑢𝑚, and 𝛼𝑚 ISSN 2338-3070 Jurnal Ilmiah Teknik Elektro Komputer dan Informatika (JITEKI) Vol. 7, No. 1, April 2021, pp. 161-174 167 p Vol. 7, No. 1, April 2021, pp. 161-174 1. Obtain the values for 𝜃, 𝜃̇, 𝑦, and 𝑦̇ from sensors' measurements 2. Compute 𝑧1, 𝑧2, 𝑧3, and 𝑧4 in (7) 3. Compute 𝜎 in (10) 4. Compute 𝛼̇ and integration of 𝛼 and 𝛽 in (17) 5. Compute 𝑢𝑒𝑞 in (14) 6. Compute 𝑢1 and integration of 𝑢2 in (16) 7. Compute 𝑢 in (13) 8. Wait for 𝑇𝑠 sampling time and then return to step 1 Initially, we choose values for all parameters but only keep 𝑐1, 𝑐2, and 𝑐3 the same in the two simulation platforms. The remaining parameters are tuned experimentally for the two simulation platforms. In the algorithm program, we cannot calculate the dynamics of the inverted pendulum system in anyKode Marilou because it is a complex system. However, we only can measure its variable states. 2.5. Implementation In Matlab/Simulink, the dynamics model can be calculated using (1) as given in Fig. 2 (left). Fig. 2. Structure of ASTA controller (left) and realistic simulation of cart-pole inverted pendulum (right). Cart-pole Inverted pendulum (Eq. 1) Controller (STA) (Eq. 16) Sliding surface (Eq. 10) Adaptive gain (Eq. 17) Coordinate Transformation (Eq. 7) Equivalent control (Eq. 14) + + x Adaptive gain (Eq. 17) Controller (STA) (Eq. 16) Cart-pole Inverted pendulum (Eq. 1) Sliding surface (Eq. 10) Fig. 2. Structure of ASTA controller (left) and realistic simulation of cart-pole inverted pendulum (right). 3. RESULT AND DISCUSSION This section will show the comparison of the present controller with other stabilizing controller methods (FSOMC and STA). In Matlab/Simulink simulation, we study the performance of controllers in different scenarios, while for anyKode Marilou simulation, we compare the effectiveness of the stabilizing controller for cart-pole inverted pendulum system from these three different control methods. The control gains and sliding surface parameters for each simulation are tuned experimentally. Those values cannot be used directly from one simulation to another simulation because of different platforms. However, the system parameters of the system plant are the same for both simulations, i.e. given by m = 0.25 kg, M = 1.0 kg, l = 0.450 m, and g = 9.8 m/s2. In all simulations, we use the initial angular position for the pendulum with 0.1 rad. 3.1. Matlab/Simulink Simulation 3 (left) for the pendulum using all three controllers can converge to zero in less than 3 seconds. The convergence of angular position and velocity for STA and ASTA methods are similar and better than FOSMC. However, Fig. 4 shows that the system response for the cart's position and velocity of the FOSMC approach has a small error compared to STA and ASTA. In Fig. 3, the angular position and velocity of the pendulum for ASTA and STA have the least chattering compared to the FOSMC approach. The angular positions in Fig. 3 (left) for the pendulum using all three controllers can converge to zero in less than 3 seconds. The convergence of angular position and velocity for STA and ASTA methods are similar and better than FOSMC. However, Fig. 4 shows that the system response for the cart's position and velocity of the FOSMC approach has a small error compared to STA and ASTA. In Fig. 3, the angular position and velocity of the pendulum for ASTA and STA have the least chattering compared to the FOSMC approach. The angular positions in Fig. 3 (left) for the pendulum using all three controllers can converge to zero in less than 3 seconds. The convergence of angular position and velocity for STA and ASTA methods are similar and better than FOSMC. However, Fig. 4 shows that the system response for the cart's position and velocity of the FOSMC approach has a small error compared to STA and ASTA. Fig. 4. Position (left) and velocity (right) of the cart for system without disturbance. Fig. 4. Position (left) and velocity (right) of the cart for system without disturbance. Fig. 5 shows the normalized control inputs for nominal system and system with disturbance, respectively. We use normalized control inputs to have fair comparisons between different controllers (FOSMC, STA, and ASTA). In the figure on the right, an impulse signal is injected at 3 sec. All three controllers can overcome this external disturbance. In comparison, we found that the FOSMC approach has more chattering compared to the other two. In the system with uncertainty, the scenario is conducted by changing its masses with 60% starting from 3 seconds, as shown in Fig. 6 (left). On the right figure, the normalized control input for all three controllers is given. 3.1. Matlab/Simulink Simulation The control gains for this simulation are ω1 = 1.5, γ1 = 2.25, ε = 1.3, αm = 0.25, μ = 0.15, um = 1, and for sliding surface parameters, we choose c1 = 1, c2 = 3.2, and c3 = 8.2. We compare the Adaptive Super-Twisting controller with other controllers using FOSMC [34] and STA [36] by conducting three different scenarios for each controllers to stabilize the cart-pole inverted pendulum system, namely, (a) for nominal system, (b) system with uncertainty, and (c) system with disturbance. To compare with other controller methods of FOSMC and STA, we recall from (13) and rewrite: 𝑢= 𝑀(𝑢𝑒𝑞+ 𝑢𝑠𝑤) (31) (31) For the FOSMC controller, the switching controller is: 𝑢𝑠𝑤= −ℎ 𝑠𝑖𝑔𝑛(𝜎) (32) ler is given by [31][32]: (32) 𝑢𝑠𝑤= −ℎ 𝑠𝑖𝑔𝑛(𝜎) (32) and for STA the switching controller is given by [31][32]: 𝑢𝑠𝑤= −ℎ 𝑠𝑖𝑔𝑛(𝜎) 𝑢𝑠𝑤= −ℎ 𝑠𝑖𝑔𝑛(𝜎) (32) A, the switching controller is given by [31][32]: and for STA, the switching controller is given by [31][32]: 𝑢𝑠𝑤= 𝑢1 + 𝑢2 𝑢1 = −𝜆 \|𝜎\| 1 2 𝑠𝑖𝑔𝑛(𝜎) (33) 𝑢𝑠𝑤= 𝑢1 + 𝑢2 (33) Design of an Adaptive Super-Twisting Control for the Cart-Pole Inverted Pendulum System (Yusie Rizal) Design of an Adaptive Super-Twisting Control for the Cart-Pole Inverted Pendulum System (Yusie Rizal) 168 Jurnal Ilmiah Teknik Elektro Komputer dan Informatika (JITEKI) ISSN 2338-3070 Vol. 7, No. 1, April 2021, pp. 161-174 Jurnal Ilmiah Teknik Elektro Komputer dan Informatika (JITEKI) Vol. 7, No. 1, April 2021, pp. 161-174 ISSN 2338-3070 Jurnal Ilmiah Teknik Elektro Komputer dan Informatika (JITEKI) Vol 7 No 1 April 2021 pp 161 174 p Vol. 7, No. 1, April 2021, pp. 161-174 𝑢̇ 2 = {−𝑢𝑠𝑤 𝑖𝑓\|𝑢𝑠𝑤\| > 𝑢𝑚 −𝑊 𝑠𝑖𝑔𝑛(𝜎) 𝑖𝑓\|𝑢𝑠𝑤\| ≤𝑢𝑚 The control gains for FOSMC is h = 1.36 and for STA are λ = 0.022, W = 0.15, and um = 1. The system parameters of c1, c2, and c3 are kept the same for all controllers (FOSMC, STA, and ASTA). Fig. 3 through Fig. 9 show these simulation results. Fig. 3. Angular position (left) and angular velocity of pendulum (right) for system without disturbance. Fig. 3. Angular position (left) and angular velocity of pendulum (right) for system without disturbance. In Fig. 3, the angular position and velocity of the pendulum for ASTA and STA have the least chattering compared to the FOSMC approach. The angular positions in Fig. Design of an Adaptive Super-Twisting Control for the Cart-Pole Inverted Pendulum System (Yusie Rizal) 3.1. Matlab/Simulink Simulation The integral of the square error (ISE) and integral of absolute error (IAE) using the following formulation: 𝐼𝑆𝐸= ∫𝑒2(𝑡)𝑑𝑡 𝐼𝐴𝐸= ∫\|𝑒(𝑡)\|𝑑𝑡 (34) (34) 170 Jurnal Ilmiah Teknik Elektro Komputer dan Informatika (JITEKI) ISSN 2338-3070 V l 7 N 1 A il 2021 161 174 ISSN 2338-3070 p Vol. 7, No. 1, April 2021, pp. 161-174 We also provide the energy delivered of the control input using normalized control input as the basis for a fair comparison. The formulation of delivered energy is as follows: We also provide the energy delivered of the control input using normalized control input as the basis for a fair comparison. The formulation of delivered energy is as follows: 𝐽= ∫ 𝑢2 𝑑𝑡 10 0 (35) (35) 0 Fig. 8. The system response of FOSMC (left) and STA (right) approaches Fig. 8. The system response of FOSMC (left) and STA (right) approaches Fig. 9. The system response for ASTA approach. Fig. 9. The system response for ASTA approach. It is shown from Table 1 that the FOSMC approach has the chattering effect for all scenarios. However, the energy delivered to the system is the least compared to STA or ASTA. In fact, between STA and ASTA approaches, these two controllers have similar delivered energy. This finding is not consistent with the results as given by Mahjoub in [27]. One possible explanation is that the obtained normal form to design controller in this paper is different from Mahjoub [27]. The second reason is the use of dual layers in Mahjoub's work for sliding surfaces. This approach is different from the designed controllers in this paper. Thus, the obtaining results in this paper are different from those presented in [27]. On average, the performance index for integral absolute error and square error for STA and ASTA is twice and 12% higher than FOSMC, respectively. It implies that the FOSMC method has less steady-state error than the other two methods (STA and ASTA). Since the steady-state error for STA and ASTA are higher than the FOSMC method, this affects the delivered energy to the system in those two methods with 58% higher than the FOSMC approach. Nevertheless, this would not be the case if the comparison of those three methods control input was not normalized. STA and ASTA methods for disturbance systems are higher than the nominal system with 10.5%, 1.62%, and 2%, respectively. 3.1. Matlab/Simulink Simulation Like the previous scenario, the FOSMC approach has more chattering than STA and ASTA while the responses of STA and ASTA approaches are the same. The angular positions of the pendulum for disturbance and uncertainty scenarios are in Fig. 7 (left) and (right), respectively. We observed that the angular position for the FOSMC approach is more chattering than the other two controllers. Although the chattering of FOSMC is consistent with previous scenarios, the transient error of this approach is also the same compared with the other controllers. Fig. 8 and Fig. 9 are the response of all states for the three control methods. All state variables (𝜃, 𝜃̇, 𝑦 and 𝑦̇) are converging to zeros which implies that the designed controller is correct because it is supported by simulation results. Design of an Adaptive Super-Twisting Control for the Cart-Pole Inverted Pendulum System (Yusie Rizal) Jurnal Ilmiah Teknik Elektro Komputer dan Informatika (JITEKI) Vol. 7, No. 1, April 2021, pp. 161-174 ISSN 2338-3070 169 p pp Fig. 5. Control input for system without disturbance (left) and with disturbance (right). Fig. 6. The system's response for mass uncertainty (left) with its control input (right). Fig. 7. Angular position of pendulum with disturbance (left) and with uncertainty (right). The graphical figures of each system and scenario have given system responses, but quantitative Fig. 5. Control input for system without disturbance (left) and with disturbance (right). Fig. 5. Control input for system without disturbance (left) and with disturbance (right). Fig. 6. The system's response for mass uncertainty (left) with its control input (right). i (l f ) i h i l i ( i h ) Fig. 6. The system's response for mass uncertainty (left) with its control input (right). Fig. 6. The system's response for mass uncertainty (left) with its control in Fig. 7. Angular position of pendulum with disturbance (left) and with uncertainty (right). Fig. 7. Angular position of pendulum with disturbance (left) and with uncertainty (right). The graphical figures of each system and scenario have given system responses, but quantitative comparisons are in Table 1. The performance indexes of each controller in respective scenarios are in Table 1. Design of an Adaptive Super-Twisting Control for the Cart-Pole Inverted Pendulum System (Yusie Rizal) 3.1. Matlab/Simulink Simulation But for integral absolute error, we have 17.4%, 5.3%, and 16.3%, respectively. On the other hand, the comparison for the uncertainty system for FOSMC and STA controllers tends to less than nominal systems with 1.5% and 10.34%, respectively. For delivered energy, the system with disturbance for the FOSMC method is 20% higher than the nominal system. Other comparisons for disturbed systems in STA and ASTA scenarios are only less than 2% compared to nominal systems. Design of an Adaptive Super-Twisting Control for the Cart-Pole Inverted Pendulum System (Yusie Rizal) Jurnal Ilmiah Teknik Elektro Komputer dan Informatika (JITEKI) Vol. 7, No. 1, April 2021, pp. 161-174 ISSN 2338-3070 171 Table 1. Comparison of performance indexes and energy delivered for three stabilizing controllers Table 1. Comparison of performance indexes and energy delivered for three stabilizing controllers Controller Scenario Performance Index Energy Criterion ( ×102 J ) ISE IAE FOSMC [34] Nominal 0.0004706 0.02564 2.755 Uncertainty 0.0004635 0.02268 2.796 Disturbance 0.0005198 0.03011 3.307 STA [36] Nominal 0.0009681 0.02807 7.132 Uncertainty 0.0008680 0.02779 7.130 Disturbance 0.0009838 0.02956 7.279 ASTA Nominal 0.0009679 0.02148 7.122 Uncertainty 0.0009679 0.02148 7.136 Disturbance 0.0009865 0.02498 7.266 3.2. ODE Simulation In this computer simulation, we aim to design a controller that can stabilize the inverted pendulum system. We can visually observe the system response as it is the realization of the system plant. If the controller gains and sliding surface parameters are tuned improperly, then the pendulum is falling. We develop ODE simulation using anyKode Marilou software with a C++ program as a programming interface to access that system plant. In C++ programming, we use the sliding parameters as given, i.e., c1 = 0.2526, c2 = 16.37, and c3 = 6.45. These three sliding parameters are for all controllers (FOSMC, STA, and ASTA). The control gains for FOSMC is h = 2.325, and for STA are λ = 0.566, and W = 0.1025, while for ASTA are α = 0.522, ω1 = 1.70, γ1 = 1.104, μ = 0.025, αm = 0.752, ε = 0.25, μ0 = 0.027, and um = 1. The sliding parameters and controller gains used in this simulation are different from those in Matlab/Simulink simulation. We tuned all parameters and control gains experimentally. Fig. 10 shows the comparison of the angular pendulum (left) and normalized control input (right) for three controllers. We observe that the chattering for the FOSMC approach larger than STA and ASTA. These results are consistent with the simulation results in Matlab/Simulink where the system converges in 3 seconds. The comparison of performance indexes and energy delivered to the system using three different controllers is in Table 2. Fig. 10. The system response of angular position of pendulum (left) and normalized control input (right). Fig. 10. The system response of angular position of pendulum (left) and normalized control input (right). In this simulation, we found that the performance indexes in Table 2 for the three controllers are almost the same except for the energy criterion in which FOSMC is slightly larger than STA and ASTA. The comparison of integral square error and absolute error for STA is 23% and 21.7%, while for ASTA is 15.4% and 16.2% smaller than FOSMC, respectively. However, for the energy criterion, the delivered energy using STA and ASTA is 16.2% and 13.1% smaller than FOSMC, respectively. These results are different from the simulation results in Matlab/Simulink simulation. From our analysis, this discrepancy is because of different mechanical structures in computed torque control input. 3.2. ODE Simulation In Matlab/Simulink, the simplified model is only for single control input u to actuate the cart, but in ODE simulation, we employ this control input u for driving four wheels of the cart. Thus, from this implementation, we found different results. Fig. 10. The system response of angular position of pendulum (left) and normalized control input (right). In this simulation, we found that the performance indexes in Table 2 for the three controllers are almost the same except for the energy criterion in which FOSMC is slightly larger than STA and ASTA. The comparison of integral square error and absolute error for STA is 23% and 21.7%, while for ASTA is 15.4% and 16.2% smaller than FOSMC, respectively. However, for the energy criterion, the delivered energy using STA and ASTA is 16.2% and 13.1% smaller than FOSMC, respectively. These results are different from the simulation results in Matlab/Simulink simulation. From our analysis, this discrepancy is because of different mechanical structures in computed torque control input. In Matlab/Simulink, the simplified model is only for single control input u to actuate the cart, but in ODE simulation, we employ this control input u for driving four wheels of the cart. Thus, from this implementation, we found different results. In this simulation, we found that the performance indexes in Table 2 for the three controllers are almost the same except for the energy criterion in which FOSMC is slightly larger than STA and ASTA. The comparison of integral square error and absolute error for STA is 23% and 21.7%, while for ASTA is 15.4% and 16.2% smaller than FOSMC, respectively. However, for the energy criterion, the delivered energy using STA and ASTA is 16.2% and 13.1% smaller than FOSMC, respectively. These results are different from the simulation results in Matlab/Simulink simulation. From our analysis, this discrepancy is because of different mechanical structures in computed torque control input. In Matlab/Simulink, the simplified model is only for single control input u to actuate the cart, but in ODE simulation, we employ this control input u for driving four wheels of the cart. Thus, from this implementation, we found different results. Design of an Adaptive Super-Twisting Control for the Cart-Pole Inverted Pendulum System (Yusie Rizal) Jurnal Ilmiah Teknik Elektro Komputer dan Informatika (JITEKI) Vol. 7, No. 1, April 2021, pp. 161-174 172 ISSN 2338-3070 Table 2. REFERENCES [1] D. Ashtiani Haghighi and S. Mobayen, "Design of an adaptive super-twisting decoupled terminal sliding mode control scheme for a class of fourth-order systems," ISA Trans., pp. 216–225, 2018. https://doi.org/10.1016/j.isatra.2018.02.006 [1] D. Ashtiani Haghighi and S. Mobayen, "Design of an adaptive super-twisting decoupled terminal sliding mode control scheme for a class of fourth-order systems," ISA Trans., pp. 216–225, 2018. https://doi.org/10.1016/j.isatra.2018.02.006 p g j [2] Y. Rizal, C.-T. Ke, and M.-T. Ho, "Point-to-point motion control of a unicycle robot: Design, implementation, and validation," in IEEE International Conference Robotics and Automation, 2015, pp. 4379–4384. https://doi.org/10.1109/ICRA.2015.7139804 p g j [2] Y. Rizal, C.-T. Ke, and M.-T. Ho, "Point-to-point motion control of a unicycle robot: Design, implementation, and validation," in IEEE International Conference Robotics and Automation, 2015, pp. 4379–4384. https://doi.org/10.1109/ICRA.2015.7139804 [3] A. Rizky Octavia, D. Nathanlius, and T. Agustinus Tamba, “Implementasi Kontrol Umpan Balik Keluaran Berbasis Tapis Kalman dan Regulator Kuadratik Linier pada Sistem Pendulum Terbalik,” J. Otomasi Kontrol dan Instrumentasi, vol. 11, no. 2, pp. 81–93, 2019. https://doi.org/10.5614/joki.2019.11.2.2 [3] A. Rizky Octavia, D. Nathanlius, and T. Agustinus Tamba, “Implementasi Kontrol Umpan Balik Keluaran Berbasis Tapis Kalman dan Regulator Kuadratik Linier pada Sistem Pendulum Terbalik,” J. Otomasi Kontrol dan Instrumentasi, vol. 11, no. 2, pp. 81–93, 2019. https://doi.org/10.5614/joki.2019.11.2.2 pp p g j ] A. U. Sambo, F. S. Bala, N. M. Tahir, and A. Y. Babawuro, "Optimal control of inverted pendulum on ca system," J. Phys. Conf. Ser., vol. 1502, no. 1, pp. 1–8, 2020. https://doi.org/10.1088/1742-6596/1502/1/012024 [5] S. Hanwate, Y. V. Hote, and A. Budhraja, "Design and implementation of adaptive control logic for cart-inverted pendulum system," Proc. Inst. Mech. Eng. Part I J. Syst. Control Eng., vol. 233, no. 2, pp. 164–178, 2019. https://doi.org/10.1177/0959651818788148 [5] S. Hanwate, Y. V. Hote, and A. Budhraja, "Design and implementation of adaptive control logic for cart-inverted pendulum system," Proc. Inst. Mech. Eng. Part I J. Syst. Control Eng., vol. 233, no. 2, pp. 164–178, 2019. https://doi.org/10.1177/0959651818788148 p g [6] L. Ovalle, H. Ríos, and M. Llama, "Robust output-feedback control for the cart–pole system: A coupled super- twisting sliding-mode approach," IET Control Theory Appl., vol. 13, no. 2, pp. 269–278, 2019. https://doi.org/10.1049/iet-cta.2018.5249 ] A. Kukker and R. Sharma, "Genetic Algorithm-Optimized Fuzzy Lyapunov Reinforcement Learning for Nonline Systems," Arab. J. Sci. Eng., vol. 45, no. 3, pp. 1629–1638, 2020. https://doi.org/10.1007/s13369-019-04126-9 [7] A. Kukker and R. Sharma, "Genetic Algorithm-Optimized Fuzzy Lyapunov Reinforcement Learning for Nonlinear Systems," Arab. J. Sci. Eng., vol. 3.2. ODE Simulation Comparison of performance indexes and energy delivered for three stabilizing controllers Controller Performance Index Energy Criterion (×103 J) ISE IAE FOSMC [34] 0.0013 0.0784 1.010 STA [36] 0.0010 0.0615 0.846 ASTA 0.0011 0.0657 0.878 Acknowledgments The authors would like to thank Pusat Penelitian dan Pengabdian Masyarakat Politeknik Negeri Banjarmasin for supporting this research under the scheme of Penelitian Pengembangan Dosen with contract no. 293/PL18/SPK-PG/2020. 4. CONCLUSION In this work, we proposed the Adaptive Super-Twisting Algorithm as an alternative to solve the stabilizing control problem for the inverted pendulum system. We present the comparison between the proposed controller with existing methods of FOSMC and Super-Twisting Algorithm. To design the controller, we obtain the normal form using the standard decoupling method. Based on this normal form, we design a sliding surface and Adaptive Super-Twisting controller. In simulation results, we provide two different approaches using standard Matlab/Simulink and ODE-based simulations. From these simulations, we found that all three controllers can stabilize the pendulum upright. However, the implementation for each simulation is different, and then, we found some discrepancies between the two simulation results, especially for performance indexes and energy criterion. In graphical figures presented in those two simulations, we observe that the STA and ASTA consistently reduced the chattering effect induced when the FOSMC approach is applied. We found that the use of STA and ASTA for stabilizing such systems is more energy- efficient, around 15% compared to FOSMC. Between the two methods (STA and ASTA), we calculate that the STA method is 3.7% more efficient than the ASTA method. However, in numerical simulation, those figures are almost similar between STA and ASTA. The findings validated Mahjoub's results [27][26], although the approach used for obtaining the system's normal form is different from those presented in Mahjoub [27][26]. We will continue to work on the same problem for the output-feedback stabilization using the same control method in future work. Unlike the approach in this paper, we assume not all state variables are available to be measured. Thus, this should be more challenging to design the controller and observer. REFERENCES Contr., vol. 47, no. 2, pp. 305–308, 2002. https://doi.org/10.1109/9.983365 pp p g [18] H. L. Pham, B. V. Adorno, V. Perdereau, and P. Fraisse, "Set-point control of robot end-effector pose using dual quaternion feedback," Robot. Comput. Integr. Manuf., vol. 52, pp. 100–110, 2018. https://doi.org/10.1016/j.rcim.2017.11.003 p g j [19] E. Coevoet et al., "Software toolkit for modeling, simulation, and control of soft robots," Adv. Robot., vol. 3 22, pp. 1208–1224, 2017. https://doi.org/10.1080/01691864.2017.1395362 [20] F. M. Noori, D. Portugal, R. P. Rocha, and M. S. Couceiro, "On 3D simulators for multi-robot systems in ROS: MORSE or Gazebo?," in 15th IEEE International Symposium on Safety, Security and Rescue Robotics, Conference, 2017, pp. 19–24. https://doi.org/10.1109/SSRR.2017.8088134 f pp p g [21] K. J. Gucwa and H. H. Cheng, "RoboSim: a simulation environment for programming virtual robots," Eng. Comput., vol. 34, no. 3, pp. 475–485, 2018. https://doi.org/10.1007/s00366-017-0553-7 p pp p g [22] Y. Rizal, "Computer Simulation of Human-Robot Collaboration in the Context of Industry Revolution 4.0," in Future of Robotics - Becoming Human with Humanoid or Emotional Intelligence, InTech, 2019. https://doi.org/10.5772/intechopen.88335 p g p [23] Y. Xu, C. Yang, J. Zhong, N. Wang, and L. Zhao, "Robot teaching by teleoperation based on visual interaction and extreme learning machine," Neurocomputing, vol. 275, pp. 2093–2103, 2017. https://doi.org/10.1016/j.neucom.2017.10.034 [24] G. G. Gentiletti, J. G. Gebhart, R. C. Acevedo, O. Yáñez-Suárez, and V. Medina-Bañuelos, “Command of a simulated wheelchair on a virtual environment using a brain-computer interface,” IRBM, vol. 30, no. 5–6, pp. 218– 225, 2009. https://doi.org/10.1016/j.irbm.2009.10.006 5] J. Zhao and M. W. Spong, "Hybrid control for global stabilization of the cart–pendulum system," Automatica, v 37, no. 12, pp. 1941–1951, 2001. https://doi.org/10.1016/S0005-1098(01)00164-9 pp p g [26] S. Mahjoub, F. Mnif, and N. Derbel, "Second-order sliding mode control applied to inverted pendulum," in International Conference on Sciences and Techniques of Automatic Control and Computer Engineering, 2013, pp. 269–273. https://doi.org/10.1109/STA.2013.6783142 p g [27] S. Mahjoub, F. Mnif, and N. Derbel, "Second-order sliding mode approaches for the control of a class of underactuated systems," Int. J. Autom. Comput., vol. 12, no. 2, pp. 134–141, 2015. https://doi.org/10.1007/s11633- 015-0880-3 [28] M. Yue, C. An, L. Ding, and Y. Zhou, "MPC motion planning-based sliding mode control for underactuated WPS vehicle via Olfati transformation," IET Control Theory Appl., vol. 12, no. 4, pp. 495–503, 2018. https://doi.org/10.1049/iet-cta.2017.0298 [29] A. Hfaiedh, A. Chemori, and A. REFERENCES 45, no. 3, pp. 1629–1638, 2020. https://doi.org/10.1007/s13369-019-04126-9 [8] E. Susanto, A. Surya Wibowo, and E. Ghiffary Rachman, "Fuzzy Swing Up Control and Optimal State Feedback Stabilization for Self-Erecting Inverted Pendulum," IEEE Access, vol. 8, pp. 6496–6504, 2020. https://doi.org/10.1109/ACCESS.2019.2963399 Systems, Arab. J. Sci. Eng., vol. 45, no. 3, pp. 1629–1638, 2020. https://doi.org/10.1007/s13369-019-04126-9 [8] E. Susanto, A. Surya Wibowo, and E. Ghiffary Rachman, "Fuzzy Swing Up Control and Optimal State Feedback Stabilization for Self-Erecting Inverted Pendulum," IEEE Access, vol. 8, pp. 6496–6504, 2020. https://doi.org/10.1109/ACCESS.2019.2963399 [9] Y. Zheng, X. Li, and L. Xu, "Balance Control for the First-order Inverted Pendulum Based on the Advantage Actor-critic Algorithm," Int. J. Control. Autom. Syst., vol. 18, no. 12, pp. 3093–3100, 2020. https://doi.org/10.1007/s12555-019-0278-z p g [10] Y. Rizal, J.-W. Syu, and M.-T. Ho, "Balance control of an inverted pendulum system using second-order s Design of an Adaptive Super-Twisting Control for the Cart-Pole Inverted Pendulum System (Yusie Rizal) 173 Jurnal Ilmiah Teknik Elektro Komputer dan Informatika (JITEKI) Vol. 7, No. 1, April 2021, pp. 161-174 ISSN 2338-3070 mode control," in International Automatic Control Conference, 2014, pp. 191–196. https://doi.org/10.1109/CACS.2014.7097186 mode control," in International Automatic Control Conference, 2014, pp. 191–19 https://doi.org/10.1109/CACS.2014.7097186 p g [11] M. Ashok Kumar and S. Kanthalakshmi, "H∞ tracking control for an inverted pendulum," JVC/Journal Vib. Control, vol. 24, no. 16, pp. 3515–3524, 2018. https://doi.org/10.1177/1077546317750977 [12] M. Ramirez-Neria, Z. Gao, H. Sira-Ramirez, R. Garrido-Moctezuma, and A. Luviano-Juarez, "Trajectory Tracking for an Inverted Pendulum on a Cart: An Active Disturbance Rejection Control Approach," in American Control Conference, 2018, vol. 2018-June, pp. 4881–4886. https://doi.org/10.23919/ACC.2018.8431712 f pp p g [13] A. K. Jayaprakash, K. B. Kidambi, W. Mackunis, S. V. Drakunov, and M. Reyhanoglu, "Finite-time state estimation for an inverted pendulum under input-multiplicative uncertainty," Robotics, vol. 9, no. 4, pp. 1–26, 2020. https://doi.org/10.3390/robotics9040087 [14] R. G. Guerra, R. Iriarte, J. Eduardo, V. Velázquez, and L. Fridman, "Robust output trajectory linearisation control for a class of linear time-varying systems," IET Control Theory Appl., no. December 2020, pp. 877–889, 2021. https://doi.org/10.1049/cth2.12090 [15] S. Mobayen, "Adaptive global sliding mode control of underactuated systems using a super-twisting scheme: an experimental study," JVC/Journal Vib. Control, vol. 25, no. 16, pp. 2215–2224, 2019. https://doi.org/10.1177/1077546319852257 p g [16] R. Olfati-Saber, "Global configuration stabilization for the VTOL aircraft with strong input coupling," IEEE Trans. Automat. Contr., vol. 47, no. 11, pp. 1949–1951, 2002. https://doi.org/10.1109/TAC.2002.804457 [17] R. Olfati-saber, "Normal forms for underactuated mechanical systems with symmetry," IEEE Trans. Automat. Design of an Adaptive Super-Twisting Control for the Cart-Pole Inverted Pendulum System (Yusie Rizal) REFERENCES Abdelkrim, "Disturbance Observer-Based Super-Twisting Control for the Inertia Wheel Inverted Pendulum," in Proceedings of the 17th International Multi-Conference on Systems, Signals and Devices, SSD 2020, 2020, pp. 747–752. https://doi.org/10.1109/SSD49366.2020.9364233 , , , pp p g [30] B. Lu, Y. Fang, and N. Sun, "Continuous Sliding Mode Control Strategy for a Class of Nonlinear Underactuated Systems," IEEE Trans. Automat. Contr., vol. 63, no. 10, pp. 3471–3478, 2018. https://doi.org/10.1109/TAC.2018.2794885 p g [31] S. Rajappa, C. Masone, H. H. Bulthoff, and P. Stegagno, "Adaptive Super Twisting Controller for a quadrotor UAV," in IEEE International Conference on Robotics and Automation, 2016, pp. 2971–2977. https://doi.org/10.1109/ICRA.2016.7487462 p g [32] M. Derbeli, M. Farhat, O. Barambones, and L. Sbita, "Control of PEM fuel cell power system using sliding mode and super-twisting algorithms," Int. J. Hydrogen Energy, vol. 42, no. 13, pp. 8833–8844, 2017. https://doi.org/10.1016/j.ijhydene.2016.06.103 p g j j y [33] A. R. Babaei, M. Malekzadeh, and D. Madhkhan, "Adaptive super-twisting sliding mode control of 6-DOF nonlinear and uncertain air vehicle," Aerosp. Sci. Technol., vol. 84, no. September, pp. 361–374, 2019. https://doi.org/10.1016/j.ast.2018.09.013 Design of an Adaptive Super-Twisting Control for the Cart-Pole Inverted Pendulum System (Yusie Rizal) 174 Jurnal Ilmiah Teknik Elektro Komputer dan Informatika (JITEKI) Vol. 7, No. 1, April 2021, pp. 161-174 ISSN 2338-3070 [34] J. Liu and J. Liu, "Chapter 9 – Sliding mode control for underactuated system with decoupling algorithm," in Sliding Mode Control Using MATLAB, Academic Press, 2017, pp. 307–327. https://doi.org/10.1016/B978-0-12- 802575-8.00009-6 [34] J. Liu and J. Liu, "Chapter 9 – Sliding mode control for underactuated system with decoupling algorithm," in Sliding Mode Control Using MATLAB, Academic Press, 2017, pp. 307–327. https://doi.org/10.1016/B978-0-12- 802575-8.00009-6 [35] Z. Feng and J. Fei, "Design and analysis of adaptive Super- Twisting sliding mode control for a microgyroscope," J. Pone, vol. 13, no. 1, pp. 1–18, 2018. https://doi.org/10.1371/journal.pone.0189457 g j [36] A. Levant, "Sliding order and sliding accuracy in sliding mode control," Int. J. Control, vol. 58, no. 6, pp. 1247– 1263, 1993. https://doi.org/10.1080/00207179308923053 Design of an Adaptive Super-Twisting Control for the Cart-Pole Inverted Pendulum System (Yusie Rizal) BIOGRAPHY OF AUTHORS Yusie Rizal is a lecturer in Electronics engineering, Politeknik Negeri Banjarmasin, Banjarmasin, Indonesia. Email: yusie.rizal@poliban.ac.id Muhammad Wahyu is a Lecturer in Electronics engineering, Politeknik Negeri Banjarmasin, Banjarmasin, Indonesia. Email: m.wahyu@poliban.ac.id Imansyah Noor is a lecturer in Electronics engineering, Politeknik Negeri Banjarmasin, Banjarmasin, Indonesia. Email: imansyah.poliban@gmail.com Joni Riadi is a lecturer in Electrical engineering, Politeknik Negeri Banjarmasin, Banjarmasin, Indonesia. Email: joni_riadi@poliban.ac.id Feriyadi is a lecturer in Electrical engineering, Politeknik Negeri Banjarmasin, Banjarmasin, Indonesia. Email: feriyadi.marzuki@gmail.com Ronny Mantala is a lecturer in Informatics management, Politeknik Negeri Banjarmasin, Banjarmasin, Indonesia. Email: rmantala@poliban.ac.id Ronny Mantala is a lecturer in Informatics management, Politeknik Negeri Banjarmasin, Banjarmasin, Indonesia. Email: rmantala@poliban.ac.id
https://openalex.org/W3022145528	https://dspace.mit.edu/bitstream/1721.1/135543.2/4/A%20brief%20review%20of%20genome%20editing%20technology%20for%20generating%20animal%20models.pdf	English	null	A brief review of genome editing technology for generating animal models	Frontiers of Agricultural Science and Engineering	2,020	cc-by	5,407	Front. Agr. Sci. Eng. 2020, 7(2): 123–128 https://doi.org/10.15302/J-FASE-2019309 Front. Agr. Sci. Eng. 2020, 7(2): 123–128 https://doi.org/10.15302/J-FASE-2019309 Available online at http://journal.hep.com.cn/fase by Higher Education Press. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0) A brief review of genome editing technology for generating animal models Haoyi WANG (✉✉)1,2,3, Sen WU (✉✉)4,5, Mario R. CAPECCHI (✉✉)6, Rudolf JAENISCH (✉✉)7,8 Haoyi WANG (✉✉)1,2,3, Sen WU (✉✉)4,5, Mario R. CAPECCHI (✉✉)6, Rudolf JAENISCH (✉✉)7,8 ✉✉)1,2,3, Sen WU (✉✉)4,5, Mario R. CAPECCHI (✉✉)6, Rudolf JAENISCH (✉✉)7,8 Haoyi WANG (✉✉)1,2,3, Sen WU (✉✉)4,5, Mario R. CAPECCHI (✉✉)6, Rudolf JAE State Key Laboratory of Stem Cell and Reproductive Biology, Institute of Zoology, Chinese Academy of Sciences, Beijing 100101, China 2 University of the Chinese Academy of Sciences, Beijing 100049, China 2 University of the Chinese Academy of Sciences, Beijing 100049, China 3 Institute for Stem Cell and Regeneration, Chinese Academy of Sciences, Beijing 100101, China 4 Beijing Advanced Innovation Center for Food Nutrition and Human Health, China Agricultural University, Beijing 100193, China 5 State Key Laboratory of Agrobiotechnology, College of Biological Sciences, China Agricultural University, Beijing 100193, China 6 Department of Human Genetics, University of Utah School of Medicine, Salt Lake City, UT 84112, USA 7 Whitehead Institute for Biomedical Research, Cambridge, MA 02142, USA , g , , 8 Department of Biology, Massachusetts Institute of Technology, Cambridge, MA 02142, USA DNA fragments become integrated with low efﬁciency into the genome of cells, leading to the addition of constructed genetic information, although at random locations. The ﬁrst demonstration that exogenous DNA could be introduced into early mammalian embryos was made by Jaenisch and Mintz in 1974[1], followed by the demonstration of germ line transmission of the exogenous DNA[2]. Subsequent efforts of Gordon et al. showed that linear DNA fragments injected into the pronucleus of mouse embryos can lead to the generation of transgenic animals[3]. This technology is still being actively used for expressing exogenous genes in the mouse and other species. While transgenic animals are useful for expressing a particular transgene and to generate insertional muta- tions[4], integration of the DNA into the genome is random and thus this approach cannot be used to disrupt speciﬁc genes. Abstract The recent development of genome editing technologies has given researchers unprecedented power to alter DNA sequences at chosen genomic loci, thereby generating various genetically edited animal models. This mini-review brieﬂy summarizes the development of major genome editing tools, focusing on the application of these tools to generate animal models in multiple species. Keywords animal model, CRISPR, genome editing, TALEN, ZFN Keywords animal model, CRISPR, genome editing, TALEN, ZFN With the coordination of zinc ions, each zinc ﬁnger domain recognizes a 3–4 bp DNA sequence, and a combination of several ﬁngers allows for the recognition of a longer sequence. Zinc ﬁnger nuclease (ZFN) is generated by fusing a zinc ﬁnger binding domain with the FokI nuclease domain. A pair of ZFNs binds to two proximal sites next to each other, leading to the homodimerization of FokI, generating site-speciﬁc DNA DSBs in the genome. ZFNs have been used to generate various indel mutations as well as to stimulate homology- directed repair (HDR) in different species including human cells[18]. Upon their injection into the nuclei of Xenopus oocytes, a pair of ZFNs binding in the opposite direction induced DNA cleavage efﬁciently, with the linker between FokI and ZF domains constraining the spacing of the two binding sites[19]. Further, by introducing heat-shock promoter-driven ZFN transgenes into the Drosophila germline, ZFNs efﬁciently knocked out endogenous genes[20] or increased the frequency of targeted integration of transgenes into the chromosome[21]. Several studies in 2008 showed that direct injection of ZFNs into embryos of Drosophila and zebraﬁsh could lead to the efﬁcient generation of gene knockout (KO) animals[22–24]. In 2009, Geurts et al. showed that microinjection of ZFN into rat zygotes led to the generation of a gene KO rat[25], demonstrating a powerful strategy to generate gene-edited animals without limitation due to HR efﬁciency or availability of germline contributing embryonic stem cells. Although a variety of strategies have been established to generate ZFPs with designed speciﬁcity, they often require Although TALENs were much easier to generate than ZFNs, CRISPR (clustered regularly interspaced short palindromic repeat DNA sequences)-Cas9 readily out- performed all previous tools for generating designed double strand breaks and quickly became the method of choice for genome editing. CRISPR-Cas systems are adaptive immune defense mechanisms protecting against invading nucleic acids such as plasmid and phage infection in a large proportion of bacterial and archaeal species[41]. Through decades of work, people dissected the mechanism of the CRISPR-Cas system[42], and the biochemical characterization of the Streptococcus pyogenes and Streptococcus thermophilus CRISPR systems demon- strated that Cas9 is an RNA guided DNA endonuclease that targets speciﬁc DNA sequences complementary to the 20 nucleotide sequence residing at the 5′ end of the guide RNA[43,44]. Upon appropriate codon optimization, CRISPR-Cas9 was able to efﬁciently target speciﬁc genomic loci in mammalian cells[45,46]. Keywords animal model, CRISPR, genome editing, TALEN, ZFN In 2009, two groups independently deciphered the mechanism of transcription activator-like effectors (TALEs) recognizing target DNA sequences and found a remarkably simple modular recognition code[26,27]. Depending on the repeat-variable di-residue (RVD), each module of the highly repetitive TALE DNA binding domain can speciﬁcally recognize one single DNA base pair. Based on the RVD-nucleotide association, a TALE DNA binding protein could be generated to recognize a sequence of choice by simply linking different modules together. The TALE nuclease (TALEN) system was readily established by replacing the ZF domain with the TALE DNA binding domain. The principle of TALEN is very similar to ZFN with the only difference being that the DNA binding speciﬁcity is determined by the array of TALE motifs. Soon afterwards, TALEN was shown to work efﬁciently in human cells[28,29]. By injecting TALEN into zygotes, animals with speciﬁc gene KO were generated in various species including non-human primates[30–34]. Co-injection of donor template into zygotes leads to efﬁcient HDR, generating animals with precise nucleotide change and targeted gene integration[35–37]. Editing mESCs using TALENs gener- ated the ﬁrst mouse containing targeted mutations on the Y chromosome[38]. TALEN was also used in livestock with well-established cloning technology to generate speciﬁc mutations in somatic cell lines which were then cloned to generate genetically-modiﬁed animals[39,40]. First in yeast and then in mammalian cells, pioneering work on DNA double-stranded break (DSB) repair demonstrated that site-speciﬁc DSB in the genome could stimulate the rate of local HR by several orders of magnitude[12,13]. Also, it was shown that, once the DSB was introduced, in addition to the HR pathway the non- homologous end joining (NHEJ) pathway could repair the DSB, often leading to the introduction of small insertions and/or deletions (indels) at target sites. These results motivated the development of methods that generate DNA DSB at a speciﬁc locus in the genome. The earlier efforts were aimed at engineering a class of rare-cutting endonuclease, meganucleases, that recognize long stretches of DNA sequences[14]. However, it has been quite difﬁcult to engineer naturally occurring meganu- cleases to bind and cleave chosen DNA target sequences and this has greatly limited their wider application. Zinc ﬁnger proteins (ZFPs) are one of the most abundant types of proteins in eukaryotes and the largest transcription factor family in human[15–17]. Keywords animal model, CRISPR, genome editing, TALEN, ZFN Animal models are invaluable tools for understanding biology and developing therapies for human diseases. The capability of making precise alterations at chosen genomic loci in cells, or the whole animal, has been established and improved upon at an unprecedented speed, thanks to the development of genome editing technology. This mini- review brieﬂy introduces the key developments in genome editing technology and its applications to animal model creation. Given that this is a brief review, it is not possible to include all important contributions here, and we apologize to those whose work has been unintentionally omitted. The seminal work of Smithies et al.[5] and Thomas et al.[6] showed that an exogenous DNA fragment could be precisely integrated into the desired genomic locus via homologous recombination (HR), a cell-intrinsic DNA repair mechanism. Although the efﬁciency of this gene- targeting technology[7] was initially low, the rare clones containing the desired targeting event could be enriched and selected by a variety of elegantly designed strategies, leading to the derivation of clonal cell lines containing the desired genetic modiﬁcation. The combination of gene- targeting and mouse embryonic stem cell (mESC) technologies[8] made it possible to generate a mouse composed entirely of cells containing the designed predetermined genetic modiﬁcation[9,10]. While this approach transformed modern biology and resulted in the award of a Nobel Prize, its application was limited to mice and more recently to rats[11], mainly because embryonic The capability of precisely editing any position in the genome has long been a dream of biologists. Recombinant DNA technology has enabled scientists to engineer DNA molecules in the test tube with great precision since the 1970s. Transgenes can be expressed by introducing engineered DNA molecules into different cells. These Received January 1, 2020; accepted January 27, 2020 Correspondences: wanghaoyi@ioz.ac.cn, swu@cau.edu.cn, mario.capecchi@genetics.utah.edu, jaenisch@wi.mit.edu Correspondences: wanghaoyi@ioz.ac.cn, swu@cau.edu.cn, mario.capecchi@genetics.utah.edu, jaenisch@wi.mit.edu Front. Agr. Sci. Eng. 2020, 7(2): 123–128 124 stem cells with germ line contribution capability have not been established in most mammalian species. Also, gene targeting in most human cell types is inefﬁcient, thereby limiting its application for genetic correction-based therapy. A means of increasing the efﬁciency of HR therefore became a key question for improving the capability of modifying genomes of all life forms. signiﬁcant expertise and intensive screening. This has limited the wider adoption of ZFN technology within the scientiﬁc community. References 1. Jaenisch R, Mintz B. Simian virus 40 DNA sequences in DNA of healthy adult mice derived from preimplantation blastocysts injected with viral DNA. Proceedings of the National Academy of Sciences of the United States of America, 1974, 71(4): 1250–1254 2. Jaenisch R. Germ line integration and Mendelian transmission of the exogenous Moloney leukemia virus. Proceedings of the National Academy of Sciences of the United States of America, 1976, 73(4): 1260–1264 3. Gordon J W, Scangos G A, Plotkin D J, Barbosa J A, Ruddle F H. Genetic transformation of mouse embryos by microinjection of puriﬁed DNA. Proceedings of the National Academy of Sciences of the United States of America, 1980, 77(12): 7380–7384 One of the advantages of the CRISPR-Cas9 system is the great ﬂexibility of repurposing its function. Cas9 has two nuclease domains, each capable of cleaving one of the target DNA strands. When either one of these domains is mutated, the Cas9-sgRNA complex becomes a sequence and strand-speciﬁc nickase; and when both nuclease domains are mutated, Cas9 becomes a programmable DNA binding protein (dCas9) without any endonuclease activity. Guided by sgRNA, dCas9 fused with different effector domains can bind to regulatory elements and regulate transcription, as well as epigenetic modiﬁca- tions[62,63]. In particular for genome editing, dCas9 or Cas9 nickase fused with single-stranded DNA deaminase was developed into programmable DNA base editors, which can make speciﬁc nucleotide changes without introducing DNA DSB or replying on HDR[64]. To construct cytosine base editors, single-stranded DNA cytosine deaminase was used to mediate C$G to T$A conversions[65,66]; to generate adenine base editors capable of converting A$T to G$C, the DNA adenine deaminase was artiﬁcially evolved from the bacterial tRNA-speciﬁc adenosine deaminase TadA[67]. Following the demonstration of efﬁcient base editing in cultured cells, multiple groups introduced base editors into the embryos of various species and demonstrated efﬁcient generation of base-edited animal models[68–73]. These studies showed that, in general, base editing generated animals carrying precise single nucleotide editing more efﬁciently than previous HDR based genome editing strategies. 4. Schnieke A, Harbers K, Jaenisch R. Embryonic lethal mutation in mice induced by retrovirus insertion into the α1(I) collagen gene. Nature, 1983, 304(5924): 315–320 5. Smithies O, Gregg R G, Boggs S S, Koralewski M A, Kucherlapati R S. Insertion of DNA sequences into the human chromosomal β- globin locus by homologous recombination. Nature, 1985, 317 (6034): 230–234 6. Keywords animal model, CRISPR, genome editing, TALEN, ZFN After development to spermatids and injection into oocytes, animals with speciﬁc genetic modiﬁcation were derived[60]. By combin- ing the gene editing and haploid stem cell technologies, an artiﬁcial sperm strategy was established as an efﬁcient method to generate gene-edited mice[61]. Whether these strategies can be transferred to other species remains to be tested. Compliance with ethics guidelines Haoyi Wang, Sen Wu, Mario R. Capecchi, and Rudolf Jaenisch declare that they have no conﬂicts of interest or ﬁnancial conﬂicts to disclose. This article is a mini-review and does not contain any studies with human or animal subjects performed by any of the authors. Keywords animal model, CRISPR, genome editing, TALEN, ZFN By simply designing a 20-nucleotide sequence within single guide RNA (sgRNA), the Cas9 nuclease-sgRNA complex can generate DNA DSBs at a speciﬁc genomic locus. The ease of use and robust performance of the CRISPR-Cas9 system resulted in its rapid adoption by the scientiﬁc community. The efﬁciency of the CRISPR-Cas9 system makes it particularly attractive for performing multiplex gene p p y y The efﬁciency of the CRISPR-Cas9 system makes it particularly attractive for performing multiplex gene Haoyi WANG et al. Review of genome editing technology 125 speciﬁc transposon systems[75,76] promise to make even more sophisticatedly-edited animal models more efﬁ- ciently in the future. editing. By co-expressing multiple sgRNAs, Cas9 can generate DSBs at multiple chosen loci within the same cell[46–48]. Following the previous principle established using ZFN and TALEN, injection of CRISPR-Cas9 into the zygote leads to highly efﬁcient generation of genetically-modiﬁed animals. Mice carrying multiple gene KOs or precise nucleotide changes can be derived within a month[47,49]. A similar approach produced genetically edited animals from ever-increasing numbers of species[50–54]. To make the process even simpler, several groups developed the protocols to deliver CRISPR-Cas9 components into rat and mouse zygotes by electroporation, which has been successfully adopted by the ﬁeld[55–59]. In addition to modifying the genome of the zygote, germline modiﬁcation has also been achieved in mice by genome editing of spermatogonial stem cells. After development to spermatids and injection into oocytes, animals with speciﬁc genetic modiﬁcation were derived[60]. By combin- ing the gene editing and haploid stem cell technologies, an artiﬁcial sperm strategy was established as an efﬁcient method to generate gene-edited mice[61]. Whether these strategies can be transferred to other species remains to be tested. editing. By co-expressing multiple sgRNAs, Cas9 can generate DSBs at multiple chosen loci within the same cell[46–48]. Following the previous principle established using ZFN and TALEN, injection of CRISPR-Cas9 into the zygote leads to highly efﬁcient generation of genetically-modiﬁed animals. Mice carrying multiple gene KOs or precise nucleotide changes can be derived within a month[47,49]. A similar approach produced genetically edited animals from ever-increasing numbers of species[50–54]. To make the process even simpler, several groups developed the protocols to deliver CRISPR-Cas9 components into rat and mouse zygotes by electroporation, which has been successfully adopted by the ﬁeld[55–59]. In addition to modifying the genome of the zygote, germline modiﬁcation has also been achieved in mice by genome editing of spermatogonial stem cells. References Repetitive zinc-binding domains in the protein transcription factor IIIA from Xenopus oocytes. EMBO Journal, 1985, 4(6): 1609–1614 16. Najafabadi H S, Mnaimneh S, Schmitges F W, Garton M, Lam K N, Yang A, Albu M, Weirauch M T, Radovani E, Kim P M, Greenblatt J, Frey B J, Hughes T R. C2H2 zinc ﬁnger proteins greatly expand the human regulatory lexicon. Nature Biotechnology, 2015, 33(5): 555–562 30. Tesson L, Usal C, Ménoret S, Leung E, Niles B J, Remy S, Santiago Y, Vincent A I, Meng X, Zhang L, Gregory P D, Anegon I, Cost G J. Knockout rats generated by embryo microinjection of TALENs. Nature Biotechnology, 2011, 29(8): 695–696 17. Takatsuji H. Zinc-ﬁnger transcription factors in plants. Cellular and Molecular Life Sciences, 1998, 54(6): 582–596 31. Sander J D, Cade L, Khayter C, Reyon D, Peterson R T, Joung J K, Yeh J R J. Targeted gene disruption in somatic zebraﬁsh cells using engineered TALENs. Nature Biotechnology, 2011, 29(8): 697–698 18. Urnov F D, Rebar E J, Holmes M C, Zhang H S, Gregory P D. Genome editing with engineered zinc ﬁnger nucleases. Nature Reviews: Genetics, 2010, 11(9): 636–646 32. Huang P, Xiao A, Zhou M, Zhu Z, Lin S, Zhang B. Heritable gene targeting in zebraﬁsh using customized TALENs. Nature Biotech- nology, 2011, 29(8): 699–700 19. Bibikova M, Carroll D, Segal D J, Trautman J K, Smith J, Kim Y G, Chandrasegaran S. Stimulation of homologous recombination through targeted cleavage by chimeric nucleases. Molecular and Cellular Biology, 2001, 21(1): 289–297 33. Qiu Z, Liu M, Chen Z, Shao Y, Pan H, Wei G, Yu C, Zhang L, Li X, Wang P, Fan H Y, Du B, Liu B, Liu M, Li D. High-efﬁciency and heritable gene targeting in mouse by transcription activator-like effector nucleases. Nucleic Acids Research, 2013, 41(11): e120 20. Bibikova M, Golic M, Golic K G, Carroll D. Targeted chromosomal cleavage and mutagenesis in Drosophila using zinc-ﬁnger nucleases. Genetics, 2002, 161(3): 1169–1175 34. Liu H, Chen Y, Niu Y, Zhang K, Kang Y, Ge W, Liu X, Zhao E, Wang C, Lin S, Jing B, Si C, Lin Q, Chen X, Lin H, Pu X, Wang Y, Qin B, Wang F, Wang H, Si W, Zhou J, Tan T, Li T, Ji S, Xue Z, Luo Y, Cheng L, Zhou Q, Li S, Sun Y E, Ji W. TALEN-mediated gene mutagenesis in rhesus and cynomolgus monkeys. References Cell Stem Cell, 2014, 14(3): 323–328 21. Bibikova M, Beumer K, Trautman J K, Carroll D. Enhancing gene targeting with designed zinc ﬁnger nucleases. Science, 2003, 300 (5620): 764 22. Beumer K J, Trautman J K, Bozas A, Liu J L, Rutter J, Gall J G, Carroll D. Efﬁcient gene targeting in Drosophila by direct embryo injection with zinc-ﬁnger nucleases. Proceedings of the National Academy of Sciences of the United States of America, 2008, 105 (50): 19821–19826 35. Remy S, Tesson L, Menoret S, Usal C, De Cian A, Thepenier V, Thinard R, Baron D, Charpentier M, Renaud J B, Buelow R, Cost G J, Giovannangeli C, Fraichard A, Concordet J P, Anegon I. Efﬁcient gene targeting by homology-directed repair in rat zygotes using TALE nucleases. Genome Research, 2014, 24(8): 1371–1383 23. Doyon Y, McCammon J M, Miller J C, Faraji F, Ngo C, Katibah G E, Amora R, Hocking T D, Zhang L, Rebar E J, Gregory P D, Urnov F D, Amacher S L. Heritable targeted gene disruption in zebraﬁsh using designed zinc-ﬁnger nucleases. Nature Biotechnology, 2008, 26(6): 702–708 36. Wefers B, Meyer M, Ortiz O, Hrabé de Angelis M, Hansen J, Wurst W, Kühn R. Direct production of mouse disease models by embryo microinjection of TALENs and oligodeoxynucleotides. Proceedings of the National Academy of Sciences of the United States of America, 2013, 110(10): 3782–3787 24. Meng X, Noyes M B, Zhu L J, Lawson N D, Wolfe S A. Targeted gene inactivation in zebraﬁsh using engineered zinc-ﬁnger nucleases. Nature Biotechnology, 2008, 26(6): 695–701 37. Bedell V M, Wang Y, Campbell J M, Poshusta T L, Starker C G, Krug R G 2nd, Tan W, Penheiter S G, Ma A C, Leung A Y H, Fahrenkrug S C, Carlson D F, Voytas D F, Clark K J, Essner J J, Ekker S C. In vivo genome editing using a high-efﬁciency TALEN system. Nature, 2012, 491(7422): 114–118 25. Geurts A M, Cost G J, Freyvert Y, Zeitler B, Miller J C, Choi V M, Jenkins S S, Wood A, Cui X, Meng X, Vincent A, Lam S, Michalkiewicz M, Schilling R, Foeckler J, Kalloway S, Weiler H, Ménoret S, Anegon I, Davis G D, Zhang L, Rebar E J, Gregory P D, Urnov F D, Jacob H J, Buelow R. Knockout rats via embryo microinjection of zinc-ﬁnger nucleases. Science, 2009, 325(5939): 433 38. References Thomas K R, Folger K R, Capecchi M R. High frequency targeting of genes to speciﬁc sites in the mammalian genome. Cell, 1986, 44 (3): 419–428 7. Capecchi M R. Gene targeting in mice: functional analysis of the mammalian genome for the twenty-ﬁrst century. Nature Reviews: Genetics, 2005, 6(6): 507–512 8. Bradley A, Evans M, Kaufman M H, Robertson E. Formation of germ-line chimaeras from embryo-derived teratocarcinoma cell lines. Nature, 1984, 309(5965): 255–256 9. Thomas K R, Capecchi M R. Site-directed mutagenesis by gene targeting in mouse embryo-derived stem cells. Cell, 1987, 51(3): 503–512 10. Doetschman T, Gregg R G, Maeda N, Hooper M L, Melton D W, Thompson S, Smithies O. Targetted correction of a mutant HPRT gene in mouse embryonic stem cells. Nature, 1987, 330(6148): 576– 578 11. Buehr M, Meek S, Blair K, Yang J, Ure J, Silva J, McLay R, Hall J, Ying Q L, Smith A. Capture of authentic embryonic stem cells from rat blastocysts. Cell, 2008, 135(7): 1287–1298 12. Rouet P, Smih F, Jasin M. Introduction of double-strand breaks into the genome of mouse cells by expression of a rare-cutting endonuclease. Molecular and Cellular Biology, 1994, 14(12): 8096–8106 Generation of genetically edited animal models is becoming much easier with rapidly-improving genome editing technology. New strategies based on recently developed prime-editing[74] and newly discovered site- 13. Rudin N, Sugarman E, Haber J E. Genetic and physical analysis of double-strand break repair and recombination in Saccharomyces Front. Agr. Sci. Eng. 2020, 7(2): 123–128 126 cerevisiae. Genetics, 1989, 122(3): 519–534 Ankoudinova I, Cost G J, Urnov F D, Zhang H S, Holmes M C, Zhang L, Gregory P D, Rebar E J. A TALE nuclease architecture for efﬁcient genome editing. Nature Biotechnology, 2011, 29(2): 143– 148 cerevisiae. Genetics, 1989, 122(3): 519–534 14. Silva G, Poirot L, Galetto R, Smith J, Montoya G, Duchateau P, Pâques F. Meganucleases and other tools for targeted genome engineering: perspectives and challenges for gene therapy. Current Gene Therapy, 2011, 11(1): 11–27 29. Hockemeyer D, Wang H, Kiani S, Lai C S, Gao Q, Cassady J P, Cost G J, Zhang L, Santiago Y, Miller J C, Zeitler B, Cherone J M, Meng X, Hinkley S J, Rebar E J, Gregory P D, Urnov F D, Jaenisch R. Genetic engineering of human pluripotent cells using TALE nucleases. Nature Biotechnology, 2011, 29(8): 731–734 15. Miller J, McLachlan A D, Klug A. References Wang H, Yang H, Shivalila C S, Dawlaty M M, Cheng A W, Zhang F, Jaenisch R. One-step generation of mice carrying mutations in multiple genes by CRISPR/Cas-mediated genome engineering. Cell, 2013, 153(4): 910–918 61. Wei L, Wang X, Yang S, Yuan W, Li J. Efﬁcient generation of the mouse model with a deﬁned point mutation through haploid cell- mediated gene editing. Journal of Genetics and Genomics, 2017, 44 (9): 461–463 48. Yang L, Güell M, Niu D, George H, Lesha E, Grishin D, Aach J, Shrock E, Xu W, Poci J, Cortazio R, Wilkinson R A, Fishman J A, Church G. Genome-wide inactivation of porcine endogenous retroviruses (PERVs). Science, 2015, 350(6264): 1101–1104 62. Adli M. The CRISPR tool kit for genome editing and beyond. Nature Communications, 2018, 9(1): 1911 63. Wang H, La Russa M, Qi L S. CRISPR/Cas9 in genome editing and beyond. Annual Review of Biochemistry, 2016, 85(1): 227–264 49. Yang H, Wang H, Shivalila C S, Cheng A W, Shi L, Jaenisch R. One-step generation of mice carrying reporter and conditional alleles by CRISPR/Cas-mediated genome engineering. Cell, 2013, 154(6): 1370–1379 64. Rees H A, Liu D R. Base editing: precision chemistry on the genome and transcriptome of living cells. Nature Reviews: Genetics, 2018, 19(12): 770–788 65. Nishida K, Arazoe T, Yachie N, Banno S, Kakimoto M, Tabata M, Mochizuki M, Miyabe A, Araki M, Hara K Y, Shimatani Z, Kondo A. Targeted nucleotide editing using hybrid prokaryotic and vertebrate adaptive immune systems. Science, 2016, 353(6305): aaf8729 50. Li W, Teng F, Li T, Zhou Q. Simultaneous generation and germline transmission of multiple gene mutations in rat using CRISPR-Cas systems. Nature Biotechnology, 2013, 31(8): 684–686 51. Li D, Qiu Z, Shao Y, Chen Y, Guan Y, Liu M, Li Y, Gao N, Wang L, Lu X, Zhao Y, Liu M. Heritable gene targeting in the mouse and rat using a CRISPR-Cas system. Nature Biotechnology, 2013, 31(8): 681–683 66. Komor A C, Kim Y B, Packer M S, Zuris J A, Liu D R. Programmable editing of a target base in genomic DNA without double-stranded DNA cleavage. Nature, 2016, 533(7603): 420–424 52. Hwang W Y, Fu Y, Reyon D, Maeder M L, Tsai S Q, Sander J D, Peterson R T, Yeh J R, Joung J K. Efﬁcient genome editing in zebraﬁsh using a CRISPR-Cas system. Nature Biotechnology, 2013, 31(3): 227–229 67. References Qin W, Dion S L, Kutny P M, Zhang Y, Cheng A W, Jillette N L, Malhotra A, Geurts A M, Chen Y G, Wang H. Efﬁcient CRISPR/ Cas9-mediated genome editing in mice by zygote electroporation of nuclease. Genetics, 2015, 200(2): 423–430 F, Yakunin A F, Garrett R A, van der Oost J, Backofen R, Koonin E V. An updated evolutionary classiﬁcation of CRISPR-Cas systems. 42. Jiang F, Doudna J A. CRISPR-Cas9 structures and mechanisms. Annual Review of Biophysics, 2017, 46(1): 505–529 57. Kaneko T, Sakuma T, Yamamoto T, Mashimo T. Simple knockout by electroporation of engineered endonucleases into intact rat embryos. Scientiﬁc Reports, 2014, 4(1): 6382 43. Gasiunas G, Barrangou R, Horvath P, Siksnys V. Cas9-crRNA ribonucleoprotein complex mediates speciﬁc DNA cleavage for adaptive immunity in bacteria. Proceedings of the National Academy of Sciences of the United States of America, 2012, 109 (39): E2579–E2586 58. Wang W, Kutny P M, Byers S L, Longstaff C J, DaCosta M J, Pang C, Zhang Y, Taft R A, Buaas F W, Wang H. Delivery of Cas9 protein into mouse zygotes through a series of electroporation dramatically increases the efﬁciency of model creation. Journal of Genetics and Genomics, 2016, 43(5): 319–327 44. Jinek M, Chylinski K, Fonfara I, Hauer M, Doudna J A, Charpentier E. A programmable dual-RNA-guided DNA endonuclease in adaptive bacterial immunity. Science, 2012, 337(6096): 816–821 59. Takahashi G, Gurumurthy C B, Wada K, Miura H, Sato M, Ohtsuka M. GONAD: genome-editing via Oviductal Nucleic Acids Delivery system: a novel microinjection independent genome engineering method in mice. Scientiﬁc Reports, 2015, 5(1): 11406 45. Mali P, Yang L, Esvelt K M, Aach J, Guell M, DiCarlo J E, Norville J E, Church G M. RNA-guided human genome engineering via Cas9. Science, 2013, 339(6121): 823–826 46. Cong L, Ran F A, Cox D, Lin S, Barretto R, Habib N, Hsu P D, Wu X, Jiang W, Marrafﬁni L A, Zhang F. Multiplex genome engineering using CRISPR/Cas systems. Science, 2013, 339 (6121): 819–823 60. Wu Y, Zhou H, Fan X, Zhang Y, Zhang M, Wang Y, Xie Z, Bai M, Yin Q, Liang D, Tang W, Liao J, Zhou C, Liu W, Zhu P, Guo H, Pan H, Wu C, Shi H, Wu L, Tang F, Li J. Correction of a genetic disease by CRISPR-Cas9-mediated gene editing in mouse spermatogonial stem cells. Cell Research, 2015, 25(1): 67–79 47. References Wang H, Hu Y C, Markoulaki S, Welstead G G, Cheng A W, Shivalila C S, Pyntikova T, Dadon D B, Voytas D F, Bogdanove A J, Page D C, Jaenisch R. TALEN-mediated editing of the mouse Y chromosome. Nature Biotechnology, 2013, 31(6): 530–532 26. Boch J, Scholze H, Schornack S, Landgraf A, Hahn S, Kay S, Lahaye T, Nickstadt A, Bonas U. Breaking the code of DNA binding speciﬁcity of TAL-type III effectors. Science, 2009, 326 (5959): 1509–1512 39. Tan W, Carlson D F, Lancto C A, Garbe J R, Webster D A, Hackett P B, Fahrenkrug S C. Efﬁcient nonmeiotic allele introgression in livestock using custom endonucleases. Proceedings of the National Academy of Sciences of the United States of America, 2013, 110(41): 16526–16531 27. Moscou M J, Bogdanove A J. A simple cipher governs DNA recognition by TAL effectors. Science, 2009, 326(5959): 1501 40. Carlson D F, Tan W, Lillico S G, Stverakova D, Proudfoot C, Christian M, Voytas D F, Long C R, Whitelaw C B, Fahrenkrug S C. Efﬁcient TALEN mediated gene knockout in livestock Proceedings 28. Miller J C, Tan S, Qiao G, Barlow K A, Wang J, Xia D F, Meng X, Paschon D E, Leung E, Hinkley S J, Dulay G P, Hua K L, Efﬁcient TALEN-mediated gene knockout in livestock. Proceedings Haoyi WANG et al. Review of genome editing technology 127 via Cas9/RNA-mediated gene targeting in one-cell embryos. Cell, 2014, 156(4): 836–843 of the National Academy of Sciences of the United States of America, 2012, 109(43): 17382–17387 of the National Academy of Sciences of the United States of America, 2012, 109(43): 17382–17387 41. Makarova K S, Wolf Y I, Alkhnbashi O S, Costa F, Shah S A, Saunders S J, Barrangou R, Brouns S J, Charpentier E, Haft D H, Horvath P, Moineau S, Mojica F J, Terns R M, Terns M P, White M F, Yakunin A F, Garrett R A, van der Oost J, Backofen R, Koonin E V. An updated evolutionary classiﬁcation of CRISPR-Cas systems. Nature Reviews: Microbiology, 2015, 13(11): 722–736 55. Hashimoto M, Takemoto T. Electroporation enables the efﬁcient mRNA delivery into the mouse zygotes and facilitates CRISPR/ Cas9-based genome editing. Scientiﬁc Reports, 2015, 5(1): 11315 Horvath P, Moineau S, Mojica F J, Terns R M, Terns M P, White M 56. References Gaudelli N M, Komor A C, Rees H A, Packer M S, Badran A H, Bryson D I, Liu D R. Programmable base editing of A$T to G$C in genomic DNA without DNA cleavage. Nature, 2017, 551(7681): 464–471 53. Xiang G, Ren J, Hai T, Fu R, Yu D, Wang J, Li W, Wang H, Zhou Q. Editing porcine IGF2 regulatory element improved meat production in Chinese Bama pigs. Cellular and Molecular Life Sciences, 2018, 75(24): 4619–4628 68. Liu Z, Lu Z, Yang G, Huang S, Li G, Feng S, Liu Y, Li J, Yu W, Zhang Y, Chen J, Sun Q, Huang X. Efﬁcient generation of mouse models of human diseases via ABE- and BE-mediated base editing. Nature Communications, 2018, 9(1): 2338 54. Niu Y, Shen B, Cui Y, Chen Y, Wang J, Wang L, Kang Y, Zhao X, Si W, Li W, Xiang A P, Zhou J, Guo X, Bi Y, Si C, Hu B, Dong G, Wang H, Zhou Z, Li T, Tan T, Pu X, Wang F, Ji S, Zhou Q, Huang X, Ji W, Sha J. Generation of gene-modiﬁed cynomolgus monkey 69. Liang P, Sun H, Sun Y, Zhang X, Xie X, Zhang J, Zhang Z, Chen Y, Ding C, Xiong Y, Ma W, Liu D, Huang J, Songyang Z. Effective gene editing by high-ﬁdelity base editor 2 in mouse zygotes. Protein Front. Agr. Sci. Eng. 2020, 7(2): 123–128 128 & Cell, 2017, 8(8): 601–611 & Cell, 2017, 8(8): 601–611 & Cell, 2017, 8(8): 601–611 & Cell, 2017, 8(8): 601–611 Zhang X, Liu M, Li D. Increasing targeting scope of adenosine base editors in mouse and rat embryos through fusion of TadA deaminase with Cas9 variants. Protein & Cell, 2018, 9(9): 814–819 70. Kim K, Ryu S M, Kim S T, Baek G, Kim D, Lim K, Chung E, Kim S, Kim J S. Highly efﬁcient RNA-guided base editing in mouse embryos. Nature Biotechnology, 2017, 35(5): 435–437 74. Anzalone A V, Randolph P B, Davis J R, Sousa A A, Koblan L W, Levy J M, Chen P J, Wilson C, Newby G A, Raguram A, Liu D R. Search-and-replace genome editing without double-strand breaks or donor DNA. Nature, 2019, 576(7785): 149–157 71. Ma Y, Yu L, Zhang X, Xin C, Huang S, Bai L, Chen W, Gao R, Li J, Pan S, Qi X, Huang X, Zhang L. Highly efﬁcient and precise base editing by engineered dCas9-guide tRNA adenosine deaminase in rats. Cell Discovery, 2018, 4(1): 39 75. Klompe S E, Vo P L H, Halpin-Healy T S, Sternberg S H. Transposon-encoded CRISPR-Cas systems direct RNA-guided DNA integration. Nature, 2019, 571(7764): 219–225 72. Liu Z, Chen M, Chen S, Deng J, Song Y, Lai L, Li Z. Highly efﬁcient RNA-guided base editing in rabbit. Nature Communica- tions, 2018, 9(1): 2717 76. Strecker J, Ladha A, Gardner Z, Schmid-Burgk J L, Makarova K S, Koonin E V, Zhang F. RNA-guided DNA insertion with CRISPR- associated transposases. Science, 2019, 365(6448): 48–53 73. Yang L, Zhang X, Wang L, Yin S, Zhu B, Xie L, Duan Q, Hu H, Zheng R, Wei Y, Peng L, Han H, Zhang J, Qiu W, Geng H, Siwko S,
https://openalex.org/W2121512815	https://bg.copernicus.org/articles/7/4037/2010/bg-7-4037-2010.pdf	English	null	Atmospheric nutrient inputs to the northern levantine basin from a long-term observation: sources and comparison with riverine inputs	Biogeosciences	2,010	cc-by	12,148	Biogeosciences, 7, 4037–4050, 2010 www.biogeosciences.net/7/4037/2010/ doi:10.5194/bg-7-4037-2010 © Author(s) 2010. CC Attribution 3.0 License. Biogeosciences, 7, 4037–4050, 2010 www.biogeosciences.net/7/4037/2010/ doi:10.5194/bg-7-4037-2010 © Author(s) 2010. CC Attribution 3.0 License. Biogeosciences Atmospheric nutrient inputs to the northern levantine basin from a long-term observation: sources and comparison with riverine inputs M. Koc¸ak1, N. Kubilay1, S. Tu˘grul1, and N. Mihalopoulos2 1Institute of Marine Sciences, Middle East Technical University, P.O. Box 28, 33731, Erdemli-Mersin, Turkey 2Environmental Chemical Processes Laboratory, Department of Chemistry, University of Crete, P.O. Box 2208, 71003 Voutes, Heraklion, Greece Received: 21 May 2010 – Published in Biogeosciences Discuss.: 1 July 2010 Revised: 22 November 2010 – Accepted: 6 December 2010 – Published: 17 December 2010 Received: 21 May 2010 – Published in Biogeosciences Discuss.: 1 July 2010 Revised: 22 November 2010 – Accepted: 6 December 2010 – Published: 17 December 2010 Received: 21 May 2010 – Published in Biogeosciences Discuss.: 1 July 2010 Revised: 22 November 2010 – Accepted: 6 December 2010 – Published: 17 December 2010 Abstract. Aerosol and rainwater samples have been col- lected at a rural site located on the coastline of the Eastern Mediterranean, Erdemli, Turkey between January 1999 and December 2007. Riverine sampling was carried out at ﬁve Rivers (Ceyhan, Seyhan, G¨oksu, Berdan and Lamas) drain- ing into the Northeastern Levantine Basin (NLB) between March 2002 and July 2007. Samples have been analyzed for macronutrients of phosphate, silicate, nitrate and ammo- nium (PO3− 4 , Sidiss, NO− 3 and NH+ 4 ). Phosphate and sili- cate in aerosol and rainwater showed higher and larger vari- ations during the transitional period when air ﬂows predomi- nantly originate from North Africa and Middle East/Arabian Peninsula. Deﬁciency of alkaline material have been found to be the main reason of the acidic rain events whilst high pH values (>7) have been associated with high Sidiss con- centrations due to sporadic dust events. In general, lowest nitrate and ammonium concentrations in aerosol and rain- water have been associated with air ﬂow from the Mediter- ranean Sea. Comparison of atmospheric with riverine ﬂuxes demonstrated that DIN and PO3− 4 ﬂuxes to NLB have been dominated by atmosphere (∼90% and ∼60% respectively) whereas the input of Si was mainly derived from riverine runoff (∼90%). N/P ratios in the atmospheric deposition (233); riverine discharge (28) revealed that NLB receives ex- cessive amounts of DIN and this unbalanced P and N inputs may provoke even more phosphorus deﬁciency. Observed molar Si/N ratio suggested Si limitation relative to nitrogen might cause a switch from diatom dominated communities to non-siliceous populations particularly at coastal NLB. 1 Introduction Koc¸ak et al.: Atmospheric nutrient inputs to the northern levantine basin from a long-term observation 4038 The Mediterranean has one of the highest concentrations of airborne mineral dust owning to its close proximity to arid regions, in particular the Saharan, Middle Eastern and Ara- bian Deserts (Lo¨ye-Pilot et al., 1986; Guerzoni et al., 1999; Kubilay et al., 2000; Koc¸ak et al., 2004a). Atmospheric de- position of desert dust over the Mediterranean supplies sol- uble or bioavailable macro (P, N) and micro nutrients (Fe) which inﬂuence ocean biogeochemistry (Herut et al., 1999, 2002, 2005; Guieu et al., 2002; Ridame and Guieu, 2002; Markaki et al., 2003, 2010; Bonnet et al., 2005; Carbo et al., 2005). The Mediterranean has one of the highest concentrations of airborne mineral dust owning to its close proximity to arid regions, in particular the Saharan, Middle Eastern and Ara- bian Deserts (Lo¨ye-Pilot et al., 1986; Guerzoni et al., 1999; Kubilay et al., 2000; Koc¸ak et al., 2004a). Atmospheric de- position of desert dust over the Mediterranean supplies sol- uble or bioavailable macro (P, N) and micro nutrients (Fe) which inﬂuence ocean biogeochemistry (Herut et al., 1999, 2002, 2005; Guieu et al., 2002; Ridame and Guieu, 2002; Markaki et al., 2003, 2010; Bonnet et al., 2005; Carbo et al., 2005). cipitation. The transition seasons, spring and autumn are of very different lengths. The relatively long spring sea- son (March–May) is characterized by unsettled winter type weather, associated with North African cyclones; the rest of this period is very similar to that in the summer. Autumn usually lasts one month (October) and is characterized by an abrupt from the summer to the unsettled weather of win- ter (Brody and Nestor, 1980). Therefore, within the annual period three seasons were deﬁned; winter, transitional, and summer. The winter period included the months November, December, January and February whereas the transitional season included the months March, April, May and October. The summer season included the months June, July, August and September. Atmospheric inputs of nutrients to the coastal system and the open ocean can take place through dry and wet (i.e. rain) deposition. According to Guerzoni et al., (1999), atmo- spheric input of inorganic nitrogen species is sufﬁcient to the nitrogen requirement in 60% of the total nitrogen enter- ing the Mediterranean from continental origin, 66% of which is via wet deposition. Kouvarakis et al. 1 Introduction (2001) suggest that the input of inorganic N species in atmospheric deposition is enough to explain nitrogen needs in the eastern Mediter- ranean Sea. Unlike N compounds which have dominant an- thropogenic sources (Spokes and Jickells, 2005) the aerosol P content and Si are of continental/natural origin (e.g. rock and soil) (Herut et al., 1999; Markaki et al., 2003; Baker et al., 2007). Aerosol and rainwater sampling was carried out at a rural site located on the coastline of the Eastern Mediterranean, Erdemli (36◦33′54′′ N and 34◦15′18′′ E), Turkey (Fig. 1). The rural sampling site is not under direct inﬂuence of any industrial activities. Its immediate vicinity is surrounded by lemon trees, cultivated land and greenhouses. The city of Mersin (population ∼800 000) is located 45 km to the east of the sampling site. The samplers were positioned on a sam- pling tower (at an altitude of ∼22 m) which is situated at the Institute of Marine Sciences, Middle East Technical Univer- sity (for more details see Kubilay and Saydam, 1995; Koc¸ak et al., 2004b). Samples were collected from January 1999 to December 2007. A great number of data on atmospheric nutrient ﬂuxes in Eastern Mediterranean have been published (Guerzoni et al., 1999; Herut et al., 1999; Kouvarakis et al., 2001; Kubilay et al., 2000; Markaki et al., 2003; Krom et al., 2004; Carbo et al., 2005). However, there is no available data on riverine ﬂuxes of nutrients except that published Ludwig et al. (2009) while possible impact of both atmospheric and riverine in- puts onto the Northeastern Levantine waters have not been evaluated yet. The present study is based on a long term aerosol, rainwater and riverine data collection in the North- ern Levantine Basin of the Eastern Mediterranean. The main aim of the current study is to enhance our knowledge of at- mospheric (dry and wet) deposition to the surface waters of the Northern Levantine Basin. Aerosol, rainwater and river- ine samples collected from the Northern Levantine Basin, will allow us to (i) deﬁne temporal variability of nutrient concentrations in aerosol and rainwater, (ii) assess the inﬂu- ence of air mass back trajectories on nutrient composition and (iii) assess and compare the atmospheric and riverine in- puts of nutrients to the Northern Levantine Basin. 1 Introduction The Mediterranean Sea has one of the most oligotrophic surface wasters in the world with Low Nutrient and Low Chlorophyll (LNLC). The average annual productivity in the Mediterranean Sea is half of the amount observed in the ultra-oligotrophic Sargasso Sea (Krom et al., 2004; Pitta et al., 2005). The main reason for this ultra-oligotrophic sta- tus is that the Mediterranean has an anti-estuarine (reverse thermohaline) circulation in which nutrient poor surface wa- ters incoming from the Atlantic is balanced by outgoing rel- atively nutrient rich deep waters of the Mediterranean Sea through the Strait of Gibraltar (Hamad et al., 2005 and ref- erences therein). Most of the nutrient inputs to the Mediter- ranean Sea are originated from the atmosphere (including dry and wet) deposition and riverine runoff (Krom et al., 2004, 2010). The Eastern Mediterranean Sea has a uniquely high N/P ratio raging from 25 to 28, compared to the Western Mediter- ranean (22) and the “normal” oceanic Redﬁeld ratio of 16 (Krom et al., 1991; Yılmaz and Tu˘grul, 1998). Thus the primary productivity in the Eastern basin is phosphorus lim- ited (Thingstad et al., 2005). In their recent study, Krom et al. (2004) budgeted ﬂuxes of N and P for the Eastern Mediter- ranean and concluded that the high N/P ratio is due primarily to the high biologically available N/P ratio in all of sources but for those particularly from the atmosphere (117:1) com- bined with regionally low denitriﬁcation rates (Krom et al., 2010). Ludwig et al. (2009) suggested that decrease in the dissolved Silica concentrations were due to a substantial re- duction in the fresh water discharges. These authors have hypothesized that Si may not necessarily reduce the produc- tivity in the Mediterranean however it can provoke a switch from diatom dominated communities to non-siliceous popu- lations. Correspondence to: M. Koc¸ak (mkocak@ims.metu.edu.tr) hypothesized that Si tivity in the Mediterr from diatom domina lations. Published by Copernicus Publications on behalf of the European Geosciences Union. Correspondence to: M. Koc¸ak (mkocak@ims.metu.edu.tr) Correspondence to: M. Koc¸ak (mkocak@ims.metu.edu.tr) Published by Copernicus Publications on behalf of the European Geosciences Union. M. 1 Introduction A total of 1525 bulk aerosol ﬁlter samples were col- lected using a high volume sampler with ﬂow rates of around 1 m3 min−1 on Whatman-41 cellulose ﬁber ﬁlters (20 cm×25 cm). Individual bulk aerosol samples were col- lected for typically 24 h (occasionally 48 h), unless inter- rupted by equipment failure (which occurred during July– December 2000, 2004 and 2005). Samples and blanks (n = 110) were kept cool at 4 ◦C until analysis conditions that our test showed no signiﬁcant change in their composition for more than a year period. Gent type PM10 stacked ﬁlter unit (SFU) sampler was used in order to collect atmospheric par- ticles in two size ranges namely, coarse and ﬁne. Twenty daily samples were collected in June 2007. Upstream of the ﬁrst holder is a preimpaction stage which intercepts particles larger than 10 µm equivalent aerodynamic diameter (EAD). The sampler operates at a ﬂow rate of 16 L min−1. It col- lects particulates which have an EAD of less than 10 mm in separate coarse (2.5–10 µm EAD) and ﬁne (<2:5 µm EAD) size fraction on two sequential (for more details see Koc¸ak et al., 2007a). It has been shown that results from SFU are in good agreement with other sampler types (Hitzenberger et al., 2004; Wieprecht et al., 2004). A total of 235 rain wa- ter samples were collected on an event speciﬁc basis by us- ing an automatic Wet/Dry sampler, Model ARS 1000, MTX Italy (for more details see ¨Ozsoy and Saydam, 2000). After each speciﬁc event, the rain water sample was immediately transferred into the laboratory for ﬁltration (0.45 µm, MFS, cellulose acetate, 47 mm diameter) and pH measurements. 2.1 Sites description and sample collection 36 The soluble nutrient measurements in aerosol, rainwater and river samples were carried out by a Technicon Model, four-channel Autoanalyzer (for more details see Yılmaz and Tu˘grul, 1998). In general, blanks were found to be less than detection limit of the instrument. The detection limits of Au- toanalyzer were 0.02, 0.10, 0.02 and 0.04 µm for phosphate, reactive silicate, nitrate and ammonium, respectively. Since this method determines reactive silicate (not polymeric sili- cates) obtained Si values will be presented as dissolved Sili- cate (Sidiss). The accuracy of the nutrient measurements was veriﬁed against Quasimeme (Quality Assurance of Informa- tion for Marine Environmental Monitoring in European Lab- oratory Performance Studies) intercalibration program sam- ples. The precisions were found to be 7.3%, 7.4%, 8.9% and 3.3% for PO3− 4 , Sidiss, NO− 3 and NH+ 4 , respectively. Subsamples of rain waters were kept frozen (−18 ◦C) until nutrient analysis (not more than two weeks). The percentage of sampling coverage is highlighted in Table 1. Sampling coverage was around 50% (except NH+ 4 in rainwater ∼45%) between January 1999 and December 2007. To complement the atmospheric samples, riverine samples for nutrient analyses were collected from ﬁve rivers (Ceyhan, Seyhan, G¨oksu, Berdan and Lamas; see Fig. 1) ﬂowing into the Northeastern Levantine Basin of the Eastern Mediter- ranean. Seasonal sample collection was carried out between March 2002 and November 2007. It should be noted that nu- trient measurements from 2000 to 2003 were very limited; dissolved ammonia was not measured, reactive phosphate, silicate, nitate+nitrite were measured merely in this period. For reliable estimation of the nutrient ﬂuxes data obtained between 2004 and 2007 will therefore be used. A total of 103 river water samples were collected from the middle of the river bed at point near the river mouths, depending on transportation facilities from land to the river mouths. There are no domestic and/or industrial waste discharges further down the sampling points other than surface runoff during heavy rains. Nutrients samples were taken in high density polyethylene bottles stored at 4 ◦C for a few hours during transport and then kept frozen until analysis. Three-cm2 of subsamples of the aerosol ﬁlters and blanks were shaken for 36 h in the dark at room temperature in pre- cleaned centrifuge tubes containing 15 ml MilliQ (18.2 ; pH∼5.6) and 50 µL chloroform (Herut et al., 2002). 2.1 Sites description and sample collection The Mediterranean climate is characterized by mild, humid winters and dry summers. Winter months are dominated by rainfall (November–February) whereas the summer months (June–September) are characterized by the lack of wet pre- Biogeosciences, 7, 4037–4050, 2010 www.biogeosciences.net/7/4037/2010/ M. Koc¸ak et al.: Atmospheric nutrient inputs to the northern levantine basin from a long-term observation 4039 Figure 1. Locations of atmospheric and riverine sampling sites. R refers to river. Fig. 1. Locations of atmospheric and riverine sampling sites. R refers to river. Table 1. Summary of statistics of the nutrients PO3− 4 , Sidiss, NO− 3 and NH+ 4 of aerosol and rainwater at Erdemli during the period Jan- uary 1999 and December 2007. P-PO−3 4 Sidiss N-NO− 3 N-NH+ 4 Aerosol (nmol m−3) Mean ± Std 0.45 ± 0.43 1.08 ± 1.53 65.3 ± 35.1 121.6 ± 63.2 Min–Max 0.03–6.40 0.04–26.57 0.2–258.8 0.1–473.2 Median 0.37 0.52 62.6 113.8 G. M.a 0.34 0.51 50.0 95.8 Nb 1525 1525 1525 1525 Coverage (%) 52 52 52 52 Rainwater (µmol L−1) WVMc 0.7 1.9 44 46 Nb 235 235 235 212 Coverage (%) 50 50 50 45 P-PO−3 4 Sidiss N-NO− 3 N-NH+ 4 Aerosol (nmol m−3) Mean ± Std 0.45 ± 0.43 1.08 ± 1.53 65.3 ± 35.1 121.6 ± 63.2 Min–Max 0.03–6.40 0.04–26.57 0.2–258.8 0.1–473.2 Median 0.37 0.52 62.6 113.8 G. M.a 0.34 0.51 50.0 95.8 Nb 1525 1525 1525 1525 Coverage (%) 52 52 52 52 Rainwater (µmol L−1) WVMc 0.7 1.9 44 46 Nb 235 235 235 212 Coverage (%) 50 50 50 45 a Geometric mean, b Number of samples, c Volume weighted mean. Average sam- pling coverage is calculated using the formulas Coverageaerosol= NSaerosol ND x100 and Coveragerain = NSrain NRE x100 . NSaerosol and NSrain refer to number of samples for aerosol and rain; ND and NRE denote number of days and rain events during sampling period. Fig. 1. Locations of atmospheric and riverine sampling sites. R refers to river. a Geometric mean, b Number of samples, c Volume weighted mean. Average sam- pling coverage is calculated using the formulas Coverageaerosol= NSaerosol ND x100 and Coveragerain = NSrain NRE x100 . NSaerosol and NSrain refer to number of samples for aerosol and rain; ND and NRE denote number of days and rain events during sampling period. 2.1 Sites description and sample collection Sam- ples were immediately analyzed for nutrients after centrifug- ing at 3500 rpm for 15 min. The homogeneity for the sub- samples was tested by applying historical data; one-eighth of each sample extracted for 45 min using 20 ml of nano- pure water and 100 µL chloroform. (Koc¸ak et al., 2004b). Figure 2 shows the regression line for two nutrients (NO− 3 and NH+ 4 ) which were obtained from different amount of ﬁlter and instruments. The slopes of the regression lines for NO− 3 and NH+ 4 were found 0.91 (R2 = 0.97) and 0.94 (R2 = 0.98), respectively. Obtained slopes suggested that NO− 3 and NH+ 4 values from Autoanalyzer (area = 3 cm2) were 9 and 6% higher than those from Ion Chromatography (area = 62.5 cm2). However, application of Wilcoxen signed- rank test showed that there were no statistical differences between two approaches for nitrate and ammonium, within 95% conﬁdence level. Furthermore, relationship between 2.4 Calculation of nutrient ﬂuxes The rain volume weighted average concentration (Cw) of nu- trient species can be calculated as follow: Cw = nP i=1 Ci ×Qi nP i=1 Qi (1) (1) where Ci is the nutrient concentration and Qi is the rainfall for the ith precipitation event. The wet and dry atmospheric ﬂuxes of nutrients were cal- culated according to the procedure explained in Herut et al. (1999, 2002). The wet atmospheric deposition ﬂuxes (Fw) were calculated from the annual amount of precipita- tion (Pannual) and the volume weighted mean concentration (Cw) of the substance of interest (Eq. 1). ned from Autoanalyzer (3 cm2) Fw = Cw ×Pannual (2) (2) d f A t Fw = Cw ×Pannual + ned from Autoanalyzer (3 cm ) aphy (62.5 cm2). The dry deposition (Fd) of nutrients can be calculated as the product of atmospheric mean nutrient concentrations (Cd) and their settling velocities (Vd), where Fd is given in units of mumol m−2 yr−1, Cd in units of mumol m−3 and Vd in units of m yr−1 (for more details see Sect. 3.5.1). g 3 4 against the concentrations acquired from Ion Chromato Fig. 2. Plot of NO− 3 and NH+ 4 concentration obtained from Autoan- alyzer (3 cm2) against the concentrations acquired from Ion Chro- matography (62.5 cm2). two approached for ammonium reveals that bacterial con- sumption of NH+ 4 was efﬁciently prevented by the addition of chloroform during 36 h extraction period. Fd = Cd ×Vd (3) (3) Fd = Cd ×Vd Annual riverine ﬂuxes (Fr, Eq. 4) were calculated by the product of Cdw and Qannual (Karakatsoulis and Ludwig, 2004). 2.3 Air mass back trajectories and cluster analysis 3-day backward trajectories to the Erdemli sampling site were computed between January 1999 and December 2007 by the HYSPLIT Dispersion Model (HybridSingle Particle Langrangian Integrated Trajectory; Draxler and Hess, 1998) and illustrated by one-hour endpoint locations in terms of lat- itude and longitude. Recently, it has been shown that bound- ary layer in the region during winter (for example, Decem- ber and January) was found to be as high as 1300 m and it ranged mostly between 700 and 800 m whereas, boundary layer in transition (for instance, March and April) reached up to 2500 m with values ranging mainly from 800 to 1400 m (Koc¸ak et al., 2010). Furthermore, with increasing altitude, particularly the % inﬂuence of Saharan airﬂow increases in the region (Koc¸ak et al., 2004a). For instance, from 1 km to 2 km its % inﬂuence increases around 10% while airﬂow from Europe remains constant. Consequently, 3-day back- ward trajectories of air masses at a height of 1000 m were selected to categorize air ﬂow. Cluster analysis was applied to categorize air mass back trajectories using the method de- scribed by Cape et al. (2000). Cdw = nP i=1 Ci ×Qi nP i=1 Qi (4) Fr = Cdw ×Qannual (5) (4) Fr = Cdw ×Qannual (5) Fr = Cdw ×Qannual (5) The discharge weighted mean concentration (Cdw, Eq. 4) was determined on the basis of n samples of instantaneous concentrations (Ci, Ci+1) and discharge values (Qi, Qi+1). The water discharges data for ﬂux calculations were provided from General Directorate of State Hydraulic Works, Turkey. The discharge weighted mean concentration (Cdw, Eq. 4) was determined on the basis of n samples of instantaneous concentrations (Ci, Ci+1) and discharge values (Qi, Qi+1). The water discharges data for ﬂux calculations were provided from General Directorate of State Hydraulic Works, Turkey. 2.2 Sample analysis The pH in rain waters was measured immediately using a Microprocessor pH meter (WTW-Model pH537) after each speciﬁc event. Calibration of the pH meter was carried out using N.B.S buffers at pH values of 4.00 and 7.00. www.biogeosciences.net/7/4037/2010/ Biogeosciences, 7, 4037–4050, 2010 M. Koc¸ak et al.: Atmospheric nutrient inputs to the northern levantine basin from a long-term observation 4040 Figure 2. Plot of NO3 - and NH4 + concentration obt against the concentrations acquired from Ion Chromato Fig. 2. Plot of NO− 3 and NH+ 4 concentration obtained from Autoan- alyzer (3 cm2) against the concentrations acquired from Ion Chro- matography (62.5 cm2). 3.1 Nutrient concentrations in aerosol and rainwater (2003) Heraklion, Crete 18 21 Sep 99–Sep 00 (41) Markaki et al. (2003) Tel Shikmona , Israel 41 25 Jan 92–Mar 98 (187) Herut et al. (1999) Ashdod, Israel 57 45 Nov 95–Mar 98 (67) Herut et al. (1999) Table 2. Comparison of nutrient concentrations in aerosol (nmol m−3) and rainwater (µmol L−1) samples for different sites of the Mediter- ranean. Table 2. Comparison of nutrient concentrations in aerosol (nmol m−3) and rainwater (µmol L−1) sample ranean. and standard deviations of 0.45 ± 0.43 and 1.08 ± 1.53, re- spectively. NO− 3 and NH+ 4 concentrations in aerosol sam- ples was found to be ranged between 0.2–258.8 and 0.1– 473.2 nmol m−3 with mean values and standard deviations of 65.3 ± 35.1 and 121.6 ± 63.2, respectively. On the other hand, volume weighted mean values for PO3− 4 , Sidiss, NO− 3 and NH+ 4 in rainwater samples were found to be 0.7, 1.9, 44 and 46 µmol L−1, respectively. and standard deviations of 0.45 ± 0.43 and 1.08 ± 1.53, re- spectively. NO− 3 and NH+ 4 concentrations in aerosol sam- ples was found to be ranged between 0.2–258.8 and 0.1– 473.2 nmol m−3 with mean values and standard deviations of 65.3 ± 35.1 and 121.6 ± 63.2, respectively. On the other hand, volume weighted mean values for PO3− 4 , Sidiss, NO− 3 and NH+ 4 in rainwater samples were found to be 0.7, 1.9, 44 and 46 µmol L−1, respectively. To our knowledge, no data of water soluble Sidiss in the aerosol over the Mediterranean and Sidiss in rainwater over the Eastern Mediterranean have been reported previously. Since Bartoli et al. (2005) reported only wet deposition inputs of Sidiss for the Western Mediterranean and hence would not be appropriate to compare these with the present Sidiss values. The mean aerosol phosphate concentration at Erdemli was comparable to levels reported for Eilat, Is- rael (Chen et al., 2007). Although phosphate concentrations were measured in seawater, highest levels over the Eastern Mediterranean were reported for Tel Shikmona (Herut et al., 2002) and Israeli coast (Carbo et al., 2005) due to the closer proximity of the sampling site to arid regions (Koc¸ak et al., 2004a). Aerosol nitrate and ammonium concentrations are in agreement with the values reported for Erdemli (Koc¸ak et al., 2004b). 3.1 Nutrient concentrations in aerosol and rainwater 37 The statistical summary for nutrients (PO3− 4 , Sidiss, NO− 3 and NH+ 4 ) measured in aerosol and rainwater samples col- lected from Erdemli between January 1999 and December 2007 were presented in Table 1. The water soluble PO3− 4 and Sidiss concentrations in aerosol ranged between 0.03– 6.40 and 0.04–26.27 nmol m−3 with arithmetic mean values Biogeosciences, 7, 4037–4050, 2010 www.biogeosciences.net/7/4037/2010/ M. Koc¸ak et al.: Atmospheric nutrient inputs to the northern levantine basin from a long-term observation 4041 Comparison of nutrient concentrations in aerosol (nmol m−3) and rainwater (µmol L−1) samples for different sites of the Mediter- Location Sidiss PO3− 4 Reference Aerosol (nmol m−3) Erdemli, Turkey 1.1 ± 1.5 0.5 ± 0.4 Jan 99–Dec 07 (1525) This Study Erdemli, Turkey − 0.3 Jan 99–Jan 00 (194) Markaki et al. (2003) Finokalia, Crete − 0.1 Sep 99–Sep 00 (85) Markaki et al. (2003) Tel Shikmona , Israel − 0.8 ± 0.5 Apr 96–Jan 99 (41) Herut et al. (2002) Tel Shikmona, Israel − 0.72 Jan 01–Apr 03 (71) Carbo et al. (2005) Eilat, Israel − 0.4 ± 0.2 Aug 03–Sep 05 (137) Chen et al. (2007) Rainwater (µmol L−1) Erdemli, Turkey 1.9 0.5 Jan 99–Dec 07 (237) This Study Erdemli, Turkey − Feb 99–Dec 99 (16) Markaki et al. (2003) Heraklion, Crete − 0.1 Sep 99–Sep 00 (41) Markaki et al. (2003) Tel Shikmona , Israel − 0.6 Jan 92–Mar 98 (187) Herut et al. (1999) Ashdod, Israel − 0.6 Nov 95–Mar 98 (67) Herut et al. (1999) Location NO− 3 NH+ 4 Reference Aerosol (nmol m−3) Erdemli, Turkey 65 ± 34 121 ± 64 Jan 99–Dec 07 (1525) This Study Erdemli, Turkey 58 118 Jan 99–Jan 00 (194) Markaki et al. (2003) Finokalia, Crete 27 ± 13 53 ± 21 Oct 96–Sep 99 (496) Kouvarakis et al. (2001) Finokalia, Crete 16 24 Sep 99–Sep 00 (85) Markaki et al. (2003) Tel Shikmona , Israel 93 ± 29a 117 ± 88 Apr 96–Jan 99 (41) Herut et al. (2002) Tel Shikmona, Israel 106 108 Jan 01–Apr 03 (71) Carbo et al. (2005) Eilat, Israel 39 ± 19 25 ± 14 Aug 03–Sep 05 (137) Chen et al. (2007) Cap Ferrat, France 63 150 May–June 92 Loye-Pilot et al. (1993) Rainwater (µmol L−1) Erdemli, Turkey 37 41 Jan 99–Dec 07 (237) This Study Erdemli, Turkey 46 − Feb 99–Dec 99 (16) Markaki et al. 3.1 Nutrient concentrations in aerosol and rainwater Observed mean aerosol nitrate and ammonium con- centrations were found to be two to four times higher than Table 2 shows the soluble nutrient concentrations in aerosol and rainwater samples obtained from different sites located around the Mediterranean. Although the values from this study and those in the literature cover different collec- tion periods (and might have different sampling and analyt- ical methodologies), comparison will be useful to evaluate spatial trends. www.biogeosciences.net/7/4037/2010/ Biogeosciences, 7, 4037–4050, 2010 M. Koc¸ak et al.: Atmospheric nutrient inputs to the northern levantine basin from a long-term observation M. Koc¸ak et al.: Atmospheric nutrient inputs to the northern levantine basin from a long-term observation 4042 Figure 3. Temporal variability of nutrient concentrations in aerosol (grey dots (black bars) samples collected from Erdemli site PO4 3-(a), Sidiss (b), NO NH4 + (d). Fig. 3. Temporal variability of nutrient concentrations in aerosol (grey dots) and rain (black bars) samples collected from Erdemli site PO3− 4 (a), Sidiss (b), NO− 3 (c) and NH+ 4 (d). those reported for Finokalia, Crete (Kouvarakis et al., 2001, Markaki et al., 2003) and Eilat, Israel (Chen et al., 2007) whereas values were comparable levels reported for Tel Shik- mona, Israel (Herut et al., 2002) and Cap Ferrat (Lo¨ye-Pilot et al., 1993). It should be noted that Erdemli and Tel Shik- mona aerosol samples were collected on Whatman 41 cel- lulose ﬁber ﬁlters whilst Finokalia and Eilat aerosol were collected on Teﬂon and polycarbonate ﬁlters, respectively. It has been reported that positive nitrate and ammonium ar- tifact can result the adsorption of gaseous HNO3 and NH3 on ﬁlter surfaces (mainly glass ﬁber and cellulose) or on al- ready collected particles (Wieprecht et al., 2004 and refer- ences therein). For instance, Keck and Wittmaack (2005) showed that the retention efﬁciencies of HNO3 and NH3 are very high up to 100 %, if the gases are presented in equimo- lar concentrations. In order to clarify this difference, aerosol samples were simultaneously collected on Whatman 41 and polycarbonate ﬁlters. Comparing nitrate and ammonium re- sults from different substrates it has been shown that NO− 3 and NH+ 4 values for Whatman 41 were 42% and 50% higher than those concentrations observed for polycarbonate ﬁlters. 3.1 Nutrient concentrations in aerosol and rainwater Therefore, it can be assumed that the measured nitrate and ammonium concentrations for the current study are equiva- lent to those of total inorganic NO− 3 and NH+ 4 plus gaseous HNO3 and NH3. In addition to dry depositions of nitrate and ammonium, alternative dry depositions will also be presented in parenthesis after multiplying by 0.58 and 0.50. Rainwater volume weighted mean phosphate, nitrate and ammonium concentrations at Erdemli were found to be com- parable to values reported for Israeli coastal sites (Herut et al., 1999) whereas lowest values were observed at Fi- nokalia (Markaki et al., 2003) since this site is categorized by natural background (distance from large pollution sources >50 km) and its proximity to arid regions located at the Mid- dle East/Arabian Peninsula. (black bars) samples collected from Erdemli site PO4 3 (a), Sidiss (b), NO NH4 + (d). Fig. 3. Temporal variability of nutrient concentrations in aerosol (grey dots) and rain (black bars) samples collected from Erdemli site PO3− 4 (a), Sidiss (b), NO− 3 (c) and NH+ 4 (d). 3.2 Variability of nutrient levels in aerosol and rainwater documented in the literature (Kubilay and Saydam, 1995; Avila et al., 1998; Moulin et al., 1998; Koc¸ak et al., 2004a) intense sporadic dust events occur over the Eastern Mediter- ranean during the transitional period when the air mass tra- jectories originate predominantly from North Africa (but rarely from the Middle East/Arabian Peninsula). Higher con- centrations and variations might be attributed to transport of PO3− 4 and Sidiss associated with mineral dust from arid desert areas. To illustrate, two distinct dust events observed 18– 20 October 2002 and 4–5 April 2003 will be used. The ﬁrst is an example of a dust event observed from 18 Oc- tober 2002 to 20 October 2002 with mean PO3− 4 and Sidiss concentrations of 2.1 and 19.5 nmol m−3, respectively. The highest PO3− 4 and Sidiss concentrations for this mineral dust episode were identiﬁed on October 19 with values of 2.6 and 26.6 nmol m−3. Figure 4a shows air mass back trajectories The temporal variations in the concentrations of nutrients (PO3− 4 , Sidiss, NO− 3 and NH+ 4 ) between January 1999 and December 2007 for aerosol and rainwater samples are pre- sented in Fig. 3. Nutrient concentrations in aerosol and rainwater samples demonstrated substantial temporal vari- ability in agreement with earlier studies carried out in the Mediterranean (Lo¨ye-Pilot and Morelli, 1988; Herut et al., 1999, 2002; Markaki et al., 2003). Concentrations of nutri- ent species were highly variable on a daily time scale and their concentrations may change an order of magnitude from day to day (see Fig. 3a, b, c, d). documented in the literature (Kubilay and Saydam, 1995; Avila et al., 1998; Moulin et al., 1998; Koc¸ak et al., 2004a) intense sporadic dust events occur over the Eastern Mediter- ranean during the transitional period when the air mass tra- jectories originate predominantly from North Africa (but rarely from the Middle East/Arabian Peninsula). Higher con- centrations and variations might be attributed to transport of PO3− 4 and Sidiss associated with mineral dust from arid desert areas. To illustrate, two distinct dust events observed 18– 20 October 2002 and 4–5 April 2003 will be used. The ﬁrst is an example of a dust event observed from 18 Oc- tober 2002 to 20 October 2002 with mean PO3− 4 and Sidiss concentrations of 2.1 and 19.5 nmol m−3, respectively. Biogeosciences, 7, 4037–4050, 2010 3.2 Variability of nutrient levels in aerosol and rainwater The highest PO3− 4 and Sidiss concentrations for this mineral dust episode were identiﬁed on October 19 with values of 2.6 and 26.6 nmol m−3. Figure 4a shows air mass back trajectories Aerosol: Although, particles are efﬁciently scavenged by rain (26% of the total annual amount of rain and 39% of the total rain events occur in transitional period) PO3− 4 and Sidiss demonstrated higher concentrations and larger varia- tions during the transitional period (spring and autumn). As Biogeosciences, 7, 4037–4050, 2010 www.biogeosciences.net/7/4037/2010/ M. Koc¸ak et al.: Atmospheric nutrient inputs to the northern levantine basin from a long-term observation M. Koc¸ak et al.: Atmospheric nutrient inputs to the northern levantine basin from a long-term observation 4043 Figure 4. Three day back trajectories for 1, 2 and 3 km altitude along with Aerosol Index on the 19 October 2002 (TOMS, a), 04 April 2003 (TOMS, b), 05 March 2004 (TOMS, c) and 16 April 2006 (OMI, d). Fig. 4. Three day back trajectories for 1, 2 and 3 km altitude along with Aerosol Index on the 19 October 2002 (TOMS, (a), 4 April 2003 (TOMS, (b), 5 March 2004 (TOMS, (c) and 16 April 2006 (OMI, (d)). Fi 4 Th d b k t j t i f 1 d 3 k l i d l i h A g y j , g on the 19 October 2002 (TOMS, a), 04 April 2003 (TOMS, b), 05 March 2004 (TOMS, c) and 16 April 2006 (OMI, d). Fig. 4. Three day back trajectories for 1, 2 and 3 km altitude along with Aerosol Index on the 19 October 2002 (TOMS, (a), 4 April 2003 (TOMS, (b), 5 March 2004 (TOMS, (c) and 16 April 2006 (OMI, (d)). for three levels (1, 2 and 3 km) along with TOMS-AI (Total Ozone Mapping Spectrometer Aerosol Index) for 19 October 2002. All air mass back trajectories for October 19 indicated air ﬂow originating from the Middle East while TOMS ex- hibits particularly a large dust plume from the Middle East region into the eastern Mediterranean with AI values ranging between 1.2 from 3. The second example from 4 April to 5 April 2003 with mean PO3− 4 and Sidiss concentrations of 0.8 and 6.4 nmol m−3, respectively. The highest PO3− 4 and Sidiss concentrations for this mineral dust episode were identiﬁed on 4 April with values of 1.0 and 6.6 nmol m−3. www.biogeosciences.net/7/4037/2010/ 3.2 Variability of nutrient levels in aerosol and rainwater Mean aerosol nutrient concentrations and volume-weighted mean (VWM) concentrations of nutrient in aerosol and rainwater, as a function of the categorized three-day air-mass back trajectories arriving at Erdemli. Species Aerosol (Mean, nmol m−3) Rainwater (VWM, µMol L−1) TUR EU MED MEAP SAH TUR EU MED MEAP SAH P-PO−3 4 0.40 ± 0.34 0.39 ± 0.30 0.33 ± 0.15 0.62 ± 0.64 0.56 ± 0.70 0.38 0.32 0.29 0.68 0.61 Sidiss 0.59 ± 0.38 0.50 ± 0.32 0.45 ± 0.33 1.76 ± 3.08 1.98 ± 2.57 0.65 0.34 0.34 1.26 1.69 N-NO− 3 65.6 ± 41.6 49.5 ± 31.6 51.5 ± 23.8 60.2 ± 30.9 68.1 ± 43.0 39.4 32.8 20.0 34.1 42.5 N-NH+ 4 133.0 ± 65.5 106.8 ± 60.5 92.9 ± 49.5 112.1 ± 51.4 104.8 ± 62.6 39.5 36.7 21.1 39.8 41.1 acidity due to damp soil around vicinity and less frequent mineral dust transport to the region. Furthermore, volume- weighted mean concentrations of Sidiss (1.8 µM) and PO3− 4 (0.8 µM) for non-acidic events were 6 and 3 times higher than those calculated for acidic events. Whereas, volume- weighted mean values for NO− 3 (50 µM) and NH+ 4 (46 µM) during non-acidic events were only 2.1 and 1.7 times higher than acidic events. For example, the most acid rain events were observed on 2 December and 3 December 2001, with the lowest pH value of 3.4. On 30 November pH was found around 5.8 with PO3− 4 , Sidiss, NO− 3 and NH+ 4 concentrations of 0.2, 0.1, 11 and 14 µM, respectively. Rain events on 2 and 3 December showed drastic decreases in pH (3.4) and Sidiss due to the higher anthropogenic inﬂuence (high NO3 levels) and the deﬁciency of neutralizing agents such as cal- cium carbonate. In contrast, on 5 March 2004 and 16 April 2006 the pH of the rainwater were found to be 7.1 (PO3− 4 ∼1; Sidiss ∼12 µm L−1) and 8.0 (PO3− 4 ∼2; Sidiss ∼9 µm L−1). Figure 4c and d illustrate air mass back trajectories along with TOMS-AI (Total Ozone Mapping Spectrometer Aerosol Index) and OMI-AI (Ozone Monitoring Instruments Aerosol Index) for 5 March 2004 and 16 April 2006, respectively. Both air mass back trajectories and AI values indicated that on 5 March 2004 and 16 April 2006 the sampling site was under the inﬂuence of air ﬂow from North Africa and Middle East/Arabian Peninsula, respectively. 3.2 Variability of nutrient levels in aerosol and rainwater 7.1% of the airﬂow (hereafter MED). Air masses originating from the Middle East/Arabian Peninsula (Cluster 6; hereafter MEAP) and Eastern Saharan (Cluster 7; hereafter SAH) rep- resent 10.1% and 10.6% of the trajectories, respectively. acidity due to damp soil around vicinity and less frequent mineral dust transport to the region. Furthermore, volume- weighted mean concentrations of Sidiss (1.8 µM) and PO3− 4 (0.8 µM) for non-acidic events were 6 and 3 times higher than those calculated for acidic events. Whereas, volume- weighted mean values for NO− 3 (50 µM) and NH+ 4 (46 µM) during non-acidic events were only 2.1 and 1.7 times higher than acidic events. For example, the most acid rain events were observed on 2 December and 3 December 2001, with the lowest pH value of 3.4. On 30 November pH was found around 5.8 with PO3− 4 , Sidiss, NO− 3 and NH+ 4 concentrations of 0.2, 0.1, 11 and 14 µM, respectively. Rain events on 2 and 3 December showed drastic decreases in pH (3.4) and Sidiss due to the higher anthropogenic inﬂuence (high NO3 levels) and the deﬁciency of neutralizing agents such as cal- cium carbonate. In contrast, on 5 March 2004 and 16 April 2006 the pH of the rainwater were found to be 7.1 (PO3− 4 ∼1; Sidiss ∼12 µm L−1) and 8.0 (PO3− 4 ∼2; Sidiss ∼9 µm L−1). Figure 4c and d illustrate air mass back trajectories along with TOMS-AI (Total Ozone Mapping Spectrometer Aerosol Index) and OMI-AI (Ozone Monitoring Instruments Aerosol Index) for 5 March 2004 and 16 April 2006, respectively. Both air mass back trajectories and AI values indicated that on 5 March 2004 and 16 April 2006 the sampling site was under the inﬂuence of air ﬂow from North Africa and Middle East/Arabian Peninsula, respectively. The Kruskal–Wallis (K–W) test was applied to test for the presence of signiﬁcant differences in nutrient concentrations. Consequently, if the nutrient concentrations are considered for each of the six air ﬂow categories (Table 3) the following general observations might be made: (a) The K-W test showed that there was a signiﬁcant dif- ference in the concentrations of PO3− 4 and Sidiss (p < 0.01). PO3− 4 and Sidiss concentrations in aerosol and rain samples were found to be higher in the SAH and MEAP than those observed for the remaining air ﬂow. 3.2 Variability of nutrient levels in aerosol and rainwater Figure 4b presents air mass back trajectories along with TOMS-AI for 4 April 2003. Both air mass back trajectory (air ﬂow from Egypt) and TOMS-AI (large dust plume from Egypt into eastern Mediterranean) demonstrated mineral dust transport to the sampling site from the Eastern Saharan region. under prevailing summer conditions in Mediterranean region (Mihalopoulos et al., 1997; Bardouki et al., 2003; Koc¸ak et al., 2004b). For instance, it has also been shown that in the eastern Mediterranean, conversion rates of SO2 to sulfate in- dicate a strong seasonal cycle with a winter minimum and a summer maximum (Erduran and Tuncel, 2001). 39 Rainwater: In general, PO3− 4 and Sidiss in rainwater showed higher concentrations in the transition periods due to dust transport from arid desert regions (e.g. North Africa and Middle East/Arabian Peninsula). It has been demonstrated that the pH of Western Mediterranean rains is strongly af- fected by the type of material scavenged from the atmosphere (Lo¨ye-Pilot et al., 1986). For instance, rain samples associ- ated with air masses from North Africa (red rains) had pH values as high as 7 as a result of the dissolution of calcium carbonate from dust. The distribution of the pH at Erdemli indicated that the largest fraction ∼70% of all events had a pH greater than 5.6 whereas acidic rain events (pH<5.6; Guerzoni et al., 1997) accounted for 30% of all samples. Vol- ume weighted mean values of PO3− 4 , Sidiss, NO− 3 and NH+ 4 in winter were 1.7, 2.1, 1.5 and 1.1 µm lower than those cal- culated for transition period. Therefore acid events (80% was observed in winter) might be attributed to the deﬁciency of alkaline material which was not sufﬁcient to neutralize the The lowest values for aerosol nutrient species were ob- served during the winter. Lower values of aerosol nutrients in winter can be attributed to efﬁcient removal of particles from the atmosphere via frequent rain events (70% total an- nual amount of rain and 55 % of the total rain events occur in winter). Aerosol NO− 3 and NH+ 4 showed higher values in summer. This variability has mainly been attributed to the absence of precipitation and active photochemical formation Biogeosciences, 7, 4037–4050, 2010 www.biogeosciences.net/7/4037/2010/ M. Koc¸ak et al.: Atmospheric nutrient inputs to the northern levantine basin from a long-term observation 4044 Table 3. 3.2 Variability of nutrient levels in aerosol and rainwater For instance, mean concentrations of PO3− 4 and Sidiss in aerosol (rain) were found approximately 1.5 (1.6) and 3 (2) times higher than the remaining, respectively. These two air ﬂows were mainly affected by crustal aerosol population associated with dust events originating from North Africa and Middle East/Arabian arid regions oc- curring particularly during the transitional period. (b) The K-W test indicated that there was a signiﬁcant difference in the concentrations of NO− 3 (p < 0.01). Aerosol nitrate concentrations were found to be lower when air ﬂows originated from Europe (EU) and Mediterranean Sea (MED) whereas concentrations for the remaining air masses were found comparable. For aerosol ammonium, the concentrations were found sta- tistically higher when air ﬂows originating from Turkey (TUR) and this might be mainly attributed to intense us- age of ammonium containing fertilizers (Koc¸ak et al., 2004b). Rainwater weighted mean values of nitrate and ammonium were found approximately two times lower when air ﬂow originated from Mediterranean Sea than those observed for the remaining. It can be sug- gested that relatively cleaner air masses associated with Mediterranean Sea during rainy seasons. Biogeosciences, 7, 4037–4050, 2010 3.4 Nutrient ﬂuxes would be logical for Sidiss since it was found to be asso- ciated mainly in coarse fractions and predominantly origi- nated from crustal material. However, it should be noted that the settling velocities used in the present study will still be a sources of uncertainty in the dry deposition calculations and the estimation might be subject to a bias of 40%. 3.3 Inﬂuence of airﬂows on nutrients By applying the cluster analysis, the daily air mass back trajectories (n > 3100) for 1 km altitude were used cover- ing the whole sampling period (1999–2007). First two clus- ters (Cluster 1 and 2; hereafter EU) shows north-westerly air ﬂows: The ﬁrst cluster includes trajectories with high wind speeds (long fetch) passing through Europe and accounting for 2.1 % of the airﬂow whereas the second cluster denotes relatively slower air ﬂow and contributes to 9.1% of the tra- jectories. The third and fourth clusters show short trajecto- ries originating from the north-west and northern Turkey and they represent 41.4% and 19.6% of the airﬂow, (Cluster 3 and 4; hereafter TUR) respectively. The ﬁfth cluster repre- sents trajectories travelling at high speeds, being maritime air masses from the western Mediterranean Sea and representing www.biogeosciences.net/7/4037/2010/ Biogeosciences, 7, 4037–4050, 2010 tmospheric nutrient inputs to the northern levantine basin from a long-term observation 4045 Table 4. Summery of dry deposition velocities (cm s−1) of the analyzed aerosol nutrients applied in the present study and the literature for different Mediterranean regions. Location PO3− 4 NO− 3 NH+ 4 Sidiss Reference Erdemli 1.56 ± 0.09 1.84 ± 0.03 0.14 ± 0.04 1.59 ± 0.12 This Study Erdemli 2 2 0.2 − Markaki et al. (2003) Finokalia 2.3a (2b) 2 0.2 − Markaki et al. (2003) Finokalia − 1 0.2 − Kouvarakis et al. (2001) Tel Shikmona , Israel 2 1.2 0.6 − Herut et al. (2002) Eilat, Israel 2 1 0.1 − Chen et al. (2007) Cap Ferrat, France − 1 0.1 − Loye-Pilot et al. (1993) Cap Ferrat, France 0.1–0.5 − − − Migon et al. (2001) a Calculated Vd from size resolved samples, b Vd value used during ﬂux calculation. a Calculated Vd from size resolved samples, b Vd value used during ﬂux calculation. a Calculated Vd from size resolved samples, b Vd value used during ﬂux calculation. 3.5 Atmospheric nutrient ﬂuxes Results from stack ﬁlter unit showed that water-soluble PO3− 4 , NO− 3 and Sidiss were associated predominantly with coarse particles (>75%) whereas NH+ 4 was mainly found in the ﬁne mode (>97%). Obtained results are in good agree- ment with previous studies conducted in the Eastern Mediter- ranean (Bardouki et al., 2003; Markaki et al., 2003; Koc¸ak et al., 2007b) An Approach adopted by Spokes et al. (2001), after Ottley and Harrison (1993) were used to estimate the settling velocities of nutrients. Aforementioned estimates for Vd based on experimental and model results are 2.0 cm s−1 for coarse particles and 0.1 cm s−1 for ﬁne (Duce et al., 1991). Settling velocities of nutrient can be calculated from relative two-sized aerosol fractions and modelled values (2.0 and 0.1 cm s−1), where Cf and Ff are percentages of coarse and ﬁne particles. Table 5 demonstrates dry, wet and total atmospheric de- positions of nutrients obtained from different sites located at the Eastern Mediterranean. Estimated dry deposition for phosphate at Erdemli was comparable to value reported for Eilat, Israel, Israel (Chen et al., 2007) whereas (as expected) phosphate dry deposition at Erdemli was approximately two times lower than those calculated for Tel Shikmona and Is- raeli coast. Estimated NO− 3 dry depositions at Erdemli, Tel Shikmona, Israeli coast and Eilat were found to be similar, whilst the highest dry deposition for ammonium was esti- mated at Tel Shikmona (Herut et al., 2002) and Israeli coast (Carbo et al., 2005) on account of applied settling veloc- ity (0.6 cm s−1). Calculated wet depositions for phosphate and nitrate at Erdemli and Tel Shikmona were two to four times higher than those of observed at Crete, respectively. Whereas, wet deposition of ammonium at Erdemli was two times higher that those estimated for Tel Shikmona. Tak- ing into account the total atmospheric depositions, phosphate demonstrated decreasing ﬂuxes in the order of Tel Shikmona > Erdemli > Crete while DIN showed declining ﬂuxes in the order of Tel Shikmona ∼Erdemli > Crete. Vd = (Cf ×2.0)+(Ff ×0.1) (6) (6) Settling velocities of nutrient to the Eastern and Western Mediterranean are presented in Table 4. Generally the Vd’s value of 2 was applied to estimate dry deposition of PO3− 4 in the Eastern Mediterranean. Migon et al. 3.5 Atmospheric nutrient ﬂuxes (2001) applied settling velocity values between 0.1 and 0.5 cm s−1 for dry deposition ﬂuxes of P in the Western Mediterranean since 90% of the P was found to be associated with anthropogenic particles. Settling velocities of nutrient to the Eastern and Western Mediterranean are presented in Table 4. Generally the Vd’s value of 2 was applied to estimate dry deposition of PO3− 4 in the Eastern Mediterranean. Migon et al. (2001) applied settling velocity values between 0.1 and 0.5 cm s−1 for dry deposition ﬂuxes of P in the Western Mediterranean since 90% of the P was found to be associated with anthropogenic particles. At Erdemli, the PO3 4, Sidiss and NH+ 4 ﬂuxes were found to be dominated by wet deposition (0.34, 0.92 and 23 mmol m−2 yr−1) with dry deposition contribu- tions amounting to 40% (0.22 mmol m−2 yr−1), 35% (0.54 mmol m−2 yr−1) and 18% (3 mmol m−2 yr−1) of their total deposition, respectively. Whereas dry and wet deposition of nitrate was comparable with a value of 22 mmol m−2 yr−1. In addition, ﬂuxes of nitrate and am- monium via wet deposition were found to be similar whilst dry deposition ﬂux of nitrate was an order of magnitude higher than those for ammonium owning to differences in their particle sizes and hence settling velocities. Vd values between 1 to 2 and 0.1 to 0.6 were applied to calculate the dry deposition ﬂuxes of NO− 3 and NH+ 4 in the Mediterranean, respectively. Nitrate at Erdemli and Fi- nokalia sites were mainly associated (>70%) with the coarse fraction due to reactions with alkaline sea salt and dust parti- cles, whilst ammonium was almost exclusively found in ﬁne fraction in the form of (NH4)HSO4 (Bardouki et al., 2003; Koc¸ak et al., 2007b). Based on our knowledge, there is no reported Vd value for Sidiss in the literature for the Mediter- ranean region. Estimated mean Vd value of 1.59 cm s−1 Biogeosciences, 7, 4037–4050, 2010 www.biogeosciences.net/7/4037/2010/ M. Koc¸ak et al.: Atmospheric nutrient inputs to the northern levantine basin from a long-term observation 4046 Table 5. Dry and wet deposition depositions of the analyzed nutrients calculated for the present study and the literature for different Mediterranean regions. 3.5 Atmospheric nutrient ﬂuxes Location Sidiss PO3− 4 NO− 3 NH+ 4 DIN Reference Dry Deposition (mmol m−2 yr−1) Erdemli, Turkeyb 0.54 0.22 38 (22) 5 (3) 43 (25) This Study Wintera 0.15 0.05 9 (5) 1 (0.5) 10 (6) This Study Transitiona 0.24 0.09 14 (8) 2 (1) 16 (9) This Study Summera 0.15 0.07 15 (9) 2 (1) 17 (10) This Study Erdemli, Turkeyb - 0.16 36 (21) 8 (4) 44 (25) Markaki et al. (2003) Finokalia, Cretec − 0.08 10 2 12 Markaki et al. (2003) Tel Shikmona, Israelb − 0.51 35 (20) 22 (11) 57 (31) Herut et al. (2002) Israeli Coastb 0.45 40 (23) 26 (13) 66 (36) Carbo et al. (2005) Eilat, Israelc − 0.25 − − 38 Chen et al. (2007) Wet Deposition (mmol m−2 yr−1) Erdemli, Turkey 0.92 0.34 22 23 45 This Study Wintera 0.54 0.18 13 16 29 This Study Transitiona 0.38 0.16 9 7 16 This Study Erdemli, Turkey − − 16 − − Markaki et al. (2003) Heraklion, Crete − 0.07 9 11 20 Markaki et al. (2003) Tel Shikmona, Israel 0.30 20 13 33 Herut et al. (1999) Atmospheric Deposition (mmol m−2 yr−1) Erdemli, Turkey 1.46 0.56 60 (44) 28 (26) 88 (70) This Study Crete − 0.15 19 13 32 Markaki et al. (2003) Tel Shikmona, Israel − 0.81 55 (40) 35 (26) 90 (76) Herut et al. (1999, 2002) a denotes seasonal ﬂuxes, b and c show Whatman 41 and Teﬂon ﬁlter. Values in parenthesis indicate alternative dry depositions of nitrate and ammonium for Whatman 41 ﬁlters after multiplying 0.58 and 0.50, respectively. Table 5. Dry and wet deposition depositions of the analyzed nutrients calculated for the present study and the literature for different Mediterranean regions. a denotes seasonal ﬂuxes, b and c show Whatman 41 and Teﬂon ﬁlter. Values in parenthesis indicate alternative dry depositions of nitrate and ammonium for Whatman 41 ﬁlters after multiplying 0.58 and 0.50, respectively. Dry deposition inputs of PO3− 4 and Sidiss in the transition period were 1.2 to 1.6 times larger than those observed dur- ing the winter and summer. NO− 3 and NH+ 4 dry deposition inputs were found comparable for the transition period and summer, whilst the lowest input was observed in winter. 3.5 Atmospheric nutrient ﬂuxes Dur- ing the winter period wet (P = 357 mm) deposition inputs of nutrients were 1.2 to 2 times higher than those calculated for the transition period (P = 133 mm) mainly due to the higher amount of precipitation. In winter, with the exception of ni- trate, nutrient inputs were dominated by wet deposition com- pared to dry deposition. For example, inputs of PO3− 4 , Sidiss and NH+ 4 via wet deposition were found to be 2 to 3 times larger than their inputs via dry deposition. In addition, inputs of nutrients were exclusively found to be originated from dry deposition in summer due to the lack of precipitation. Table 6. Discharge weighted mean nutrient concentrations (µM) and discharges (m3 s−1) for studied Rivers. River Sidiss PO3− 4 NH+ 4 NO− 3 Q Seyhan 117 5.6 16 83 168 Ceyhan 161 1.9 19 105 144 G¨oksu 112 3.3 7 58 45 Berdan 91 4.8 34 85 6 Lamas 113 0.4 1 101 3 the Northern Levantine Basin with very low ammonium and phosphate concentrations. On the other hand, concentrations of ammonium and phosphate for the remaining imply that these fresh water sources are substantially inﬂuenced by in- dustrial, domestic and agricultural activities. 3.5.1 Riverine nutrient ﬂuxes 10, c Ludwig et al., 2009. NLB: Northeastern Levantine Basin, Total: Riverine + Atmospheric inputs, EMED: Eastern Mediterranean, A: Northern Levantine Basin. Seasonal riverine inputs of nu- trients exhibited a decrease in values in the order of Tran- sitional > Winter > Summer. During the transition period nutrient inputs were found to be two to ﬁve times higher than those in winter and summer, respectively. This of course is not unexpected owing to the higher ﬂuvial discharge in the transitional period. tribution being more than 90%. Riverine phosphate ﬂux (40%) had a substantial contribution to the phos- phate pool in the NLB, however the atmosphere was found to be the main source to the surface waters with a mean contribution of 60%. Unlike inorganic nitrogen and phosphate, the NBL Si pool was almost exclusively dominated by riverine ﬂuxes (90%) and only 10% of the Si was attributed to atmospheric source. 3.5.2 Comparison between atmospheric and riverine nutrient ﬂuxes (b) Riverine molar N/P ratios ranged from 18 to 279 with a mean value of 28 and in contrast the molar Si/N ra- tios were found to range from 0.8 to 1.7, with a mean value of 1.3. Obtained riverine N/P and Si/N ratios sug- gested that riverine sources in the region are deﬁcient in phosphate compare to DIN and Si. Atmospheric (dry and wet) molar mean N/P ratios were found to be order of magnitude higher than former ratio whereas riverine Si/N ratio was 100 times greater than those observed for atmospheric inputs. It appears that both sources were deﬁcient in phosphorus compared to nitrogen. In other words, the Northeastern Levantine Basin of the Mediter- ranean Sea receives excessive amounts of DIN; higher than those required by autotrophic organisms. Consid- ering N/P ratio it might be suggested that unbalanced phosphorus and nitrogen inputs may provoke even more phosphorus deﬁciency in NLB. Although Si inputs have no effect on phosphorus limitation, total Si/N ratio sug- gests that Si deﬁciency relative to nitrogen might cause a switch from diatom dominated phytoplankton popula- tion to non-siliceous communities particularly at coastal areas in NLB. It has been shown that coastal sites of the Cilician Basin are dominated by diatom population (Uysal and K¨oksalan, 2010). Recently, our seasonal Krom et al. (2004, 2010) and Ludwig et al. (2009) calculated the atmospheric and the riverine nutrient ﬂuxes in Eastern Mediterranean, respectively. As can be deduced from Ta- ble 7, current atmospheric phosphorous and nitrogen ﬂuxes are in good agreement with the values reported by Krom et al. (2004, 2010) for EMED. In spite of different approaches during the estimation of the riverine nutrient ﬂuxes in NLB, the current riverine ﬂuxes and reported ﬂuxes by Ludwig et al. (2009) were found in the same magnitude. Our estimated riverine ﬂuxes for silicate, phosphorous and nitrogen were 1.8, 3 and 4.6 times lower than those calculated by Ludwig et al. (2009) for NLB. This discrepancy might be attributed to difference in study period. As suggested by Ludwig et al. (2009) local patterns of rivers which are not sufﬁciently captured by their method might also be as a source of this difference. 3.5.1 Riverine nutrient ﬂuxes Seyhan and Ceyhan Rivers (Tables 6, 7) were found to be main fresh water sources and more than 85 % of the PO3− 4 , Sidiss, NO− 3 and NH+ 4 (DIN as well) originated from these two rivers. In addition, the contribution from the G¨oks¨u River was 10, 11, 8 and 5 % to PO3− 4 , Sidiss, NO− 3 and NH+ 4 ﬂuxes, respectively. Although, ammonium inputs from rivers showed substantial contribution (15%), nitrate inputs were the primary component of the DIN pool (85%) to the Table 6 shows discharge weighted mean nutrient concentra- tions and discharges for the studied rivers. Annual mean wa- ter discharge for Seyhan, Ceyhan, G¨oksu, Berdan and Lamas were found to be 168, 144, 45, 6 and 3 m3 s−1, respectively. Discharges of Rivers show similar seasonality with highest values during spring. Table 6 clearly indicates that Lamas River is a typical example for the least polluted rivers in Biogeosciences, 7, 4037–4050, 2010 www.biogeosciences.net/7/4037/2010/ M. Koc¸ak et al.: Atmospheric nutrient inputs to the northern levantine basin from a long-term observation 4047 Table 7. Comparison of riverine and atmospheric nutrient inputs (109 mol km−2 yr−1) to the Northeastern Levantine Basin of the Eastern Mediterranean and the literature for the Eastern Mediterranean region. River Sidiss PO3− 4 NO− 3 NH+ 4 DIN N/P Si/N Seyhan 0.620 0.030 0.44 0.08 0.62 18 1.2 Ceyhan 0.730 0.009 0.48 0.09 0.73 65 1.3 G¨oksu 0.160 0.005 0.08 0.01 0.16 20 1.7 Berdan 0.020 0.001 0.02 0.01 0.02 25 0.8 Lamas 0.010 0.00004 0.01 0.0001 0.01 279 1.1 Winter 0.465 0.014 0.258 0.058 0.366 26 1.3 Transition 0.770 0.020 0.514 0.093 0.604 30 1.3 Summer 0.305 0.010 0.252 0.036 0.242 24 1.3 NLB-Ra 1.54 0.04 1 0.2 1.2 28 1.3 NLB-Aa 0.16 0.06 7 3 10 233 0.01 Totala 1.70 0.10 8 3.2 11 145 0.1 EMED-Aa 2.44 0.93 100 47 147 145 0.1 EMED-Ab − 0.95 − − 111 117 − NLB-Rc 2.77 0.12 − − 5.5 46 0.5 a Current study, b Krom et al., 2004, 2010, c Ludwig et al., 2009. NLB: Northeastern Levantine Basin, Total: Riverine + Atmospheric inputs, EMED: Eastern Mediterranean, A: Atmospheric ﬂux and R: Riverine ﬂux. a Current study, b Krom et al., 2004, 2010, c Ludwig et al., 2009. NLB: Northeastern Levantine Basin, Total: Riverine + Atmospheric inputs, EMED: Eastern Mediterranean, A: Atmospheric ﬂux and R: Riverine ﬂux. References Avila, A., Alarcon, M., and Queralt-Mitjans, I.: The chemical com- position of dust transported in red rains its contribution to the biogeochemical cycle of a holm oak forest in Catalonia, Spain, Atmos. Environ., 32, 179–191, 1998. Nutrient concentrations in aerosol and rainwater samples exhibited substantial temporal variability, up to an order of magnitude on daily basis. PO3− 4 and Sidiss in aerosol and rainwater exhibited higher and larger variations during the transitional period since their concentrations are affected by sporadic dust events originating from North Africa and Mid- dle East/Arabian Peninsula. Their mean concentrations were at least 1.5 times higher during airﬂows originating from the Saharan and Middle East/Arabian Peninsula than those ob- served for the remaining air ﬂows. Deﬁciency of alkaline ma- terial were found to be the main reason of acidic rain events whereas, alkaline rain events were observed when air mass back trajectories originated from arid and semi-arid desert re- gions. Higher aerosol nitrate and ammonium concentrations in the summer were due to the lack of precipitation and ac- tive photochemical formation. In general, lowest nitrate and ammonium concentrations in aerosol and rainwater were as- sociated with air ﬂow from Mediterranean Sea. Baker, A. R., Weston, K., Kelly, S. D., Voss, M., Streu, P., and Cape, J. N.: Dry and wet deposition of nutrients from the tropical Atlantic atmosphere: Links to primary productivity and nitrogen ﬁxation, Deep-Sea Res. Pt. II, 54, 1704–1720, 2007. Bardouki, H., Liakakou, H., Economou, C., Sciare, J., Smolik, J., Zdimal, V., Eleftheriadis, K., Lazaridis, M., Dye, C., and Mi- halopoulos, N.: Chemical composition of size resolved atmo- spheric aerosols in the eastern Mediterranean during summer and winter, Atmos. Environ., 37, 195–208, 2003. Bartoli, G., Migon, C., and Losno, R.: Atmospheric input of dis- solved inorganic phosphorus and silicon to the coastal northwest- ern Mediterranean Sea: Fluxes, variability and possible impact on phytoplankton dynamics, Deep-Sea Res. Pt. I, 52, 2005–2016, 2005. Bonnet, S., Guieu, C., Chiaverini, J., Ras, J., and Stock, A.: Ef- fect of atmospheric nutrients on the autotrophic communities in a low nutrient, low chlorophyll system, Limnol. Oceanorg., 50(6), 1810–1819, 2005. Brody, L. R. and Nestor, M. J. R.: Handbook for Forecasters in the Mediterranean, Part 2. Regional forecasting aids for the Mediter- ranean Basin. Naval Environmental Prediction Research Facility (Technical Report TR 80-10), pp. VII-1; VII-13, Monterey, Cal- ifornia, 1980. 4 Conclusions In this study, factors controlling nutrient composition in aerosol and rainwater, differences in sea-water and pure- water solubilities of nutrients and atmospheric and riverine nutrient inputs to the Northeastern Levantine Basin of the Eastern Mediterranean have been investigated. 3.5.2 Comparison between atmospheric and riverine nutrient ﬂuxes On the basis of annual atmospheric and riverine inputs (see Table 7) the following general observation might be made: (a) Comparison of the atmospheric and riverine ﬂuxes (annual and seasonal) reveals that inorganic nitrogen species (DIN = NO− 3 + NH+ 4 ) ﬂuxes to NLB were dom- inated by the atmospheric pathway with a mean con- www.biogeosciences.net/7/4037/2010/ Biogeosciences, 7, 4037–4050, 2010 M. Koc¸ak et al.: Atmospheric nutrient inputs to the northern levantine basin from a long-term observation M. Koc¸ak et al.: Atmospheric nutrient inputs to the northern levantine basin from a long-term observation 4048 M. Koc¸ak et al.: Atmospheric nutrient inputs to the northern levantine basin from a long-term observation Koc¸ak et al.: Atmospheric nutrient inputs to the northern levantine basin from a long-term observatio studies in this area have revealed that Si might be as important as P at coastal sites. For instance, in February and March Si/N ratios were found to be ranged from 0.2 to 0.8 during diatom bloom which imply Si deﬁciency proportionate to nitrogen is likely to inﬂuence on phy- toplankton composition on seasonal basis. Acknowledgements. This work has been supported European Union (through CITYZEN and SESAME projects), Republic of Turkey Ministry of Environment (through MEDPOLL), Turkish Scientiﬁc and Technical Research Council (SINHA, 107Go66). The authors would like to thank the 3 anonymous reviewers for their comments which help to improve the manuscript. Edited by: M. Dai References D.: The role of dust in sup- plying nitrogen and phosphorus to the Southeast Mediterranean, Limnol. Oceanorg., 47, 870–878, 2002. Kubilay, N., Nickovic, S., Moulin, C., and Dulac, F.: An illustration of the transport and deposition of mineral dust onto the eastern Mediterranean, Atmos. Environ., 34, 1293–1303, 2000. Herut, B., Zohary, T., Krom, M. D., Mantoura, R. F. C., Pitta, P., Psarra, S., Rassoulzadegan, F., Tanaka, T., and Thingstad, T. F.: Response of east Mediterranean surface water to Saharan dust: On-board microcosm experiment and ﬁeld observations, Deep- Sea Res. Pt. II, 52, 3024–3040, 2005. Lo¨ye-Pilot, M. D., Martin, J. M., and Morelli, J.: Inﬂuence of Saha- ran dust on the rain acidity and atmospheric input to the Mediter- ranean, Nature, 321, 427–428, 1986. Lo¨ye-Pilot, M. D. and Morelli, J.: Fluctuations of ionic composition of precipitations collected in Corsica related to changes in the origin of incoming aerosols, J. Aerosol. Sci., 19, 577–585, 1988. Hitzenberger, R., Berner, A., Galambos, Z., Maenhaut, W., Cafmeyer, J., Schwarz, J., M¨uller, K., Spindler, G., Wieprecht, W., Acker, K., Hillamo, R., and Makela, T.: Intercomparison of methods to measure the mass concentration of the atmospheric aerosol during INTERCOMP2000-inﬂuence of instrumentation and size cuts, Atmos. Environ., 38, 6467–6476, 2004. Lo¨ye-Pilot, M. D., Klein, C., and Martin, J. M.: Major Inorganic elements in North Western Mediterranean Aerosols, Eros 2000 4th Workshop on the North Western Mediterranean Sea, Water Pollution Report, 30, 271–277, 1993. Ludwig, W., Dumont, E., Meybeck, M., and Heussner, S.: River discharges of water and nutrients to the Mediterranean and Black Sea: Major drivers for ecosystem changes during past and future decades?, Prog. Oceanogr., 80(3–4), 199–217, 2009. Karakatsoulis, P. and Ludwig, W.: Guidelines for River (in- cluding estuaries) pollution monitoring programme for the Mediterranean region, UNEP/MAP Technical report Series 151, UNEP/MAP Athens, 2004. Markaki, Z., Oikonomou, K., Koc¸ak, M., Kouvarakis, G., Chanio- taki, A., Kubilay, N., and Mihalopoulos, N.: Atmospheric de- position of inorganic phosphorus in the Levantine basin, eastern Mediterranean: spatial and temporal variability and its role in seawater productivity, Limnol. Oceanorg., 48, 1557–1568, 2003. Keck, L. and Wittmaack, K.: Laboratory studies on the retention of nitric acid, hydrochloric acid and ammonia on aerosol ﬁlters, Atmos. Environ., 39, 2157–2162, 2005. Koc¸ak, M., Nimmo, M., Kubilay, N., and Herut, B.: Spatio- temporal aerosol trace metal concentrations and sources in the Levantine basin of the eastern Mediterranean, Atmos. Environ., 38, 2133–2144, 2004a. Markaki, Z., Lo¨ye-Pilot, M. References Koc¸ak et al.: Atmospheric nutrient inputs to the northern levantine basin from a long-term observation M. Koc¸ak et al.: Atmospheric nutrient inputs to the northern levantine basin from a long-term observation 4049 Mediterranean, Atmos. Environ., 41(18), 3806–3818, 2007b. Mediterranean, Atmos. Environ., 41(18), 3806–3818, 2007b. Guerzoni, S., Molinaroli, E., and Chester, R.: Saharan dust inputs to the western Mediterranean Sea: depositional patterns, geochem- istry and sedimentological implications. Deep-Sea Res. Pt. II, Topical Studies in Oceanography, 44(3–4), 631–654, 1997. Koc¸ak, M., Theodosi, C., Zarmpas, P., Im, U., Bougia- tioti., A., Yenig¨un, O., and Mihalopoulos, N.: Particu- late matter (PM10) in Istanbul: Origin, source areas and potential impact on surrounding regions, Atmos. Environ., doi:10.1016/j.atmosenv.2010.10.007, 2010. Guerzoni, S., Chester, R., Dulac, F., Herut, B., Lo¨ye-Pilot, M. D., Measures, C., Migon, C., Molinaroli, E., Moulin, C., Rossini, P., Saydam, C., Soudine, A., and Ziveri, P.: The role of atmo- spheric deposition in the biogeochemistry of the Mediterranean sea, Prog. Oceanogr., 44, 147–190, 1999. Kouvarakis, G., Mihalopoulos, N., Tselepides, A., and Stavrakaki, S.: On the importance of atmospheric inputs of inorganic nitro- gen species on the productivity of the eastern Mediterranean Sea, Global Biogeochem. Cy., 15, 805–817, 2001. Guieu, C., Loye-Pilot, M.-D., Ridame, C., and Thomas, C.: Chem- ical characterization of the Saharan dust end-member: Some biogeochemical implications for the western Mediterranean Sea, J. Geophys. Res., 107(D15), 4258, doi:10.1029/2001JD000582, 2002. Krom, M. D., Kress, N., Brenner, S., and Gordon, L. I.: Phosphorus limitation of primary productivity in the eastern Mediterranean, Limnol. Oceanorg., 36, 424–432, 1991. Krom, M. D., Herut, B., and Mantoura, R. F. C.: Nutrient bud- get for the Eastern Mediterranean: implications for P limitation, Limnol. Oceanorg., 49, 1582–1592, 2004. Hamad, W., Millot, C., and Taupier-Letage, I.: A new hypothesis about the surface circulation in the eastern basin of the mediter- ranean sea, Prog. Oceanogr., 662–4, 287–298, 2005. Krom, M. D., Emeis, K. C., and Van Cappellen, P.: Why is the East- ern Mediterranean phosphorus limited? Prog. Oceanogr., 85(3– 4), 236–244, 2010. Herut, B., Krom, M. D., Pan, G., and Mortimer, R.: Atmospheric in- put of nitrogen and phosphorus to the Southeast Mediterranean: sources, ﬂuxes and possible impact, Limnol. Oceanorg., 44, 1683–1692, 1999. Kubilay, N. and Saydam, C.: Trace elements in atmospheric par- ticulates over the Eastern Mediterranean: concentration, sources and temporal variability, Atmos. Environ., 29, 2289–2300, 1995. Herut, B., Collier, R., and Krom, M. References Dry and wet deposition ﬂuxes were found comparable for PO3− 4 . Sidiss and NH+ 4 ﬂuxes were found to be dominated by wet deposition (∼60%) whilst dry deposition was a main source of nitrate ﬂux (∼80%). Seyhan and Ceyhan were found to be the main fresh water over the study the region with nutrient contributions more than 85% of the total river- ine nutrient inputs. Riverine DIN pool was found to be domi- nated by nitrate input (75%). Comparison of the atmospheric and riverine ﬂuxes demonstrated that DIN and PO3− 4 ﬂuxes to NLB were dominated by atmospheric pathway (∼90% and ∼60%). However, the Si pool in the NLB was almost ex- clusively originated from riverine runoff (∼90%). Consider- ing molar N/P ratios from the atmosphere (236) and riverine (22) sources it is clear that the NLB of the Eastern Mediter- ranean Sea receives excessive amounts of DIN; more than the amounts required by autotrophic organisms and this un- balanced P and N inputs may provoke even more phosphorus deﬁciency. Observed molar Si/N ratio suggested that Si lim- itation relative to N and it might cause a switch from diatom dominated communities to non-siliceous populations partic- ularly at coastal areas where in NLB. Cape, J. N., Methven, J., and Hudson, L. E.: The use of trajec- tory cluster analysis to interpret trace gas measurements at Mace Head, Ireland, Atmos. Environ., 34, 3651–3663, 2000. Carbo, P., Krom, M. D., Homoky, W. B., Benning, L. G., and Herut, B.: Impact of atmospheric deposition on N and P geochemistry in the southeastern Levantine basin, Deep-Sea Res. Pt. II, 52, 3041–3053, 2005. Chen, Y., Mills, S., Street, J., Golan, D., Post, A., Jacobson, M., and Paytan, A.: Estimates of atmospheric dry deposition and associ- ated input of nutrients to Gulf of Aqaba seawater, J. Geophys. Res., 112, D04309, doi:10.1029/2006JD007858, 2007. Draxler, R. R. and Hess, G. D.: An overview of the HYSPLIT 4 modeling system for trajectories, dispersion and deposition, Aus- tralian Meteorological Magazine, 47, 295–308, 1998. Duce, R. A., Liss, P. S., Merrill, J. T., Atlas, E. L., and Buat-Menard, P.: The atmospheric input of trace species to the world ocean, Global Biogeochem. Cy., 5, 193–259, 1991. Erduran, S. M. and Tuncel, S. G.: Gaseous and particulate air pol- lutants in the Northeastern Mediterranean Coast., Sci. Total En- viron., 281, 205–215, 2001. Biogeosciences, 7, 4037–4050, 2010 www.biogeosciences.net/7/4037/2010/ M. Biogeosciences, 7, 4037–4050, 2010 M. Koc¸ak et al.: Atmospheric nutrient inputs to the northern levantine basin from a long-term observation 4050 References D., Violaki, K., Benyahya, L., and Mihalopoulos, N.: Variability of atmospheric deposition of dis- solved nitrogen and phosphorus in the Mediterranean and possi- ble link to the anomalous seawater N/P ratio, Mar. Chem., 120, 187–194, 2010. Koc¸ak, M., Kubilay, N., and Mihalopoulos, N.: Ionic composition of lower tropospheric aerosols at a northeastern Mediterranean site: implications regarding sources and long-range transport, Atmos. Environ., 38, 2067–2077, 2004b. Migon, C., Sandroni, V., and B´ethoux, J. P.: Atmospheric input of anthropogenic phosphorus to the northwest Mediterranean under oligotrophic conditions, Mar. Environ. Res., 52, 413–426, 2001. Koc¸ak, M.: Comprehensive chemical characterization of aerosols in the Eastern Mediterranean: Sources and Long range transport, Ph. D. thesis, Middle East Technical University, Institute of Ma- rine Sciences, Erdemli, Turkey, 2006. Mihalopoulos, N., Stephanou, E., Kanakidou, M., Pilitsidis, S., and Bousquet, P.: Tropospheric aerosol ionic composition in the E. Mediterranean region, Tellus, 49B, 1–13, 1997. Koc¸ak, M., Mihalopoulos, N., and Kubilay, N.: Chemical composi- tion of the ﬁne and coarse fraction of aerosols in the northeastern Mediterranean, Atmos. Environ., 41(34), 7351–7368, 2007a. Moulin, C., Lambert, E., Dayan, U., Masson, V., Ramonet, M., Bousquet, P., Legrand, M., Balkanski, Y. J., Guelle, W., Marti- corena, B., Bergametti, G. and Dulac, F.: Satellite climatology of African dust transport in the Mediterranean atmosphere, J. Geo- Koc¸ak, M., Mihalopoulos, N., and Kubilay, N.: Contributions of natural sources to high PM10 and PM2.5 events in the eastern Biogeosciences, 7, 4037–4050, 2010 www.biogeosciences.net/7/4037/2010/ M. Koc¸ak et al.: Atmospheric nutrient inputs to the northern levantine basin from a long-term observation Koc¸ak et al.: Atmospheric nutrient inputs to the northern levantine basin from a long-term observatio phys. Res., 103, 13137–13144, 1998. Thingstad, T. F., Krom, M. D., Mantoura, R. F. C., Flaten, G. A. F., Groom, S., Herut, B., Kress, N., Law, C. S., Pasternak, A., Pitta, P., Psarra, S., Rassoulzadegan, F., Tanaka, T., Tselepides, A., Wassmann, P., Woodward, E.M.S., Wexels Riser, C., Zodiatis, G., and Zohary, T.: Nature of phosphorus limitation in the ultra- oligotrophic eastern Mediterranean, Science, 309, 1068–1071, 2005. phys. Res., 103, 13137–13144, 1998. Ottley, C. J. and Harrison, R. M.: Atmospheric dry deposition ﬂux of metallic species to the North Sea, Atmos. Environ., 27A, 685– 695, 1993. ¨Ozsoy, T. and Saydam, C.: Acidic and alkaline precipitation in the Cilician Basin, north western Mediterranean sea, Sci. Total Env- iron., 352, 93–109, 2000. Uysal, Z. and K¨oksalan, I.: Synechococcus dynamics in the Levan- tine basin shelf waters (Northeastern Mediterranean), submitted to Medit. Mar. Sci., 11(2), 277–294, 2010. Pitta, P., Stambler, N., Tanaka, T., Zohary, T., Tselepides, A., and Rassoulzadegan, F.: Biological response to P addition in the Eastern Mediterranean Sea: the microbial race against time, Deep-Sea Res. Pt. II, 52, 2961–2974, 2005. Wieprecht, W., Br¨uggemann, E., M¨uller, K., Acker, K., Spindler, G., ten Brink, H.M., Hitzenberger, R., and Maenhaut, W.: IN- TERCOMP2000: Ionic constitution and comparison of ﬁlter and impactor samples and its analysis, Atmos. Environ., 38, 6477– 6486, 2004. Ridame, C. and Guieu, C.: Saharan input of phosphate to the olig- otrophic water of the open western Mediterranean Sea, Limnol. Oceanorg., 47(3), 856–869, 2002. Spokes, L. J. and Jickells, T. D.: Is the atmosphere really an impor- tant source of reactive nitrogen to coastal waters?, Cont. Shelf Res., 25, 2022–2035, 2005. Yılmaz, A. and Tu˘grul, S.: The effect of cold- and warm-core eddies on the distribution and stoichiometry of dissolved nutrients in the northeastern Mediterranean, J. Mar. Sys., 16, 253–268, 1998. Spokes, L. J., Jickells, T. D., and Jarvis, K.: Atmospheric inputs of trace metals to the northeast Atlantic Ocean: the importance of south-easterly ﬂow, Mar. Chem., 76, 319–330, 2001. Biogeosciences, 7, 4037–4050, 2010 www.biogeosciences.net/7/4037/2010/
https://openalex.org/W3187407507	https://zenodo.org/records/7104193/files/37%2025738.pdf	English	null	Novel deep learning model for vehicle and pothole detection	Indonesian journal of electrical engineering and computer science	2,021	cc-by	4,419	Indonesian Journal of Electrical Engineering and Computer Science Vol. 23, No. 3, September 2021, pp. 1576~1582 ISSN: 2502-4752, DOI: 10.11591/ijeecs.v23.i3.pp1576-1582 Indonesian Journal of Electrical Engineering and Computer Science Vol. 23, No. 3, September 2021, pp. 1576~1582 ISSN: 2502-4752, DOI: 10.11591/ijeecs.v23.i3.pp1576-1582 Indonesian Journal of Electrical Engineering and Computer Science Vol. 23, No. 3, September 2021, pp. 1576~1582 ISSN: 2502-4752, DOI: 10.11591/ijeecs.v23.i3.pp1576-1582  1576 Corresponding Author: Thangavelu S Dept of Computer Science & Engineering, Amrita School of Engineering Coimbatore Amrita Vishwa Vidyapeetham Amrita University, India Email: s_thangavel@cb.amrita.edu Keywords: Faster R-CNN Inception model Transfer learning This is an open access article under the CC BY-SA license. This is an open access article under the CC BY-SA license. Corresponding Author: Novel deep learning model for vehicle and pothole detection Gayathri K, Thangavelu S Department of Computer Science and Engineering, Amrita School of Engineering, Coimbatore, Amrita Vishwa Vidyapeetham, Amrita University, India Gayathri K, Thangavelu S Department of Computer Science and Engineering, Amrita School of Engineering, Coimbatore, Amrita Vishwa Vidyapeetham, Amrita University, India ABSTRACT The most important aspect of automatic driving and traffic surveillance is vehicle detection. In addition, poor road conditions caused by potholes are the cause of traffic accidents and vehicle damage. The proposed work uses deep learning models. The proposed method can detect vehicles and potholes using images. The faster region-based convolutional neural network (CNN) and the inception network V2 model are used to implement the model. The proposed work compares the performance, accuracy numbers, detection time, and advantages and disadvantages of the faster region-based convolution neural network (Faster R-CNN) with single shot detector (SSD) and you only look once (YOLO) algorithms. The proposed method shows good progress than the existing methods such as SSD and YOLO. The measure of performance evaluation is Accuracy. The proposed method shows an improvement of 5% once compared with the previous methods such as SSD and YOLO. Article history: Received Apr 14, 2021 Revised Jul 8, 2021 Accepted Jul 18, 2021 Journal homepage: http://ijeecs.iaescore.com 1. INTRODUCTION YOLO also has difficulty in finding small things or objects. g y p p g y g g j A lot of researchers are working on this issue. The high volume of vehicular traffic causes traffic blockage, unpleasant delays, financial loss, accidents, road damages, the passage of emergency vehicles, pollution, violations of traffic, and health concerns. There is a necessity for an enhanced system that would optimize the efficiency of traffic flow. Prasad et al. [1] developed a real-time model for estimating traffic density at a junction with the help of image processing methods. To estimate traffic density, the author used techniques such as local binary pattern and histogram of oriented gradient-based approaches. Choudekar et al. [2] proposed a model that controls the traffic light with the help of an image processing technique. This system captures Image sequences. With the help of picture matching and a reference picture of the vacant road, the captured pictures are matched sequentially. The edge detection operator – prewitt operator has been used to perform the detection of edge. The duration of the lights is then adjusted depending upon the percentage of traffic matching. Kastrinaki et al. [3] presented a summary of the image processing tool and analysis tool which are applied in the applications. The authors linked those two tools to finalize traffic-related models. Furthermore, the authors [3] have distinguished between the static and mobile camera cases. The goal of the work is original. As a result, the paper serves three purposes. In the first step [3], used image processing techniques were classified. The next step is to list the benefits as well as the drawbacks of those algorithms. And in the final and third step, an assessment of flaws is attempted from the integrated consideration. Mednis et al. [4] have proposed a model which is based on an accelerometer. Algorithms such Z- DIFF, G-ZERO, and Z-THRESH are used in this accelerometer-based model. The quality of the model was evaluated using a 90 percent true positive value. For the detection of a pothole, Lin et al. [5] used a non-linear support vector machine classification tool. Pereira et al. [6] have utilized this Neural Network called CNN. Pereira et al. [6] showed that their proposed method gives good progress and good performance around 99.80% than the existing methods like support vector machine. Subbiah et al. 1. INTRODUCTION Economic development and social well-being are both dependent on the road system. Surveillance, intelligent traffic control, public safety and security, and autonomous driving all require vehicle detection. Due to overloaded vehicles and poor drainage systems, road maintenance is difficult. Potholes can also damage lower-riding vehicles, causing flat tyres and wheel damage. Potholes can also cause vehicle collisions and accidents. As a result, detecting potholes is one of the most essential tasks in the road management system. It will aid in the betterment of road conditions, particularly in developing countries with limited resources. This paper proposes a novel approach for detecting vehicles and potholes using deep learning. The performance, accuracy numbers, detection time, and benefits and drawbacks about the algorithm faster region-based convolution neural network (Faster R-CNN) with the single shot detector (SSD) algorithm and you only look once (YOLO) are discussed and compared in the proposed work. Only highly reliable computing machines are suitable for algorithms like SSD and YOLO. These algorithms require more time to complete the training process. In the proposed work, the Faster-RCNN compares with SSD and YOLO algorithms. The Faster R-CNN is a convolutional neural network-based object detection algorithm that is unified, faster, and more accurate. The algorithm SSD is not suitable to identify small objects. The SSD improves detection by adding the layers of various features into this end network. And mainly when this comes to the case of smaller objects, then the SSD performs significantly worse than this Faster R-CNN. Foremost reasons of this weakness are that higher-resolution layers in the algorithm SSD are Journal homepage: http://ijeecs.iaescore.com Indonesian J Elec Eng & Comp Sci 1577 ISSN: 2502-4752  in complete control of detecting small objects. Another drawback of the SSD object detection algorithm is that the SSD training needs a very large quantity of the data. Depending on the application, this can be costly and time-consuming. in complete control of detecting small objects. Another drawback of the SSD object detection algorithm is that the SSD training needs a very large quantity of the data. Depending on the application, this can be costly and time-consuming. YOLO on the other hand, provides an end-to-end network. The YOLO algorithm has a lower recall and a higher localization error than the Faster R-CNN algorithm. It also has trouble detecting close objects because each grid can only propose two bounding boxes. 1. INTRODUCTION [7] demonstrated that the you only look once the architecture and the variants of YOLO are the most efficient when evaluating metrics in a real-time system. A new image dataset has been created by Dharneeshkar et al. [8]. The data was gathered from Indian highways. You only look once was used to annotating and training this dataset. YOLO is a phrase that means "you only look once". The mAP, precision, and recall are used to evaluate the results. The authors tested this proposed model on a variety of images containing potholes, and the results were accurate in detecting them. Ajay et al. [9] use aerial images to find small vehicles in a restricted environment. The experiment makes use of the vehicle detection in aerial imagery (VEDAI) dataset. The two phases of this model's architecture are the training and detection phases. This study uses the feature extraction techniques singular value decomposition (SVD) and histogram of oriented gradients (HOG). Ajay et al. [9] concluded that the proposed SVD features are comparable to existing HOG features in terms of performance. Pradeep et al. [10] develop two improvised models for detecting vehicles during the nighttime. For this, the authors used an optical camera that is facing forward. In the case of a dual tail-light scenario, method one is used to improve accuracy. In the case of single-tail-light functionality, they use the second technique to improve accuracy. The algorithms are tested in real-life traffic scenarios. 2. RESEARCH METHOD Research method for the project involves three main parts-data collection, training, and detection. The proposed vehicle and pothole detection technique detect vehicles and potholes in images with the transfer learning technique. Faster R-CNN and inception v2 models are the techniques used in the proposed model. The ultimate benefit of this proposed method is least time for training along with very high accuracy.  1578  1578 ISSN: 2502-4752 The new dataset fetches to fine-tune the convolutional neural networks (CNNs) [11], [12]. And if latest datasets are similar to an actual datasets, then selfsame weight is used to extract the features. Here this dataset is vastly divergent from these original datasets. Earlier layers of this CNN found a generic property. The layers of CNN [13], [14], on the other hand, become increasingly particular to the specifics of classes in the actual or real dataset. As a result, previous layers will aid in the extraction of current data features. Because of the fewer amount of data, it fixes the earlier layers. Rest of these layers are retrained. 2.1. Transfer learning technique Transfer learning is a technique that involves training a neural network model on a problem that is similar to the solved one. The layers of the trained model use a new model to train on the problem of interest. The implementation of gained knowledge while resolving one problem to another but related problem is known as the transfer learning technique. It is possible to save a significant amount of time with the transfer learning method. Any pre-trained model can be chosen and used in this method as a first step. A model that has had all of its parameters trained refers to a pre-trained model. The final step will be to fine-tune the parameters. Novel deep learning model for vehicle and pothole detection (Gayathri K) Indonesian J Elec Eng & Comp Sci The following steps are performed to obtain the trained faster region-based convolutional neural network architecture:  Training of RPN; To propose the expected region, the region proposal network architecture is first trained with the dataset.  Training of RPN; To propose the expected region, the region proposal network architecture is first trained with the dataset.  Fast R-CNN training; Faster region-based convolutional neural Network is an amalgamation of the region proposal network and the fast region-based convolutional neural network, to make [18], [19] a faster region-based convolutional neural network, a fast region-based convolutional neural network is trained with the proposals obtained by region proposal network. This is done after the training.  Fast R-CNN training; Faster region-based convolutional neural Network is an amalgamation of the region proposal network and the fast region-based convolutional neural network, to make [18], [19] a faster region-based convolutional neural network, a fast region-based convolutional neural network is trained with the proposals obtained by region proposal network. This is done after the training.  Fixing convolutional layers; Region proposal network is fine-tuned for unique layers.  Fixing convolutional layers; Fast region-based convolutional neural network’s fully connected layers are then fine-tuned.  Fixing convolutional layers; Fast region-based convolutional neural network’s fully connected layers are then fine-tuned. 2.3. Architecture of inception network V2 The Inception network v2 is an improved type in the series of the inception networks. Inception network v2 is a complicated architecture of convolutional neural networks. First version of the inception network series is the Inception network v1. Convolution executes in Inception network v1. In addition, maximum pooling takes place. The outputs are all connected. The outputs are then forwarded to the next inception module. Figure 2 depicts an Inception network v1 module. Computationally, these deep learning models are expensive [20], [21]. To restrict the number of input channels includes an extra convolution. The model uses this inception module to develop the GoogLeNet neural network architecture. There are nine such inception modules in the GoogLeNet neural network architecture [22], [23]. Also, it introduces the inception v2 architecture to improve computational speed, as shown in Figure 3. Also, introduce a new method for improving performance. Filter size convolution nxn in this model combines filter size convolution 1xn and filter size convolution nx1. The model is 33% less expensive. Figure 2. Inception network v1 Figure 3. Architecture of inception-v2 Figure 2. Inception network v1 Figure 3. Architecture of inception-v2 Figure 2. Inception network v1 Figure 2. Inception network v1 2.2. Architecture of F-RCNN The faster region-based convolutional neural network is also known as F-RCNN or Faster R-CNN. First of all, it is necessary to understand about the Fast R-CNN before moving on to faster RCNN. Fast RCNN (fast region-based CNN) is a detector. Fast region-based convolutional neural network uses external proposal. External selective search, convolutional neural network with maximum pooling, fully connected layers, region of interest pooling layer, then output layers are among the components. With the help of convolutional neural network and pooling layers - max-pooling layers, Fast RCNN receives an input image. It extracts the convolutional feature map. The region of interest pooling layer plays a salient role. These fully connected layer accepts fixed sizes only. As a result, the ROI pooling layers convert CNN's output into a predetermined size. Fast RCNN becomes faster region-based convolutional neural network when fast RCNN combines along with region proposal network. So, the main dissimilarity among fast region-based convolutional neural network and faster region-based convolutional neural network is only about the region proposal. Fast region-based convolutional neural network has an external selective search [15]. Region proposal network combines with the Architecture in faster region-based convolutional neural network. region proposal network also chooses where to look in the image to reduce computational complexity and requirements. It mainly scans this input image. It then generates k output boxes, all with two scores representing probability for an object's availability [16], [17]. Figure 1 depicts the F-RCNN architecture. Figure 1. F-RCNN architecture Figure 1. F-RCNN architecture Indonesian J Elec Eng & Comp Sci, Vol. 23, No. 3, September 2021: 1576 - 1582 1579 ISSN: 2502-4752 Indonesian J Elec Eng & Comp Sci   1580  1580 ISSN: 2502-4752 2.6. Detection of vehicles and potholes After the pre-processing of the data, a model to detect vehicles and potholes has created. The model has fine-tuned using the Inception network v2 architecture [24], [25]. Training requires a configuration file. There are nine different classes in this configuration file. Ambulance, auto, bus, car, motorcycle, pothole, potholeGroup, truck, and van are the nine classes. The proposed system will take some time to train model, depending on the hardware configuration. Once the training completes, the testing process begins. During testing, the model performed admirably. It can, however, be improved. 2.5. Data pre-processing There are over a thousand images in the dataset. The vehicle detection and pothole detection tasks consider as uncomplicated only if region in those pictures containing road has cropped. Later it is utilized in the detection. Also, it requires extra time in order to crop one thousand or more images. All of the images are labeled. Then a comma-separated values (CSV) file is created. This CSV file includes the name of the image, path, dimension of the image, and coordinate of the vehicles and potholes. Then TfRecords for the dataset is created. It feeds and fine-tunes the Inception model. The image size is 3680x2760 pixels. 2.4. Dataset The datasets identified and used in this proposed work are, traffic surveillance dataset and the Pothole detection dataset. The Traffic Surveillance dataset is composed of 400 images that include different types of vehicles on the road. This dataset is for vehicle detection. This dataset provides images to perform vehicle detection. The Pothole detection dataset is an image dataset that contains around 665 images of roads with potholes. This dataset is from Kaggle. This dataset mainly includes images with different types of potholes on roads. This dataset is for detection models and can work as proof of concept (POC)/proof of value (POV) for the maintenance of roads. For this project, the necessary data and images of both vehicles and potholes are gathered from various sources. Novel deep learning model for vehicle and pothole detection (Gayathri K) Indonesian J Elec Eng & Comp Sci  Figure 6 compares the accuracy numbers for different object sizes, such as a small object, medium object, and large object, for the three algorithms: Faster R-CNN, YOLO, and SSD. From Figure 6, it is evident that when it comes to the case of smaller objects, then the algorithm SSD performs significantly worse than this Faster R-CNN. Ultimate reason for such weakness is that the higher-resolution layers of SSD have complete control over small object detection. Another drawback of this method is that the training of SSD necessarily needs a very large quantity of data. Depending on the application, this can be costly and time-consuming. YOLO also struggles to detect smaller objects. As a result, the Faster R-CNN algorithm performs much better than SSD and YOLO in terms of accuracy numbers. Hence the Faster R-CNN is identified as the best of three algorithms. Table 1. Performance comparison Table 2. Detection time taken No. Model mAP (%) 1 Existing Model - SSD 78.70 2 Existing Model – YOLO 81.71 3 Proposed Model – Faster R-CNN 86.41 No. Model Detection Time Taken (ms) 1 SSD 2.9 ms 2 YOLO 2.5 ms 3 Faster R-CNN 2 ms Figure 6. For different sizes: faster R-CNN vs SSD vs YOLO Figure 6. For different sizes: faster R-CNN vs SSD vs YOLO 4. CONCLUSION This paper proposes a new model for detecting vehicles and potholes in images. The proposed model employs the convolutional neural network's inception network v2. The results of the experiments show that the proposed method outperforms others. The model accurately detects vehicles and potholes. And in the proposed work, the performance, accuracy numbers, detection time, and advantages and disadvantages of Faster R-CNN with SSD algorithm and YOLO are identified, discussed, and compared. Novel deep learning model for vehicle and pothole detection (Gayathri K) 3. RESULTS AND DISCUSSION 3, September 2021: 1576 - 1582 1581 ISSN: 2502-4752 Indonesian J Elec Eng & Comp Sci 3. RESULTS AND DISCUSSION The result analysis of an algorithm or model relies on a set of parameters. The majority of the evaluation depends on parameters. They are accuracy, detection time, performance, and other parameters. The percentage of the desired result produced by the algorithm is one such important parameter that measures the model's or algorithm's promising element. Another parameter is the time taken which illustrates the time taken by the model or algorithm to give the outcome. Then the performance is also one such important parameter that determines how well the model or algorithm performs. Figure 4 and Figure 5 show the outcomes of the testing process. The object detection metrics [26] are precision and recall, intersection over union (IoU), average precision (AP), and mean average precision (mAP). Precision is a metric used for determining whether a prediction is correct or not. From all of the correctly predicted data, the recall gives the actual predictions. The IoU metric can be used to find differences between ground truth annotations and predicted bounding boxes. The mean AP or mAP provides precision for the entire model. Here mAP is used to obtain the percentage of correct predictions in the system. Figure 4. Detection of vehicles Figure 5. Detection of pothole Figure 4. Detection of vehicles Figure 5. Detection of pothole The mean AP is the average value of AP. Existing models such as SSD and YOLO had mAP values of 78.70 percent and 81.71 percent, respectively, based on the calculation of mean average precision. And the proposed model showed an mAP value of 86.41%. There has been an increase of around 5% in mean average precision for this proposed system. Table 1 shows a comparison about the performance of the model. When the most common and widely used parameters are applied to object detection algorithms, the results show that Faster R-CNN is significantly faster than the SSD and YOLO algorithms. The proposed regions are taken into account by an independent network in the Faster R-CNN algorithm. Other methods, such as SSD and YOLO, are slower and slower than the Faster R-CNN method. Table 2 represents a comparison of these three methods or models in terms of detection time taken in ms. It is clear that Faster R-CNN algorithm outperforms the other two algorithms from Table 2. Indonesian J Elec Eng & Comp Sci, Vol. 23, No. REFERENCES [1] D. Prasad, K. Kapadni, A. Gadpal, M. Visave, and K. Sultanpure, "HOG, LBP and SVM based Traffic Density Estimation at Intersection," IEEE Pune Section International Conference (PuneCon), Pune, India, pp. 1-5, 2020, doi: 10.1109/PuneCon46936.2019.9105731. [2] P. Choudekar, S. Banerjee, and M. K. Muju, "Implementation of Image Processing in Real Time Traffic Light Control," 3rd International Conference on Electronics Computer Technology, Kanyakumari, India, 2011, pp. 94-98, [3] V. Kastrinaki, M. Zervakis, and K. Kalaitzakis, "A Survey of Video Processing Techniques for Traffic Applications," Image and Vision Computing, vol. 21, no. 4, pp. 359-381, 2003, doi: 10.1016/S0262-8856(03)00004-0. [4] A. Mednis, G. Strazdins, R. Zviedris, G. Kanonirs, and L. Selavo, "Real time pothole detection using Android smartphones with accelerometers," International Conference on Distributed Computing in Sensor Systems and Workshops (DCOSS), Barcelona, Spain, 2011, pp. 1-6, doi: 10.1109/DCOSS.2011.5982206. [5] J. Lin and Y. Liu, "Potholes Detection Based on SVM in the Pavement Distress Image," Ninth International Symposium on Distributed Computing and Applications to Business, Engineering and Science, Hong Kong, China, 2010, pp. 544-547, doi: 10.1109/DCABES.2010.115. [6] V. Pereira, S. Tamura, S. Hayamizu and H. Fukai, "A Deep Learning-Based Approach for Road Pothole Detection in Timor Leste," IEEE International Conference on Service Operations and Logistics, and Informatics (SOLI), Singapore, 2018, pp. 279-284, doi: 10.1109/SOLI.2018.8476795. Novel deep learning model for vehicle and pothole detection (Gayathri K)  1582 ISSN: 2502-4752 [7] U. Subbiah, D. K. Kumar, S. K. Thangavel, and L. Parameswaran, "An Extensive Study and Comparison of the Various Approaches to Object Detection using Deep Learning," Third International Conference on Smart Systems and Inventive Technology (ICSSIT), Tirunelveli, India, 2020, pp. 183-194, doi: 10.1109/ICSSIT48917.2020.9214185. [8] J. Dharneeshkar, V. Soban Dhakshana, S. A. Aniruthan, R. Karthika and L. Parameswaran, "Deep Learning based Detection of potholes in Indian roads using YOLO," International Conference on Inventive Computation Technologies (ICICT), Coimbatore, India, 2020, pp. 381-385, doi: 10.1109/ICICT48043.2020.9112424. [9] A. Ajay, V. Sowmya, and K. P. Soman, "Vehicle detection in aerial imagery using eigen features", International Conference on Communication and Signal Processing (ICCSP), Chennai, India, 2018, pp. 1620-1624, doi: 10.1109/ICCSP.2017.8286664. [10] C. S. Pradeep and R. Ramanathan, "An improved technique for Night-time Vehicle detection," International Conference on Advances in Computing, Communications and Informatics (ICACCI), Bangalore, India, 2018, pp. 508- 513, doi: 10.1109/ICACCI.2018.8554712. [11] S. Ren, K. He, R. Girshick and J. REFERENCES Sun, "Faster RCNN: Towards Real-Time Object Detection with Region Proposal Networks," IEEE Transactions on Pattern Analysis and Machine Intelligence, pp. 1-14, 2017, doi: 10.1109/TPAMI.2016.2577031 [12] N. F. Sahidan, A. K. Juha, and Z. Ibrahim, "Evaluation of basic convolutional neural network and bag of features for leaf recognition," Indonesian Journal of Electrical Engineering and Computer Science, vol. 14, no. 1, pp. 327-332, 2019, doi: 10.11591/ijeecs.v14.i1.pp327-332. [13] S. Shindo, T. Goto, T. Kirishima, and K. Tsuchida, "An Optimization of Facial Feature Point Detection Program by Using Several Types of Convolutional Neural Network," Indonesian Journal of Electrical Engineering and Computer Science, vol. 16, no. 2, pp. 827-834, 2019, doi: 10.11591/ijeecs.v16.i2.pp827-834. [14] M. Z. Osman, M. A. Maarof, M. F. Rohani, and N. N. A. Sjarif, "A multi-color based features from facial images for automatic ethnicity identification model," Indonesian Journal of Electrical Engineering and Computer Science, vol. 18, no. 3, pp. 1383-1390, 2020, doi: 10.11591/ijeecs.v18.i3.pp1383-1390. [15] P. Patel and A. Thakkar, "The upsurge of deep learning for computer vision applications," Indonesian Journal of Electrical Engineering and Computer Science, vol. 10, no. 1, pp. 538-548, 2020, doi: 10.11591/ijece.v10i1.pp538-548. [16] R. Rassetiadi and S. Suharjito, "Foreign exchange prediction based on indices and commodities price using convolutional neural network," Indonesian Journal of Electrical Engineering and Computer Science, vol. 18, no. 1, pp. 494-501, 2019, doi: 10.11591/ijeecs.v18.i1.pp494-501. pp j pp [17] N. N. A. A. Hamid, R. A. Razali, and Z. Ibrahim, "Comparing bags of features, conventional convolutional neural network and alexnet for fruit recognition", Indonesian Journal of Electrical Engineering and Computer Science, vol. 14, pp. 333-339, 2019, doi: 10.11591/ijeecs.v14.i1.pp333-339. pp j pp [18] H. F. Pardede, A. R. Yuliani, and R. Sustika, "Convolutional Neural Network and Feature Transformation for Distant Speech Recognition," Indonesian Journal of Electrical Engineering and Computer Science, vol. 8, no. 6, p. 5381, 2018, doi: 10.11591/ijece.v8i6.pp.5381-5388. p j pp [19] K. S. Gautam and S. K. Thangavel, "Video analytics based intelligent surveillance system for smart buildin Springer Soft computing, vol. 23, no. 8, pp. 2813-2837, 2019, doi: 10.1007/s00500-019-03870-2. [20] T. Arathi and L. Parameswaran, "Slantlet Transform and Phase Congruency based Image Compression," International Journal of Computer Applications (IJCA), Proceedings on Amrita International Conference of Women in Computing, 2013, pp. 12-16. p g pp [21] L. Suhao, L. Jinzhao, L. Guoquan, B. Tong, W. Huiqian, and P. Yu, "Vehicle type detection based on deep learning in traffic scene," Procedia Computer Science, pp. REFERENCES 564-572, 2018, doi: 10.1016/j.procs.2018.04.281. [22] W. Lumchanow and S. Udomsiri, "Image classification of malaria using hybrid algorithms: convolutional neural network and method to find appropriate K for K-Nearest neighbor," Indonesian Journal of Electrical Engineering and Computer Science, vol. 16, no. 1, pp. 382-388, 2019, doi: 10.11591/ijeecs.v16.i1.pp382-388. [23] P S. Gupta, P. Sharma, D. Sharma, and C. Gupta, "Detection and localization of potholes in thermal images using deep neural networks," Multimedia Tools and Applications, Springer, vol. 79, pp. 35-36, 2020, doi: 10.1007/s11042-020-09293-8. [24] D. A. Jasm, M. M. Hamad and A. T. H. Alrawi, "Deep image mining for convolution neural network," Indonesian Journal of Electrical Engineering and Computer Science, vol. 20, no. 1, pp. 347-352, 2020, doi: 10.11591/ijeecs.v20.i1.pp347-352. [25] P. K. Sethy, N. K. Barpanda, A. K. Rath, and S. K. Behera, "Rice false smut detection based on Faster R-CNN," Indonesian Journal of Electrical Engineering and Computer Science, vol. 19, no. 3, pp. 1590-1595, 2020, doi: 10.11591/ijeecs.v19.i3.pp1590-1595. [26] M. A. Anwer, S. M. Shareef, and A. M. Ali, "Accident vehicle types classification: a comparative study between different deep learning models," Indonesian Journal of Electrical Engineering and Computer Science, vol. 21, no. 3, pp. 1474-1484, 2021, doi: 10.11591/ijeecs.v21.i3.pp1474-1484. Indonesian J Elec Eng & Comp Sci, Vol. 23, No. 3, September 2021: 1576 - 1582
https://openalex.org/W2803370420	https://ojs.bibl.u-szeged.hu/index.php/analecta/article/download/144/95	English	null	Kiskunfélegyháza City Hospital and Clinic, Bath and Rehabilitation Center development and some context of patient satisfaction and logistics	Analecta	2,014	cc-by	3,288	ABSTRACT The topic niche processing, rehabilitation domestic situation analysis has been very little. The results are set out proposals to increase the efficiency of operation. Keywords: rehabilitation, logistics, musculoskeletal rehabilitation, therapy, post operation treatment Keywords: rehabilitation, logistics, musculoskeletal rehabilitation, therapy, post operation treatment 2014/1 2014/1 ISSN 2064-7964 Analecta ABSTRACT The study of Kiskunfélegyháza hospital rehabilitation department of the launch and operation are summarized in the light of the results data. Started in May 2005, public investment is achieved, containing functional reconstruction of partial hospitalization as a result of enlargement, was delivered in September 2006, was built to exploit the thermal rehabilitation unit, which includes a 40-bed department on the first floor, the ground floor clinic rooms, the spa offers therapeutic and physiotherapy service rooms and swimming pools, the basement of mechanical rooms. A further problem in question – the subject-oriented guide to healing, health, illness, and a definition of rehabilitation took place – according to the literature. It will be determined in the prevention, injury, disability and concepts of disability in the light of the logistics. Write the process that accompanies the service, taking into account the economic aspects of the market. Have been processed under the rehabilitated patient population, patient registration data by age between 2007 and 2012. Considering the cases of the total care days are the number of transferees from other classes, social and home placement per capita average care days in the bed occupancy and the number of death cases were surveyed on written questionnaires. Rehabilitation hospital care is most effective with measurable - which can be seen in the analysis - that of the average length of more than 18 days, the bed occupancy of 100% and a mortality rate of 2%. However, the situation logistical aspects of the process can be improved further by examining the quantitative and qualitative indicators of well. The topic niche processing, rehabilitation domestic situation analysis has been very little. The results are set out proposals to increase the efficiency of operation. Have been processed under the rehabilitated patient population, patient registration data by age between 2007 and 2012. Considering the cases of the total care days are the number of transferees from other classes, social and home placement per capita average care days in the bed occupancy and the number of death cases were surveyed on written questionnaires. Rehabilitation hospital care is most effective with measurable - which can be seen in the analysis - that of the average length of more than 18 days, the bed occupancy of 100% and a mortality rate of 2%. However, the situation logistical aspects of the process can be improved further by examining the quantitative and qualitative indicators of well. 3. SUPPLY OPTIONS, SWIMMING COMPLEX MEDICAL TREATMENT Offered facilities and by Kiskunfélegyháza City Hospital and Clinic, Bath and Rehabilitation Center A) medicinal water pool Mud pack, carbonic acid bath, water jet massage / underwater jet massage / massage, weight bath, underwater gymnastics group. B) physiotherapy treatments Iontoforésis (using various gels rheumatism), ultrasound treatment, landing, four compartments and other galvanic treatments, TENS therapy to strengthen the purpose of pain, nerve stimulation, diadynamic current, Bemer and magnet therapy. C) physiotherapy: Physiotherapy is the most special tournaments and the Mc. Kenzei- gymnastics, the suspension grid) and joint use of moving machinery. (A. S. LEON ET AL. 2005; M. F. PIEPOLI ET AL. 2010; E.G. 2012; S. GEROLD ET AL. 2007) C) physiotherapy: Physiotherapy is the most special tournaments and the Mc. Kenzei- gymnastics, the suspension grid) and joint use of moving machinery. (A. S. LEON ET AL. 2005; M. F. PIEPOLI ET AL. 2010; E.G. 2012; S. GEROLD ET AL. 2007) The assigned team members are, physical therapists, physiotherapists, masseurs, lifeguards, specialized doctors, assistants, patients, and last but not least, providing transportation and carrying out staff, patient transport family members by car or public transport, train, bus, and not insignificantly electric wheelchair, using a bicycle take the smooth operation of the unit. Everything has to count the cost of which lies directly or indirectly, but the patient. (Szegedi Z. - Prezenszki J. 2010; Szabó Gy., Renner A. 2004) The specialist doctor ordered therapy according to the disease, like number and the composition of the treatment and patients get treatment sheet. It is a spa treatments medical prescription. Then, the patient controls the form and the combination of vehicles visiting the dispatcher service for the purpose of accounting for compensation cost and schedule of treatments. Part o the treatments paid by TB (medical insurance) as a fixed portion and the rest is paid by the patient. In the case of public health beneficiaries, the service is free of charge. Timing of schedule starts, chosen the earliest possible time, leaving no empty space in view of the high degree of efficiency and demand. Generally, waiting for treatment the average of days is between 10 days and two weeks. Having been ready administration treatment logistics starts, using above mentioned tools, and knowledge of the exact dates. 1. INTRODUCTION Kiskunfélegyháza is lucky thermal characteristic features, which provides public bathing facilities for the general public for decades. People - in folk medicine - realized the benefits of, and then this option is left unused for a long time. During the health care systems, is an excellent opportunity offered itself as a mineral water based uptake rehabilitation, intensive development in the city. Since the first of January 2011 has started in Kiskunfélegyháza hospital musculoskeletal rehabilitation clinics funded by the National Health Insurance (OEP) 9 hours a week, which is controlled by two specialist medical doctor. 48 2014/1 ISSN 2064-7964 Analecta Direct territorial coverage obligation of consultation covers 52,000 inhabitants in Kiskunfélegyháza and its surroundings, as well as indirectly and partially Bács-Kiskun and Csongrád counties’ population. How to get in movement or activities disabled patients – through general medical doctor – to a special rehabilitation center from their home? How many hurdles to overcome not only because of the disability, but the access generally as well? After following the general medical care, or previous clinical care, follow-up examination clinic reference to quotations made by the dispatching computer program to book appointment (exact month, day, hour graduations) for 15-minute intervals to share by your personal contact or telephone appointment negotiation call. Direct territorial coverage obligation of consultation covers 52,000 inhabitants in Kiskunfélegyháza and its surroundings, as well as indirectly and partially Bács-Kiskun and Csongrád counties’ population. How to get in movement or activities disabled patients – through general medical doctor – to a special rehabilitation center from their home? How many hurdles to overcome not only because of the disability, but the access generally as well? After following the general medical care, or previous clinical care, follow-up examination clinic reference to quotations made by the dispatching computer program to book appointment (exact month, day, hour graduations) for 15-minute intervals to share by your personal contact or telephone appointment negotiation call. 2. MATERIALS AND METHODS The rehabilitation activities and related logistics processed and a short summary of the literature. Questionnaire survey among patients treated in the hospital department. Summary and some forward thinking wording. 2014/1 ISSN 2064-7964 Analecta Analecta The above figure shows that over the past six years – during the development of the department's image – an average of nearly 1,000 patients treated per year. Ensuring continuity of patient care mentioned earlier could only be a result of well-organized logistics process. Carefully planned work in liaison with the surgeon and acute departments who want their patients’ rehabilitation in Kiskunfélegyháza. Patients that they are the class after a telephone consultation, and may place patients rehabilitated made redundant at the scheduled time of the next patient bed occupancy, in addition to the overall economy is. In addition, home of the patients in the pre-notification ticket is sought, up to one week prior to the recording by phone call. Perform some tests before the shooting – which are essential for successful rehabilitation – outpatient needs to be, that the cost of the hospital's budget is not burdened (x-rays, lab etc.) too. It is important that the available recording time of the patients known drugs, the medication should be smooth. This requires a range of medicines to ensure the accuracy of the hospital logistic processes by which the suppliers created as a result. In the last two years has increased the territorial obligation, since 1st October 2010 from Kecskemét County Hospital to Kiskunfélegyháza 32 musculoskeletal rehabilitation beds have been moved, so increased the capacity of the musculoskeletal rehabilitation department. In the past 20 musculoskeletal beds capacity has increased up to 52 beds and the existing 20 bed cardiovascular unit that has 72 beds, which are the total number of beds in the rehabilitation department. With the number of patients around 1/3-dal increased compared to 2010. (FIGURE 1) However, neither the The above figure shows that over the past six years – during the development of the department's image – an average of nearly 1,000 patients treated per year. Ensuring continuity of patient care mentioned earlier could only be a result of well-organized logistics process. Carefully planned work in liaison with the surgeon and acute departments who want their patients’ rehabilitation in Kiskunfélegyháza. Patients that they are the class after a telephone consultation, and may place patients rehabilitated made redundant at the scheduled time of the next patient bed occupancy, in addition to the overall economy is. 3. SUPPLY OPTIONS, SWIMMING COMPLEX MEDICAL TREATMENT Control compulsory for the patient, in order that check efficiency, effectiveness in order to assess the importance of logistics process is repeated test within 3 weeks after completion of treatment. 49 2014/1 4. A REPRESENTATIVE SURVEY OF PATIENTS REHABILITATION OF A WORKING OF CLASS (SURVEY) 2013 survey done by the department treated patients in order to measure the work quality of the medical staff. A random selection of patients responding occurred, highlighting patients to different age, gender, educational level and social level of the database. 100 questionnaires were issued, of which 61 were processed was because the only answer to all your questions on tax questionnaires could be evaluated. The sampling period was three weeks because he planned to spend three weeks in most of the patients in the ward. p p p The survey aims to provide a cross-sectional picture of the operation of the department of users of opinion that the standard of the personnel work of the well-being of patients, refer to the state of healing. Received particular attention was in the compilation of the questionnaire aspects simple to use, comprehensible station, clarity, ease of answering and interpretability. The questionnaires were in the days before discharge rates in patients after completion of the final report is handed over we got back. Specific questions were as follows: Specific questions were as follows:  Why did you choose your class for the purpose of rehabilitation management?  The x-th time away in our department?  The x-th time away in our department?  Satisfaction of operating personnel (doctor, nurse, physiotherapist, physiotherapists, bath with crew), and the efficacy of treatments?  If you have been lying in our hospital, the present management and the effectiveness of previous treatment how do you judge?  Does the department received information in a clear explanation of the objectives of the state of the disease, and the results of medical treatment studies?  Do you want the option to continue rehabilitation at our institution to continue?  Do you recommend to our class to others? y The answer is choice (yes-no) and 1 to 10 points for evaluation occurred. 2014/1 In addition, home of the patients in the pre-notification ticket is sought, up to one week prior to the recording by phone call. Perform some tests before the shooting – which are essential for successful rehabilitation g – outpatient needs to be, that the cost of the hospital's budget is not burdened (x-rays, lab etc.) too. It is important that the available recording time of the patients known drugs, the medication should be smooth. This requires a range of medicines to ensure the accuracy of the hospital logistic processes by which the suppliers created as a result. In the last two years has increased the territorial obligation, since 1st October 2010 from Kecskemét County Hospital to Kiskunfélegyháza 32 musculoskeletal rehabilitation beds have been moved, so increased the capacity of the musculoskeletal rehabilitation department. In the past 20 musculoskeletal beds capacity has increased up to 52 beds and the existing 20 bed cardiovascular unit that has 72 beds, which are the total number of beds in the rehabilitation department. With the number of patients around 1/3-dal increased compared to 2010. (FIGURE 1) However, neither the human nor the technical resources are not increased, thus significantly hard-working staff, which requires using the capacity of the human resources available. 731 934 925 706 1114 1118 0 200 400 600 800 1000 1200 2007 2008 2009 2010 2011 2012 patients/year Figure 1 Patient turnover 2007-2012 patient/year Source: Authors’ own data Figure 1 Patient turnover 2007-2012 patient/year Source: Authors’ own data This is the point where it is very thought-provoking insights to strategically what further action is possible for the department to meet growing patient care needs. Given that the infrastructure does not allow the inclusion of more beds, overcrowding would become inevitable, as is the case in many other domestic institutions is already visible. The increase in 50 2014/1 ISSN 2064-7964 Analecta staff numbers, mainly due to known medical deficiency, on the other hand the non- medical staff wage bill increase further undermine the position of the class. 6. SUMMARY Should be aware that the patient's recovery will not be completed for a successful surgery. The patient should be encouraged to her old life, back to work, which requires the help of experts organized within the framework of institutional and outpatient care. 10 years ago there was no rehabilitation unit in the hospital, initially tried to establish the conditions for a successful recovery, within their own wards. In May of 2005 started, addressed public investment realized, partial hospitalization reconstruction containing functional expansion as a result, in September 2006, was handed over to the thermal-recovery built rehabilitation unit, which includes the 40-bed department participated in the floors, the ground floor clinic rooms, the spa offers therapeutic and physiotherapy service rooms and the swimming pools, the basement of the horseradish rooms and mechanical rooms. 100 questionnaires were issued, of which 61 has been working on, which represented the class works. The respondents were slightly more than a quarter of men, the average age of nearly 66 years, which shows that more were treated in the older age group. Half were treated as described in our institution, which is not considered a bad number ratio. What is also good that, half of the patients directed to other institutes. As to how many times you have been lying in our department, the average was close to 3.5, the majority (34 %) 1-2 times taken advantage of our services. Maximum level of positive responses is to the operating personnel and the efficacy of treatments given. The efficacy of physiotherapy was held just under 90 % of. In connection with their condition has improved to 90 % of the response was yes . Over 90% of new management took effective as it was earlier. 100% of respondents would like to further continue their treatment at our institution and recommend to others. The investigation confirmed expectations, because their experience has shown that, over the years, the patients took a liking to the institution, regardless of age, gender, social status as to, satisfied with their services. The rate of re-admissions are high, many people are patiently waiting for the subscription period already in excess of half a year now. Currently, the waiting list of more than 800 patients have pre-registered, which is a very large number. The urgent recruitment needs, and suffered an acute event, other regional treatment sites insist to Kiskunfélegyháza hospital. 5. RESULTS AND DISCUSSION The 61 interviewed patients, 17 men, 44 women were, this ratio is broadly representative of the patient's circulation is observed, the composition of sex. The mean age of patients 65.7 years, the youngest 31, the oldest was 83 years old. Half of the patients were previously informed from the rehabilitation department (31 patients) or other institution offered or heard in the general population, or read about it in the media. 1/3- respondents (22 patients), other inpatient institute sent after acute events, as cardiac surgery, heart attack, hip nailing, prosthesis implantation them. Sample one-third of the number of cases examined accounts. (The questionnaires were processed with the help of Excel.) Therefore it may be legitimate for sub-acute rehabilitation name. 36 patients it was in order to institute rehabilitation to one care (hospital active clinic department, orthopedics, rheumatology, traumatology, surgery, cardiology, neurology) initiated the referral were a surprisingly large proportion of those (25 patients), whose family doctor suggested to the clinics in the rehabilitation care needs. A number of returning patients are usually among the beneficiaries, on average, three times so far lay there, the answers were between 1 and 10 times. Data presented in the previous section of the paper clearly show that, since the musculoskeletal rehabilitation participants exceeds 50% of cases each year, and cardiovascular patients in 30%. 51 2014/1 ISSN 2064-7964 Analecta Positive responses were received almost maximum satisfaction issues. The management staff, they were all satisfied, the potential 10-point average of 9.8 doctors and 9.5 nurses, physiotherapists 9.6, 9.3 physiotherapists, the spa personnel received 9.3 points. The effectiveness of the treatments were mostly pleased with the 9.4 gym, physiotherapy 8.9, the spa treatment was 9.3 points. During the rehabilitation period, 88% (54 patients) felt that they had improved to about 12% (7 patients) left unchanged. To the question that you have received this information due to illness, the state of the tests and their results, the medical treatment, all patients answered yes. Almost all respondents happy to return to and would recommend it to other patients Kiskunfélegyháza rehabilitation clinic department. 6. SUMMARY Rehabilitation hospital care is most effective in the analysis – which can be measured by looking at well-defined – that the average number of nursing for more than 18 days, the bed occupancy of 100% and a mortality rate of 2%. Errors, that rehabilitation is very little analysis on the domestic situation, which could increase the efficiency of the operation. 52 2014/1 ISSN 2064-7964 ISSN 2064-7964 Analecta REFERENCES A. S. Leon et al. (2005): Cardiac Rehabilitation and Secondary Prevention of Coronary Heart Disease An American Heart Association Scientific Statement From the Council on Clinical Cardiology (Subcommittee on Exercise, Cardiac Rehabilitation, and Prevention) and the Council on Nutrition, Physical Activity, and Metabolism (Subcommittee on Physical Activity), in Collaboration With the American Association of Cardiovascular and Pulmonary Rehabilitation. Circulation. 2005; 111: 369-376. y E.G. (2012) European Guidelines on cardiovascular disease prevention in clinical practice: The Fifth Joint Task Force of the European Society of Cardiology and Other Societies on Cardiovascular Disease Prevention in Clinical Practice (constituted by representatives of nine societies and by invited experts). Developed with the special contribution of the European Association for Cardiovascular Prevention & Rehabilitation . Eur Heart J, 2012 33: 1635-1701 M. F. Piepoli et al. (2010): Secondary prevention through cardiac rehabilitation: from knowledge to implementation. A position paper from the Cardiac Rehabilitation Section of the European. Association of Cardiovascular Prevention and Rehabilitation. E. J. Cardiovasc. Prev. and Rehab. S. Gerold et al. (2007): The International Classification of Functioning, Disability and Health: A unifying model for the conceptual description of the rehabilitation strategy. J. Rehab. Med. 2007, 39: 279. Szegedi Z. - Prezenszki J. (2010): Logisztika-Menedzsment (Logistics-Management). Kossuth. Budapest Szabó Gy., Renner A. (2004): A rehabilitáció gyakorlata (Rehabilitation is Practice). Medicina. Budapest 53
https://openalex.org/W2982631755	https://readersinsight.net/jmi/article/download/967/790	English	null	Employability and Competence Skills of Graduating Students in a Private Higher Educational Institution in Taguig City, Metro Manila, Philippines	Journal of management info	2,019	cc-by	6,987	City, Metro Manila. The study aims to answer the following research questions: City, Metro Manila. The study aims to answer the following research questions: 1. What is the profile of the graduating students in terms of selected socio-demographic characteristics? Entering the labor market is one of the major life-defining events in a graduate’s professional life after graduation. However, there still exists a gap between the level of employability skills of graduates and their competency skills acquired during their stay in college and often, work requirements do not meet with these skills (Ranasinghe, 1992; Lindsay, 2002). 2. What are the work-related values which are considered important in gaining employment? 3. What are the levels of importance and competence of employability skills of graduating students in terms of problem-solving and analytic, learning, personal organization and time management, creativity, innovation, and change, and personal strengths? 3. What are the levels of importance and competence of employability skills of graduating students in terms of problem-solving and analytic, learning, personal organization and time management, creativity, innovation, and change, and personal strengths? Higher educational institutions should make sure that they know the status of their graduates. Tracer studies are in fact encouraged to determine the employment status of the graduates. One way of measuring the effectiveness of the service delivery of universities and colleges is by looking at the employability of their graduates (Orence & Laguador, 2013; Loquias, 2015). Keywords: Competency Skills, Employability, College Students, Philippines Keywords: Competency Skills, Employability, College Students, Philippines Keywords: Competency Skills, Employability, College Students, Philippines © Readers Insight Publication Copyright © 2019 Authors. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. 13 Abstract Policy makers in the Philippines are focusing on the gaps between the knowledge and skills provided by higher educational institutions and the job opportunities available in the labor market. Graduates are often faced with challenges in looking for jobs right after graduation, thus, schools play a crucial role in ensuring that students are equipped with knowledge and competency skills. q pp g p y This study examines the level of employability and competence skills of the 309 graduating students taking up different degree programs of The Fisher Valley College located in Taguig City, Metro Manila. About 17.8% of the respondents are currently working and 21.7% wanted to work in the service sector. More than half of the total respondents wanted to work in the government. Most of the respondents will look for a job immediately after graduation. Communication skills are the most important of skills in order for them to get hired in their first job applications. The most important factor identified when selecting an organization or company for work is when it provides ample opportunities for them for career advancement. Among work- related values, respondents considered love for God and perseverance and hard work as the most important. p Five components of employability skills were identified: (1) problem-solving and analytic; (2) learning; (3) personal organization and time management; (4) creativity, innovation, and change; and (5) personal strengths. Overall, students assessed that these skills are very important in employment. However, they reported lower competency levels on these skills particularly on solving problems, identifying essential components of the problem, sorting out the relevant data to solve the problem, and contributing to group problem solving. p g It is very important for all sectors, i.e., schools, government, and industry, to prepare the students acquire the necessary skills and provide them with opportunities as they enter the labor market. Employability and Competence Skills of Graduating Students in a Private Higher Educational Institution in Taguig City, Metro Manila, Philippines Marlon B. Raquel1, Henry B. Raquel2, Melven B. Raquel3, Beryl T. Carloto4 1,2,3,4College of Business Administration and Accountancy, The Fisher Valley College, Taguig City, Philippines Corresponding author: raquel.marlon@yahoo.com ARTICLE INFORMATION ARTICLE INFORMATION Journal of Management Info 6(3); 13-18 Journal of Management Info 6(3); 13-18 Journal of Management Info (JMI) ISSN:2313-3376 www.readersinsight.net/jmi LITERATURE REVIEW g Policy makers in the Philippines are focusing on the gaps between the knowledge and skills provided by higher educational institutions and the job opportunities available in the labor market. Graduates are often faced with challenges in looking for jobs right after graduation, thus, schools play a crucial role in ensuring that students are equipped with knowledge and competency skills. Employability is notoriously complex to define, but for the purposes of this review we have looked at research that uses the term ‘employability’ and which intersects with one of the key concepts which are contained with the Higher Education Academy’s (HEA’s) Framework for embedding employability. The framework was used both to provide keywords for the literature review, and to help in structuring our report. Higher education providers are under considerable pressure from policymakers, students and employers to ensure that graduates emerge from higher education ready for the Thus, this study examines the levels of employability and competence skills of graduating students in The Fisher Valley College, a private higher educational institution located in Taguig 13 Journal of Management Info 6(3); 13-18 RESULTS AND FINDINGS Table 1. Profile of the Respondents Variables Frequency Percentage Age Groups 18-19 20-21 22 and above 114 114 81 36.9 36.9 26.2 Sex Male Female 162 147 52.4 47.6 Civil Status Single Married 292 17 94.5 5.5 Religion Roman Catholic Non-Roman Catholic 259 59 80.9 19.1 College CBAA CCS COE COHM 109 75 42 83 35.3 24.3 13.6 26.9 Monthly Family Income P10,000 and below P10,001-P20,000 P20,001-P30,000 P30,001-P40,000 P40,001-P50,000 P50,001-P60,000 P60,001-P70,000 Over P70,000 141 53 63 4 26 8 10 4 45.6 17.2 20.4 1.3 8.4 2.6 3.2 1.3 Employment Status Currently Working Not Working 54 249 19.4 80.6 Total 309 100.0 RESULTS AND FINDINGS labour market. The imminent implementation of the Teaching Excellence Framework (TEF) looks set to increase this pressure (HEA, 2016). labour market. The imminent implementation of the Teaching Excellence Framework (TEF) looks set to increase this pressure (HEA, 2016). labour market. The imminent implementation of the Teaching Excellence Framework (TEF) looks set to increase this pressure (HEA, 2016). Frequency Percentage It is also possible to argue that a focus on employability is a moral duty for higher education providers. LITERATURE REVIEW Students invest their time and money in accessing higher education with the expectation that it will offer them access to greater life chances than they would have obtained if they had not attended HE. For the most part, this expectation is fulfilled with graduates generally doing better in the labor market than non-graduates. However, it is not simply holding a degree that results in these outcomes (although there is an important signaling effect), rather it is critical that higher education develops students in ways that support them to be successful in the future. Academic interest in employability appears to be growing. Academics are wrestling with the nature of employability, its political implications, the ways in which it is delivered and the relative efficacy of each of them. Academics approach these questions from a range of political, theoretical, methodological and professional perspectives. Research Design and Sample Size This study uses a descriptive, quantitative research design. It is primarily descriptive because it describes the characteristics of the respondents. A total of 320 graduating students for second semester, academic year 2018-2019 were given survey questionnaires but only 309 questionnaires were completely answered. Eleven questionnaires were considered spoiled due to many items left unanswered, thus, the response rate is 95.24%. There were 109 graduating students from the College of Business Administration and Accountancy who are taking up Bachelor of Science in Accountancy and Bachelor of Science in Business Administration major in Management. Eighty three (83) graduating students are from the College of Hospitality Management taking up Bachelor of Science in Hotel and Restaurant Management. From the College of Computer Studies, 75 graduating students participated in the study who are taking up Bachelor of Science in Information Technology and Bachelor of Science in Computer Science. There were 42 respondents who are taking up Bachelor of Elementary Education and Bachelor of Secondary Education under the College of Education. Table 1 shows the frequency and percentage distributions of respondents by selected socio-demographic characteristics. Results show that 18 to 21 years old respondents comprise 73.8% of the total respondents while 26.2% of the students are 22 years old and above. There are more males than females (52.4% vs. 47.6%). An overwhelming majority of the respondents are single (94.5%) while 5.5% of the respondents are married. In this survey, “single” refers to those who have never married and “married” denotes those who had a partner (including common-law marriage). Of the 309 respondents, 250 (80.9%) students are Roman Catholics while only 59 (19.1%) are in other religions. Copyright © 2019 Authors. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. 14 Instrumentation and Data Analysis Survey questionnaire is the main instrument used in this study. Questionnaire items included in the instrument were selected socio- demographic characteristics of the respondents, employment-related data, competencies acquired at school, work-related values, and levels of importance and competence of employability skills of the respondents. One hundred nine (109) students are from the College of Business Administration and Accountancy, 83 students are from the College of Hospitality Management, 75 students are taking up programs under the College of Computer Studies while the remaining 42 students are taking up programs under the College of Education. There are 141 respondents or 45.6% of the total number of respondents who have a monthly family income of P10,000.00 or less. About 17.2% of the respondents have monthly income between P10,001 to P20,000.00 while 20.4% has income between P20,001 to P30,000. The remaining percentage has income above P30,000.00. As to employment status, 249 respondents or 80.6% are not working while 17.8% of the total respondents are currently working. Questions are designed on a 5-point Likert scale rating anchored on ‘1’ with range of means 1.00-1.79 for not important/not competent, ‘2’ with range of means 1.80-2.59 for less important/less competent, ‘3’ with range of means 2.60-3.39 for moderately important/moderately competent, ‘4’ with means 3.40-4.19 for important/competent, and ‘5’ with means 4.20-5.00 for very important/very competent. Frequencies, percentage distributions, weighted means, verbal interpretations, and ranks were employed to analyze the results of the study. Table 2. What occupation category do you want to work after graduation? Table 2. What occupation category do you want to work after graduation? th C ti C Att ib ti Li hi h it t i t d di t ib ti Table 2. What occupation category do you want to work after graduation? What occupation category do you want to work after graduation? Freq uency Percent age Services Technical Operations Business Process Outsourcing 67 30 6 13 21.7 9.7 1.9 4.2 14 Journal of Management Info 6(3); 13-18 Banking and Finance Human Resources Academe/Teaching Telecommunications Trade/Industry Others 43 35 51 2 5 57 13.9 11.3 16.5 0.6 1.6 18.4 Total 309 100.0 Do not have any idea yet Total Respondents were als shown in Table 5. Aroun decided to look for a job students will continue wo (11) t d t t t d th t Respondents were also asked about their plans after graduation as shown in Table 5. Instrumentation and Data Analysis Around 249 or 80.6% of the respondents were decided to look for a job immediately while 5.2% of the graduating students will continue working in their present employment. Eleven (11) students stated that they will migrate or go abroad while 10 students will pursue further studies, i., will be taking master’s degrees. Seven (7) students will join family businesses and five (5) students will start their own businesses. There are 10 students who do not have any idea yet what to do while one student stated that he will not do anything but rest. Table 2 indicates the respondents’ choice of occupation category after graduation. The service sector is the top choice of the respondents (21.7%), followed by academe or teaching (16.5%), banking and finance (13.9%), and human resources (11.3%). Others wanted to work in the technical, business process outsourcing, operations, and trade or industry. The remaining 18.4% of the respondents choose to work in other fields. The result confirms that majority of respondents were willing to look for a job immediately or after graduation in order to utilize their level of education. Both male and female respondents regardless of course programs convinced that college education is an important investment that helps people build their skills and prepare for high- skilled jobs. Historically, those who have made this investment have received a substantial economic benefit that lasts over their lifetime (Abel et al., 2014). From the table above, many students wanted to work in occupations which are outlined within the business administration and accountancy programs. This can be attributed to the fact that there is a high demand in the local and global workplace and diverse business environments that produces intangible goods, more precisely services for business graduates. Countries with economies focused on the service sector are considered more advanced than other markets. This perception is paralleled by Uhlig, et al (2017) when they affirmed that employees with business administration and accountancy degrees have higher earnings, job security and job satisfaction. However, there have been reports of newly minted college graduates who are unsuccessful at finding jobs suited to their level of education and those who do find work are often confined to low-wage positions (Vedder et al., 2013). Instrumentation and Data Analysis Moreover, some factors may hinder in finding a job is due to a mismatch between the aspirations of graduates and employment opportunities available and the influence of career choice decisions for an individual’s career (Nyamwange, 2016). Table 3. What type of company/organization do you want to work? What type of company/organization Table 3. What type of company/organization do you want to work? What type of company/organization do you want to work? Frequency Percentage Private Government 124 185 40.1 59.9 Total 309 100.0 Table 6. How much is your expected gross monthly salary as new graduate? Table 6. How much is your expected gross monthly salary as new graduate? graduate? What is your plan after graduation? Frequency Percentage P10,000 and below P10,001-P12,000 P12,001-P15,000 P15,001-P17,000 P17,001-P20,000 P20,001-P22,000 P22,000-P25,000 P25,001-P27,000 Over P27,000 249 16 5 7 10 11 1 10 80.6 5.2 1.6 2.3 3.2 3.6 0.3 3.2 Total 309 100.0 The person’s response to employability concern is determined by the type of organization he is engaged in, hence, variable was examined by the researchers and data relating to the type of organization is depicted in Table 3. Surprisingly, majority of the respondents (59.9%) wanted to work in the government sector as contrast to 40.1% who wanted to get an employment in private companies. Many graduating students are enthusiastic to take their career paths in the government due to the benefits and job security. These factors make government positions beneficial. Table 4. Is there any member of your family who attended in this school? Table 4. Is there any member of your family who attended in this school? Is there any member of your family who attended in this school? Freque ncy Percent age Yes No 139 170 45.0 55.0 Total 309 100.0 Table 6 shows the expected gross monthly salary of respondents as new graduates. it is clear that the most number of respondents belongs to the salary range of 10,000 and below that accounts to 80.6%, 16 or 5.2% from 10,001-12,000, 11 or 3.6% from 20,001- 22,000, 10 or 3.2% from 17,001-20,000 and 25,001-27,000, respectively. Seven respondents or 2.3% from 15,001-17,000 salary range, 5 or 1.6% from 12,001-15,000, 1 or 0.3% from 22,000-25,000 and the monthly salary of over 27,000 is the lowest point or insignificant. Students were asked whether or not there are family members who attended in the same institution. Instrumentation and Data Analysis Solving problems 4.372 .923 V I 1 4.220 .832 V C 5 2. Identifying problems 4.311 .901 V I 3. 5 4.356 .881 V C 2 3. Prioritizing problems 4.307 .915 V I 4 4.367 .902 V C 1 4. Identifying essential components of the problem 4.311 .943 V I 3. 5 4.223 .999 V C 4. 5 5. Sorting out the relevant data to solve the problem 4.288 .821 V I 5 4.223 .949 V C 4. 5 6. Contributing to group problem solving 4.366 .879 V I 2 4.288 .921 V C 3 Composite Mean Score 4.326 VI 4.280 V C Learning Level of Importance Level of Competence SD V I R SD V I R 1. Keeping up-to-date on developments in the field 4.502 .771 V I 3 4.372 .879 V C 3 2. Gaining new 4.557 .744 V 1 4.482 .820 V 2 Table 10. Levels of Importance and Competence of Employability Skills of the Graduating Students Table 10. Levels of Importance and Competence of Employability Skills of the Graduating Students Problem-solving and Analytic Level of Importance Level of Competence SD V I R SD V I R 1. Solving problems 4.372 .923 V I 1 4.220 .832 V C 5 2. Identifying problems 4.311 .901 V I 3. 5 4.356 .881 V C 2 3. Prioritizing problems 4.307 .915 V I 4 4.367 .902 V C 1 4. Identifying essential components of the problem 4.311 .943 V I 3. 5 4.223 .999 V C 4. 5 5. Sorting out the relevant data to solve the problem 4.288 .821 V I 5 4.223 .949 V C 4. 5 6. Contributing to group problem solving 4.366 .879 V I 2 4.288 .921 V C 3 Composite Mean Score 4.326 VI 4.280 V C In recent decades, career development has seen a shift in the way it is approached. Traditionally, it was up to an organization to ensure that its employees had the skills to meet the company’s long-term goals. This shift has changed the way that organizations handle career development. Employer invests in their employees via career development in order to improve organizations’ competitive ultimately. They undertook that career development should be nuanced and tailored to individual needs and stages of career. Instrumentation and Data Analysis Which of the following factors is the most important to you when selecting an organization or company for work? Which of the following factors is the most important to you when selecting an organization or company for work? Freq uency Perce ntage Provides ample opportunities for career advancement Has opening for my preferred field of interest Presence of friends or relatives in the company Near my residence Others 189 48 8 59 5 61.2 15.5 2.6 19.1 1.6 Total 309 100.0 work Professional Integrity 4.495 .907 MI Punctuality 4.437 .908 MI Supportiveness 4.359 .949 MI Tolerance 4.346 .977 MI Unity 4.453 1.058 MI Composite Mean Score 4.425 MI Table 9 exhibits work related values which are very m considered in finding employment among graduating students. L for God and perseverance and hard work are very important weighted means of 4.563 and 4.528, respectively. The rest of work-related values are considered moderately impor Interestingly, obedience to superior and efficiency got the lo mean scores of 4.311 and 4.246, respectively. Refozar, Vasquez and Luistro (2017) found that hard work ra second among work related values which contribute to meeting Human Relations Skills Information Technology Skills Problem-solving Skills 17 16 18 5.5 5.2 5.8 6 5 4 Total 309 100.0 Table 8. Which of the following factors is the most important to you when selecting an organization or company for work? Which of the following factors is the most important to you when selecting an organization or company for work? Freq uency Perce ntage Provides ample opportunities for career advancement Has opening for my preferred field of interest Presence of friends or relatives in the company Near my residence Others 189 48 8 59 5 61.2 15.5 2.6 19.1 1.6 Total 309 100.0 Table 9 exhibits work related values which are very much considered in finding employment among graduating students. Love for God and perseverance and hard work are very important with weighted means of 4.563 and 4.528, respectively. The rest of the work-related values are considered moderately important. Interestingly, obedience to superior and efficiency got the lowest mean scores of 4.311 and 4.246, respectively. Refozar, Vasquez and Luistro (2017) found that hard work ranked second among work-related values which contribute to meeting the demands of employment among business and computer management graduates in a Philippine university. The consistency of the results is not surprising given the fact that Filipinos are hardworking individuals. Instrumentation and Data Analysis They wanted to help their families who have supported them in their studies. Table 8 shows the list of factors which are considered the most important for the respondents when selecting an organization or company for work. Out of the 309 respondents, 189 (61.2%) indicated that they will choose a company that provides ample opportunities for career advancement while about 19.1% will consider a company if it is near their places of residence. About 15.5% of the respondents will choose an organization that provides job opening for their preferred field of interest while 2.6% or eight respondents indicated that the presence of friends or relatives in the company is the most important factor. Meanwhile, five (5) respondents mentioned other reasons. Based on the present study, love for God continues as the dominant work-related values or very important to the job placement of the graduates. In this today’s complex, fast-faced, demanding and volatile business environment this is most appealing. Reave (2005) detailed that spiritual values (including Love for God) animates transcendent quality of work. When individuals possess such values are more likely to increased productivity, lowered rates of turnover, greater sustainability, and able to promote positive work relationships. It is interesting to note that an individual regardless of caste or occupation embraces their sense of spirituality able to handle the conflicts, paradoxes and ambiguities (Silver, 2013). It is interesting to note that when growth opportunities are abundant and environment is conducive, individuals work with considerably more vigor. This is possible to have an influence on the responses of the respondents. Research has shown unequivocally that greater career development leads to increased productivity and commitment. Li et al. (2016) examined the impact of career development on employee commitment among part-time faculty in Hong Kong University. They found out that career development have a significantly positive influence on employee commitment. Kraimer et al. (2011) asserted that perceived growth opportunities offered have been linked to higher job performance and lower turnover rate. Therefore, aligning an employee’s values specifically love of God coupled with graduate’s perseverance and hard work will surely foster loyalty and a bond between employer and employee (Buenviaje, et al., 2015). Table 10. Levels of Importance and Competence of Employability Skills of the Graduating Students Problem-solving and Analytic Level of Importance Level of Competence SD V I R SD V I R 1. Instrumentation and Data Analysis Lee and Bruvold (2003) claimed that equipping employees with competencies, new knowledge and skills means investment in career development that can enable them to accept new challenges outside or within the organization. Copyright © 2019 Authors. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. 16 Table 9. Means, Standard Deviations, Verbal Interpretations, and Ranks of Work-related Values Work-related Values Means Std. Dev. Verbal Interpretations Ran ks Courage 4.485 .796 MI 5 Creativity and innovativeness 4.418 .816 MI 9 Efficiency 4.246 .956 MI 15 Fairness and Justice 4.401 .894 MI 10 Honesty and love for truth 4.492 .763 MI 4.5 Leadership 4.492 .917 MI 4.5 Love for Co-workers and others 4.440 .777 MI 7 Love for God 4.563 .926 VI 1 Nationalism 4.340 .969 MI 13 Obedience to superior 4.311 1.054 MI 14 Perseverance and hard 4.528 .877 VI 2 y g essential components of the problem I 5 C 5 5. Sorting out the relevant data to solve the problem 4.288 .821 V I 5 4.223 .949 V C 4. 5 6. Contributing to group problem solving 4.366 .879 V I 2 4.288 .921 V C 3 Composite Mean Score 4.326 VI 4.280 V C Learning Level of Importance Level of Competence SD V I R SD V I R 1. Keeping up-to-date on developments in the field 4.502 .771 V I 3 4.372 .879 V C 3 2. Gaining new 4.557 .744 V 1 4.482 .820 V 2 Table 9. Means, Standard Deviations, Verbal Interpretations, and Ranks of Work-related Values Table 9. Means, Standard Deviations, Verbal Interpretations, and Ranks of Work-related Values Work-related Values Means Std. Dev. Verbal Interpretations Ran ks Courage 4.485 .796 MI 5 Creativity and innovativeness 4.418 .816 MI 9 Efficiency 4.246 .956 MI 15 Fairness and Justice 4.401 .894 MI 10 Honesty and love for truth 4.492 .763 MI 4.5 Leadership 4.492 .917 MI 4.5 Love for Co-workers and others 4.440 .777 MI 7 Love for God 4.563 .926 VI 1 Nationalism 4.340 .969 MI 13 Obedience to superior 4.311 1.054 MI 14 Perseverance and hard 4.528 .877 VI 2 Copyright © 2019 Authors. Copyright © 2019 Authors. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. Instrumentation and Data Analysis While more than half (55.0%) of the respondents stated in the negative, the remaining 45% who said that there are family members who also attended in the same institution is considered a high percentage. The analysis validates findings in previous research that newly graduates would accept jobs with low salaries since they do not have professional experiences yet and they can change jobs within a year (Barea and Vasudeva, 2015). Salary is not a constraint for the right people. Recent graduates expressed that they would desire to work at a firm with a satisfied social atmosphere and accept a lower salary, rather than receive high salaries with less entertainment. Table 5. What is your plan after graduation? What is your plan after graduation? Frequen cy Percenta ge Look for a job immediately Currently working, continue with my present employment Start my own business Join family business Pursue further studies Migrate Rest 249 16 5 7 10 11 1 80.6 5.2 1.6 2.3 3.2 3.6 0.3 Table 5. What is your plan after graduation? Table 7. Which of the following competencies is the most important to you in order to be hired in a company? Which of the following competencies is the most important to you in order to be hired in a company? Freq uency Perce ntage Rank Communication Skills Critical Thinking Skills Entrepreneurial Skills 171 27 60 55.3 8.7 19.4 1 3 2 Table 7. Which of the following competencies is the most important to you in order to be hired in a company? Copyright © 2019 Authors. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. Copyright © 2019 Authors. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. 15 Journal of Management Info 6(3); 13-18 work Professional Integrity 4.495 .907 MI 3 Punctuality 4.437 .908 MI 8 Supportiveness 4.359 .949 MI 11 Tolerance 4.346 .977 MI 12 Unity 4.453 1.058 MI 6 Composite Mean Score 4.425 MI Human Relations Skills Information Technology Skills Problem-solving Skills 17 16 18 5.5 5.2 5.8 6 5 4 Total 309 100.0 Table 8. Instrumentation and Data Analysis This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. 16 Journal of Management Info 6(3); 13-18 knowledge in areas outside the immediate job I C 3. Gaining new knowledge from everyday experiences 4.521 .816 V I 2 4.521 .828 V C 1 Composite Mean Score 4.526 VI 4.459 V C Personal Organization and Time Management Level of Importance Level of Competence SD V I R SD V I R 1. Allocating time efficiently 4.447 .883 V I 2 4.372 .901 V C 2 2. Setting priorities 4.499 .812 V I 1 4.346 .922 V C 3 3. Managing / overseeing several tasks at once 4.275 .956 V I 4 4.220 .992 V C 4 4. Meeting deadlines 4.324 .970 V I 3 4.395 .893 V C 1 Composite Mean Score 4.386 VI 4.333 V C Creativity, Innovation and Change Level of Importance Level of Competence SD V I R SD VI R 1. Adapting to situations of change 4.366 .950 V I 2 4.346 .901 V C 4 2. Keeping up-to-date with external realities related to your company’s success 4.375 .909 V I 1 4.401 .834 V C 2 3. Re-conceptualizing your role in response to changing corporate realities 4.324 .900 V I 4 4.392 .844 V C 3 4. Initiating change to enhance productivity 4.363 .911 V I 3 4.414 .885 V C 1 5. Providing novel solutions to problems 4.262 .999 V I 5 4.249 1.006 V C 5 Composite Mean Score 4.338 V I 4.361 V C Personal Strengths Level of Importance Level of Competence SD V I R SD VI R 1. Functioning well in stressful situations 4.453 .898 V I 2 4.314 .988 V C 5 2. Functioning at an optimal level of performance 4.353 1.007 V I 6 4.275 .993 V C 6 3. Responding positively to constructive criticism 4.379 .982 V I 5 4.356 .910 V C 4 4. Maintaining a positive attitude 4.469 .913 V I 1 4.385 .962 V C 2 5. Maintaining a high energy level 4.440 .943 V I 3 4.369 1.054 V C 3 6. Ability to work independently 4.427 .939 V I 4 4.398 .919 V C 1 Table 10 shows the levels of importance and competence of employability skills of graduating students. Instrumentation and Data Analysis Re-conceptualizing your role in response to changing corporate realities 4.324 .900 V I 4 4.392 .844 V C 3 4. Initiating change to enhance productivity 4.363 .911 V I 3 4.414 .885 V C 1 5. Providing novel solutions to problems 4.262 .999 V I 5 4.249 1.006 V C 5 Composite Mean Score 4.338 V I 4.361 V C Personal Strengths Level of Importance Level of Competence SD V I R SD VI R 1. Functioning well in stressful situations 4.453 .898 V I 2 4.314 .988 V C 5 2. Functioning at an optimal level of performance 4.353 1.007 V I 6 4.275 .993 V C 6 3. Responding positively to constructive criticism 4.379 .982 V I 5 4.356 .910 V C 4 4. Maintaining a positive attitude 4.469 .913 V I 1 4.385 .962 V C 2 5. Maintaining a high energy level 4.440 .943 V I 3 4.369 1.054 V C 3 6. Ability to work independently 4.427 .939 V I 4 4.398 .919 V C 1 Composite Mean Score 4.425 VI 4.350 V C Overall, graduating students consider all employability skills as very important with a composite mean score of 4.40. They also believe that they are very competent in these areas with a composite mean score of 4.356. CONCLUSION Maintaining a positive attitude 4.469 .913 V I 1 4.385 .962 V C 2 5. Maintaining a high energy level 4.440 .943 V I 3 4.369 1.054 V C 3 6. Ability to work independently 4.427 .939 V I 4 4.398 .919 V C 1 Composite Mean Score 4.425 VI 4.350 V C While students reported a relatively high importance and competence in all employability skills, it is noteworthy that they reported lowest competency levels in terms of problem-solving and analytical skills. These sets of skills are important in examining situations to identify the strengths and weaknesses so that they would be able to make recommendations how to solve problems in their respective companies where they will be employed. CONCLUSION Creativity, Innovation and Change Level of Importance Level of Competence SD V I R SD VI R 1. Adapting to situations of change 4.366 .950 V I 2 4.346 .901 V C 4 2. Keeping up-to-date with external realities related to your company’s success 4.375 .909 V I 1 4.401 .834 V C 2 3. Re-conceptualizing your role in response to changing corporate realities 4.324 .900 V I 4 4.392 .844 V C 3 4. Initiating change to enhance productivity 4.363 .911 V I 3 4.414 .885 V C 1 5. Providing novel solutions to problems 4.262 .999 V I 5 4.249 1.006 V C 5 Composite Mean Score 4.338 V I 4.361 V C Level of Importance Level of Competence One in every five students is currently working while studying. The concept of work-study balance has been an ongoing struggles for many young Filipinos who do not want to stop their college studies. Universities, private companies, and government agencies, therefore, have to implement programs that would help these working students cope up this daunting task before them. Most of the graduating students wanted to work in the government. This is quite interesting because in the past decades, Filipinos did not want to join government because of low salaries. The trend was reversed in the past few years due to increases of salaries among government employees which encourages many young professionals to join the government. These future professionals also want to engage in the service sector. Services sector posted the highest share of GDP in the fourth quarter of 2018 which is 56.2% (DTI, 2018). Communication skills are considered the most important of all skills, thus, schools should strengthened the English curriculum, particularly oral and writing skills, so that students would be ready to face job interviews. In addition, since Filipinos are considered religious, schools have to reinforce teaching religious and moral values which are important in the workplaces. Work values should be integrated in all aspects of classroom teaching and trainings and seminars. Personal Strengths Level of Importance Level of Competence SD V I R SD VI R 1. Functioning well in stressful situations 4.453 .898 V I 2 4.314 .988 V C 5 2. Functioning at an optimal level of performance 4.353 1.007 V I 6 4.275 .993 V C 6 3. Responding positively to constructive criticism 4.379 .982 V I 5 4.356 .910 V C 4 4. Copyright © 2019 Authors. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. Instrumentation and Data Analysis Learning skills are considered very important and at the same time, respondents reported very competent in these areas. It is very important for them to gain new knowledge in areas outside of their immediate job and from everyday experiences and keeping up-to-date on the developments in the field. considered very important and at the same time, respondents reported very competent in these areas. It is very important for them to gain new knowledge in areas outside of their immediate job and from everyday experiences and keeping up-to-date on the developments in the field. Among the five skills, problem-solving and analytical skills ranked fifth in terms of importance and competence as perceived by the respondents. While verbal interpretations suggest that these are very important and respondents claimed that they are very competent in identifying and solving problems, this is worthy of taking into scrutiny. Level of Importance Level of Competence Personal Organization and Time Management Level of Importance Level of Competence SD V I R SD V I R 1. Allocating time efficiently 4.447 .883 V I 2 4.372 .901 V C 2 2. Setting priorities 4.499 .812 V I 1 4.346 .922 V C 3 3. Managing / overseeing several tasks at once 4.275 .956 V I 4 4.220 .992 V C 4 4. Meeting deadlines 4.324 .970 V I 3 4.395 .893 V C 1 Composite Mean Score 4.386 VI 4.333 V C One striking result in this study is that the levels of competence of all employability skills are lower than the levels of importance except for creativity, innovation, and change. This means that while students believe that these skills are very important, they reported lower levels of competence. This can be attributed to the fact that they still lack the necessary experiences which are very important in getting employed. 3. Managing / overseeing several tasks at once 4.275 .956 V I 4 4.220 .992 V C 4 4. Meeting deadlines 4.324 .970 V I 3 4.395 .893 V C 1 Composite Mean Score 4.386 VI 4.333 V C Creativity, Innovation and Change Level of Importance Level of Competence SD V I R SD VI R 1. Adapting to situations of change 4.366 .950 V I 2 4.346 .901 V C 4 2. Keeping up-to-date with external realities related to your company’s success 4.375 .909 V I 1 4.401 .834 V C 2 3. REFERENCES Bagaoisan, A. J. (2015). Why Filipinos are among world’s most religious. In Global Attitudes Survey Pew Research Center. Retrieved at https://news.abs-cbn.com/focus/01/07/16/why-filipinos-are-among- worlds-most-religious Barea, D. and Vasudeva, P. (2015). Will Today’s Graduates Want to Work For You? London: Accenture. Baring, R. & Cacho, R. (2015). Contemporary engagements and challenges for Catholic religious education in Southeast Asia. In Global Perspectives on Catholic Religious Education in Schools. Edited by Michael Buchanan and Adrian Gellel. Basel: Springer Publishing, pp. 143–53. p g g, pp Buenviaje, M.G., Galicano, V.M., Añonuevo, F. & Martinez, M. (2015). Employability of Business and Computer Management Graduates of Buenviaje, M.G., Galicano, V.M., Añonuevo, F. & Martinez, M. (2015). Employability of Business and Computer Management Graduates of 17 Journal of Management Info 6(3); 13-18 one Higher Education Institution in the Philippines, Asia Pacific Journal of Academic Research in Business Administration, 3(5): 63-71 one Higher Education Institution in the Philippines, Asia Pacific Journal of Academic Research in Business Administration, 3(5): 63-71 Higher Education Academy (2016). Framework for Embedding employability in higher education. Retrieved from https://www.heacademy.ac.uk/system/files/downloads/embedding- employability-in-he.pdf Jones, C. G. (2011). Written and Computer-Mediated Accounting Communication Skills An Employer Perspective. Business Communication Quarterly, 74(3), 247-271. y Kraimer, M. L., Seibert, S. E., Wayne, S. J., & Liden, R. C. (2011). Antecedents and outcomes of organizational support for development: The critical role of career opportunities. Journal of Applied Psychology, 96, Kraimer, M. L., Seibert, S. E., Wayne, S. J., & Liden, R. C. (2011). Antecedents and outcomes of organizational support for development: The critical role of career opportunities. Journal of Applied Psychology, 96, 485–500 Kraimer, M. L., Seibert, S. E., Wayne, S. J., & Liden, R. C. (2011). Antecedents and outcomes of organizational support for development: The critical role of career opportunities. Journal of Applied Psychology, 96, 485–500. y gy Lee, C. H., & Bruvold, N. T. (2003). Creating value for employees: investment in employee development. The International Journal of Human Resource Management, 14(6), 981-1000. Lindsay, C. (2002). Long-term unemployment and the “employability gap”: Priorities for renewing Britain’s New Deal. Journal of European Industrial Training, 26(9), 411-419. g Li, I. W., Mahuteau, S., Dockery, A., M., Junankar, P. N. & Mavromaras, K. (2016). Labour market outcomes of Australian university graduates from equity groups. Perth, WA: National Centre for Student Equity in Higher Education (NCSEHE) at Curtin University. https://www.ncsehe.edu.au/wp-content/uploads/2016/03/Labour- Market-Outcomes-ofAustralian-University-Graduates-from-Equity- Groups.pdf Loquias, R. (2015). Copyright © 2019 Authors. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. 18 Copyright © 2019 Authors. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. Copyright © 2019 Authors. This is an open access article distributed under th and reproduction in any medium, provided the original work is properly cited. REFERENCES Employability of the Bachelor of Science in Electronics Engineering Graduates of Camarines Sur Polytechnic Colleges. Asia Pacific Journal of Multidisciplinary Research, 3(4), 33–40. Nyamwange, J. (2016). Influence of Student’s Interest on Career Choice among First Year University Students in Public and Private Universities in Kisii County, Kenya. Journal of Education and Practice, 7(4), 96-102. ( ) Orence, A. & Laguador, J.M. (2013). Employability of Maritime Graduates of Lyceum of the Philippines University from 2007-2011, International Journal of Research in Social Science, 3(3): 142-157. Philippine National Accounts, 2015-2018 (Q4). Retrieved from https://www.dti.gov.ph/resources/statistics/gross-domestic-product-gdp Ranasinghe, S.W. (1992). Human resource development in Sri Lanka: Present trends and future Perspectives. In Salleh, S. and Gurung, S.B. (Eds.), Human Resource Development in South Asia. Asian and Pacific Development Centre, Malaysia. Refozar, R.F., Velasquez, J.E. & Luistro, E.J. (2017). Employability of BS Business and Computer Management Graduates from 2013 to 2015 in one Academic Institution in the Philippines, Asia Pacific Journal of Academic Research in Business Administration, 3(3): 1-9 Academic Research in Business Administration, 3(3): 1-9 Reave, L. (2005). Spiritual values and practices related to leadership effectiveness. The Leadership Quarterly, 16(5), 655-687. Silver, C. (2012). Coping with the consequences of ‘too many lawyers’: Securing the place of international graduate law students. International Journal of the Legal Profession, 19(2-3), 227-245. Uhlig, R., Mehta, K., Silverstone, S., and Mossavar-Rahmani, F., (2015). The Value of Business Education: A Case Study of Business Administration Graduates of a Large University, Journal of Business and Behavioral Sciences, Vol. 7, Issue 1, p. 89-106, Spring and Behavioral Sciences, Vol. 7, Issue 1, p. 89-106, Spring Vedder, R., Denhart, C., & Robe, J. (2013). Why are recent college graduates underemployed? University enrollments and labor-market realities. Washington, D.C.: Center for College Affordability and Productivity. 18
https://openalex.org/W3190847372	https://epub.ub.uni-muenchen.de/91019/1/s12969-021-00653-8.pdf	English	null	Infection with SARS-CoV-2 causes flares in patients with juvenile idiopathic arthritis in remission or inactive disease on medication	Pediatric rheumatology online journal	2,021	cc-by	4,137	RESEARCH ARTICLE Open Access © The Author(s). 2021 Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated in a credit line to the data. Hügle et al. Pediatric Rheumatology (2021) 19:163 https://doi.org/10.1186/s12969-021-00653-8 Hügle et al. Pediatric Rheumatology (2021) 19:163 https://doi.org/10.1186/s12969-021-00653-8 Infection with SARS-CoV-2 causes flares in patients with juvenile idiopathic arthritis in remission or inactive disease on medication Boris Hügle* , Manuela Krumrey-Langkammerer and Johannes-Peter Haas Background l d acute risk of dying [3]. In adults with rheumatic dis- eases, survey data and a retrospective study in a La- tino cohort has shown an increased risk of flare after COVID-19 [4–6]. Juvenile idiopathic arthritis (JIA) is the most common chronic rheumatic disease in children, with a reported prevalence that varies between 16 and 150 per 100,000 [1]. Children are among the least affected by the disease, with COVID-19 less frequent and less aggressive in this age group [7]. Rare cases of an inflammatory multisys- tem syndrome following infection with SARS-CoV-2 have been describes in children, however (PIMS-TS) [8]. About the risk of flare in JIA due to coronavirus infec- tion, little is known. The course of JIA may have multiple periods of in- active disease and flares of inflammatory activity [2]. The COVID-19 pandemic, caused by SARS-CoV-2 infection, has seen at least 190 million confirmed cases and over 4 million deaths worldwide recorded until July 2021. In most people the infection will cause a mild-to-moderate flu-like illness character- ized by fever, cough, and loss of taste and smell, among other symptoms; some patients require hospitalization and ventilatory support and are at We report five cases of patients with JIA in remission or long-term inactive disease on medication who devel- oped flares of their disease in close temporal correlation with a confirmed prior diagnosis of SARS-CoV-2 infection. * Correspondence: huegle.boris@rheuma-kinderklinik.de German Center for Pediatric and Adolescent Rheumatology, Gehfeldstrasse 24, 82467 Garmisch-Partenkirchen, Germany Abstract Background: Flares of juvenile idiopathic arthritis (JIA) have been described in the context of various infections. Flares of rheumatic diseases in adults have been described following infection with SARS-CoV-2 in several cohorts. So far, the effect of infection with SARS-CoV-2 on the course of JIA is unknown. Methods: The database of the German Center for Pediatric and Adolescent Rheumatology was searched for patients with confirmed infection with SARS-CoV-2 and subsequent disease flare, admitted from July 2020 until June 2021. cJADAS-27, ESR and C-reactive protein, as well as uveitis activity, medication at the time of flare and treatment of flare was extracted. Patient cases were described individually. Results: Out of 988 patients admitted, five patients with remission off medication (n = 2) or inactive disease on medication (n = 3) were identified, with flare symptoms up to four weeks after infection with SARS-CoV-2. Conclusions: Flares can occur after infection with SARS-CoV-2 in patients with JIA in remission or inactive disease on medication. Treating physicians need to be aware of this fact, especially when counseling patients with rheumatic diseases about the respective dangers of COVID-19 and vaccination against SARS-CoV-2. * Correspondence: huegle.boris@rheuma-kinderklinik.de German Center for Pediatric and Adolescent Rheumatology, Gehfeldstrasse 24, 82467 Garmisch-Partenkirchen, Germany Page 2 of 5 Page 2 of 5 Hügle et al. Pediatric Rheumatology (2021) 19:163 Hügle et al. Pediatric Rheumatology (2021) 19:163 Hügle et al. Pediatric Rheumatology Methods in 178 cases. Of these, 13 samples were positive. Six pa- tients had no signs of arthritis flare but a recent history of other symptoms compatible with infection with SARS-CoV-2. Two had signs of arthritis flare after dis- continuation of medication following the diagnosis of COVID-19. The remaining five patients with flare after infection with SARS-CoV-2, despite no change in medi- cation, are presented here (Table 1). Laboratory data and treatment of the flare is given in Table 2. No patient showed any sign of uveitis at time of flare. Prior to ad- mission, the patients had been seen by their local pediatric rheumatologists on a regular, usually three- monthly basis, where inactive disease was confirmed. The database of the German Center for Pediatric and Adolescent Rheumatology in Garmisch-Partenkirchen was searched for patients with serologically confirmed coronavirus infection and subsequent flares of their rheumatic disease. A flare was defined as new signs of arthritis or uveitis, i.e. an increase in the total joint count or JADAS or increase in anterior chamber cells or flare requiring a modification in therapy, following a period of inactive disease. Inclusion criteria were: 1) a diagnosis of juvenile idio- pathic arthritis according to ILAR criteria [9], 2) flare of the disease in close temporal correlation, i.e. within three months, to a confirmed infection with SARS-CoV-2, and 3) no change to treatment after the last flare-up of the disease until the current flare including no pauses in medication documented in the patient file. Explicitly ex- cluded were patients who reduced or stopped medica- tion as a consequence of diagnosis of infection with SARS-CoV-2, and subsequently developed a flare of their rheumatic disease. The following data was ex- tracted from the registry: cJADAS-27, ESR and C- reactive protein at the time of flare and at the last assess- ment prior to the flare, uveitis activity according to SUN criteria, medication at time of flare and treatment of the flare. As a matter of routine, all patients were required to show negative antigen testing (SARS-CoV-2 Rapid Antigen Test, Roche, Grenzach-Wyhlen, Germany) and PCR testing (Vivalytic SARS-CoV-2 PCR-Test, Bosch Healthcare, Waiblingen, Germany) prior to admission, and patients in whom a preceding infection was sus- pected received quantitative determination of SARS- CoV-2-IgG by ELISA. Patient cases Case 1 is a seven-year-old girl diagnosed with extended oligoarthritis at the age of two years and ten months with arthritis in knees, ankles and, in the course of the disease, the left subtalar joint and both temporoman- dibular joints. She had been treated with methotrexate and later with adalimumab. Medication was stopped ap- proximately 18 months ago and she continued well with- out clinical signs of arthritis. She presented, off all medications, with swelling of the right ankle. Approximately three weeks prior to admis- sion, several cases of COVID-19 were diagnosed in the family but the patient had no clinical signs and repeat- edly negative antigen tests. Ankle swelling started shortly thereafter. On admission, she tested highly positive for SARS-CoV-2-IgG (> 120.00 relative units/ml), but nega- tive for IgM antibodies and PCR and antigen testing. Case 2 is a 17-year-old adolescent male who had been diagnosed with rheumatoid-factor negative polyarthritis at the age of 11 years and six months. He was treated with intraarticular steroid injections, methotrexate and in the course of his disease, etanercept, adalimumab and tocilizumab. He was switched to baricitinib 6 months prior to admission and experienced a complete response, Results In 988 cases with JIA admitted from July 2020 until June 2021 to the German Center for Pediatric and Adolescent Rheumatology, serology for COVID-19 was determined Table 1 Demographic and clinical data of the patients Case number Age Diagnosis according to ILAR criteria Age at Diagnosis (years) Age at last visit (years) Duration of Remission (months) Clinical signs of SARS-CoV-2 infection Affected joins Estimated time between infection and arthritis flare Medication 1 7 extended Oligoarthritis 2.8 7.6 40.8 none right ankle 3 weeks none 2 17 Rheumatoid- factor negative Polyarthritis 11.5 18.5 5.8 none both ankles and subtalar joints 3 weeks baricitinib 3 9 Seronegative Polyarthritis 0.7 8.9 12.2 fever, diarrhea both knees, right ankle 1 week methotrexate, etanercept 4 17 Enthesitis- associated arthritis 10.9 18.1 41.5 respiratory symptoms, loss of taste right knee, ankles, wrists, fingers 4 weeks none 5 15 extended Oligoarthritis 2.9 15.6 20.5 fever, respiratory symptoms wrists, hips, knees, ankles 4 weeks golimumab, etanercept Hügle et al. Pediatric Rheumatology (2021) 19:163 Page 3 of 5 Hügle et al. Pediatric Rheumatology Table 2 Clinical and laboratory values, and treatment of flare of the patients Case number Age CRP on last admission prior to flare (mg/dl) CRP at time of flare (mg/ dl) ESR on last admission prior to flare (mm/h) ESR at time of flare (mm/ h) cJADAS −27 on last admission prior to flare cJADAS- 27 at time of flare Treatment 1 7 0,09 0,83 7 11 1 3 IACSI, NSAID 2 17 0,02 0,2 2 18 N/A 7 methylprednisolone pulses × 3 3 9 < 0,06 0,11 10 10 1 15 methylprednisolone pulses × 3 4 17 0,35 18,11 3 73 6 20 methylprednisolone pulses × 3, IACSI, addition of methotrexate 5 15 0 0,22 5 31 1 10 methylprednisolone pulses × 3, IACSI, switch from golimumab to abatacept with confirmed inactivity by the treating physician. However, he presented with arthritis in both ankle and subtalar joints three weeks after COVID-19, with in- apparent clinical symptoms diagnosed by antigen testing after his brother had developed a symptomatic infection with SARS-CoV-2. then due to antibodies to both, to tocilizumab. Due to ongoing arthritis she was switched to golimumab at age 13 which, together with methotrexate, led to lasting re- mission. Discussion Here we describe five cases of patients with JIA, in clinical remission or inactive on medication, who de- veloped a flare of their disease shortly after a docu- mented infection with SARS-CoV-2. Time between infection and flare can only be roughly estimated, as patients were asymptomatic in two cases and arthritis was in most cases assessed by a trained rheumatolo- gist at a later date. In general, the time span between infection and start of joint symptoms appears to be between one and four weeks. Patients were on various medications, or off medication completely, but all were inactive for at least six months prior to the viral infection with SARS-CoV-2. She presented on etanercept (approx. 0.7 mg/kg) with arthritis in both knees and the left ankle, both ongoing for three months. This had developed shortly after COVID-19 of the whole family (confirmed by antigen testing at that time) where she had fever and mild diarrhea. Case 4 is an 18-year-old adolescent male who had been diagnosed with enthesitis associated arthritis at age 11 years. He was treated with methotrexate, hydroxy- chloroquine and low-dose prednisolone until medication was stopped at the age of 17 years. He continued well without medication until he developed an infection with SARS-CoV-2, with positive antigen test, with signs of in- fection in the upper respiratory tract, vertigo and loss of taste. He developed increasing joint pains shortly there- after and presented three months later with confirmed arthritis in the right knee, both wrists, ankles and several finger joints. Frequency and clinical symptoms of acute respiratory infections in patients with chronic disease have been poorly studied. Respiratory infections in children with cystic fibrosis was not more frequent, but associated with higher morbidity compared to healthy controls [10]. Infectious diseases have, however, long been identi- fied as a possible source of flare in JIA, especially M. pneumoniae and C. jejuni [11]. Flares have been associ- ated with acute respiratory infections and influenza-like illness, although the majority of infections do not cause flares [12]. Case 5 is a 15-year-old adolescent female who had been diagnosed with extended oligoarthritis starting at age three. She was treated with multiple intraarticular steroid joint injections and methotrexate, but was started on etanercept at age nine due to ongoing arthritis. Results She presented approximately 18 months later, four weeks after COVID-19, confirmed by antigen test, with arthritis in both wrists, hips, knees, ankles and the right temporomandibular joint. Case 3 is a nine-year-old girl who had been diagnosed with rheumatoid-factor negative arthritis at the age of ten months. She was initially treated with methotrexate, intraarticular steroid injections, and etanercept was added at age three years. She had occasional mild flares, the last one at age seven years where dosage was adjusted. Competing interests h h The authors are not aware of any conflict of interest. Received: 3 August 2021 Accepted: 16 November 2021 Received: 3 August 2021 Accepted: 16 November 2021 Abbreviations COVID: Coronavirus Disease; JIA: Juvenile Idiopathic Arthritis; ELISA: Enzyme- Linked Immunosorbent Assay; SARS-CoV-2: Severe Acute Respiratory Syndrome Coronavirus-2; IACSI: Intra-articular Corticosteroid Injection; NSAI D: Non-steroidal Anti-inflammatory Drug Availability of data and materials The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request. Discussion At age ten she developed uveitis of the left eye, when she was switched first to adalimumab, then infliximab, and Even without prior infections, flares are frequent oc- currences in JIA patients in inactive disease, with 42.5% of patients developing a flare within one year of achiev- ing inactive disease [2]. However, there are indications that flare rate in rheumatic disease is increased after Page 4 of 5 Page 4 of 5 Page 4 of 5 Hügle et al. Pediatric Rheumatology (2021) 19:163 Hügle et al. Pediatric Rheumatology (2021) 19:163 SARS-CoV-2 infection: in adults with rheumatic disease, 19% of patients with rheumatic disease hospitalized for SARS-CoV-2 showed signs of flare of their underlying rheumatic disease [13]. A recent study from Italy de- scribes an increased flare rate in a single center in pa- tients with JIA, with the flare rate overall rising from 6.3 to 16.9% compared to the previous year [14]. In a survey conducted in the German registry, disease activity mea- sured by numeric rating scale increased more than one point in five of 32 paediatric patients with rheumatic dis- ease [15]. Little is known regarding flares in JIA after COVID-19 vaccination, even if other vaccinations have been shown not to be associated with an increased risk [16]. A preliminary study from Greece did not show any disease exacerbation in 21 adolescents with JIA on eta- nercept following vaccination [17]. No flares after vac- cination were observed in this institution. Authors’ contributions Patients infected with SARS-CoV-2 have shown to ex- hibit a variety of elevated cytokine levels, most promin- ently IL-6 [18, 19]. The efficacy of treatments targeting these cytokines in COVID-19 has been demonstrated [20]. It has also been shown that the dysregulation of the immune system can last for a long time after the infec- tion with SARS-CoV-2, and can cause long-term damage in different organ systems [21–23]. The role of cytokines in the pathogenesis of JIA has been established, and more recent data points at various roles especially of IL- 6 in the different subtypes of JIA [24, 25]. It is therefore conceivable that a change in cytokine patters induced by SARS-CoV-2 leads to a flare of disease in an already pre- disposed person. BH and JPH designed the study and performed the statistical analysis. MKL participated in the design of the study and helped draft the manuscript. All authors read and approved the final manuscript. Funding This work Funding This work was not supported by any outside source. Funding This work was not supported by any outside source. Conclusions In conclusion, we describe flares after infection with SARS-CoV-2 in patients with JIA in remission or in- active disease on medication. Paediatric rheumatologists need to be aware of this fact in their daily practice, and the possibility of flares despite adequate medication might add emphasis to the recommendation to vaccinate against COVID-19. However, more studies are necessary to establish the true risk of flares in JIA from COVID- 19, including the risk of vaccination against SARS-CoV- 2. Competing interests The authors are not aware of any conflict of interest. Competing interests The authors are not aware of any conflict of interest. Ethics approval and consent to participate Ethics approval and consent to participate The study was approved by the Ethics Committee of the Medical Faculty, Ludwig-Maximilians University Munich, Germany (Nr. 21–0828). The study was approved by the Ethics Committee of the Medical Faculty, Ludwig-Maximilians University Munich, Germany (Nr. 21–0828). p No individual patient data is reported in this manuscript. No individual patient data is reported in this manuscript. An early survey from Italy in adults showed the benefit of continuing rheumatic medications to stabilize the dis- ease and prevent flares [26]. The American College of Rheumatology Guidance for COVID-19 vaccination in patients with rheumatic diseases also recommends con- tinuation of most medications in JIA after vaccination, with the exception of methotrexate and abatacept, where a pause of one week is warranted [27]. Given that add- itional patients were seen in this cohort with flare after discontinuation of medication following a diagnosis of COVID-19, it seems a sensible course of action to con- tinue antirheumatic medications after an infection with SARS-CoV-2. In fact, it would be worthwhile to consider optimizing the dosage of conventional synthetic and bio- logic DMARDS after confirmed infection with SARS- CoV-2. Consent for publication Consent for publication No individual patient data is reported in this manuscript. Co se t o pub cat o No individual patient data is reported in this manuscript. Publisher’s Note 9. Petty RE, Southwood TR, Manners P, Baum J, Glass DN, Goldenberg J, et al. International League of Associations for Rheumatology.et al, International league of associations for rheumatology classification of juvenile idiopathic arthritis: second revision, Edmonton, 2001. J Rheumatol. 2004;31(2):390–2. 9. Petty RE, Southwood TR, Manners P, Baum J, Glass DN, Goldenberg J, et al. International League of Associations for Rheumatology.et al, International league of associations for rheumatology classification of juvenile idiopathic arthritis: second revision, Edmonton, 2001. J Rheumatol. 2004;31(2):390–2. Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. 10. van Ewijk BE, van der Zalm MM, Wolfs TFW, Fleer A, Kimpen JLL, Wilbrink B, et al. Prevalence and impact of respiratory viral infections in young children with cystic fibrosis: prospective cohort study. Pediatrics. 2008;122(6):1171–6. https://doi.org/10.1542/peds.2007-3139. 11. Aslan M, Kasapcopur O, Yasar H, Polat E, Saribas S, Cakan H, et al. Do infections trigger juvenile idiopathic arthritis? Rheumatol Int. 2011;31(2):215– 20. https://doi.org/10.1007/s00296-009-1253-4. 11. Aslan M, Kasapcopur O, Yasar H, Polat E, Saribas S, Cakan H, et al. Do infections trigger juvenile idiopathic arthritis? Rheumatol Int. 2011;31(2):215– 20. https://doi.org/10.1007/s00296-009-1253-4. 12. Carvalho LM, de Paula FE, Silvestre RVD, Roberti LR, Arruda E, Mello WA, et al. Prospective surveillance study of acute respiratory infections, influenza-like illness and seasonal influenza vaccine in a cohort of juvenile idiopathic arthritis patients. Pediatr Rheumatol Online J. 2013;11(1):10. https://doi.org/10.1186/1546-0096-11-10. 12. Carvalho LM, de Paula FE, Silvestre RVD, Roberti LR, Arruda E, Mello WA, et al. Prospective surveillance study of acute respiratory infections, influenza-like illness and seasonal influenza vaccine in a cohort of juvenile idiopathic arthritis patients. Pediatr Rheumatol Online J. 2013;11(1):10. https://doi.org/10.1186/1546-0096-11-10. p g 13. Ye C, Cai S, Shen G, Guan H, Zhou L, Hu Y, et al. Clinical features of rheumatic patients infected with COVID-19 in Wuhan, China. Ann Rheum Dis. 2020;79(8):1007–13. https://doi.org/10.1136/annrheumdis-2020-217627. 14. Naddei R, Alfani R, Bove M, Discepolo V, Mozzillo F, Guarino A, et al. Increased relapse rate during COVID-19 lockdown in an Italian cohort of children with juvenile idiopathic arthritis. Arthritis Care Res (Hoboken). 2021. https://doi.org/10.1002/acr.24768. 15. Sengler C, Eulert S, Minden K, Niewerth M, Horneff G, Kuemmerle-Deschner J, et al. Clinical manifestations and outcome of SARS-CoV-2 infections in children and adolescents with rheumatic musculoskeletal diseases: data from the National Paediatric Rheumatology Database in Germany. RMD Open. 2021;7(2):(2). https://doi.org/10.1136/rmdopen-2021-001687. 16. Silva VB, Faquin G, Nicácio A, Regacini R, Lederman H, Hilário MO, et al. References Paediatric inflammatory multisystem syndrome temporally associated with COVID-19 (PIMS-TS): a narrative review and the viewpoint of the Latin American Society of Pediatric Intensive Care (SLACIP) Sepsis committee. BMJ Paediatr Open. 2021;5(1):e000894. https://doi.org/1 0.1136/bmjpo-2020-000894. 8. Fernandez-Sarmiento J, et al. Paediatric inflammatory multisystem syndrome temporally associated with COVID-19 (PIMS-TS): a narrative review and the viewpoint of the Latin American Society of Pediatric Intensive Care (SLACIP) Sepsis committee. BMJ Paediatr Open. 2021;5(1):e000894. https://doi.org/1 0.1136/bmjpo-2020-000894. References 1. Ravelli A, Martini A. Juvenile idiopathic arthritis. Lancet. 2007;369(9563):767– 78. https://doi.org/10.1016/S0140-6736(07)60363-8. 2. Guzman J, Oen K, Huber AM, Watanabe Duffy K, Boire G, Shiff N, et al. The risk and nature of flares in juvenile idiopathic arthritis: results from the ReACCh-out cohort. Ann Rheum Dis. 2016;75(6):1092–8. https://doi.org/1 0.1136/annrheumdis-2014-207164. 3. Hyrich KL, Machado PM. Rheumatic disease and COVID-19: epidemiology and outcomes. Nat Rev Rheumatol. 2021;17(2):71–2. https://doi.org/10.1038/ s41584-020-00562-2. 4. Barbhaiya M, Levine JM, Bykerk VP, Jannat-Khah D, Mandl LA. Systemic rheumatic disease flares after SARS-CoV-2 vaccination among rheumatology outpatients in new York City. Ann Rheum Dis. 2021;80(10):1352–4. https:// doi.org/10.1136/annrheumdis-2021-220732. This is a retrospective case series, with no definitive proof of pathophysiological correlation between docu- mented infection and observed flare. Further studies in a larger cohort are necessary to determine the risk of flare in the JIA population and possible ways to prevent this, beyond the obvious benefits of vaccination. 5. Felten R, Scherlinger M, Guffroy A, Poindron V, Meyer A, Giannini M, et al. Incidence and predictors of COVID-19 and flares in patients with rare autoimmune diseases: a systematic survey and serological study at a national reference center in France. Arthritis Res Ther. 2021;23(1):188. https://doi.org/10.1186/s13075-021-02565-0. 6. Fike A, Hartman J, Redmond C, Williams SG, Ruiz-Perdomo Y, Chu J, et al. Risk factors for COVID-19 and rheumatic disease flare in a US cohort of Latino patients. Arthritis Rheumatol. 2021;73(7):1129–34. https://doi.org/10.1 002/art.41656. Page 5 of 5 Page 5 of 5 Page 5 of 5 Hügle et al. Pediatric Rheumatology (2021) 19:163 Hügle et al. Pediatric Rheumatology (2021) 19:163 Hügle et al. Pediatric Rheumatology 7. Licciardi F, Giani T, Baldini L, Favalli EG, Caporali R, Cimaz R. COVID-19 and what pediatric rheumatologists should know: a review from a highly affected country. Pediatr Rheumatol Online J. 2020;18(1):35. https://doi.org/1 0.1186/s12969-020-00422-z. 26. Favalli EG, Agape E, Caporali R. Incidence and clinical course of COVID-19 in patients with connective tissue diseases: a descriptive observational analysis. J Rheumatol. 2020;47(8):1296. https://doi.org/10.3899/jrheum.200507. 26. Favalli EG, Agape E, Caporali R. Incidence and clinical course of COVID-19 in patients with connective tissue diseases: a descriptive observational analysis. J Rheumatol. 2020;47(8):1296. https://doi.org/10.3899/jrheum.200507. 27. Curtis JR, Johnson SR, Anthony DD, Arasaratnam RJ, Baden LR, Bass AR, et al. American College of Rheumatology Guidance for COVID-19 vaccination in patients with rheumatic and musculoskeletal diseases: version 1. Arthritis Rheumatol. 2021;73(7):1093–107. https://doi.org/10.1002/art.41734. 8. Fernandez-Sarmiento J, et al. Publisher’s Note Association between the ultrasonographic and clinical findings in the hips of patients with juvenile idiopathic arthritis. Rev Bras Reumatol. 2013;53(4): 322–7. 17. Dimopoulou D, et al. Safety and tolerability of the COVID-19 mRNA-vaccine in adolescents with juvenile idiopathic arthritis on treatment with TNF- inhibitors. Arthritis Rheumatol. 2021. https://doi.org/10.1002/art.41977. 18. Merza MY, Hwaiz RA, Hamad BK, Mohammad KA, Hama HA, Karim AY. Analysis of cytokines in SARS-CoV-2 or COVID-19 patients in Erbil city, Kurdistan Region of Iraq. PLoS ONE. 2021;16(4):e0250330. https://doi.org/1 0.1371/journal.pone.0250330. 19. Qin C, Zhou L, Hu Z, Zhang S, Yang S, Tao Y, et al. Dysregulation of immune response in patients with coronavirus 2019 (COVID-19) in Wuhan, China. Clin Infect Dis. 2020;71(15):762–8. https://doi.org/10.1093/cid/ciaa248. 20. Rubio-Rivas M, Forero CG, Mora-Luján JM, Montero A, Formiga F, Homs NA, et al. Beneficial and harmful outcomes of tocilizumab in severe COVID-19: a systematic review and Meta-analysis. Pharmacotherapy. 2021. https://doi. org/10.1002/phar.2627. 21. Heneka MT, Golenbock D, Latz E, Morgan D, Brown R. Immediate and long- term consequences of COVID-19 infections for the development of neurological disease. Alzheimers Res Ther. 2020;12(1):69. https://doi.org/1 0.1186/s13195-020-00640-3. 22. Jing Y, Luo L, Chen Y, Westerberg LS, Zhou P, Xu Z, et al. SARS-CoV-2 infection causes immunodeficiency in recovered patients by downregulating CD19 expression in B cells via enhancing B-cell metabolism. Signal Transduct Target Ther. 2021;6(1):345. https://doi.org/10.1 038/s41392-021-00749-3. 22. Jing Y, Luo L, Chen Y, Westerberg LS, Zhou P, Xu Z, et al. SARS-CoV-2 infection causes immunodeficiency in recovered patients by downregulating CD19 expression in B cells via enhancing B-cell metabolism. Signal Transduct Target Ther. 2021;6(1):345. https://doi.org/10.1 038/s41392-021-00749-3. 23. McDonald LT. Healing after COVID-19: are survivors at risk for pulmonary fibrosis? Am J Physiol Lung Cell Mol Physiol. 2021;320(2):L257–l265. https:// doi.org/10.1152/ajplung.00238.2020. 23. McDonald LT. Healing after COVID-19: are survivors at risk for pulmonary fibrosis? Am J Physiol Lung Cell Mol Physiol. 2021;320(2):L257–l265. https:// doi.org/10.1152/ajplung.00238.2020. 24. Akioka S. Interleukin-6 in juvenile idiopathic arthritis. Mod Rheumatol. 2019; 29(2):275–86. https://doi.org/10.1080/14397595.2019.1574697. 24. Akioka S. Interleukin-6 in juvenile idiopathic arthritis. Mod Rheumatol. 2019; 29(2):275–86. https://doi.org/10.1080/14397595.2019.1574697. 25. Woo P. Cytokines and juvenile idiopathic arthritis. Curr Rheumatol Rep. 2002;4(6):452–7. https://doi.org/10.1007/s11926-002-0050-9. 25. Woo P. Cytokines and juvenile idiopathic arthritis. Curr Rheumatol Rep. 2002;4(6):452–7. https://doi.org/10.1007/s11926-002-0050-9.
https://openalex.org/W2319381059	https://www.nature.com/articles/srep23806.pdf	English	null	Genetic variant in IL-33 is associated with idiopathic recurrent miscarriage in Chinese Han population	Scientific reports	2,016	cc-by	5,770	Jun Yue1,2,, Yu Tong3,4,, Lan Xie1,2, Tao Ma1,2 & Jiyun Yang2,5,6 Jun Yue1,2,, Yu Tong3,4,, Lan Xie1,2, Tao Ma1,2 & Jiyun Yang2,5,6 Recurrent miscarriage (RM) is the occurrence of repeated pregnancies that end in miscarriage of the fetus before 20 weeks of gestation. At least 50% of the RM patients are considered idiopathic. High IL-33 levels are critical in early pregnancy and impact the outcome of subsequent pregnancies. However, the association of polymorphisms of IL-33 with idiopathic RM is still unclear. The present study was initiated to investigate whether IL-33 polymorphisms are risk factors for idiopathic RM in Chinese Han population. Study subjects comprised of 321 cases and 384 controls. Five polymorphisms (rs10435816, rs16924159, rs16924171, rs1929992, rs1332290) in IL-33 and serum IL-33 concentrations were assessed. rs16924159 variant exhibits significant association with RM in additive and recessive genetic model (additive model P = 0.015, recessive model P = 0.007). In contrast, rs10435816, rs16924171, rs1929992 and rs1332290 are not significantly associated with RM. Serum IL-33 levels are significantly lower in RM cases than in control (173.51 ± 94.12 versus. 200.97 ± 110.06 (pg/ml), P = 4.57 × 10−4). There are lower levels of serum IL-33 in rs16924159 homozygous mutant (AA) than homozygous wild-type (GG) in this study population, including cases and control groups (172.18 ± 103.01 versus. 205.82 ± 119.01 (pg/ml), P = 0.006). Reduced IL-33 levels and rs16924159 IL-33 variant may contribute to the pathogenesis of idiopathic RM in Chinese Han population. Recurrent miscarriage (RM) is the loss of three or more consecutive pregnancies before 20 weeks of gestation. RM affects 1–5% of women who attempt to bear children1. Various factors have been identified for miscarriage, including uterine anomaly, chromosomal abnormalities, endocrine dysfunction, thrombophilia, immune disor- ders, life style factors and maternal infections1. However, At least 50% of the RM patients have no deviations from any applied diagnostic test. The underlying causes remain unknown and are considered idiopathic. Idiopathic RM is characterized by multifactorial etiology. Although much work has been done to identify the cause of idiopathic RM, its reason remains unknown2. There is growing evidence that RM has also genetic susceptibility2–4. In most conducted genetic association studies targeting RM, the most frequently addressed genes in the context of RM are associated with immunotolerance, inflammation and changes of maternal metabolism and blood coagulation3. Successful pregnancy depends on the cytokine environment, which can either be protective or harmful to the conceptus5. www.nature.com/scientificreports www.nature.com/scientificreports www.nature.com/scientificreports Genetic variant in IL-33 is associated with idiopathic recurrent miscarriage in Chinese Han population d: 07 January 2016 ed: 15 March 2016 OPEN received: 07 January 2016 accepted: 15 March 2016 Published: 30 March 2016 received: 07 January 2016 accepted: 15 March 2016 Published: 30 March 2016 Jun Yue1,2,, Yu Tong3,4,, Lan Xie1,2, Tao Ma1,2 & Jiyun Yang2,5,6 Scientific Reports \| 6:23806 \| DOI: 10.1038/srep23806 Materials and Methodsh Subjects. This study has been carried out in accordance with The Code of Ethics of the World Medical Association (Declaration of Helsinki) for experiments involving humans. The Institutional Review Boards of the Sichuan Provincial People’s Hospital and West China Second University Hospital, Sichuan University approved this study. All subjects provided informed consent before participating in the study. y j p p p g y Cases comprised 321 women from Sichuan Provincial People’s Hospital and West China Second University Hospital between January 2011 and June 2013, who had experienced three or more pregnancy losses with the same partner, and which had occurred from between the beginning of pregnancy to the 20th week of gestation; gestational age was calculated as the time between the first day of the last normal menstrual period and the first signs of pregnancy loss. The work-up included endometrial biopsies for evaluating luteal phase defects;pelvic ultrasound scans to assess ovarian morphology and the uterine cavity; hysteroscopy to evaluate uterine anatomic abnormalities; karyotyping of peripheral blood to evaluate chromosomal aberrations in both patient and partner. These procedures were performed on all patients. In addition, trisomy and triploidy of fetal tissues and were evaluated using fluorescence in situ hybridization (FISH). Data on FISH of fetal tissues were available for 266 (82.87%) patients.i Exclusion criteria included older age (40 years or older at first miscarriage), parental and fetal chromosomal aberrations, Rh incompatibility, anatomical abnormalities, preclinical miscarriages, endocrine disorders (includ- ing diabetes), liver function abnormalities, and abnormal thyroid function, thyroid antibodies, hyperprolactine- mia prior to luteal phase defects, and fetomaternal alloimmune thrombocytopenia, infections (toxoplasmosis, human cytomegalovirus, rubella, human immunodeficiency virus, Group B streptococci, Chlamydia trachomatis, hepatitis B and C and bacterial vaginosis). Patients were also excluded if they reported systemic autoimmune dis- ease, such as anti-phospholipid syndrome, systemic lupus erythematosus, multiple sclerosis, rheumatoid arthritis. Control subjects composed of 384 women who were recruited from Sichuan Provincial People’s Hospital and West China Second University Hospital between January 2011 and June 2013. Age and ethnical group were matched with patients. All subjects had at least one live birth and no miscarriages, and did not have a family history of RM. IL-33 genotyping. Blood samples were taken from all participants in EDTA containing tube for total genomic DNA extraction. Total genomic DNA was extracted by silica columns isolation kits (Tiangen BioTech CO.LTD, Beijing China). Jun Yue1,2,, Yu Tong3,4,, Lan Xie1,2, Tao Ma1,2 & Jiyun Yang2,5,6 The balance of pro-inflammatory and anti-inflammatory cytokines was suggested to be critical for successful pregnancy6. Maternal immunological tolerance of the allogenic fetus in pregnancy is linked with imbalance between T regulatory cells (Tregs) and T helper 17 cells (Th17)7,8. There is an elevation of Treg subset in the normal human pregnancy. However, a decrease in the number of Tregs is associated with miscarriage9,10. 1Department of Gynecology & Obstetrics, Hospital of the University of Electronic Science and Technology of China and Sichuan Provincial People’s Hospital, Chengdu, China. 2Prenatal Diagnosis Center, Hospital of the University of Electronic Science and Technology of China and Sichuan Provincial People’s Hospital, Chengdu, China. 3Laboratory of Early Developmental and Injuries, West China Institute of Woman and Children’s Health, West China Second University Hospital, Sichuan University, Chengdu, China. 4Key Laboratory of Gynecologic & Obstetric and Pediatric Diseases and Birth Defects of Ministry of Education, Chengdu, China. 5Sichuan Provincial Key Laboratory for Human Disease Gene Study, Hospital of the University of Electronic Science and Technology of China and Sichuan Provincial People’s Hospital, Chengdu, China. 6School of Medicine, University of Electronic Science and Technology of China, Chengdu, China. These authors contributed equally to this work. Correspondence and requests for materials should be addressed to J.Y. (email: yangjiyun@yeah.net) Scientific Reports \| 6:23806 \| DOI: 10.1038/srep23806 1 www.nature.com/scientificreports/ In contrast to the Tregs cells, the proportion of Th17 cells was significantly higher in patients with RM than in normal pregnant11–13. p g IL-33 is a member of the IL-1 family. Members of this family are known to play important roles in host defense, immune regulation, neuronal injury and inflammation14. IL-33 is mainly expressed in epithelial, endothelial cells, activated Th2 cells and mast cells14–17. IL-33 appears to be a cytokine with dual function, performing both as a tra- ditional cytokine through activation of the IL-33/ST2 pathway and as a nuclear factor involved in transcriptional repression and regulation of genes18,19. IL-33 has been reported to be the specific ligand for suppression of tumor- igenicity 2(ST2), which is a member of the IL-1 receptor superfamily. IL-33 significantly stimulated the produc- tion of inflammatory cytokines, such as IL-1β , TNF-α , IL-8 and IL-620. IL-33 enhances Tregs cells differentiation. Furthermore, it provides a necessary signal for Tregs cells accumulation and maintenance in inflamed tissues13,21. Jun Yue1,2,, Yu Tong3,4,*, Lan Xie1,2, Tao Ma1,2 & Jiyun Yang2,5,6 p y g gl Recent research shows that IL-33 expression was found in endothelial and smooth muscle cells in the placenta, chorioamniotic membranes, and umbilical cord22. Human endometrial stromal cells (HESCs) rapidly release IL-33. IL-33 knockdown in undifferentiated HESCs abrogate this pro-inflammatory decidual response23. Further, sequential activation of the IL-33/ST2L/sST2 axis was disordered in decidualizing HESCs from women with recurrent pregnancy loss. Signals from HESC cultures prolonged the implantation window but also caused sub- sequent pregnancy failure in mice23. Serum IL-33 levels of normal pregnancies were highly variable in first eleven weeks of early pregnancy, with no statistical differences identified between each gestational week24. However, there is a significant increase in serum IL-33 levels from patients who went on to abortion compared with normal control pregnancies24. Above data suggest that dysregulated levels of serum IL-33 indicate miscarriage in live early pregnancies.h p g However, the association of polymorphisms of IL-33 with idiopathic RM is still unclear. The present study was initiated to evaluate association of the IL-33 gene polymorphisms with idiopathic RM in Chinese Han population. Five polymorphisms (rs10435816, rs16924159, rs16924171, rs1929992, rs1332290) in IL-33 gene were selected for genotyping, and then we analyzed the association between the five SNPs and the risk of idiopathic RM. Materials and Methodsh Characteristic Cases Controls P value Mean age (years) 31.35 ± 5.04 31.95 ± 4.32 0.089 Smoking (%) 25 (7.79) 33 (8.59) 0.584 Mean BMI (kg/m2) 21.31 ± 2.99 20.95 ± 3.06 0.366 Menarche(years) 12.88 ± 1.32 13.19 ± 1.18 0.013 Irregular menstrual history (%) 39 (12.15) 28 (7.29) 0.038 Number of pregnancies 4.91 ± 1.15 3.01 ± 1.28 < 0.0001 Miscarriage 3.83 ± 0.725 0 ± 0 < 0.0001 Table 1. Clinical Characteristics of the Study Population. P versus controls; BMI, Body Mass Index. Serum IL-33 Measurements. Because pregnancy influences IL-33 levels, blood drawing criteria were used for patients and control women. This included interval (4–6 months) from the last event (miscarriage or delivery) and time of blood draw (8:00–10:00 AM). Blood samples were taken from all participants in plain tubes (no pre- servatives) for serum preparation. Serum was prepared by centrifugation of coagulated blood tubes at 2000 g for 10 min at room temperature, and was stored in 200 μl aliquots at − 80 °C. Serum IL-33 were measured using a human IL-33 enzyme-linked immunosorbent assay (catalogue number ADI-900-201; Enzo Life Sciences., Shanghai China) according to the manufacturer’s instructions. Assay sensitivity was 1.7 pg/ml (range: 7.8–500 pg/ml). Samples were tested in triplicate at the same time in the same assay. Statistical analysis. Statistical analysis was performed by the software SPSS for Windows (version 13.0). Continuous variables were expressed as mean ± SD. Statistical differences were performed using t test or Mann– Whitney U test. Then differences in genotype distribution of the polymorphism between case and control sub- jects were compared by χ 2 test. Three models were used for statistical analysis, including dominant model, the recessive genetic model and the additive model. Logistic regression analysis was used to evaluate the association between genotype of SNPs and idiopathic RM. Subsequently, Pearson χ 2 statistics were calculated and P values were computed by comparing the statistic to a χ 2 distribution with 1 or 2 degrees of freedom for the allelic and genotypic tests. P value of less than 0.05 was considered statistically significant. Gene variants were tested for Hardy-Weinberg equilibrium using Haploview version 4.2 (http://www.broad.mit.edu/mpg/haploview). We used CaTS Power Calculator (www.sph.umich.edu/csg/abecasis/cats) to calculate the power to detect an association between IL-33 variants and RM in our cohort. Materials and Methodsh The parameters used were: 321 RM cases and 384 control women, genotypic relative risk for heterozygotes (1/2) and minor allele homozygote (2/2), and the MAF for RM cases and controls for the five tested SNPs, and assumed 5% population prevalence of RM. Assuming these parameters, this sample size provided 82%, 83%, 82%, 82% and 83% power for rs10435816, rs16924159, rs16924171, rs1929992 and rs1332290, respectively. Results The clinical characteristics of study participants. The clinical characteristics of patients and con- trols are listed in Table 1. There are no significant differences in mean age, smoking (%) and mean BMI (kg/m2) between cases and control (P > 0.05). While lower age at menarche, higher irregular menstrual history (%) and number of pregnancies were seen in the RM group. Although they did not constitute strong risk factors of RM, they were selected as the covariates that were controlled for in subsequent analyses. IL-33 genotype distribution. Genotype distribution of five (rs10435816, rs16924159, rs16924171, rs1929992 and rs1332290) SNPs are in Hardy-Weinberg equilibrium in cases and controls (P > 0.05).if Table 2 summarizes the distribution of IL-33 genotypes between patients and control. Significant differ- ences in the distribution of rs16924159 genotype are observed between patients and controls. rs16924159 vari- ant exhibits significant association with RM in additive and recessive genetic model (additive model P = 0.015, recessive model P = 0.007). In contrast, rs10435816, rs16924171, rs1929992 and rs1332290 are not significantly associated with RM in any of three genetic models. As shown in Table 3, after adjustments for menarche age, irregular menstrual history and number of pregnancies by Binary Logistic Regression, rs16924159 showed sig- nificant association with RM in two genetic models (adjusted odds ratio (OR) = 2.037, 95% confidence interval (CI) = 1.135–3.656, P = 0.017 in an additive model; OR = 1.975, 95% CI = 1.161–3.357, P = 0.012 in a recessive model; OR = 1.926, 95% CI = 1.232–3.012, P = 0.014 in a homozygous model; OR = 1.199, 95% CI = 0.863– 1.667, P = 0.299 in a heterozygous model). Mean serum IL-33 levels. As shown in Fig. 1A, serum IL-33 levels are significantly lower in RM cases than in control (173.51 ± 94.12 versus 200.97 ± 110.06 pg/ml, P = 4.57 × 10−4). As shown in Fig. 1B, there are lower levels of serum IL-33 in rs16924159 homozygous mutant (AA) than homozygous wild-type (GG) in this study population, including cases and control groups (172.18 ± 103.01 versus 205.82 ± 119.01 pg/ml, P = 0.006). Materials and Methodsh Genotype data of IL-33 for the chromosomal region 9:6215149–6257983 were download from the HapMap website (CHB database, Hapmap Data Rel 27 Phase II III, Feb09, on NCBI B36 assembly, dbSNP b126). In total, 37 SNPs with a frequency > 0.10 were identified in CHB (Han Chinese in Beijing). The criteria for all the referenced SNPs was that minor allele frequency (MAF) set at > 10% and r2 > 0.8. The five tag-SNPs (rs10435816, rs16924159, rs16924171, rs1929992, rs1332290) were selected by running the tagger program in Haploview 4.2. These five SNPs in 5 tests capture 37 of 37 (100%) alleles with a mean r2 of 0.957. rs10435816, rs16924159 and rs16924171 are located in the 5′ upstream region of IL-33 (NM_001199640.1), rs1929992 and rs1332290 are located in the intron region of IL-33 (NM_001199640.1). IL-33 genotyping was performed using real time PCR assays (TaqMan probe). TaqMan® SNP Genotyping Assay kit were from Applied Biosystems. Cat. C__30088929_10 (rs10435816), C__34291196_10 (rs16924159), C__34291189_10 (rs16924171), C__11716827_10 (rs1929992), C___2762144_10 (rs1332290). Replicate blinded quality control samples were included to assess the reproducibility of the genotyping procedure. The successful genotyping rate was 99.0%. In brief, 24 individuals of cases and controls were randomly selected to sequence for confirmation of five SNPs genotyping. The confirmation rate was 100%. Scientific Reports \| 6:23806 \| DOI: 10.1038/srep23806 2 www.nature.com/scientificreports/ Characteristic Cases Controls P value Mean age (years) 31.35 ± 5.04 31.95 ± 4.32 0.089 Smoking (%) 25 (7.79) 33 (8.59) 0.584 Mean BMI (kg/m2) 21.31 ± 2.99 20.95 ± 3.06 0.366 Menarche(years) 12.88 ± 1.32 13.19 ± 1.18 0.013 Irregular menstrual history (%) 39 (12.15) 28 (7.29) 0.038 Number of pregnancies 4.91 ± 1.15 3.01 ± 1.28 < 0.0001 Miscarriage 3.83 ± 0.725 0 ± 0 < 0.0001 Table 1. Clinical Characteristics of the Study Population. P versus controls; BMI, Body Mass Index. Characteristic Cases Controls P value Mean age (years) 31.35 ± 5.04 31.95 ± 4.32 0.089 Smoking (%) 25 (7.79) 33 (8.59) 0.584 Mean BMI (kg/m2) 21.31 ± 2.99 20.95 ± 3.06 0.366 Menarche(years) 12.88 ± 1.32 13.19 ± 1.18 0.013 Irregular menstrual history (%) 39 (12.15) 28 (7.29) 0.038 Number of pregnancies 4.91 ± 1.15 3.01 ± 1.28 < 0.0001 Miscarriage 3.83 ± 0.725 0 ± 0 < 0.0001 Table 1. Clinical Characteristics of the Study Population. P versus controls; BMI, Body Mass Index. Results Adjustments for menarche age, irregular menstrual history and number of pregnancies by Binary Logistic Regression; OR, odds ratio; CI, confidence interval. P value Wald value OR 95% CI Additive 0.017 5.69 2.037 1.135–3.656 Dominant 0.288 1.127 1.244 0.831–1.862 Recessive 0.012 6.33 1.975 1.161–3.357 Homozygous 0.014 5.87 1.926 1.232–3.012 Heterozygous 0.299 1.036 1.199 0.863–1.667 Table 3. Associated between rs16924159 variant with risk for idiopathic recurrent miscarriage after adjustment for Binary Logistic Regression. Adjustments for menarche age, irregular menstrual history and number of pregnancies by Binary Logistic Regression; OR, odds ratio; CI, confidence interval. significantly difference is observed in serum IL-33 levels according to rs16924159 genotypes among controls (AA versus GG, 187.57 ± 105.90 versus 211.96 ± 129.34 pg/ml, P = 0.189) (Fig. 1D). ignificantly difference is observed in serum IL-33 levels according to rs16924159 genotypes among controls (AA ersus GG, 187.57 ± 105.90 versus 211.96 ± 129.34 pg/ml, P = 0.189) (Fig. 1D). Haplotype analysis. Five-locus (IL-33 rs10435816, rs16924159, rs16924171, rs1929992, rs1332290) IL-33 haplotypes were constructed based on the prevalence of individual SNPs and linkage disequilibrium between them. Nine haplotypes were found to be common and were included in subsequent analysis. No haplotypes exhibits significant association with RM (Table 4). Results However, There is no significantly difference in homozygous wild-type (GG), compared with homozygous wild-type (GG) and heterozygous (A/G) genotype carriers (172.18 ± 103.01 versus 191.61 ± 103.91 pg/ml, P = 0.067).i Moreover, there is a significant decline in serum IL-33 levels according to rs16924159 genotypes in cohorts (AA versus GG, 160.57 ± 100.03 versus 196.90 ± 102.12 pg/ml, P = 0.024) (Fig. 1C). However Scientific Reports \| 6:23806 \| DOI: 10.1038/srep23806 3 www.nature.com/scientificreports/ SNP Genotype Case (n = 321) Control (n = 384) Additive Dominant Recessive P value Corrected P P value Corrected P P value Corrected P rs10435816 AA 103 (0.321) 118 (0.307) AG 142 (0.442) 174 (0.453) GG 76 (0.237) 92 (0.240) 0.926 1 0.699 1 0.93 1 rs16924159 GG 104 (0.324) 151 (0.393) GA 152 (0.474) 184 (0.479) AA 65 (0.202) 49 (0.128) 0.006 0.03 0.057 0.285 0.007 0.035 rs16924171 AA 91 (0.283) 115 (0.299) AT 159 (0.495) 183 (0.477) TT 71 (0.221) 86 (0.224) 0.867 1 0.93 1 0.642 1 rs1929992 AA 67 (0.209) 98 (0.255) AG 172 (0.536) 185 (0.482) GG 82 (0.255) 101 (0.263) 0.264 1 0.819 1 0.147 0.735 rs1332290 AA 58 (0.181) 71 (0.185) AC 148 (0.461) 189 (0.492) CC 115 (0.358) 124 (0.323) 0.602 0.324 0.886 Table 2. Association between SNPs and risk for idiopathic recurrent miscarriage in Chinese Han population. P versus controls; P value by χ 2 test. Corrected P = P value × 5 (the number of genotyped SNPs). Table 2. Association between SNPs and risk for idiopathic recurrent miscarriage in Chinese Han l i P l P l b 2 C d P P l ( h b f d SN Table 2. Association between SNPs and risk for idiopathic recurrent miscarriage in Chinese Han population. P versus controls; P value by χ 2 test. Corrected P = P value × 5 (the number of genotyped SNPs). population. P versus controls; P value by χ test. Corrected P = P value × 5 (the number of genotyped SNPs). P value Wald value OR 95% CI Additive 0.017 5.69 2.037 1.135–3.656 Dominant 0.288 1.127 1.244 0.831–1.862 Recessive 0.012 6.33 1.975 1.161–3.357 Homozygous 0.014 5.87 1.926 1.232–3.012 Heterozygous 0.299 1.036 1.199 0.863–1.667 Table 3. Associated between rs16924159 variant with risk for idiopathic recurrent miscarriage after adjustment for Binary Logistic Regression. Scientific Reports \| 6:23806 \| DOI: 10.1038/srep23806 Discussionh The role of IL-33 polymorphisms in idiopathic RM pathogenesis remains unknown. In this case–control study, we selected five tag SNPs in IL-33 to investigate the association of IL-33 polymorphisms with idiopathic RM. We found that a tag SNP rs16924159, located in the regulatory region of IL-33, was associated with idiopathic RM. The homozygous mutant (AA) genotype of rs16924159 is risk factor in idiopathic RM patients. This is the first study to assess the contribution of IL-33 polymorphisms in hereditary susceptibility of idiopathic RM.i y p y p y p y p We found that serum IL-33 levels are significantly lower in idiopathic RM cases than in control. Recently, it has been reported that activation of the IL-33/ST2 pathway in human endometrial stromal cells (HESCs) is crit- ical for a successful pregnancy23. Autocrine IL-33 signaling in HESCs facilitates embryo implantation in mice23. However, IL-33 knockdown in decidualizing HESCs impairs embryo implantation23. Women experiencing with RM have deficit in Treg cell number and function, compared with normal pregnant women25. These studies sug- gest that high IL-33 levels are critical in early pregnancy and impact the outcome of subsequent pregnancies23. However, compared with normal control pregnancies, there is a significant increase in serum IL-33 levels at six weeks’ gestation in patients who destined to abort24. The biological significance of a rise of serum IL-33 levels is uncertain in pregnant woman destined to miscarry. As IL-33 is a crucial role in promoting Tregs cells differenti- ation and adaptation to the inflammatory environment. One possible explanation is that the pregnancy is failing due to sensitization of the maternal immune system, and that the rise is a compensatory response to attempting to rescue the pregnancy24. Although IL-33 plays an exact role in embryo-maternal interactions, especially during early stages of implantation, our data support the notion that miscarriage may be attributed, at least in part, to low level of IL-33 expression and secretion. In this study, serum IL-33 measurements were performed in non preg- nant women. We did not compare serum IL-33 levels between pre-pregnancy with post-pregnancy in the same Scientific Reports \| 6:23806 \| DOI: 10.1038/srep23806 4 www.nature.com/scientificreports/ Figure 1. Comparison of serum IL-33 levels among IL-33 genotypes. (A) Serum IL-33 concentrations in cases and controls. IL-33 levels were significantly lower in patients with RM than in controls. (B) Serum IL-33 levels according to rs16924159 genotypes among cases and controls. Discussionh Women with AA genotype showed significantly lower IL-33 levels than GG genotypes. (C) Serum IL-33 levels according to rs16924159 genotypes among cases. Idiopathic RM patients with the AA genotype showed significantly lower serum IL-33 levels than those with GG genotypes. (D) Serum IL-33 levels according to rs16924159 genotypes among controls. No statistical difference of serum IL-33 levels was found between women with those with GG genotypes. Figure 1. Comparison of serum IL-33 levels among IL-33 genotypes. (A) Serum IL-33 concentrations in cases and controls. IL-33 levels were significantly lower in patients with RM than in controls. (B) Serum IL-33 levels according to rs16924159 genotypes among cases and controls. Women with AA genotype showed significantly lower IL-33 levels than GG genotypes. (C) Serum IL-33 levels according to rs16924159 genotypes among cases. Idiopathic RM patients with the AA genotype showed significantly lower serum IL-33 levels than those with GG genotypes. (D) Serum IL-33 levels according to rs16924159 genotypes among controls. No statistical difference of serum IL-33 levels was found between women with those with GG genotypes. Figure 1. Comparison of serum IL-33 levels among IL-33 genotypes. (A) Serum IL-33 concentrations i Figure 1. Comparison of serum IL-33 levels among IL-33 genotypes. (A) Serum IL-33 concentrations in cases and controls. IL-33 levels were significantly lower in patients with RM than in controls. (B) Serum IL-33 levels according to rs16924159 genotypes among cases and controls. Women with AA genotype showed significantly lower IL-33 levels than GG genotypes. (C) Serum IL-33 levels according to rs16924159 genotypes among cases. Idiopathic RM patients with the AA genotype showed significantly lower serum IL-33 levels than those with GG genotypes. (D) Serum IL-33 levels according to rs16924159 genotypes among controls. No statistical difference of serum IL-33 levels was found between women with those with GG genotypes. Haplotypea Totalb Casesb Controlb P value GGTAC 0.376 0.389 0.36 0.2637 AGAGA 0.269 0.272 0.266 0.8109 AAAGA 0.187 0.182 0.192 0.6504 AATAC 0.034 0.031 0.037 0.5395 AAAAC 0.023 0.022 0.024 0.8519 GAAGA 0.017 0.017 0.017 0.9973 GGAAC 0.016 0.013 0.02 0.276 AGTAC 0.015 0.016 0.015 0.8443 AGAGC 0.012 0.011 0.013 0.7764 Table 4. IL-33 Haplotypes. Note: aIL-33 rs10435816/rs16924159/rs16924171/rs1929992/rs1332290 haplotypes. bHaplotype frequency. Scientific Reports \| 6:23806 \| DOI: 10.1038/srep23806 Acknowledgementsh g This work has been supported by National Natural Science Foundation of China (Grant No. 81271048, To JY.Y.), Foundation for Transfer of Scientific and Technological Achievements of Sichuan Province (Grant No. 15010118, To JY.Y.). g This work has been supported by National Natural Science Foundation of China (Grant No. 81271048, To JY.Y.), Foundation for Transfer of Scientific and Technological Achievements of Sichuan Province (Grant No. 15010118, To JY.Y.). References Immunity 23, 479–490, doi: 10.1016/j.immuni.2005.09.015 (2005).l y y j 21. Morita, H. et al. An Interleukin-33-Mast Cell-Interleukin-2 Axis Suppresses Papain-Induced Allergic Inflammation by Prom Regulatory T Cell Numbers Immunity 43 175–186 doi: 10 1016/j immuni 2015 06 021 (2015) y y j 21. Morita, H. et al. An Interleukin-33-Mast Cell-Interleukin-2 Axis Suppresses Papain-Induced Aller Regulatory T Cell Numbers. Immunity 43, 175–186, doi: 10.1016/j.immuni.2015.06.021 (2015). 21. Morita, H. et al. An Interleukin-33-Mast Cell-Interleukin-2 Axis Suppresses Papain-Induced Allergic Inflammation by Promoting Regulatory T Cell Numbers. Immunity 43, 175–186, doi: 10.1016/j.immuni.2015.06.021 (2015). Regulatory T Cell Numbers. Immunity 43, 175–186, doi: 10.1016/j.immuni.2015.06.021 (2015). 22. Topping, V. et al. Interleukin-33 in the human placenta. J Matern Fetal Neonatal Med 26, 327–338, doi: 10.3109/14767058.2012.73 g y y j 2. Topping, V. et al. Interleukin-33 in the human placenta. J Matern Fetal Neonatal Med 26, 327–338, doi: 10.3109/14767058.2012.735724 (2013). 22. Topping, V. et al. Interleukin-33 in the human placenta. J Matern Fetal Neonatal Med 26, 327–338, doi: (2013). 23. Salker, M. S. et al. Disordered IL-33/ST2 activation in decidualizing stromal cells prolongs uterine receptivity in women with recurrent pregnancy loss. PLoS One 7, e52252, doi: 10.1371/journal.pone.0052252 (2012). p g y j p 24. Kaitu’u-Lino, T. J., Tuohey, L. & Tong, S. Maternal serum interleukin-33 and soluble ST2 across early pregnancy, and their association with miscarriage. J Reprod Immunol 95, 46–49, doi: 10.1016/j.jri.2012.06.003 (2012).h g p j j 5. La Rocca, C., Carbone, F., Longobardi, S. & Matarese, G. The immunology of pregnancy: regulatory T cells control materna immune tolerance toward the fetus. Immunol Lett 162, 41–48 (2014). Discussionh Haplotypea Totalb Casesb Controlb P value GGTAC 0.376 0.389 0.36 0.2637 AGAGA 0.269 0.272 0.266 0.8109 AAAGA 0.187 0.182 0.192 0.6504 AATAC 0.034 0.031 0.037 0.5395 AAAAC 0.023 0.022 0.024 0.8519 GAAGA 0.017 0.017 0.017 0.9973 GGAAC 0.016 0.013 0.02 0.276 AGTAC 0.015 0.016 0.015 0.8443 AGAGC 0.012 0.011 0.013 0.7764 Table 4. IL-33 Haplotypes. Note: aIL-33 rs10435816/rs16924159/rs16924171/rs1929992/rs1332290 haplotypes. bHaplotype frequency. Table 4. IL-33 Haplotypes. Note: aIL-33 rs10435816/rs16924159/rs16924171/rs1929992/rs1332290 haplotypes. bHaplotype frequency. case. In later study, more experiments are also needed to elucidate the association of serum IL-33 levels between pre-pregnancy with post-pregnancy in idiopathic recurrent miscarriage patients. case. In later study, more experiments are also needed to elucidate the association of serum IL-33 levels between pre-pregnancy with post-pregnancy in idiopathic recurrent miscarriage patients. Scientific Reports \| 6:23806 \| DOI: 10.1038/srep23806 5 www.nature.com/scientificreports/ Our data also indicate that there are lower levels of serum IL-33 in rs16924159 homozygous mutant (AA) than homozygous wild-type (GG) in this study population, including cases and control groups. Moreover, there is a progressive decline in serum IL-33 levels according to rs16924159 genotypes in RM cohorts. rs16924159 variant, located in the regulatory region of IL-33, was associated with reduced serum IL-33 levels. This may be linked with defective transcriptional processing of IL-33 mRNA.i p p g To our knowledge, we first found that genetic variant in IL-33 is associated with idiopathic RM, and the AA genotype of rs16924159 is associated with lower serum IL-33 levels. However, our study has some limitations. It was performed only in Chinese Han population. Additional more large-scale population studies from different ethnic backgrounds are needed to confirm the association of IL-33 variant rs16924159 with RM. Functional experiments are also needed to elucidate the association. p In conclusion, our study demonstrated that a tag SNP rs16924159 was associated with idiopathic RM and serum IL-33 levels are significantly lower in idiopathic RM cases than in control. There are lower levels of serum IL-33 in rs16924159 homozygous mutant (AA) than homozygous wild-type (GG) in this study population. It may be concluded that IL-33 acts as a risk factor in the pathogenesis of idiopathic RM. References Mol Hum Reprod 10, 347–353, doi: 10.1093/molehr/gah044 (2004).h p g 11. Wang, W. J. et al. Increased prevalence of T helper 17 (Th17) cells in peripheral blood and decidua in spontaneous abortion patients. J Reprod Immunol 84, 164–170, doi: 10.1016/j.jri.2009.12.003 (2010).h p g 11. Wang, W. J. et al. Increased prevalence of T helper 17 (Th17) cells in peripheral blood and decidua in unexplained recu spontaneous abortion patients J Reprod Immunol 84 164 170 doi: 10 1016/j jri 2009 12 003 (2010) p p p j j 12. Liu, Y. S. et al. Study on the relationship between Th17 cells and unexplained recurrent spontaneous abortion. Am J Reprod Immunol 65, 503–511, doi: 10.1111/j.1600-0897.2010.00921.x (2011).h j 13. Schiering, C. et al. The alarmin IL-33 promotes regulatory T-cell function in the intestine. Nature 513, 564–568, doi: 10.1 nature13577 (2014). 13. Schiering, C. et al. The alarmin IL-33 promotes regulatory T-cell function in the intestine. Nature 513, 564–568, doi: 10.1038/ nature13577 (2014). 14 Lefrancais E et al Central domain of IL 33 is cleaved by mast cell proteases for potent activation of group 2 innate lymphoid cells 14. Lefrancais, E. et al. Central domain of IL-33 is cleaved by mast cell proteases for potent activation of group-2 innate lymphoid Proc Natl Acad Sci USA 111, 15502–15507, doi: 10.1073/pnas.1410700111 (2014). p 15. Balato, A. et al. IL-33 is regulated by TNF-alpha in normal and psoriatic skin. Arch Dermatol Res 306, 299–304, doi: 10.1007/s00403- 014-1447-9 (2014). k l l k d k h d h l l ll bl Th b 6. Stojkovic, S. et al. Interleukin-33 induces urokinase in human endothelial cells–possible impact on angiogenesis. J Thromb Haemos 12, 948–957, doi: 10.1111/jth.12581 (2014). j 7. Tung, H. Y., Plunkett, B., Huang, S. K. & Zhou, Y. Murine mast cells secrete and respond to interleukin-33. J Interferon Cytokine Re 34, 141–147, doi: 10.1089/jir.2012.0066 (2014). j 8. Haraldsen, G., Balogh, J., Pollheimer, J., Sponheim, J. & Kuchler, A. M. Interleukin-33 - cytokine of dual function or novel alarmin Trends Immunol 30, 227–233, doi: 10.1016/j.it.2009.03.003 (2009). j 9. Carriere, V. et al. IL-33, the IL-1-like cytokine ligand for ST2 receptor, is a chromatin-associated nuclear factor in vivo. Proc Nat Acad Sci USA 104, 282–287, doi: 10.1073/pnas.0606854104 (2007). p 20. Schmitz, J. et al. IL-33, an interleukin-1-like cytokine that signals via the IL-1 receptor-related protein ST2 and induces T helper type 2-associated cytokines. References 1. Branch, D. W., Gibson, M. & Silver, R. M. Clinical practice. Recurrent miscarriage. N Engl J Med 363, 1740–1747, doi: 10.1056 NEJMcp1005330 (2010). p 2. George, L., Granath, F., Johansson, A. L., Olander, B. & Cnattingius, S. Risks of repeated miscarriage. Paediatr Perinat Epidemiol 20 119–126, doi: 10.1111/j.1365-3016.2006.00703.x (2006). j ( ) 3. Sierra, S. & Stephenson, M. Genetics of recurrent pregnancy loss. Semin Reprod Med 24, 17–24, doi: 10.1055/s-2006-931797 (2006). 4 R ll K N i j L & L M G ti f t i i h ll t k l d f t di ti F t G t j rra, S. & Stephenson, M. Genetics of recurrent pregnancy loss. Sem j 3. Sierra, S. & Stephenson, M. Genetics of recurrent pregnancy loss. Semin Reprod Med 24, 17–24, doi: 10.1055/s-2006-931797 (2006). 4. Rull, K., Nagirnaja, L. & Laan, M. Genetics of recurrent miscarriage: challenges, current knowledge, future directions. Front Genet 3 34 d i 10 3389/f 2012 00034 (2012) 3. Sierra, S. & Stephenson, M. Genetics of recurrent pregnancy loss. Semin Reprod Med 24, 17–24, doi: 10.1055/s-2006-931797 (2006) 4. Rull, K., Nagirnaja, L. & Laan, M. Genetics of recurrent miscarriage: challenges, current knowledge, future directions. Front Gene 3, 34, doi: 10.3389/fgene.2012.00034 (2012).l g 5. Chatterjee, P., Chiasson, V. L., Bounds, K. R. & Mitchell, B. M. Regulation of the Anti-Inflammatory Cytokines Interleukin-4 and Interleukin-10 during Pregnancy. Front Immunol 5, 253, doi: 10.3389/fimmu.2014.00253 (2014). g g yi 6. Choi, Y. K. & Kwak-Kim, J. Cytokine gene polymorphisms in recurrent spontaneous abortions: a comprehensive review. Am J Reprod Immunol 60, 91–110, doi: 10.1111/j.1600-0897.2008.00602.x (2008). Reprod Immunol 60, 91–110, doi: 10.1111/j.1600-0897.2008.00602.x p j 7. Lee, S. K. et al. An imbalance in interleukin-17-producing T and Foxp3(+ ) regulatory T cells in women with idiopathic recu pregnancy loss. Hum Reprod 26, 2964–2971, doi: 10.1093/humrep/der301 (2011). p g y p p 8. Aluvihare, V. R., Kallikourdis, M. & Betz, A. G. Regulatory T cells mediate maternal tolerance to the fetus. Nat Immunol 5, 266– doi: 10.1038/ni1037 (2004). 9. Somerset, D. A., Zheng, Y., Kilby, M. D., Sansom, D. M. & Drayson, M. T. Normal human pregnancy is associated with an elevation in the immune suppressive CD25+ CD4+ regulatory T-cell subset. Immunology 112, 38–43, doi: 10.1111/j.1365-2567.2004.01869.x (2004). ( ) 10. Sasaki, Y. et al. Decidual and peripheral blood CD4+ CD25+ regulatory T cells in early pregnancy subjects and spontaneous abortion cases. Author Contributions Conceived and designed the experiments: JY.Y. and J.Y. Performed the experiments: Y.T. Collected samples: J.Y., L.X. and T.M. Analyzed the data: JY.Y., J.Y. and Y.T. Drafted the manuscript: JY.Y. and J.Y. Revised the manuscript and gave final approval of the version to be published: JY.Y. 6 Scientific Reports \| 6:23806 \| DOI: 10.1038/srep23806 www.nature.com/scientificreports/ How to cite this article: Yue, J. et al. Genetic variant in IL-33 is associated with idiopathic recurrent miscarriage in Chinese Han population. Sci. Rep. 6, 23806; doi: 10.1038/srep23806 (2016). Scientific Reports \| 6:23806 \| DOI: 10.1038/srep23806 This work is licensed under a Creative Commons Attribution 4.0 International License. The images or other third party material in this article are included in the article’s Creative Commons license, unless indicated otherwise in the credit line; if the material is not included under the Creative Commons license, users will need to obtain permission from the license holder to reproduce the material. To view a copy of this license, visit http://creativecommons.org/licenses/by/4.0/ Additional Informationi Additional Information Competing financial interests: The authors declare no competing financial interests. Competing financial interests: The authors declare no competing financial interests. How to cite this article: Yue, J. et al. Genetic variant in IL-33 is associated with idiopathic recurrent miscarriage in Chinese Han population. Sci. Rep. 6, 23806; doi: 10.1038/srep23806 (2016). This work is licensed under a Creative Commons Attribution 4.0 International License. The images or other third party material in this article are included in the article’s Creative Commons license, unless indicated otherwise in the credit line; if the material is not included under the Creative Commons license, users will need to obtain permission from the license holder to reproduce the material. To view a copy of this license, visit http://creativecommons.org/licenses/by/4.0/ Scientific Reports \| 6:23806 \| DOI: 10.1038/srep23806 7
https://openalex.org/W3198918408	https://ressat.org/index.php/ressat/article/download/559/106	English	null	Transgender student experiences of online education during COVID-19 pandemic era in rural Eastern Cape area of South Africa: A descriptive phenomenological study	Research in social sciences and technology	2,021	cc-by	11,317	Transgender student experiences of online education during COVID-19 pandemic era in rural Eastern Cape area of South Africa: A descriptive phenomenological study Transgender student experiences of online education during COVID-19 pandemic era in rural Eastern Cape area of South Africa: A descriptive phenomenological study Azwihangwisi Helen Mavhandu-Mudzusi1, Tshimangadzo Selina Mudau2, Thulile Shandu1 & Nthomeni Dorah Ndou3 wihangwisi Helen Mavhandu-Mudzusi1, Tshimangadzo Selina Mudau2, Thulile Shandu1 & Nthomeni Dora dou3 Corresponding Author: mmudza@unisa.ac.za 1. University of South Africa, South Africa 2. Sefako Makgatho Health Sciences University, South Africa 3. University of Venda, South Africa Received : 2021-06-14 Revised : 2021-07-20 Accepted : 2021-08-04 Corresponding Author: mmudza@unisa.ac.za 1. University of South Africa, South Africa 2. Sefako Makgatho Health Sciences University, South Africa 3. University of Venda, South Africa Received : 2021-06-14 Revised : 2021-07-20 Accepted : 2021-08-04 10.46303/ressat.2021.14 How to cite this paper: Mavhandu-Mudzusi, A.H., Mudau, T.S., Shandu, T. & Ndou, N.D. (2021). Transgender student experiences of online education during COVID-19 pandemic era in rural Eastern Cape area of South Africa: A descriptive phenomenological study. Research in Social Sciences and Technology, 6(2), 110-128. https://doi.org/10.46303/ressat.2021.14 This is an Open Access article distributed under the terms of the Creative Commons Attribution 4.0 International license (https://creativecommons.org/licenses/by/4.0/). This is an Open Access article distributed under the terms of the Creative Commons Attribution 4.0 International license (https://creativecommons.org/licenses/by/4.0/). Research in Social Sciences and Technology Volume 6 Issue 2, 2021 Research in Social Sciences and Technology Volume 6 Issue 2, 2021 Research in Social Sciences and Technology Volume 6 Issue 2, 2021 110 Mavhandu-Mudzusi, A.H. et al. Transgender student experiences of online education during COVID-19 pandemic era in rural Eastern Cape area of South Africa: A descriptive phenomenological study Volume 6 Issue 2, 2021 Transgender student experiences of online education during COVID-19 pandemic era in rural Eastern Cape area of South Africa: A descriptive phenomenological study Azwihangwisi Helen Mavhandu-Mudzusi1, Tshimangadzo Selina Mudau2, Thulile Shandu1 & Nthomeni Dorah Ndou3 Corresponding Author: mmudza@unisa.ac.za 1. University of South Africa, South Africa 2. Sefako Makgatho Health Sciences University, South Africa 3. University of Venda, South Africa Received : 2021-06-14 Revised : 2021-07-20 Accepted : 2021-08-04 How to cite this paper: Mavhandu-Mudzusi, A.H., Mudau, T.S., Shandu, T. & Ndou, N.D. (2021). Transgender student experiences of online education during COVID-19 pandemic era in rural Eastern Cape area of South Africa: A descriptive phenomenological study. Research in Social Sciences and Technology, 6(2), 110-128. https://doi.org/10.46303/ressat.2021.14 This is an Open Access article distributed under the terms of the Creative Commons Attribution 4.0 International license (https://creativecommons.org/licenses/by/4.0/). 10.46303/ressat.2021.14 Introduction This paper reports on transgender students’ experiences of online education during COVID-19 Era in rural areas of the Eastern Cape. When Coronavirus disease of 2019 (COVID-19) hit global communities in 2020, the world was thrust into turmoil and had to deal with various types of change. Chief of these were the restrictions that had to be imposed as countries went into lockdown. South Africa was not immune to these changes. On 23 March 2020, the President of South Africa, Cyril Ramaphosa, announced that the country’s lockdown would begin on 26 March 2020 (SANews, 2020). The lockdown protocols detailed various measures to curb the spread of the virus, including the restrictions imposed on the movement of and contact between individuals. The pandemic and its concomitant restrictions, therefore, meant that various sectors of society had to find different ways of carrying on with their lives. Educational institutions, then, as one of the main areas where gatherings are mandated, frequent and prolonged, were generally affected. They consequently had to find innovative ways to ensure that students, of various kinds and from various backgrounds, continued to have access to education (Daniel, 2020; Dube, 2020). One such was online education implemented by most of the institutions of higher education in South Africa, regardless of students’ socioeconomic environment. Therefore, it was important for the researchers to explore the experiences of transgender students in the rural area of Eastern Cape towards online education, especially because of the community’s attitudes associated with this population. Abstract COVID-19 affected education in many ways. As a response, various strategies were introduced to ensure students’ access to education, including online education. For most of the students, fulltime online education brought diverse challenges. This descriptive phenomenological study explored the experiences of transgender students regarding online education during the COVID-19 pandemic in Buffalo City Metro Municipality, South Africa. Data were collected by means of individual telephonic interviews with eight purposively selected transgender students using the snowballing technique. Data were thematically analysed using the interpretative phenomenological analysis framework for data analysis. The findings indicate that transgender students faced barriers in accessing online education, including (i) limited interaction, (ii) unconducive home environment because of stigmatisation, abuse and disruptions, and (iii) lack of access to the internet owing to the centralisation of internet access points, the unaffordability of data, unstable internet connections and an intermittent electricity supply. The findings further highlighted that transgender students face stigmatisation and abuse which hinder their learning. The situation is exacerbated by a lack of sources of income for transgender students, especially those who must make means for money to buy educational resources such as data for internet connection needed for online classes. Based on the findings, the researchers recommend continuous support for transgender students to ensure their continued engagement in online education, amidst the challenges they face. To this end, educational institutions should ensure that students have alternative means of accessing education so that those from diverse populations, settings and socioeconomic statuses are reached. Keywords: COVID-19; descriptive phenomenology; online education; transgender students; Rural Eastern Cape 2021 Research in Social Sciences and Technology E-ISSN: 2468-6891 ressat.org Research in Social Sciences and Technology Volume 6 Issue 2, 2021 111 Mavhandu-Mudzusi, A.H. et al. Transgender student experiences of online education during COVID-19 pandemic era in rural Eastern Cape area of South Africa: A descriptive phenomenological study Research in Social Sciences and Technology E-ISSN: 2468-6891 ressat.org Literature review COVID-19 affected education globally (Mailizar, Almanthari, Maulina & Bruce, 2020). The result of the various restrictions meant that students and teachers could not have face-to-face contact in class owing to restrictions on gatherings. As a response, educational institutions have been seeking innovative alternatives to ameliorate the effects of the disruptions. To this end, a quick Google Scholar search of recent literature indicates how “throughout the COVID-19 pandemic, there has been a growing reliance on the use of technology to learn, live, and stay connected” (Goldschmidt, 2020). What comes out prominently, though, is how technology is touted as an enabler for access to education during lockdown restrictions. Technology can widen access, increase flexibility and reduce the need for on-site teaching accommodation (Kumar, Saxena & Baber, 2021; Dube, 2020; Subedi & Subedi, 2020). It stands to reason, then, that technology platforms have been the platforms of choice for education under constraints. In other words, technology has provided access to education for students who would otherwise have had to miss out. “In education, the term access typically refers to the ways in which educational institutions and policies ensure – or at least strive to ensure – that students have equal and equitable opportunities to take full advantage of their education” (EdGlossary, 2014). At the core of access, then, is the issue of student support, in that institutions should strive to ensure learning opportunities for students. Student support, in this regard, refers to institutional (administrative and policy), cognitive (quality content and efficient pedagogy) and affective (coping strategies, mental wellness and conducive learning environments) strategies (Williams, McIntosh & Russell, 2021; Mailizar et al., 2020). Student support, therefore, is the foundation and defining character of access, as aptly argued by Williams et al., (2021) and Waters and Hensley (2020) who stated that access to learning is anchored on student technological support and use of various opportunities available and preferred by students. The 2021 Research in Social Sciences and Technology Volume 6 Issue 2, 2021 112 Research in Social Sciences and Te Volume 6 Issue 2, 2021 Mavhandu-Mudzusi, A.H. et al. Transgender student experiences of online education during COVID-19 pandemic era in rural Eastern Cape area of South Africa: A descriptive phenomenological study versatility of technology, then, provides the access and support so that students can have opportunities for successful learning. What is key, though, is that such opportunities should be equitable and equal (EdGlossary, 2014). Literature review While the benefits of technology cannot be disputed, it remains to be seen if technology in education is “an infallible means to develop profound knowledge and wide range of language skills” (Ahmad, 2016: 118), particularly in relation to students who might be regarded as those on the periphery. There remains a dearth of research on, for example, the role of technology in relation to providing access to education for transgender students, particularly in the time of COVID-19. Even outside the pandemic era, transgender students have often been left out of discussions and decisions regarding access to education, even though they face various adversities and impediments to accessing education (Omercajic & Wayne, 2020). Scholars argue that issues of access to education for transgender students are issues of equity and social justice (Omercajic & Wayne, 2020). According to Parents and Friends of Lesbians and Gays (PFLAG, 2004), transgender people are “those whose gender identity or gender expression differs from conventional expectations for their physical sex”. They face a myriad of challenges because they are different and their behaviour and expression of self, differs from what society expects or associates with that gender (Enke, 2016). According to Goldberg (2018), “many transgender students experience discrimination and harassment at college, which may have implications for their academic success and retention”. In fact, the negative experiences of transgender students are sometimes described as hostile (Omercajic & Wayne, 2020; Wayne, Kassen & Omercajic, 2020; Daniels et al., 2019). In addition to societal attitudes, transgender populations face adverse conditions such as poverty (Kia, Robinson, MacKay & Ross, 2020; Suissa & Sullivan, 2021). In view of this literature review, considering the experiences of transgender students of transphobic discrimination, bullying, poverty, transgender students still have to access online education like all other students during COVID-19 lockdown, this paper asks the following question: What might be the transgender students’ experiences of online education during COVID-19 Era in rural areas of the Eastern Cape? Thus, this study explores transgender students’ experiences of online education during COVID-19 Era in rural areas of the Eastern Cape. Research in Social Sciences and Technology E-ISSN: 2468-6891 ressat.org Design This study employed descriptive phenomenology to describe experiences of transgender students regarding online education during the COVID-19 lockdown. According to Moser and Korstjens (2018), phenomenological research is concerned with the essence of the participants’ meaning through its iterative nature. The researchers used descriptive phenomenology because it was deemed a suitable design to gain an understanding of the lived experiences of transgender students by digging deeper into their real-life world (Sloan & Bowe, 2014) when accessing online education during the COVID-19 pandemic in South Africa. This is because, according to Husserl, reality cannot be studied separately from the consciousness of those experiencing it (Morrow, Rodriguez & King, 2015). Theoretical framework. Perhaps the most relevant theoretical framework for exploring access is one that describes what the access to education entails – interaction. According to Moore (1989), a successful teaching and learning environment involves bi-directional interaction of various kinds. In view of the remote learning that has been associated with the pandemic restrictions, interaction theory, as encapsulated in the theory of transactional distance, is a worthwhile lens because it refers to “the universe of teacher-learner relationships that exist when students and instructors are separated by space and/or by time" (Moore, 1993). In this case, when one considers technology as a solution to transgender students’ access to education, one must consider Moore’s (1989) argument for student–student, student–teacher and student–content interaction. It should be noted that human–human interaction is not merely for the purposes of socialising, or an affective outlet, but is meant for cognitive activities such as discussions, explanations, cognitive exchanges, argumentation and the negotiation of meaning (Kumar et al., 2021; Saxena, Baber & Kumar, 2020). In addition to human–human interaction, one needs 2021 Research in Social Sciences and Technology Volume 6 Issue 2, 2021 113 Mavhandu-Mudzusi, A.H. et al. Transgender student experiences of online education during COVID-19 pandemic era in rural Eastern Cape area of South Africa: A descriptive phenomenological study to consider the fact that students interact with course content and assessment through technology (Aguilera-Hermida, 2020). This means that access to education must also explore transgender students’ educational experiences in relation to these types of interaction. Such access should also be considered based on the availability of technological resources for successful education (Williams et al., 2021). There is thus room to explore whether technology provides an enabling or disabling environment for the transgender students’ access to education. Research in Social Sciences and Technology E-ISSN: 2468-6891 ressat.org Data collection instrument These interviews were guided by an interview guide drafted by the researchers. Before actual data collection, the interview guide was tested through interviewing two transgender women who are not part of the sample. This was done to ensure that the questions are clear and that the language used is not discriminative, judgmental or sensitive as the transgender women are among the vulnerable population. After first pilot interview, the audio recording was listened to by all authors, and discussed the process, cautioning against using leading and closed ended questions which can affect the reflective and descriptive nature of the study. After the comments, second test interview was conducted, and discussion made. All the researchers were satisfied with the process followed during second interview and the probes used. The interview guide was then finalised. Ethical considerations To conduct this study, an ethical clearance was obtained from the University of South Africa’s (UNISA) Department of Health Studies Research Ethics Committee. Each participant informed about all the ethical related issues were provided to enable them to make verbal consent as the study was conducted at the height of global COVID-19 pandemic where South Africa was not spared. Furthermore, the study aims, objectives, and the process of data collection to the participants were explained to participants. Ethical principles such as the right to privacy and confidentiality, right to withdraw anytime during the study were explained to the participants (Flick, 2018). To minimise harm, which for this study was mostly the spread or contracting COVID-19 for both the researched and the researcher, the researcher ensured that special precautionary measures were followed by conducting telephonic interviews than physical meetings (Flick, 2018; Gray, Grove, & Sutherland, 2017; South African Government 2020). Table 1. Biographical data of participants Table 1. Biographical data of participants Table 1. Biographical data of participants Research in Social Sciences and Technology E-ISSN: 2468-6891 ressat.org Setting, population and sample The study was conducted at the Buffalo City Metro Municipality of the Eastern Cape Province, South Africa. The municipality had experienced high rates of COVID-19 infections and death rates during the second wave of the COVID-19 resurgence in South Africa. Participants were recruited using snowballing technique. To ensure homogeneity of the sample, the study inclusion criteria were participants who were 18 years and older; resided in the Buffalo City Metro Municipality for at least two years; who identified themselves as a transgender woman, or have been assigned male sex at birth and currently living as a woman, and have registered as a student before 2020, and still registered for current academic year (being at the higher education for at least two years). The recruitment process was initiated through the link of one transgender woman who met the inclusion criteria. The women provided the fourth author who was responsible for recruiting the participants and collecting data the cellphone numbers of other potential participants. The total sample size was eight transgender women. The number was determined by data saturation. As Kincheloe (2005) asserted, quality findings in qualitative research are not determined by number of participants but depth and prolonged engagement with the researched. Table 1 presents the participants’ biographical data. The names used for the participants are pseudonyms which were given in order to protect their identity. 2021 Research in Social Sciences and Technology E-ISSN: 2468-6891 ressat.org Research in Social Sciences and Technology Volume 6 Issue 2, 2021 Research in Social Sciences and Technology Volume 6 Issue 2, 2021 114 Mavhandu-Mudzusi, A.H. et al. Transgender student experiences of online education during COVID-19 pandemic era in rural Eastern Cape area of South Africa: A descriptive phenomenological study Mavhandu-Mudzusi, A.H. et al. Transgender student experiences of online education during COVID-19 pandemic era in rural Eastern Cape area of South Africa: A descriptive phenomenological study Table 1. Biographical data of participants Name Age Degree Level of study Gender identity disclosure at home Rudo 25 Psychology 3rd Not disclosed Thoby 23 Education 4th Not disclosed Bessy 23 Education 3rd Disclosed Nana 25 Social Work 3rd Disclosed Belinda 22 Information Technology 4th Disclosed Fanisa 23 Social Work 3rd Disclosed Oniah 24 Psychology 3rd Disclosed Metsi 25 Microbiology 3rd Disclosed, Data collection The researcher conducted individual in-depth telephonic interviews from volunteering transgender students guided by the final interview guide. All interviews were digitally audio recorded. The following free attitude open-ended question was posed to each participant to initiate the discussion. “Kindly share with me your experiences of online education during COVID-19 pandemic.” The initial question was then followed by further probes to explore further the single response given by the participants. This was to allow the researcher to Research in Social Sciences and Technology E-ISSN: 2468-6891 ressat.org 2021 Research in Social Sciences and Technology Volume 6 Issue 2, 2021 Research in Social Sciences and Technology Volume 6 Issue 2, 2021 115 Research in Social Sciences and Te Volume 6 Issue 2, 2021 Mavhandu-Mudzusi, A.H. et al. Transgender student experiences of online education during COVID-19 pandemic era in rural Eastern Cape area of South Africa: A descriptive phenomenological study exhaust all possible areas in the description of the participants’ experiences. The following were some of the probes used: What is your view about technology teaching? What did you like most about that? The researcher asked further probing questions such as: How did it affect you as an individual? What could be the best solution? Probing was to ensure that the hidden experiences are carefully mined out from participants as the experiential experts of the experiences and the essence of meaning of their experiences is unearthed (Flick, 2018; Morrow et al., 2015). Probing was blended with continuous reflections between the researchers and the researched to explore contextual meaning of the item picked for reflection. Through the reflections, the researcher ensured rich engagement to dig deeper the meaning-making of the participants’ experiences (Moser & Korstjens, 2018). Though the researcher could not observe the non-verbal cues such as body language, the researcher made field notes recording aspects such as tone of voice and silence during the interviews (Preethi, 2019). Interviews lasted from 45 minutes to an hour. At the end of every interview, the researcher marked all the different recordings with a participant number. Interviews were conducted until data saturation was reached (Gray et al., 2017). Data analysis Data were transcribed verbatim by first author to promote the voice of the participants (Castillo, 2018; Peat, Rodriguez & Smith, 2019). Two of the researchers individually analyzed the data guided by the steps of interpretative phenomenological analysis framework for data analysis as follows: The step-by-step approach involved reading and rereading followed by noting of common themes through circling to ensure emersion with data. Thereafter, the researcher matched emergent themes with field notes per individual case. A connection between emergent themes was sought. The researcher ensured a continuous reflective process before moving to the next step while seeking for similar patterns across all cases. The last step involved peer critiquing by other researchers together with consulting available literature to compare findings with participants’ verbatim quotes for a coherent analytical account of the whole process. Finally, a table was created to summarise superordinate themes, sub-themes, and themes of the findings as indicated below in Table 2: Research in Social Sciences and Technology E-ISSN: 2468-6891 ressat.org Results Based on the analysis of the participants’ description of their experiences, two superordinate themes, composed of several themes and subthemes, namely, psychosocial environment and limited interaction, emerged from the data analysis, as shown in Table 2. 2021 Research in Social Sciences and Technology Volume 6 Issue 2, 2021 116 Mavhandu-Mudzusi, A.H. et al. Transgender student experiences of online education during COVID-19 pandemic era in rural Eastern Cape area of South Africa: A descriptive phenomenological study Table 2: Experiences of transgender students regarding online teaching and learning during the COVID-19 lockdown Superordinate themes Themes Sub-themes Unconducive psychosocial environment Stigmatisation Stigmatisation by parents Name-calling by extended family members Exposure to abuse Sexual abuse by stepparents Abuse by partner Disruptions Noise Involvement in household chores Internet-related challenges Centralised internet Unaffordability of data Intermittent electricity supply Limited interaction Lack of opportunity to interact with classmates Classmates leaving the virtual space Immediate disconnection to save data No time to interact with lecturers Pre-recorded issue Sparing the data for the next lesson Unconducive psychosocial environment This superordinate theme highlights the psychosocial environment in which transgender students found themselves during the COVID-19 lockdown. It is composed of three themes: stigmatisation, exposure to abuse and disruptions. Exposure to abuse I do not think that I will pass (Nana). Besides the abuse, there are disruptions in the home environment which hinder transgender women from fully participating in online education. Stigmatisation This theme relates to the stigmatisation faced by transgender women while they were at home under lockdown: At home, my parents used to know me as a straight boy. But, since I left home to go to the university, I have realised that I am a transwoman and have to live as such. However, at home, the moment they saw me putting on my makeup, my father mentioned that, he has only one wife and a son, not a girl and request me to urgently wipe off the nonsense from my face. This make me so uncomfortable and disturbed my concentration during online lessons (Rudo). Besides stigmatisation by parents, transgender women are also stigmatised by other extended family members: Because I was on campus, I have bought a new wardrobe to dress comfortably. Now I no longer have any male clothes. As I have not disclosed my gender identity, every time 2021 Research in Social Sciences and Technology E-ISSN: 2468-6891 ressat.org Research in Social Sciences and Technology Volume 6 Issue 2, 2021 117 Mavhandu-Mudzusi, A.H. et al. Transgender student experiences of online education during COVID-19 pandemic era in rural Eastern Cape area of South Africa: A descriptive phenomenological study my cousins see me dressing up, they call me names like, sister, queen and mammy. I cannot even dress up without being laughed at But this traumatised me making me to be unable to concentrate with my school work, unlike when I am on campus where I dress freely without being stigmatised. In fact, on campus they are used to me. It is not even an issue of how I dress because we are so many (Thoby). Apart from the stigmatisation experienced in the home environment, which affect transgender students’ concentration, participants also reported the issue of abuse. Exposure to abuse This theme focuses on the abuse experienced by transgender women because of being at home and being expected to attend online classes. The transgender students complained of different forms of abuse. One of the participants mentioned the abuse she experienced from a stepparent: Since I passed Grade 12, and joined the university, at least I was away from my stepfather. Even during holidays, I did not go home but remain in the hostel, or visit a friend. But now, with Covid-19 and online education, I was forced to come back home. My stepfather continues to molest me. The problem is that I do not want to make a scene, because my mother loves him, and she is at least happy for the first time in her life, after she was physically abused by my biological father because of my gender identity. But I know that, as soon as my mother leaves home, he will sexually abuse me. I cannot concentrate when I study as I know that, anytime he can come and drag me to his room and force himself on me (Belinda). The abuse experienced comes not only from parents but also from sexual partners, as highlighted in the following excerpt: When I left secondary school, I started to live openly as a transwoman. My parents became aware when my sister show them my picture and they said that I should not come home, as they have given birth to a “boy”, so they do not want to see me anyhow. I therefore had to move in with my boyfriend, who is a gay. He is very abusive. He expected me to do what he wants in his own time although I had to join the online class. I have missed several virtual classes. I do not think that I will pass (Nana). When I left secondary school, I started to live openly as a transwoman. My parents became aware when my sister show them my picture and they said that I should not come home, as they have given birth to a “boy”, so they do not want to see me anyhow. I therefore had to move in with my boyfriend, who is a gay. He is very abusive. He expected me to do what he wants in his own time although I had to join the online class. I have missed several virtual classes. Centralised internet Access to the network is a problem for some of the transgender students. This is due to a lack of network access in certain areas, as shown by the following statement: When I at home, because it is in rural village, I cannot access internet. For me to get the internet access, I am expected to go to the shopping mall. But this is impossible due to lockdown. And even when I am at the shopping mall, I can only download tutorials and save. I cannot join the classes due to noise. I cannot even write examination at the mall (Rudo). Besides the issues relating to localisation of internet in certain areas, some participants cannot even afford to buy data for accessing the internet. Internet-related challenges This superordinate theme focuses on the issues related to internet which influence the accessibility of transgender women access to online education. Results indicate that transgender women face diverse challenges related to internet access. The challenges range from the internet only being accessible in a centralised place, the unaffordability of data and unstable internet connections. Disruptions In addition to the noise that hinders transgender women, some of the participants were expected to help out with the household chores: Nobody understand that I am at school. They expect me to do the house chores daily. I must cook, clean the house and do laundry. If there is shortage of something, they will send me to town to buy. When they need electricity, they send me. When I tell them that I am supposed to be in class they do not understand. They assume that I just want to be playing with my laptop. I end up cooking while my laptop is on (Fanisa). Nobody understand that I am at school. They expect me to do the house chores daily. I must cook, clean the house and do laundry. If there is shortage of something, they will send me to town to buy. When they need electricity, they send me. When I tell them that I am supposed to be in class they do not understand. They assume that I just want to be playing with my laptop. I end up cooking while my laptop is on (Fanisa). Besides disruptions at home caused by noise or involvement in household chores, participants raised the issue of internet access as a particular hindrance to online education by transgender women. Disruptions This theme highlights the disruptions experienced by transgender women when trying to engage in online education at home. The disruptions include noise in the environment and expectations to carry out household chores. One of the participants highlighted the inability to concentrate on online classes due to noise, as indicated in the following quotation: My uncle likes very loud music. I cannot even tell him to reduce the volume as it is his home. I therefore try to study but cannot concentrate. I cannot join the classes due to loud noise (Bessy). 2021 Research in Social Sciences and Technology E-ISSN: 2468-6891 ressat.org Research in Social Sciences and Technology Volume 6 Issue 2, 2021 Mavhandu-Mudzusi, A.H. et al. Transgender student experiences of online education during COVID-19 pandemic era in rural Eastern Cape area of South Africa: A descriptive phenomenological study 118 The noise emanates not only from the home but also from the neighbourhood: Research in Social Sciences and Technology Volume 6 Issue 2, 2021 118 Mavhandu-Mudzusi, A.H. et al. Transgender student experiences of online education during COVID-19 pandemic era in rural Eastern Cape area of South Africa: A descriptive phenomenological study The noise emanates not only from the home but also from the neighbourhood: The noise emanates not only from the home but also from the neighbourhood: As most of the people are at home, my neighbours play loud music from morning up to sunset. This is so disturbing. I wish we could be allowed to go back to the university. We shall do the lockdown in the hostel, because online classes at home are not even feasible. Unfortunate I cannot even tell anyone to keep quite as a transgender person, we are voiceless in rural community. Any irritation to the community may be calling death to myself (Belinda). As most of the people are at home, my neighbours play loud music from morning up to sunset. This is so disturbing. I wish we could be allowed to go back to the university. We shall do the lockdown in the hostel, because online classes at home are not even feasible. Unfortunate I cannot even tell anyone to keep quite as a transgender person, we are voiceless in rural community. Any irritation to the community may be calling death to myself (Belinda). Limited interaction This superordinate theme focuses on the limited interaction inherent to online education. Participants raised concerns that online education limits their interaction. The interaction highlighted refers to interaction with fellow students and with the lecturers. Unaffordability of Wi-Fi data Being at home is an issue for some of the participants as they do not have money to buy data. This is compounded by the problem of a lack of money-making opportunities due to the COVID- 19 lockdown: When I am at school I do not have a problem of accessing data. The university give us a bit of data, which cannot even last for a week. This means that I should go and buy data. 2021 Research in Social Sciences and Technology E-ISSN: 2468-6891 ressat.org Research in Social Sciences and Technology Volume 6 Issue 2, 2021 119 Mavhandu-Mudzusi, A.H. et al. Transgender student experiences of online education during COVID-19 pandemic era in rural Eastern Cape area of South Africa: A descriptive phenomenological study But I do not have money. Because of lockdown and fear of Covid-19 infection, I cannot even hassle. Before Covid-19 pandemic, I could do the household chores for other people and get paid. Sometimes I would exchange sex for money and can buy data and whatever I need. But now I am stuck. I do not have any plan. I am missing a lot at school (Belinda). But I do not have money. Because of lockdown and fear of Covid-19 infection, I cannot even hassle. Before Covid-19 pandemic, I could do the household chores for other people and get paid. Sometimes I would exchange sex for money and can buy data and whatever I need. But now I am stuck. I do not have any plan. I am missing a lot at school (Belinda). Besides the unaffordability of data, there is the challenge of the instability of the internet connection. Unstable internet Fortunately, I have Wi-Fi network, but the quality is not good. Most of the time it is difficult to hear what the lecturer is saying. Sometimes I get disconnected in the middle of session (Oniah). Fortunately, I have Wi-Fi network, but the quality is not good. Most of the time it is difficult to hear what the lecturer is saying. Sometimes I get disconnected in the middle of session (Oniah). Unstable internet is also related to an intermittent electricity supply. Unstable internet is also related to an intermittent electricity supply. Intermittent electricity supply Participants mentioned that they sometimes fail to access the internet due to a lack of continuity in the electricity supply which is mostly in the form of intermittent load shedding: The online test and examination are challenging. We need to adjust. With the network, we get frustrated because of load shedding. If I fail to finish due to network, I must write another test, which becomes very unfair (Fanisa). Load shedding not only affects internet access, but some student cannot even start their computers due to lack of power: Load shedding not only affects internet access, but some student cannot even start their computers due to lack of power: There is load shedding most of the time. When there is load shedding, the network also disappears. I end up failing to connect. So, I usually miss lessons. Sometimes we spend more than 24 hours without internet and even electricity. This means that you cannot even study. Load shedding is rare while we are on campus because there is even backup of electricity in the form of generators (Metsi). There is load shedding most of the time. When there is load shedding, the network also disappears. I end up failing to connect. So, I usually miss lessons. Sometimes we spend more than 24 hours without internet and even electricity. This means that you cannot even study. Load shedding is rare while we are on campus because there is even backup of electricity in the form of generators (Metsi). Besides the internet-related challenges, participants also raised the issues of limited interaction. Limited interaction Inability to interact with fellow students Besides trying to spare data, some participants feel uncomfortable about initiating online communication with students they do not know: Besides trying to spare data, some participants feel uncomfortable about initiating online communication with students they do not know: The problem for online thing is that you cannot ask anyone as you do not know the people. It is better to be taught physically in class. Online learning without knowing people is difficult as we do not even know their attitudes towards us transgender people. (Musi) Apart from the failure to discuss with the fellow classmates, participants could not interact freely with the lecturer for clarification. Apart from the failure to discuss with the fellow classmates, participants could not interact freely with the lecturer for clarification. Inability to interact with fellow students Participants mentioned that online education hinders them from interacting with fellow students. One the reasons for not interacting with other students is that after formal lectures, their classmates leave the virtual space. We fail to discuss the content because there is no proper interaction with colleagues. Immediately after the online session, everybody leave the online platform. Unlike with the face-to-face lecture where when the educator leaves the class, we remain discussing until our next class start. We can even go to the student centres, or our 2021 Research in Social Sciences and Technology E-ISSN: 2468-6891 ressat.org Research in Social Sciences and Technology Volume 6 Issue 2, 2021 Mavhandu-Mudzusi, A.H. et al. Transgender student experiences of online education during COVID-19 pandemic era in rural Eastern Cape area of South Africa: A descriptive phenomenological study 120 residential areas and continue with the discussions. But now, such an opportunity is not there (Rudo). Research in Social Sciences and Technology Volume 6 Issue 2, 2021 120 Mavhandu-Mudzusi, A.H. et al. Transgender student experiences of online education during COVID-19 pandemic era in rural Eastern Cape area of South Africa: A descriptive phenomenological study residential areas and continue with the discussions. But now, such an opportunity is not there (Rudo). residential areas and continue with the discussions. But now, such an opportunity is not there (Rudo). Participants further mentioned that interaction online is also determined by data affordability as shown by the following statement: Sometimes people might have an opportunity of joining the discussion groups organised by one of the student in class. But because we want to spare data as much as possible, we do not join such discussions so that we can be able to attend the formal classes. With this online system, you find that data can be exhausted quickly (Belinda). Sometimes people might have an opportunity of joining the discussion groups organised by one of the student in class. But because we want to spare data as much as possible, we do not join such discussions so that we can be able to attend the formal classes. With this online system, you find that data can be exhausted quickly (Belinda). Limited time to interact with the lecturer This theme focuses on the participants’ inability to interact with the lecturer. Failure to interact was due to pre-recorded presentations and data-related issues. The participants mentioned that most of the lessons are pre-recorded: The lessons are pre-recorded making it difficult to ask the question or interrupt the lecturer during the lesson. This is a problem because, sometimes the whole lesson ends without me understanding anything. It is very difficult to ask the questions directly just using chat (Rudo). The lessons are pre-recorded making it difficult to ask the question or interrupt the lecturer during the lesson. This is a problem because, sometimes the whole lesson ends without me understanding anything. It is very difficult to ask the questions directly just using chat (Rudo). In other instances, the student failed to interact because of the types of platform used Most of the time during online lessons, it is only the facilitator’s platform which is active, we cannot interrupt if there is something you do not understand. For you to be given chance, is at the end of the lesson where the facilitators link controller can unmute your mike, but the time is limited, because some platforms just close when the scheduled lesson time is over (Thoby). In addition to the way in which lessons are provided, participants also referred to data as a challenge to continue interacting with the lecturer when the formal lesson is over. In addition to the way in which lessons are provided, participants also referred to data as a challenge to continue interacting with the lecturer when the formal lesson is over. We would like to interact and ask more questions to the facilitators after the formal lesson. But when one think of the data, we just disconnect. Otherwise data will get depleted before the next lesson (Metsi). Discussion The advent of COVID-19 seems to have intensified the psychosocial challenges that transgender students had been going through prior to the pandemic. These challenges related mainly to The advent of COVID-19 seems to have intensified the psychosocial challenges that transgender students had been going through prior to the pandemic. These challenges related mainly to Research in Social Sciences and Technology E-ISSN: 2468-6891 ressat.org 2021 Research in Social Sciences and Technology Volume 6 Issue 2, 2021 Mavhandu-Mudzusi, A.H. et al. Transgender student experiences of online education during COVID-19 pandemic era in rural Eastern Cape area of South Africa: A descriptive phenomenological study 121 unconducive home environments where stigmatisation, abuse and disruptions were experienced. Research in Social Sciences and Technology Volume 6 Issue 2, 2021 121 Mavhandu-Mudzusi, A.H. et al. Transgender student experiences of online education during COVID-19 pandemic era in rural Eastern Cape area of South Africa: A descriptive phenomenological study unconducive home environments where stigmatisation, abuse and disruptions were experienced. unconducive home environments where stigmatisation, abuse and disruptions were experienced. The findings have indicated that the psychosocial environment where the transgender women find themselves in, especially during lockdown, is not conducive for them to attend online education. The environment includes their homes. Participants mention that, when they are on campus, it was easier for them to live freely and openly as transgender. But because of lockdown, participants live under pressure because their parents forced them to act as a “straight individual”. This situation affects the transgender women psychologically, as they are traumatized, making them to fail to concentrate during online education. The study findings indicate that, though online education is one of the innovative ways to ensure that students continue to have access to education as stated by Daniel (2020) and Dube (2020), the situation was different for transgender students in the rural areas of the Eastern Cape, South Africa during CODID-19 lockdown. Furthermore, for transgender women, going back to their home environment exposes them to sexual abuse, which they have previously averted by being on campus. But currently, being back at home makes them to be abused by people such as their step fathers. Having to suffer such sexual abuse, it affected their concentration during the online education. The challenge is wanting to put other’s happiness first, as person like Belinda chose to suffer in silence in order to protect her mother’s relationship. Research in Social Sciences and Technology E-ISSN: 2468-6891 ressat.org Discussion This meant that no matter how much their concentration levels were disturbed, lockdown meant that they were locked down to their homes without the many alternatives they would have had outside the pandemic. The reluctance of transgender students to confront relatives and neighbours due to fear of stigma and discrimination attest to findings by Enke, (2016) mentioning that transgender individuals face a lot of challenges based on how they behave or express themselves which is contrary to what society expects from a specific gender. Besides being disturbed by noise, people at home expected transgender women to be involved also on the house chores such as cooking, cleaning the house and also running all the family errands like shopping. People at home did not understand that students can attend school at home. They consider non-involvement to home chores while a person has not gone to class as laziness. This situation prohibited transgender women from being engaged fully to online education. In addition, the fact that the participants were required to stop participating in online classes so that they could attend to household chores meant added stress. This was mainly due to the mostly African perspective where certain household chores are gendered. Because transgender students are different and express themselves differently from the expectations of society in relation to specific genders (Enke, 2016), they had to deal with the double challenge of missing classes and conforming to expectations that did not speak to what they wanted to do. Such behavior of conforming to societal expectations concurs with findings by Hentschel, et al., (2019) that transgender women are less assertive and have poor leadership skills. The situation of forcing all the students to go home during COVID-19 lockdown, and initiating online education without considering the environmental factors of students such as transgender and other marginalized students showed limitation on the part of universities to provide adequate support to students as according to Williams et al., (2021) and Mailizar et al., (2020), student support also include ensuring conducive learning environments for the students. Another major environmental factor which affected transgender women from benefiting from online education, is the challenges related to technical issues. Discussion The situation prohibits her from concentrating to the studies. These findings attest to Enke (2016), Omercajic and Wayne (2020), Wayne et al. (2020), mentioning that transgender students experience diverse challenges including discrimination, harassment and other hostile treatments. Furthermore, such decision to stay in the abusive relations aligns with findings by Hentschel, Heilman and Peus (2019) that transgender women are less assertive than male counterparts. As some transgender women were living openly on campus, but “pretending to be cisgender men” at home, going back home makes them to be “outed” by some of their relatives who have seen them such as siblings. The situation caused people like Nana to be disowned by the parents because of gender identity. That compelled her to move out and go and stay with the sexual partner. Unfortunately, the sexual partner was also abusive, not allowing her to join online classes. That made her to miss several online classes. Apart from direct emotional and sexual abuse experienced by some of the transgender women which negatively affected their engagement to online education, findings indicate that there was noise in the environment. The noise was either from loud volume of music from home which they have no control of or from neighbours. People like Bessy and Belinda failed to concentrate during the online classes because of loud music played at home and also by neighbours. Due to voicelessness of transgender women because of fear of stigma and discrimination based on how the gender nonconforming individuals are viewed in the community, transgender women could not request relatives or neighbours to lower their volume. While these challenges are also generally experienced by employees working from home (Palumbo, 2020) and other students (Adnan & Anwar, 2020; Dube, 2020, Daniels et al., 2019), the transgender students’ plight was particularly hard because of the existing tensions related to the stigma of their sexuality. For example, when the transgender students were disturbed by noise in their homes, they had few places to escape to since most faced hostility 2021 Research in Social Sciences and Technology Volume 6 Issue 2, 2021 122 Research in Social Sciences and Te Volume 6 Issue 2, 2021 Mavhandu-Mudzusi, A.H. et al. Transgender student experiences of online education during COVID-19 pandemic era in rural Eastern Cape area of South Africa: A descriptive phenomenological study from their neighbours and communities. Research in Social Sciences and Technology E-ISSN: 2468-6891 ressat.org Discussion The other source of her income was transactional sex, where she would provide sexual services to clients in order to get money. Because of COVID-19 lockdown restrictions, she could not go to places like brothels, restaurants, shebeens or any places where she can find the clients. The situation hinders her from accessing the online education as she cannot afford to buy data. The lack of financial resources by transgender students attest to the finding by Kia et al., (2020) and Suissa and Sullivan, (2021) indicating that transgender populations face adverse conditions such as poverty. For other transgender women, the challenge was not affordability of data, but the issue was that the internet connectivity was unstable, thus, hindering them from joining the online classes. Even when sometimes they manage to connect, poor quality of network prohibits them from hearing what is said in the class, or continually disconnected. This situation makes them not to benefit from online education. The situation of not having data to access education indicates that the institutions of higher education, though they initiated online education, did not provide the necessary tools for marginalised and economically disadvantaged students to access education. Not providing data is in contrast with the recommendation by Williams et al., (2021) and Waters and Hensley (2020) who stated that access to learning is anchored on student technological support and use of various opportunities available and preferred by students. The unstable internet was exacerbated by intermittent supply of electricity. Transgender women mention that they failed to join online classes due to electricity load shedding. This is the situation where the electricity will be switched off for at least two hours or even up to 24 hours in some areas as the country’s means of rationing electricity supply. During load shedding, the internet is also affected, which blocks the students from joining online classes. This is the situation which, when students are on campus, they rarely experience. This situation means that they fail to connect to internet. They could also not read the downloaded materials as for most of the laptops used by students, the batteries can last only for four hours if not connected to the electricity source. Discussion The technical challenges in this study relate to the affordability and availability of the internet as well as the intermittent availability of electricity due to load shedding (a system by the South African power utility where electricity is cut at intermittent intervals to lessen the load or demand for the dwindling power supply). Participants raised the issue of being unable to access internet, as the area where they stay, there is no internet. Internet access is only found in the centralised place such as shopping complex. This means that, for transgender women to be able to access online education, they have to go to the nearby malls. However, because of lockdown regulations, they cannot go to the mall to access class or even to download educational materials. That situation also makes students such as Rudo, who stays in the internet-less area to even fail to write the examination as it was supposed to be conducted online. Failure to stay connected to internet defeats the requirement of online education which requires students to stay connected in order to learn as indicated by Goldschmidt (2020). In some instances, the challenge was not the unavailability of internet, but the price of data which students could not afford due to lack of money. The lack of money became the issue for transgender women like Belinda, whose source of income was piece job where she would go 2021 Research in Social Sciences and Technology Volume 6 Issue 2, 2021 123 esearch in Social Sciences and T Volume 6 Issue 2, 2021 Mavhandu-Mudzusi, A.H. et al. Transgender student experiences of online education during COVID-19 pandemic era in rural Eastern Cape area of South Africa: A descriptive phenomenological study to people’s houses to assist them with laundry and other household chores so that she can get paid. The other source of her income was transactional sex, where she would provide sexual services to clients in order to get money. Because of COVID-19 lockdown restrictions, she could not go to places like brothels, restaurants, shebeens or any places where she can find the clients. The situation hinders her from accessing the online education as she cannot afford to buy data. The lack of financial resources by transgender students attest to the finding by Kia et al (2020) to people’s houses to assist them with laundry and other household chores so that she can get paid. Research in Social Sciences and Technology E-ISSN: 2468-6891 ressat.org Discussion Lack of the interaction with educators and fellow-students denies the students from benefiting from the human–human interaction, which according to the theoretical framework used in this study, is important not only for socialising and affective outlet, but also meant for cognitive activities such as discussions, explanations, cognitive exchanges, argumentation and the negotiation of meaning (Kumar et al., 2021; Saxena, Baber & Kumar, 2020). This is because of immediate disconnecting from virtual space as soon as the lecturer finished the presentation. This is different from what they used to do following face to face learning, where after formal lesson, students will remain interacting with each other, and even take discussions further in the student learning centres. The follow-up discussion made the participants to seek clarification and also more information regarding the lessons. But now, as all students want to conserve data for future lessons, there is no time to remain online and form discussion groups. Lack of the interaction with educators and fellow-students denies the students from benefiting from the human–human interaction, which according to the theoretical framework used in this study, is important not only for socialising and affective outlet, but also meant for cognitive activities such as discussions, explanations, cognitive exchanges, argumentation and the negotiation of meaning (Kumar et al., 2021; Saxena, Baber & Kumar, 2020). The other challenge is that participants do not have time to know each other, making it difficult to identify the person they can contact for further discussions. They cannot just blindly contact any person based on attendance list as they are not sure of the person’s attitudes towards transgender due to fear of being victimised or exposed to violence. The situation of transgender students’ reluctance to engage in group discussion is not unique to the COVID era, but concurs with Omercajic and Wayne (2020) documenting that transgender students are usually excluded from discussions and decisions regarding access to education. Some of the lessons are pre-recorded, making it difficult for the students to interject during the lesson to ask questions. This means that, if the learner fails to understand specific aspects in the beginning of the lecture, it can affect the understanding of the entire lesson. Which is not there while students are in physical class as they can raise their hands and ask for clarification immediately. Discussion While lack of data and poor internet connectivity are challenges which ultimately cause uncertainty for virtual learning for many students in South Africa (Dube, 2020), the transgender students in this study were particularly affected because they face adverse conditions and poverty (Kia et al., 2020; Daniels et al; 2019), which affected their ability to afford data to enable them to participate in online activities such as completing assessment tasks. To this end, the participants noted that if they had been on campus, their studies would not have been affected by data and electricity limitations. This does not negate the literature which exposes the discrimination and harassment that these students face on campuses (Goldberg, 2018), but it seems campuses offer more learning opportunities without the restrictions of technology. While the unconducive psychosocial environment and technical challenges were expressed emphatically in this study, it became clear that at the core of the struggles with online learning for transgender students was a crucial gap in interaction in its various facets. The lockdown regulations prohibited the students from interacting in person for discussions with lecturers and other students, towards understanding the content. Participants raised the challenge inherent to how online education is structured such as limited interaction with fellow students. 2021 Research in Social Sciences and Technology Volume 6 Issue 2, 2021 124 Research in Social Sciences and Te Volume 6 Issue 2, 2021 Mavhandu-Mudzusi, A.H. et al. Transgender student experiences of online education during COVID-19 pandemic era in rural Eastern Cape area of South Africa: A descriptive phenomenological study This is because of immediate disconnecting from virtual space as soon as the lecturer finished the presentation. This is different from what they used to do following face to face learning, where after formal lesson, students will remain interacting with each other, and even take discussions further in the student learning centres. The follow-up discussion made the participants to seek clarification and also more information regarding the lessons. But now, as all students want to conserve data for future lessons, there is no time to remain online and form discussion groups. Research in Social Sciences and Technology E-ISSN: 2468-6891 ressat.org Discussion This is in contrast with Moore (1989), mentioning that successful teaching and learning environment involves bi-directional interaction of various kinds. It also contravenes interaction theory, as encapsulated in the theory of transactional distance (Moore, 1993). The other hinderance to interaction during online education is the type of platforms utilised where only the lecturer’s platform is active during online teaching while all the attendees are muted. This disables participants from interacting with the facilitators. The only time the link controller unmutes the participants is towards the end of the lesson, when there is limited time to interact because those platforms automatically close when the scheduled lesson time is over. Therefore, these findings where students fail to fully access education based on the platforms used and saving data are in contrast with the assertion by Kumar, Saxena, and Baber (2021) and Dube (2020) that the use of online education increases flexibility and access to education. In this study, transgender women prefer onsite physical teaching unlike what is recommended by Kumar et al. (2021) and Dube (2020) mentioning that the use of technology can reduce the need for onsite teaching. It stands to reason, then, that technology platforms have been the platforms of choice for education under constraints. However, the issue of trying to save data also affects the continuous interaction during online teaching as students disconnect as soon as the formal presentation is over, so that they can still have data for the next session. All the above situations make transgender women to continually experience challenges in accessing education. 2021 Research in Social Sciences and Technology Volume 6 Issue 2, 2021 125 esearch in Social Sciences and T Volume 6 Issue 2, 2021 Mavhandu-Mudzusi, A.H. et al. Transgender student experiences of online education during COVID-19 pandemic era in rural Eastern Cape area of South Africa: A descriptive phenomenological study In other words, lockdown restrictions affected student–student, student–lecturer and student– content interaction (Moore, 1989; 1993). Instead of in-person discussions with lecturers and group discussions with peers, students had to rely on virtual meetings that were limited by time, technology, data and electricity. The transgender students, thus, preferred to attend physical classes, rather than virtual learning. Limitations Because of the COVID-19 lockdown regulations, in-depth interviews were conducted telephonically. This hindered the researchers from observing some of the non-verbal cues which may have assisted in describing and understanding the data. The data collection was also affected by the Wi-Fi network, as sometimes during the interviews the participants were disconnected or became inaudible. Then, when attempting to resume the interview, some participants were no longer available. Owing to telephonic interviews and limited internet access, participants were given information-related research and other ethical aspects verbally to obtain informed consent. Regardless of these limitations, the researchers did everything they could to ensure trustworthiness and adhere to all ethical principles. As this study followed descriptive phenomenology, the researchers suggest a future quantitative study, comparing the transgender students’ view of online education between rural and urban areas to check if online education is not worsening the gap between the urban and rural transgender students. Discussion This calls into question, then, the growing “reliance on technology” (Goldschmidt, 2020: 88), assertions of technology as widening access to education (Kumar et al., 2021; Dube, 2020) and the view of technology as “an infallible means to develop profound knowledge and wide range of language skills” (Ahmad, 2016: 91). While these technology affordances are true, it seems there is a need for conscious adjustments that will increase opportunities for bi-directional interaction. This requires creativity on the part of the teachers, institutional support and, probably, government support. What is clear is that technology should be used so that students can have equal and equitable opportunities to education (EdGlossary, 2014). The negative experiences of students regarding online education continuously violate transgender students’ access to education, even when scholars such as Omercajic and Wayne (2020) regard accessing education for transgender an issue of equity and social justice. Research in Social Sciences and Technology E-ISSN: 2468-6891 ressat.org Conclusion The issue of online education seems to be a challenge for all the students, especialy some of the transgender students during COVID-19 lockdown. Apart from the psychosocial environment where transgender students face stigma and discrimination, some of the challenges faced by these students are not limited to them alone, but appear to apply to other students in similar settings, in particular the issue related to internet access and limited interaction. However, if the situation is not addressed with the main focus on marginalised groups such as transgender students, the access to education for this population could be continuously compromised. This may further lead to the worsening of marginalisation and perpetuating inequality, poverty and further exploitation among transgender individuals and scourge of gender based violence. To avert the above situation, the researchers therefore recommend the following: Relentlessly fighting for the betterment of the education and lives of transgender through increased awareness of the community in relation to transgender individuals to increase the understanding of transgender issues. This might increase the 2021 Research in Social Sciences and Technology Volume 6 Issue 2, 2021 126 search in Social Sciences and T Volume 6 Issue 2, 2021 Mavhandu-Mudzusi, A.H. et al. Transgender student experiences of online education during COVID-19 pandemic era in rural Eastern Cape area of South Africa: A descriptive phenomenological study acceptance of the transgender individual by the family and community members. Regarding internet access, institutions offering online education should ensure that students are provided with adequate data. When there is lockdown, the students could be given the choice to remain on campus if their home environment is not conducive to learning, with measures in place to comply with health requirements. https://doi.org/10.1016/j.ijedro.2020.100011. Ahmad, J. (2016). Technology assisted language learning is a silver bullet for enhancing language competence and performance: A case study. International Journal of Applied Linguistics and English Literature, 5(7), 118–131. Ahmad, J. (2016). Technology assisted language learning is a silver bullet for enhancing language competence and performance: A case study. International Journal of Applied Linguistics and English Literature, 5(7), 118–131. Daniel, J. (2020). Education and the COVID-19 pandemic. Prospects, 49(1), 91–96. Daniels, J., Struthers, H., Maleke, K., Catabay, et al. (2019). Rural school experiences of South African gay and transgender youth. Journal of LGBT Youth, 16. 1-25. https://www.doi.org/10.1080/19361653.2019.1578323. References Adnan, M., & Anwar, K. (2020). Online learning amid the COVID-19 pandemic: Students' perspectives. Online Submission. Journal of Pedagogical Sociology and Psychology, 2(1), 45–51. Aguilera-Hermida, A. P. (2020). College students’ use and acceptance of emergency online learning due to Covid-19. International Journal of Educational Research Open, 1. h //d / / d https://www.doi.org/10.1080/19361653.2019.1578323. Dube, B. (2020). Rural online learning in the context of Covid 19 in South Africa: Evoking an inclusive education approach. Multidisciplinary Journal of Educational Research, 10(2), 135–157. EdGlossary. (2014). https://www.edglossary.org/access/ Enke, A. (2016). Introduction: Transfeminist perspective in and beyond transgender and gender studies, In A. Enke (Ed). 1–15. Philadelphia, PA: Temple University Press. Flick, U. (Ed). 2018. SAGE handbook of qualitative data collection. Los Angeles, CA: Sage Publications. Goldschmidt, K. (2020). The COVID-19 pandemic: Technology use to support the wellbeing of children. Journal of Pediatric Nursing. Goldberg, A.E. (2018). Transgender students in higher education. California: UCLA. https://www.escholarship.org/content/qt4p22m3kx/qt4p22m3kx.pdf. Gray, J., Grove, S.K,. & Sutherland, S. (2017). Burns and Grove's the practice of nursing research: Appraisal, synthesis, and generation of evidence, 8th edition. St. Louis, MO: Elsevier. Hentschel, T., Heilman M. E., & Peus C. V. (2019). The Multiple Dimensions of Gender Stereotypes: A Current Look at Men’s and Women’s Characterizations of Others and Themselves. Frontiers in Psychology, 10, 11. https://.doi.org/10.3389/fpsyg.2019.00011. Kia, H., Robinson, M., MacKay, J., & Ross, L.E. (2020). Poverty in lesbian, gay, bisexual, transgender, queer, and two-spirit (LGBTQ2S+) populations in Canada: An intersectional review of the literature. Journal of Poverty and Social Justice, 28(1), 21– 54. 2021 Research in Social Sciences and Technology E-ISSN: 2468-6891 ressat.org Research in Social Sciences and Technology Volume 6 Issue 2, 2021 127 Mavhandu-Mudzusi, A.H. et al. Transgender student experiences of online education during COVID-19 pandemic era in rural Eastern Cape area of South Africa: A descriptive phenomenological study Kincheloe, J.L. (2005). On to the next level: Continuing the conceptualization of the bricolage. Qualitative Inquiry, 11(3), 323–350. Kincheloe, J.L. (2005). On to the next level: Continuing the conceptualization of the bricolage. Qualitative Inquiry, 11(3), 323–350. Kumar, P., Saxena, C. & Baber, H. (2021). Learner-content interaction in e-learning- the moderating role of perceived harm of COVID-19 in assessing the satisfaction of students. Smart Learn. Environ. 8- 5, https://doi.org/10.1186/s40561-021-00149-8. Mailizar, Almanthari, A., Maulina, S., & Bruce, S. (2020). Secondary school mathematics teachers’ views on e-learning implementation barriers during the Covid-19 pandemic: The case of Indonesia. Eurasia Journal of Mathematics, Science and Technology Education, 16(7), em1860. Moore, M.G. 1989. Editorial: Three types of interaction. American Journal of Distance Education, 3(2), 1–7. https://doi.org/10.1080/08923648909526659. Moore, M.G. (1993). Theory of transactional distance. In D. Keegan (Ed.), Theoretical principles of distance education, 22–38. London: Routledge. Moser, A., & Korstjens, I. (2018). Series: Practical guidance to qualitative research. Research in Social Sciences and Technology E-ISSN: 2468-6891 ressat.org https://www.doi.org/10.1080/19361653.2019.1578323. Part 3: Sampling, data collection and analysis. European Journal of General Practice, 24(1), 9– 18. Morrow, R., Rodriguez, A., & King, N. (2015). Colaizzi’s descriptive phenomenological method. The Psychologist, 28(8), 643-644. Omercajic K., & Wayne, M. (2020). Supporting Transgender Inclusion and Gender Diversity in Schools: A Critical Policy Analysis. Frontiers in Sociology, 5. DOI? Palumbo, R. (2020). Let me go to the office! An investigation into the side effects of working from home on work-life balance. International Journal of Public Sector Management, 33(6/7), 771–790. Parents and Friends of Lesbians and Gays (PFLAG). (2004). Our trans children: A publication of the Transgender Network of Parents, Families, and Friends of Lesbians and Gays (PFLAG), 4th ed. Hardy, VA: PFLAG. Peat, G., Rodriguez, A., & Smith, J. (2019). Interpretive phenomenological analysis applied to healthcare research. Evid Based Nurs, 22(1),7-9. Https://www.doi:10.1136/ebnurs- 2018-103017. Preethi S. (2019). Remote consent clinical research. Clin Trial Pract Open J, 1(1), 39-41. Doi: 10.17140/CTPOJ-1-109. Republic of South Africa. Department of Health. (2020). COVID-19 clinical case management (Version 1-April 2020). Pretoria. Government Printers SANews. (2020). March 23. https://www.sanews.gov.za/south-africa/president-ramaphosa- announces-nationwide-lockdown.. Saxena, C., Baber, H., & Kumar, P. (2020). Examining the moderating effect of perceived benefits of maintaining social distance on e-learning quality during COVID-19 pandemic. Journal of educational technology systems, 49 (4), 532-554. Https://www.doi.org/10.1177/0047239520977798. Sloan, A., & Bowe, B (2014). Phenomenology and hermeneutic phenomenology: the philosophy, the methodologies and using hermeneutic phenomenology to investigate lecturers' experiences of curriculum design. Quality & Quantity, 48(3), 1291-1303. 2021 Research in Social Sciences and Technology E-ISSN: 2468-6891 ressat.org Research in Social Sciences and Technology Volume 6 Issue 2, 2021 128 Mavhandu-Mudzusi, A.H. et al. Transgender student experiences of online education during COVID-19 pandemic era in rural Eastern Cape area of South Africa: A descriptive phenomenological study Subedi, D., & Subedi, R. (2020). Practicing Self Learning of ICT for Resilience Amidst the COVID-19 Outbreak: Experiences from Kathmandu Valley. Research in Educational Policy and Management, 2(2), 78-96. https://doi.org/10.46303/repam.2020.5 Subedi, D., & Subedi, R. (2020). Practicing Self Learning of ICT for Resilience Amidst the COVID-19 Outbreak: Experiences from Kathmandu Valley. Research in Educational Policy and Management, 2(2), 78-96. https://doi.org/10.46303/repam.2020.5 Subedi, D., & Subedi, R. (2020). Practicing Self Learning of ICT for Resilience Amidst the COVID-19 Outbreak: Experiences from Kathmandu Valley. Research in Educational Policy and Management, 2(2), 78-96. https://doi.org/10.46303/repam.2020.5 Policy and Management, 2(2), 78 96. https://doi.org/10.46303/repam.2020.5 Suissa, J., & Sullivan, A. (2021), The Gender Wars, Academic Freedom and Education. https://www.doi.org/10.1080/19361653.2019.1578323. Journal of Philosophy of Education, 55: 55-82. https://doi.org/10.1111/1467-9752.12549. Waters, S., & Hensley, M. (2020). Measuring Rural P-12 Teachers’ Attitudes, Perceptions, and Utilizations of Social Media. Research in Social Sciences and Technology, 5 (3), 25-54. Suissa, J., & Sullivan, A. (2021), The Gender Wars, Academic Freedom and Education. Journal of Philosophy of Education, 55: 55-82. https://doi.org/10.1111/1467-9752.12549. Suissa, J., & Sullivan, A. (2021), The Gender Wars, Academic Freedom and Education. Journal of Philosophy of Education, 55: 55-82. https://doi.org/10.1111/1467-9752.12549. Waters, S., & Hensley, M. (2020). Measuring Rural P-12 Teachers’ Attitudes, Perceptions, and Utilizations of Social Media. Research in Social Sciences and Technology, 5 (3), 25-54. https://www.doi.org/10.46303/ressat.05.03.2. Waters, S., & Hensley, M. (2020). Measuring Rural P-12 Teachers’ Attitudes, Perceptions, and Utilizations of Social Media. Research in Social Sciences and Technology, 5 (3), 25-54. https://www.doi.org/10.46303/ressat.05.03.2. Wayne M., Kassen, J., & Omercajic, K. (2020). Supporting transgender students in schools: beyond an individualist approach to trans inclusion in the education system. Educational Review. https://www.doi.org/10.1080/00131911.2020.1829559. Williams, T.K., McIntosh, R.W., & Russell, W.B. (2021). Equity in distance education during COVID-19. Research in Social Sciences and Technology, 6(1), 1-24. https://doi.org/10.46303/ressat.2021.1. Williams, T.K., McIntosh, R.W., & Russell, W.B. (2021). Equity in distance education during COVID-19. Research in Social Sciences and Technology, 6(1), 1-24. https://doi.org/10.46303/ressat.2021.1. 2021
W2059500881.txt	https://zenodo.org/records/1593022/files/article.pdf	de		Angaben über eine möglichst vollkommene Isolation elektrischer Messinstrumente	Fresenius' Zeitschrift für Analytische Chemie	1,916	public-domain	482	52 Berieht: Allgemeine analytisehe Methoden etc. Belieben ein paralleles oder konvergent gemachtes Lichtstrahlenbiindel benutzt werden, das durch eine Blonde in den Glastrog fiillt, in welchen die zu untersuchende I,Ssung gef~llt wird. Der Trog ist durch Mikrometerschrauben innerhalb eines lichtdichten Kastens in horizontaler ~nd vertikaler Richtung beliebig verschiebbar. W. M e c k l e n b u r g hat mit dem yon S c h m i d t and H a e n s e h gebauten Apparat verschiedene Messungsreihenl) durchgeft~hrt and zuniichst an verschiedeu konzentrierten Zim~siuresolen gezeigt, dass man einen fiir die jeweils untersuchte LOsung cMrakteristischen Helligkeitswert erh~lt, wenn man unmittelbar unter der Fltissigkeitsoberfl~ehe einen T y n d a l l s t r e i f e n erzeugt and seine Helligkeit an mehreren, verschieden welt yon der Einfullsstelle des Lichtes entfernten Punkten misst, da sich dana dureh graphische Interpolation der erhaltenen Werte der schwiichende Einfluss tier L5sung eliminieren l~tsst. Weitere, mit einfarbigen, tells parullelen tells konvergenten Lichtstratden angestcllte Yersuche dienten zum Nachweise der Richtigkeit tier R a y l e i g h ' s c h e n Theorie for derartige Beugungserscheinungen; auch wurde das Instrument Zu Messungen betreffend die Abb~tngigkeit der Fluoreszenz yon der Konzentration 2) benutzt~ auf die indessen, ebenso wie auf die letztgenannten Arbeiten~ wegen ihres vorwiegend physikalischen Interesses bier nur hingewiesen werden kann. Angaben iiber eine m~gliohst vollkommene Isolation elektrischer Messinstrumente hat W a l t e r P. W h i t e ~) gemacht. Der Verfasser weist darauf hin~ dass zwar unter gewSbnlichen Umstiinden die iibliehe Art der Aufstellung eine hinreichende Iso]ierung der Messstromkreise sichere, dass aber bei feuchtem W e t t e r usw. unliebsame Fehler durch vagabundierende StrSme yon im Untersuchungsraum vorhandenen Lichtodor Kraftleitungen verursacht werden kSnnen, Hiergegen schfitzt man sieh vOllig, indem man auf den Tisehen and an allen anderen Stellen, an denen der Messstromkreis mit ~usseren Gegenst~nden in Bertihrung kommt, aus durch Driihte verbundenen Metallplatten einen Sehirm herstellt, tier iiberalt nahezu gleiches Potential hat, da die grossen Metallfl~chen nile Unterschiede sofort ausgleiehen. Eine derartige J~quipotentialfl~ehe h~lt bei richtiger Anlage alle fiusseren StSrungen ab, sie wirkt am siehersten, wenn sie nieht geerdet wird. Auf die weiteren ausftihrlichen Angaben des Verfassers, wie man durch Anbringen eines kastenartigen Schirmes auch eine Isolation gegen die aus etwa vorhandenen, auf mehr als 1300 o erbitzten elektrischen 0fen strSmende, :stark ionisierte Luft herbeiftihren kann, and wie man bei thermoelektrischen Messungen auch die ¥ergleichsbatterie entsprechend abschirmen kann~ - - beides VorsichtsmaSregeln, die nur bei ganz ausser, gew0hnlieh genauen and empfindlichen Messungen in Betraeht kommen diirften - - kann ich an dieser Stelle nur hinweisen. l) Kolloid-Ztschrft. lt. 172 (1913); 15.149 (1914); 16, 97 (1915).- ~) i~ e c kl e n, b u r g u V a l e n t i n e r , Physik. Ztschrft. 15, 267 (1914!. -- 3) Journ. Americ. Chem. Soc. ~6, 2011 (1914).
https://openalex.org/W4392237077	https://link.springer.com/content/pdf/10.1007/s11083-024-09663-z.pdf	English	null	Poset Ramsey Number $$R(P,Q_n)$$. II. N-Shaped Poset	Order	2,024	cc-by	13,121	B Christian Winter christian.winter@kit.edu Abstract Given partially ordered sets (posets) (P, ≤P) and (P′, ≤P′), we say that P′ contains a copy of P if for some injective function f : P →P′ and for any A, B ∈P, A ≤P B if and only if f (A) ≤P′ f (B). For any posets P and Q, the poset Ramsey number R(P, Q) is the least positive integer N such that no matter how the elements of an N-dimensional Boolean lattice are colored in blue and red, there is either a copy of P with all blue elements or a copy of Q with all red elements. We focus on the poset Ramsey number R(P, Qn) for a ﬁxed poset P and an n-dimensional Boolean lattice Qn, as n grows large. It is known that n + c1(P) ≤R(P, Qn) ≤c2(P)n, for positive constants c1 and c2. However, there is no poset P known, for which R(P, Qn) > (1 + ϵ)n, for ϵ > 0. This paper is devoted to a new method for ﬁnding upper bounds on R(P, Qn) using a duality between copies of Qn and sets of elements that cover them, referred to as blockers. We prove several properties of blockers and their direct relation to the Ramsey numbers. Using these properties we show that R(N, Qn) = n + (n/ log n), for a poset N with four elements A, B, C, and D, such that A < C, B < D, B < C, and the remaining pairs of elements are incomparable. Keyword Poset Ramsey · Boolean lattice · Induced subposet Keyword Poset Ramsey · Boolean lattice · Induced subposet Order (2024) 41:401–418 https://doi.org/10.1007/s11083-024-09663-z Order (2024) 41:401–418 https://doi.org/10.1007/s11083-024-09663-z ( ) https://doi.org/10.1007/s11083-024-09663-z 1 Karlsruhe Institute of Technology, Karlsruhe, Germany Maria Axenovich1 · Christian Winter1 Received: 30 November 2022 / Accepted: 15 February 2024 / Published online: 28 February 2024 © The Author(s) 2024 Poset Ramsey Number R(P, Qn). II. N-Shaped Poset Maria Axenovich1 · Christian Winter1 1 Introduction A partially ordered set, shortly a poset, is a set P equipped with a relation ≤P that is transitive, reﬂexive, and antisymmetric. For any non-empty set Z, let Q(Z) be the Boolean lattice of dimension \|Z\| on a ground set Z, i.e. the poset consisting of all subsets of Z equipped with the inclusion relation ⊆. We use Qn to denote a Boolean lattice with an arbitrary n-element ground set. We refer to a poset either as a pair (P, ≤P), or, when it is clear from context, simply as a set P. When clear from context, we shall write A ≤B instead of A ≤P B, and A < B when A ≤B and A ̸= B. When A and B are not comparable, we write A ∥B. The elements of P are often called vertices. A poset P1 is an (induced) subposet of P2 if P1 ⊆P2 and for every X1, X2 ∈P1, X1 ≤P1 X2 if and only if X1 ≤P2 X2. An (induced) copy of a poset P1 in P2 is an induced subposet P′ of P2, isomorphic to P1. We shall be considering three special posets: V, V , and 123 402 Order (2024) 41:401–418 Fig. 1 Hasse diagrams of posets V, V , and N Fig. 1 Hasse diagrams of posets V, V , and N N, see Fig. 1. The poset V has three vertices A, B, and C, C < A, C < B, and A ∥B. The poset V has three vertices A, B, and C, C > A, C > B, and A ∥B. The poset N has four vertices A, B, C, D and relations A < C, B < D, B < C, A ∥B, A ∥D, and C ∥D. Extremal properties of posets and their induced subposets have been investigated in recent years and mirror similar concepts in graphs. Carroll and Katona [4] initiated the consideration of so called Turán-type problems for induced subposets. Most notable is a result by Methuku and Pálvölgyi [16] which provides an asymptotically tight bound on the maximum size of a subposet of a Boolean lattice that does not have a copy of a ﬁxed poset P, for general P. Their statement has been reﬁned for several special cases, see e.g. Lu and Milans [13], and Méroueh [15]. 1 Introduction Further Turán-type results are, for example, given by Methuku and Tompkins [17], and Tomon [18]. Note that Turán-type properties are also investigated in depth for non-induced, so called weak subposets, which are not considered here. Besides that, saturation-type extremal problems are studied for induced and weak subposets, see a recent survey of Keszegh, et al. [12]. In this paper we are dealing with Ramsey-type properties of induced subposets in Boolean lattices. Consider an assignment of two colors, blue and red, to the vertices of posets. Such a coloring c: P →{blue, red} is a blue/red coloring of P. A colored poset is monochromatic if all of its vertices share the same color. A monochromatic poset whose vertices are blue is called a blue poset. Similarly deﬁned is a red poset. Extending the classical deﬁnition of graph Ramsey numbers, Axenovich and Walzer [1] introduced the poset Ramsey number which is deﬁned as follows. For posets P and Q, let R(P, Q) = min{N ∈N: every blue/red coloring of QN contains either a blue copy of P or a red R(P, Q) = min{N ∈N: every blue/red coloring of QN contains either a blue copy of P or a red co Further known bounds on poset Ramsey numbers include results of Chen et al. [6], [7], Chang et al. [5] as well as Falgas-Ravry et al. [10]. Further known bounds on poset Ramsey numbers include results of Chen et al. [6], [7], Chang et al. [5] as well as Falgas-Ravry et al. [10]. It is unknown whether there exists a poset P such that R(P, Qn) ≥(1 + c)n for some c > 0. Therefore it is natural to consider the value of R(P, Qn) −n and determine its asymptotic behaviour. We say that a tight bound on R(P, Qn) is a function f (n) such that R(P, Qn) = n + ( f (n)). A tight bound is only known for a handful of posets, see for example Theorem 2, and Winter [21]. A poset is trivial if it does not contain a copy of either V or V . Otherwise we refer to it as non-trivial. For trivial posets P the trivial lower bound is asymptotically tight. Theorem 3 (Axenovich-Winter [2]) If P is a trivial poset, then R(P, Qn) = n + c(P) where c(P) is a constant only depending on P. If P is a non-trivial poset, then R(P, Qn) ≥ R(V, Qn) ≥n + n 15 log n . For non-trivial posets P, there are only two known approaches for upper bounds on R(P, Qn). The ﬁrst one was introduced by Grosz, Methuku and Tompkins [11] for an upper bound on R(Q2, Qn) and is based on the following idea. In a blue/red coloring of the host lattice, there is either a red copy of Qn (and we are done) or there are many blue chains. Then these chain counting arguments can be applied for ﬁnding a monochromatically blue structure. An alternative approach is given in [2] by the present authors for proving an upper bound on R(V, Qn). With a careful analysis of the blue subposet of a hosting lattice with forbidden red Qn one can obtain much more information than the existence of many chains. In this paper we will elaborate on the second approach and formulate the central, intermediate step as a theorem for general P. This approach involves so-called blockers, posets that contain a vertex from each copy of Qn from a special, easier to analyse subclass. R(P, Q) = min{N ∈N: every blue/red coloring of QN contains either a blue copy of P or a red co One of the central questions in this area is to determine R(Qn, Qn). The best bounds currently known are 2n + 1 ≤R(Qn, Qn) ≤n2 −n + 2, see listed chronologically Walzer [20], Axenovich and Walzer [1], Cox and Stolee [8], Lu and Thompson [14], Bohman and Peng [3]. It should be highlighted that the upper bound on R(Qn, Qn) shows that R(P, Q) is well-deﬁned for any P and Q because any poset is contained as a copy in a Boolean lattice Qn for sufﬁciently large n. One subject of research on poset Ramsey numbers is the off-diagonal setting R(P, Qn) for a ﬁxed poset P and large n. As general bounds the ﬁrst author and Walzer [1] showed the following. The height h(P) of a poset P is deﬁned as the size of a longest chain in P. The 2-dimension dim2(P) of a poset P is the dimension of the smallest Boolean lattice containing a copy of P. It is an easy observation that the 2-dimension is well-deﬁned for any P. Proposition 1 (Axenovich-Walzer [1]) Let P be a ﬁxed poset. Then n + h(P) −1 ≤R(P, Qn) ≤h(P)n + dim2(P). Proposition 1 (Axenovich-Walzer [1]) Let P be a ﬁxed poset. Then n + h(P) −1 ≤R(P, Qn) ≤h(P)n + dim2(P). 123 123 Order (2024) 41:401–418 403 Here, the lower bound is trivial and is obtained by a coloring of Qn+h(P)−2 in which all vertices in each layer {X ∈Qn+h(P)−2 : \|X\| = ℓ}, 0 ≤ℓ≤n + h(P) −2, have the same color, red or blue, and there are n red layers and h(P) −1 blue layers. For the off-diagonal setting R(Qm, Qn) with m ﬁxed and n large, an exact result is only known if m = 1. It is easy to see that R(Q1, Qn) = n + 1. For m = 2, it was shown in [1] that R(Q2, Qn) ≤2n + 2. This was improved by Lu and Thompson [14] to R(Q2, Qn) ≤(5/3)n + 2, and by Grósz, Methuku, and Tompkins [11] who showed for ϵ > 0 and sufﬁciently large n ∈N: n + 3 ≤R(Q2, Qn) ≤n + (2+ϵ)n log n . Finally, the present authors [2] proved a lower bound asymptotically matching the upper one: Theorem 2 (Grosz-Methuku-Tompkins [11], Axenovich-Winter [2]) Theorem 2 (Grosz-Methuku-Tompkins [11], Axenovich-Winter [2]) R(Q2, Qn) = n + n log n . R(P, Q) = min{N ∈N: every blue/red coloring of QN contains either a blue copy of P or a red co We show in Theorem 9 that extremal properties of P-free blockers immediately give an upper bound on R(P, Qn). Our result on R(N, Qn) then follows: 2 Main Deﬁnitions and Tools A vertex Z of a poset F is a minimum of F if it is the unique minimal element of F, i.e. Z ≤F for every F ∈F. Similarly, a maximum of F is a unique maximal vertex of F. Given a ﬁxed poset P, a poset F is P-free if it contains no (induced) copy of P. Let X and Y be disjoint sets. For a subposet F ⊆Q(Y), the X-shift of F is the poset F′ with vertices {Y ∪X : Y ∈F} ordered by inclusion. Note that F′ is isomorphic to F. Let F1 and F2 be two disjoint posets. The parallel composition of F1 and F2 is the poset on vertices F1 ∪F2 such that pairs of vertices in F1, as well as pairs of vertices in F2 are comparable if and only if they are likewise comparable in F1 or F2, respectively, and any two F1 ∈F1 and F2 ∈F2 are incomparable. In the literature this poset is also referred to as the independent union of F1 and F2. If for a poset F there exists a partition F = F1 ∪F2 into non-empty subposets F1 and F2 such that F is the parallel composition of F1 and F2, we say that F is disconnected. Otherwise, we say that F is connected. A weak homomorphism of a poset F1 into another poset F2 is a function φ : F1 →F2 such that for any two A, B ∈F1 with A ≤F1 B, we have φ(A) ≤F2 φ(B). Similarly, a function φ : F1 →F2 is a strong homomorphism if for any A, B ∈F1, A ≤F1 B if and only if φ(A) ≤F2 φ(B). An injective weak [strong] homomorphism is a weak [strong] embedding of F1 into F2. Here we exclusively consider strong embeddings and weak homomorphisms, so we usually simply refer to them as “embeddings” and “homomorphisms”, respectively. Throughout this paper, we consider a set Z as the ground set of our hosting lattice Q(Z) where \|Z\| = N for some integer N. We then partition Z into two disjoint sets X and Y, \|Y\| ̸= ∅, such that \|X\| = n and \|Y\| = k for some integers n and k, i.e. N = n +k. A (strong) embedding ψ : Q(X) →Q(Z) is X-good if for every X ⊆X, ψ(X) ∩X = X. Theorem 4 n + n 15 log n ≤R(N, Qn) ≤n + (1 + o(1))n log n . Here and throughout the paper, ‘log’ refers to the logarithm with base 2. The lower bound follows immediately from Theorem 3, so the focus of this paper is on the upper bound. Here and throughout the paper, ‘log’ refers to the logarithm with base 2. The lower bound follows immediately from Theorem 3, so the focus of this paper is on the upper bound. The paper is structured as follows. In Section 2 main deﬁnitions and basic results are given. Section 3 deals with the main tool used - blockers. In Section 4 a proof of Theorem 4 is given. 404 Order (2024) 41:401–418 2 Main Deﬁnitions and Tools We say that a copy Q of Qn in Q(Z) is X-good if there exists an X-good embedding of Q(X) with image Q. See Fig. 2 (a) for a {1, 2}-good copy of Q2 in Q({1, 2, x1, x2}). Moreover, we say that X is a deﬁning set for a copy of Qn if this copy is X-good. One of the main structural observations we have is the following: Lemma 5 (Axenovich-Walzer [1]) Let n ∈N. Any copy of Qn in Q(Z) is X-good for some subset X ⊆Z with \|X\| = n. We shall also need some deﬁnitions to describe N-free posets. Let F1 and F2 be two disjoint posets. The series composition of F1 below F2 is the poset on vertices F1 ∪F2, where pairs of vertices in F1, as well as pairs of vertices in F2 are comparable if and only if they are likewise comparable in F1 or F2, respectively, and F1 < F2 for any F1 ∈F1 and F2 ∈F2. A poset is series-parallel if either it is a 1-element poset or it is obtained by series composition or parallel composition of two series-parallel posets. Valdes [19] showed the following characterization. Theorem 6 (Valdes [19]) A non-empty poset is N-free if and only if it is series-parallel. 3.1 Definition and Examples of Y-blockers Outline of the Main Idea The deﬁnition of R(P, Qn) implies that there is a coloring of Q(Z), \|Z\| ≤R(P, Qn) −1, in blue and red such that the blue vertices “cover" all copies of Qn, i.e. there is a blue vertex in each copy of Qn and there is no copy of P having only blue 123 123 Order (2024) 41:401–418 405 vertices, i.e. the set of blue vertices is P-free. We shall classify all copies of Qn according to their deﬁning sets and consider the set of only those blue vertices that “cover" copies of Qn with a speciﬁc ﬁxed deﬁning set X. We refer to the poset induced by blue vertices as a Y-blocker, where Y = Z \ X. We shall derive several properties of general Y-blockers and those that are P-free. Then we will bound R(P, Qn) in terms of blockers. This generalises an approach used in [2], where V -free Y-blockers were considered and called Y-shrubs. Deﬁnition 7 Let Y and Z be two non-empty sets such that Y ⊆Z. A Y-blocker in Q(Z) is a subposet F in Q(Z) which contains a vertex from every X-good copy of Q(X), where X = Z \ Y. We say that a Y-blocker F in Q(Z) is critical if for any vertex F ∈F the subposet F \ {F} is not a Y-blocker in Q(Z). Note that for any Y ⊆Z, a Y-blocker in Q(Z) exists, for example take F = Q(Z). Later on we consider “thinner” Y-blockers satisfying special properties, in particular being P-free. Example Let Z = {1, 2, x1, x2}, Y = {1, 2}, and X = {x1, x2}. Let F be the {x1}-shift of Q(Y), see Fig. 2 (a). To show that F is a Y-blocker, consider an arbitrary X-good copy Q of Q(X) in Q(Z), with a corresponding X-good embedding ψ : Q(X) →Q(Z). Then ψ({x1}) = {x1} ∪Y for some Y ⊆Y, and hence ψ({x1}) ∈F. Thus, F is a Y-blocker in Q(Z). Fig. 2 (b) also depicts a Y-blocker, which we shall verify by Theorem 11. Theorem 9 Let P be a poset and let n ∈N be an integer. Then R(P, Qn) ≤min{N : there is no P-free Y-blocker in Q([N]) for some Y ⊆[N], \|Y\| = N −n}. Proof Let N be the smallest integer such that for some Y ⊆[N], \|Y\| = N −n, there is no P-free Y-blocker in Q([N]). Consider an arbitrarily blue/red colored Boolean lattice Q([N]) and let B be the induced subposet of Q([N]) consisting of all blue vertices. We shall show that there is either a blue copy of P or a red copy of Qn. Let X = [N] \ Y. If in Q([N]) there is a monochromatic red copy of Qn which is X-good, the proof is complete. Otherwise each X-good copy of Qn contains a blue vertex, i.e. the blue subposet B is a Y-blocker. By the deﬁnition of N, B is not P-free. Thus there is a blue copy of P in Q([N]). It remains to show that this minimum is well-deﬁned, i.e. we shall ﬁnd an integer N such that there is no P-free Y-blocker in Q([N]), where Y ⊆[N] with \|Y\| = N −n. In order to show this, we bound the size \|F\| of a P-free Y-blocker F in Q([N]) from above and from below. On the one hand, by a result of Methuku and Pálvölgyi [16] the size of the P-free subposet F ⊆Q([N]) is bounded by \|F\| ≤c(P) N N/2 ≤c′(P) · 2N √N/2 , where c and c′ are constants depending only on P. On the other hand, Lemma 8 provides that where c and c′ are constants depending only on P. On the other hand, Lemma 8 provides that \|F\| ≥2\|Y\| = 2N−n. For sufﬁciently large N, we have that √N/2 c′(P) > 2n, which implies that there is no P-free Y-blocker \|F\| in Q([N]). ⊓⊔ ⊓⊔ Deﬁnition 10 For a subposet F of Q(Z) and Y ⊆Z, we say that a (weak) homomorphism φ : F →Q(Y) is Y-hitting if there exists some F ∈F with φ(F) = F ∩Y. Conversely, φ is Y-avoiding if φ(F) ̸= F ∩Y for every F ∈F. Remark In the following we show that the existence of a Y-blocker is equivalent to the non- existence of a Y-avoiding homomorphism. One can think of an homomorphism φ : F → Q(Y) as a “recipe” encoding an X-good copy of Qn in Q(X ∪Y). 3.2 General Properties of Y-blockers Lemma 8 (i) Let Z be a setwith \|Z\| > n. A blue/red colored Boolean lattice Q(Z)contains no red copy of Qn if and only if for each X ⊆Z of size \|X\| = n, there is a Z \X-blocker with all vertices blue. (ii) Let F be a Y-blocker where Y ̸= ∅and let Y ⊆Y. Then there is a vertex Z ∈F with Z ∩Y = Y. In particular, if Z is a minimum of F, then Z ∩Y = ∅; and if Z is a maximum of F, then Z ∩Y = Y. (iii) If F is a Y-blocker, then \|F\| ≥2\|Y\|. (iii) If F is a Y-blocker, then \|F\| ≥2\|Y\|. Proof Part (i) follows immediately from Lemma 5 and the deﬁnition of a Y-blocker. For (ii), let F be a Y-blocker in Q(Z) and X = Z \ Y. Observe that F contains a vertex U with U ∩Y = Y for every Y ⊆Y, because otherwise the Y-shift of Q(X) is an X-good copy of Q(X) that does not contain a vertex from F. Considering Y = ∅, we see that there is U ∈F such that U ∩Y = ∅. If Z is a minimum of F, then it has Y-part Z ∩Y ⊆U ∩Y = ∅. Similarly, if there is a maximum Z of F, it has Y-part Z ∩Y = Y. For (iii), since there are 2\|Y\| subsets of Y, part (ii) immediately implies that \|F\| ≥2\|Y\|. ⊓⊔ Fig. 2 Two {1, 2}-blockers in Q({1, 2, x1, x2}) Fig. 2 Two {1, 2}-blockers in Q({1, 2, x1, x2}) Fig. 2 Two {1, 2}-blockers in Q({1, 2, x1, x2}) Fig. 2 Two {1, 2}-blockers in Q({1, 2, x1, x2}) 123 Order (2024) 41:401–418 406 Theorem 9 Let P be a poset and let n ∈N be an integer. Then Theorem 9 Let P be a poset and let n ∈N be an integer. Then Consider the function φ : F →Q(Y) given by φ(F) := ψ(F ∩X) ∩Y for each F ∈F. Using the properties of ψ, it is easy to see that if F ⊆F′ for F, F′ ∈F, then φ(F) ⊆φ(F′), so φ is a homomorphism. Thus φ is Y-hitting and we ﬁnd some Z ∈F with φ(Z) = Z ∩Y. Then ψ(Z ∩X)∩Y = φ(Z) = Z ∩Y. Since ψ is X-good, we know that ψ(Z ∩X)∩X = Z ∩X. Therefore ψ(Z ∩X) = Z. Since the image of ψ is Q, we obtain Z = ψ(Z ∩X) ∈Q, thus F and Q have the vertex Z in common. Proof of Theorem 11 Let Y ⊆Z be a non-empty subset and let X = Z \ Y. For the ﬁrst part of the proof let F be a subposet in Q(Z) such that every homomorphism φ : F →Q(Y) is Y-hitting. We shall show that F is a Y-blocker. Let Q be an arbitrary X-good copy of Q(X) in Q(Z) with a corresponding X-good embedding ψ : Q(X) →Q(Z). Consider the function φ : F →Q(Y) given by φ(F) := ψ(F ∩X) ∩Y for each F ∈F. Using the properties of ψ, it is easy to see that if F ⊆F′ for F, F′ ∈F, then φ(F) ⊆φ(F′), so φ is a homomorphism. Thus φ is Y-hitting and we ﬁnd some Z ∈F with φ(Z) = Z ∩Y. Then ψ(Z ∩X)∩Y = φ(Z) = Z ∩Y. Since ψ is X-good, we know that ψ(Z ∩X)∩X = Z ∩X. Therefore ψ(Z ∩X) = Z. Since the image of ψ is Q, we obtain Z = ψ(Z ∩X) ∈Q, thus F and Q have the vertex Z in common. From now on let F be a subposet in Q(Z) for which there exists a Y-avoiding homomor- phism φ : F →Q(Y). We shall show that F is not a Y-blocker. For that we shall construct an X-good embedding ψ : Q(X) →Q(Z) such that the image of ψ does not contain a vertex from F. Fix some X ∈Q(X), now we deﬁne ψ(X) iteratively: Informally spoken, in step i we introduce a set fi(X) ⊆Y and check whether X ∪fi(X) is a “feasible” choice for ψ(X); and if not, we extend fi(X) to its strict superset fi+1(X) and repeat. Let f0(X) = ∅. Theorem 9 Let P be a poset and let n ∈N be an integer. Then For i ∈N, let Fi(X) = {Z ∈F : Z ⊆X ∪fi−1(X)} be the down-set of X ∪fi−1(X) and let fi(X) = Z∈Fi(X) φ(Z). Clearly fi(X) ⊆Y. Note that ∅= f0(X) ⊆f1(X), thus F1(X) ⊆F2(X) and so f1(X) ⊆f2(X). Iteratively, we obtain that Fi(X) ⊆Fi+1(X) and fi(X) ⊆fi+1(X) ⊆Y, see Fig. 3 (a). Thus after ﬁnitely many steps f j(X) = f j+1(X) for some j ∈N, i.e. this set is “feasible”, and let j(X) be the minimal such index j. Observe that f j(X)(X) = f j(X)+1(X) = f j(X)+2(X) = . . . as F j(X)+1(X) = F j(X)+2(X) = . . . . We set ψ(X) := X ∪f j(X)(X). Theorem 9 Let P be a poset and let n ∈N be an integer. Then Recall that a Y-blocker is deﬁned as a poset which has a vertex in common with every X-good copy of Qn. If a “recipe” φ is Y-hitting, then a vertex F ∈F with φ(F) = F ∩Y is contained in the X-good copy encoded by φ, i.e. the Y-blocker F has a vertex in common with this copy. However, there is no 1-to-1 correspondence between homomorphisms φ and embeddings ψ of X-good copies of Qn, and the presented constructions building ψ from φ as well as φ from ψ are not inverse of each other. Theorem 11 Let Y be a non-empty subset of a set Z. A subposet F of a Boolean lattice Q(Z) is a Y-blocker if and only if every (weak) homomorphism φ : F →Q(Y) is Y-hitting. Theorem 11 Let Y be a non-empty subset of a set Z. A subposet F of a Boolean lattice Q(Z) is a Y-blocker if and only if every (weak) homomorphism φ : F →Q(Y) is Y-hitting. Example Let Z = {1, 2, x1, x2} and Y = {1, 2}. In the Boolean lattice Q({1, 2, x1, x2}) consider the subposet F on vertices ∅, {1, x1}, {1, 2, x1}, {2, x2}, {1, 2, x2}, see Figure 2 (b). Then F is a {1, 2}-blocker. In order to prove this, we can use Theorem 11. Assume towards a contradiction that there is a homomorphism φ : F →Q(Y) such that for every F ∈F we have φ(F) ̸= F ∩Y. Then φ(∅) ∩Y ̸= ∅, say without loss of generality 1 ∈φ(∅) ∩Y. Since φ is a homomorphism φ(∅) ⊆φ({1, x1}), so 1 ∈φ({1, x1}) ∩Y. Now, because φ({1, x1}) ∩Y ̸= {1}, we obtain that φ({1, x1}) ∩Y = {1, 2}. Then using that φ({1, x1}) ⊆φ({1, 2, x1}), we obtain φ({1, 2, x1}) ∩Y = {1, 2}, a contradiction. 123 123 Order (2024) 41:401–418 407 Proof of Theorem 11 Let Y ⊆Z be a non-empty subset and let X = Z \ Y. For the ﬁrst part of the proof let F be a subposet in Q(Z) such that every homomorphism φ : F →Q(Y) is Y-hitting. We shall show that F is a Y-blocker. Let Q be an arbitrary X-good copy of Q(X) in Q(Z) with a corresponding X-good embedding ψ : Q(X) →Q(Z). Claim 1 ψ is an X-good embedding of Q(X). Proof of Claim 1. Note that for every X ∈Q(X), we have f j(X)(X) ⊆Y and so ψ(X) ∩ X = X. Thus it remains to show that ψ is an embedding in order to prove the claim. Let X1, X2 ∈Q(X). We shall show that X1 ⊆X2 if and only if ψ(X1) ⊆ψ(X2). Proof of Claim 1. Note that for every X ∈Q(X), we have f j(X)(X) ⊆Y and so ψ(X) ∩ X = X. Thus it remains to show that ψ is an embedding in order to prove the claim. Let X1, X2 ∈Q(X). We shall show that X1 ⊆X2 if and only if ψ(X1) ⊆ψ(X2). First suppose that X1 ⊆X2. Then X1 ∪f0(X) = X1 ⊆X2 = X2 ∪f0(X), so F1(X1) ⊆ F1(X2). This implies that f1(X1) ⊆f1(X2), so X1 ∪f1(X) ⊆X2 ∪f1(X), see Fig. 3 (b). Iteratively, Fi(X1) ⊆Fi(X2) and fi(X1) ⊆fi(X2). We obtain that 1, 2 Q( ) 1 ⊆ 2 y ψ( 1) ⊆ψ( 2) First suppose that X1 ⊆X2. Then X1 ∪f0(X) = X1 ⊆X2 = X2 ∪f0(X), so F1(X1) ⊆ F1(X2). This implies that f1(X1) ⊆f1(X2), so X1 ∪f1(X) ⊆X2 ∪f1(X), see Fig. 3 (b). Iteratively, Fi(X1) ⊆Fi(X2) and fi(X1) ⊆fi(X2). We obtain that f j(X1)(X1) = fmax{ j(X1), j(X2)}(X1) ⊆fmax{ j(X1), j(X2)}(X2) = f j(X2)(X2), f j(X1)(X1) = fmax{ j(X1), j(X2)}(X1) ⊆fmax{ j(X1), j(X2)}(X2) = f j(X2)(X2), thus ψ(X1) ⊆ψ(X2). Now suppose that ψ(X1) ⊆ψ(X2). Then in particular X1 = ψ(X1)∩ X ⊆ψ(X2) ∩X = X2, so X1 ⊆X2. Claim 2 The image of ψ contains no vertex from F. Claim 2 The image of ψ contains no vertex from F. Proof of Claim 2. Let X ∈X and assume that ψ(X) ∈F. We shall ﬁnd a contradiction by consideringφ(ψ(X)).Observethatψ(X) = X∪f j(X)(X) ∈F j(X)+1(X)andsoφ(ψ(X)) ⊆ f j(X)+1(X). Since φ is Y-avoiding, φ(ψ(X)) ̸= ψ(X) ∩Y = f j(X)(X) = f j(X)+1(X). Consequently, there exists an element a ∈f j(X)+1(X) \ φ(ψ(X)). By deﬁnition of fi(X), we ﬁnd a vertex Z ∈F j(X)(X) ⊆F with a ∈φ(Z). Now, since Z ∈F j(X)(X), we obtain that Z ⊆X ∪f j(X)(X) = ψ(X) whereas element a witnesses φ(Z) ⊈φ(ψ(X)). This contradicts the fact that φ is a homomorphism. So, indeed, F is not a Y-blocker. ⊓⊔ This concludes the proof of Theorem 11. In the following we use the characterization from Theorem 11 to analyse properties of critical blockers. Recall that for Y ⊆Z, a Y-blocker F in Q(Z) is critical if for any vertex F ∈F the subposet F \ {F} is not a Y-blocker in Q(Z). Order (2024) 41:401–418 408 Fig. 3 (a) Construction of ψ(X), (b) Iteration in Claim 1 Fig. 3 (a) Construction of ψ(X), (b) Iteration in Claim 1 Fig. 3 (a) Construction of ψ(X), (b) Iteration in Claim 1 Lemma 12 Let F be a critical Y-blocker for a non-empty set Y. Then F is a connected poset. Lemma 12 Let F be a critical Y-blocker for a non-empty set Y. Then F is a connected pos Proof Assume that F is the parallel composition of two non-empty posets F1 and F2, i.e. F1 and F2 are vertex-wise incomparable in F. Then each of F1 and F2 is not a Y-blocker by criticality of F. Thus there are Y-avoiding homomorphisms φ1 : F1 →Q(Y) and φ2 : F2 → Q(Y). Now the function ψ : F →Q(Y), ψ(F) = φ1(F), if F ∈F1 φ2(F), if F ∈F2. is a homomorphism of F and Y-avoiding. Recall that F is a Y-blocker, so this is a contra- diction to Theorem 11. ⊓⊔ is a homomorphism of F and Y-avoiding. Recall that F is a Y-blocker, so this is a contra- diction to Theorem 11. ⊓⊔ ⊓⊔ Lemma 13 Let F be a critical Y-blocker for a non-empty set Y. Let U1, U2 ∈F with U1 ̸= U2. If either U1 ∩Y = ∅= U2 ∩Y or U1 ∩Y = Y = U2 ∩Y, then U1 and U2 are not comparable. Proof Assume thatU1∩Y = ∅= U2∩Y andU1 ⊆U2. As F is a critical Y-blocker, the poset F′ = F \ {U2} is not a Y-blocker, so by Theorem 11 we ﬁnd a Y-avoiding homomorphism φ : F′ →Q(Y). Let U = {U ∈F′ : U ⊂U2}, note that U ̸= ∅, see Fig. 4 (a). We extend φ to a function ψ : F →Q(Y) by deﬁning ψ(F) = φ(F), if F ̸= U2 U∈U φ(U), if F = U2. ψ(F) = φ(F), if F ̸= U2 U∈U φ(U), if F = U2. In order to reach a contradiction, it remains to show that ψ is a Y-avoiding homomorphism. We shall show that ψ is a homomorphism by considering any two F1, F2 ∈F such that F1 ⊆F2 and verifying that ψ(F1) ⊆ψ(F2). We need to consider cases whether either of F1 or F2 is equal to U2. We repeatedly use the fact that φ is a homomorphism: f F1 ̸= U2 and F2 ̸= U2, then ψ(F1) = φ(F1) ⊆φ(F2) = ψ(F2). ◦If F1 ̸ U2 and F2 ̸ U2, then ψ(F1) φ(F1) ⊆φ(F2) ψ(F2). ◦If F1 = U2, then ψ(F1) = ψ(U2) = U∈U φ(U) ⊆ U∈U φ(F2) ⊆φ(F2) = ψ(F2). Lemma 12 Let F be a critical Y-blocker for a non-empty set Y. Then F is a connected pos Here we used the property that for any U ∈U U ⊆U ⊆F ◦If F1 = U2, then ψ(F1) = ψ(U2) = U∈U φ(U) ⊆ U∈U φ(F2) ⊆φ(F2) = ψ(F2). Here we used the property that for any U ∈U, U ⊆U2 ⊆F2. ◦If F1 = U2, then ψ(F1) = ψ(U2) = U∈U φ(U) ⊆ U∈U φ(F2) ⊆φ(F2) = ψ(F2). Here we used the property that for any U ∈U, U ⊆U2 ⊆F2. p p y y ◦If F2 = U2, then ψ(F1) = φ(F1) ⊆ U∈U φ(U) = ψ(U2) = ψ(F2). Here, we used that F1 ∈U and thus F1 ⊆ U∈U U. Therefore, ψ is a homomorphism. ◦If F2 = U2, then ψ(F1) = φ(F1) ⊆ U∈U φ(U) = ψ(U2) = ψ(F2). Here, we used that F1 ∈U and thus F1 ⊆ U∈U U. Therefore, ψ is a homomorphism. U∈U To show that ψ is Y-avoiding, we need to verify that ψ(F) ̸= F ∩Y for any F ∈F. First consider F ∈F with F ̸= U2, i.e. F ∈F′. Since φ is Y-avoiding, ψ(F) = φ(F) ̸= F ∩Y. For F = U2, ψ(U2) = φ(U1) ̸= U1 ∩Y since φ is Y-avoiding. We selected U1 and U2 such U∈U To show that ψ is Y-avoiding, we need to verify that ψ(F) ̸= F ∩Y for any F ∈F. First consider F ∈F with F ̸= U2, i.e. F ∈F′. Since φ is Y-avoiding, ψ(F) = φ(F) ̸= F ∩Y. For F = U2, ψ(U2) = φ(U1) ̸= U1 ∩Y since φ is Y-avoiding. We selected U1 and U2 such 123 123 Order (2024) 41:401–418 409 Fig. 4 (a) Setting in Lemma 13, (b) Setting in Lemma 14 Fig. 4 (a) Setting in Lemma 13, (b) Setting in Lemma 14 Fig. 4 (a) Setting in Lemma 13, (b) Setting in Lemma 14 that U1 ∩Y = ∅= U2 ∩Y, thus ψ(U2) ̸= U2 ∩Y. We conclude that ψ is Y-avoiding. This contradicts Theorem 11 and the fact that F is a Y-blocker. that U1 ∩Y = ∅= U2 ∩Y, thus ψ(U2) ̸= U2 ∩Y. We conclude that ψ is Y-avoiding. This contradicts Theorem 11 and the fact that F is a Y-blocker. Under the assumption that U1 ∩Y = Y = U2 ∩Y and U1 ⊆U2, a symmetric proof holds for U = {U ∈F \ {U1}: U ⊃U1} and ψ : F →Q(Y) with ψ(F) = φ(F), if F ̸= U1 U∈U φ(U), if F = U1. ⊓⊔ Lemma 14 Let F be a critical Y-blocker where Y ̸= ∅. Let F1 ⊆{U ∈F : U ∩Y = ∅} such that F is a series composition of F1 below F \ F1, then \|F1\| ≤1. Similarly, let F2 ⊆{U ∈F : U ∩Y = Y} such that F is a series composition of F \ F2 below F2, then \|F2\| ≤1. Lemma 14 Let F be a critical Y-blocker where Y ̸= ∅. Let F1 ⊆{U ∈F : U ∩Y = ∅} such that F is a series composition of F1 below F \ F1, then \|F1\| ≤1. Similarly, let F2 ⊆{U ∈F : U ∩Y = Y} such that F is a series composition of F \ F2 below F2, then \|F2\| ≤1. Proof For the ﬁrst part, assume towards a contradiction that there are two distinct vertices U1, U2 ∈F1. Since F is a critical Y-blocker, there is a Y-avoiding homomorphism φ : F \ {U2} →Q(Y). Let ψ : F →Q(Y) such that ψ(F) = φ(F), if F ̸= U2 φ(U1), if F = U2. ψ(F) = φ(F), if F ̸= U2 φ(U1), if F = U2. ψ(F) = φ(F), if F ̸= U2 φ(U1), if F = U2. We shall prove that ψ is a Y-avoiding homomorphism of F. By Lemma 13, F1 is an antichain. In order to show that ψ is a homomorphism, we consider two arbitrary F1, F2 ∈F with F1 ⊆F2 and show that ψ(F1) ⊆ψ(F2). 123 ◦If F2 = U2, then in particular F2 ∈F1 and so F1 ∈F1, because F is a series composition of F1 below F \ F1. Since F1 is an antichain, we obtain that F1 = U2 = F2. Then trivially ψ(F1) = ψ(U2) = ψ(F2). ◦If F1 = U2 and F2 ̸= U2, we know that F2 ∈F \ F1 because F1 is an antichain. Then U1 ⊆F2. Because φ is a homomorphism and by deﬁnition of ψ, we obtain that ψ(U2) = φ(U1) ⊆φ(F2). ◦If F1 ̸= U2 and F2 ̸= U2, then ψ(F1) = φ(F1) ⊆φ(F2) = ψ(F2). Thus ψ is a homomorphism of F. ◦If F1 ̸= U2 and F2 ̸= U2, then ψ(F1) = φ(F1) ⊆φ(F2) = ψ(F2). Thus ψ is a homomorphism of F. 123 123 410 Order (2024) 41:401–418 Order (2024) 41:401–418 For every F ∈F \ {U2}, we know that ψ(F) = φ(F) ̸= F ∩Y. Furthermore, ψ(U2) = φ(U1) ̸= U1 ∩Y = ∅= U2 ∩Y. Thus ψ is Y-avoiding, a contradiction. If we assume that there are distinct U1, U2 ∈F2, a symmetric argument considering the same function ψ : F →Q(Y), ψ(F) = φ(F), if F ̸= U2 φ(U1), if F = U2. yields a contradiction. ⊓⊔ Lemma 15 Let X and Y be two disjoint sets with \|Y\| = 1. Let F be a critical Y-blocker in Q(X ∪Y). Then F is a chain consisting of two vertices X1, X2 ∪Y, where X1 ⊆X2 ⊆X. Proof Since \|Y\| = 1, we ﬁnd that for every Z ∈F either Z ∩Y = ∅or Z ∩Y = Y. Consider subposets F1 = {Z ∈F : Z ∩Y = ∅} and F2 = {Z ∈F : Z ∩Y = Y} partitioning F. Lemma 8 provides that F1 ̸= ∅and F2 ̸= ∅. By Lemma 12, F is connected, so in particular there are two vertices from F1 and from F2 which are comparable. Let these vertices be X1 ∈F1 and X2 ∪Y ∈F2, where X1, X2 ⊆X. Then X1 ⊆X2 ∪Y, so X1 ⊆X2. Next we need to show that F = {X1, X2∪Y}. Consider the subposet F′ = {X1, X2∪Y} ⊆ F. We show that F′ is a Y-blocker in Q(X ∪Y), i.e. by Theorem 11 we shall show that there is no Y-avoiding homomorphism from F′ to Q(Y). 123 A homomorphism φ : F′ →Q(Y) is Y-avoiding only if φ(X1) = Y and φ(X2 ∪Y) = ∅, but such a homomorphism does not exist, since φ(X1) ⊆φ(X2 ∪Y) because of X1 ⊆X2 ∪Y. We obtain that F′ is a Y-blocker, therefore F = {X1, X2 ∪Y} since F is critical. ⊓⊔ Next we need to show that F = {X1, X2∪Y}. Consider the subposet F′ = {X1, X2∪Y} ⊆ F. We show that F′ is a Y-blocker in Q(X ∪Y), i.e. by Theorem 11 we shall show that there is no Y-avoiding homomorphism from F′ to Q(Y). A homomorphism φ : F′ →Q(Y) is Y-avoiding only if φ(X1) = Y and φ(X2 ∪Y) = ∅, but such a homomorphism does not exist, since φ(X1) ⊆φ(X2 ∪Y) because of X1 ⊆X2 ∪Y. We obtain that F′ is a Y-blocker, therefore F = {X1, X2 ∪Y} since F is critical. ⊓⊔ ⊓⊔ Lemma 16 Let Y be a set of size at least 2 and let a ∈Y. Let F be a Y-blocker. Then the induced subposets {F ∈F : a ∈F} and {F ∈F : a /∈F} are (Y \ {a})-blockers. Proof Let F′ = {F ∈F : a ∈F}. Assume that F′ is not a (Y \{a})-blocker, i.e. by Theorem 11 there is a (Y \ {a})-avoiding homomorphism φ : F′ →Q(Y \ {a}). We ﬁnd a Y-avoiding homomorphism of F in order to reach a contradiction. Let ψ : F →Q(Y) with ψ(F) = φ(F) ∪{a}, if F ∈F′ {a}, if F /∈F′. Observe that ψ is a homomorphism, because {a} ⊆φ(F) ∪{a} for all F ∈F′ and φ is a homomorphism. For every F ∈F \ F′, note that a ∈ψ(F) but a /∈F ∩Y, thus ψ(F) ̸= F ∩Y. On the other hand, recall that φ is (Y \ {a})-avoiding. Hence for every F ∈F′ we know that φ(F) ̸= F ∩(Y \ {a}) where a /∈φ(F) and a /∈F ∩(Y \ {a}). This implies ψ(F) = φ(F) ∪{a} ̸= F ∩(Y \ {a}) ∪{a} = F ∩Y. As a result, ψ is a Y-avoiding homomorphism of F, which is a contradiction. The second part of the lemma follows from a symmetric argument for F′′ = {F ∈F : a /∈ F} using the function ψ : F →Q(Y), ψ(F) = φ(F), if F ∈F′′ Y \ {a}, if F /∈F′′. 123 ψ(F) = φ(F), if F ∈F′′ Y \ {a}, if F /∈F′′. ⊓⊔ ⊓⊔ 123 Order (2024) 41:401–418 411 3.3 Properties of N -free Y-blockers In the ﬁrst case we obtain that Y1 ̸= ∅, thus Y1 = Y and hence Y2 = Y (because Let Y1 = ( F∈F1 F) ∩Y be the Y-part of the union of all vertices in F1 and let Y2 = ( F∈F2 F) ∩Y be the Y-part of the intersection of all vertices in F2. Clearly, Y1 ⊆Y2 ⊆Y. Let Y1 = ( F∈F1 F) ∩Y be the Y-part of the union of all vertices in F1 and let Y2 ( F∈F2 F) ∩Y be the Y-part of the intersection of all vertices in F2. Clearly, Y1 ⊆Y2 ⊆ 2 First assume that Y1 /∈{∅, Y}. Then there are a ∈Y1 and b ∈Y \ Y1. Lemma 8 provides that the Y-blocker F contains a vertex U with U ∩Y = {b}. Then U /∈F1 since b ∈U while b /∈Y1, but also U /∈F2 as a /∈U while a ∈Y1 ⊆Y2. We arrive at a contradiction, hence Y1 ∈{∅, Y}. Symmetrically, Y2 ∈{∅, Y}. Take an arbitrary Y-part Y ⊆Y such that Y /∈{∅, Y}. By Lemma 8 there is a vertex Z ∈F with Z ∩Y = Y. Then Z ∈F1 or Z ∈F2. In the ﬁrst case we obtain that Y1 ̸= ∅, thus Y1 = Y and hence Y2 = Y (because Y1 ⊆Y2 ⊆Y). In the second case Y2 ̸= Y, so Y1 = Y2 = ∅. ) ̸ Thus either Y1 = Y2 = ∅or Y1 = Y2 = Y. For the rest of the proof we suppose that Y1 = Y2 = ∅. If Y1 = Y2 = Y, a symmetric argument holds. Because Y1 = ( F∈F1 F) ∩Y = ∅, we obtain F ∩Y = ∅for every F ∈F1. By Lemma 14, there is at most one vertex in F1. The unique vertex Z ∈F1 is the unique minimal vertex of F and Z ∩Y = Y1 = ∅. In the case that Y1 = Y2 = Y, we can argue symmetrically and obtain that Z is a maximum of F and Z ∩Y = Y2 = Y. ⊓⊔ 3.4 Construction of the Family {(FS, ZS, AS, BS) : S ∈S} 3.3 Properties of N -free Y-blockers Theorem 17 Let X and Y be disjoint sets with Y ̸= ∅. Let F be an N-free, critical Y-blocker in Q(X ∪Y). Then F has at least one of a minimum vertex or a maximum vertex. Theorem 17 Let X and Y be disjoint sets with Y ̸= ∅. Let F be an N-free, critical Y-blocker in Q(X ∪Y). Then F has at least one of a minimum vertex or a maximum vertex. Proof of Theorem 17 Since Y ̸= ∅, Lemma 8 implies that \|F\| ≥21. By Theorem 6, F is series-parallel, so it can be partitioned into two disjoint, non-empty posets F1 and F2 such that F is either the parallel composition of F1 and F2 or the series composition of F1 below F2. The former could not happen by Lemma 12. Thus, F can be partitioned into two disjoint, non-empty posets F1 and F2 such that for every F1 ∈F1 and F2 ∈F2, F1 ⊆F2. p y p y Let Y1 = ( F∈F1 F) ∩Y be the Y-part of the union of all vertices in F1 and let Y2 = ( F∈F2 F) ∩Y be the Y-part of the intersection of all vertices in F2. Clearly, Y1 ⊆Y2 ⊆Y. First assume that Y1 /∈{∅, Y}. Then there are a ∈Y1 and b ∈Y \ Y1. Lemma 8 provides that the Y-blocker F contains a vertex U with U ∩Y = {b}. Then U /∈F1 since b ∈U p y p y Let Y1 = ( F∈F1 F) ∩Y be the Y-part of the union of all vertices in F1 and let Y2 = ( F∈F2 F) ∩Y be the Y-part of the intersection of all vertices in F2. Clearly, Y1 ⊆Y2 ⊆Y. First assume that Y1 /∈{∅, Y}. Then there are a ∈Y1 and b ∈Y \ Y1. Lemma 8 provides that the Y-blocker F contains a vertex U with U ∩Y = {b}. Then U /∈F1 since b ∈U while b /∈Y1, but also U /∈F2 as a /∈U while a ∈Y1 ⊆Y2. We arrive at a contradiction, hence Y1 ∈{∅, Y}. Symmetrically, Y2 ∈{∅, Y}. Take an arbitrary Y-part Y ⊆Y such that Y /∈{∅, Y}. By Lemma 8 there is a vertex Z ∈F with Z ∩Y = Y. Then Z ∈F1 or Z ∈F2. 3.4 Construction of the Family {(FS, ZS, AS, BS) : S ∈S} In the following proof we will deﬁne posets and vertices indexed by ordered sets. Deﬁnition 18 An ordered set S is a sequence S = (y1, . . . , ym) of distinct elements yi, i ∈[m]. Given a set Y, S is an ordered subset of Y if yi ∈Y for all i ∈[m]. We denote the empty ordered set by ∅o = (). The underlying unordered set of S is denoted by S, and \|S\| = \|S\| is the size of S. For an ordered set S = (y1, . . . , ym) and an element ym+1 /∈S, we write (S, ym+1) for the ordered set (y1, . . . , ym, ym+1). We say that an ordered set S′ is a preﬁx of S if \|S′\| ≤\|S\| and each of the ﬁrst \|S′\| members of S coincides with the respective member of S′. Note that ∅o is a preﬁx of every ordered set. For i ∈{0, . . . , \|S\|}, we denote by S[i] the unique preﬁx of S of size i. A preﬁx S′ of S is strict if S′ ̸= S. For a set Y and an ordered subset S of Y, we denote the set of all elements of Y that are not in S by Y −S = Y \ S. In the following we analyse the structure of an N-free critical Y-blocker by selecting smaller and smaller subposets which are critical Y′-blockers for some Y′ ⊆Y. Recall that Theorem 17 implies that any critical Y′-blocker has either a minimum or a maximum vertex, we call such a vertex a root of the blocker. Note that the blocker could have both a minimum vertex and a maximum vertex. In this case we select one of them to be the assigned root of the blocker and ignore the second root. Construction 19 Let Y be a set with \|Y\| = k. Let F be an N-free, critical Y-blocker in Q(Z), Y ⊆Z. Let S be the set of all ordered subsets of Y of size at most k −1. In the 123 Order (2024) 41:401–418 412 following we recursively construct a family {(FS, ZS, AS, BS) : S ∈S}, where FS is a critical (Y −S)-blocker, FS ⊆F, and ZS is the root of FS. 3.4 Construction of the Family {(FS, ZS, AS, BS) : S ∈S} In addition AS ∪BS = S, where each element of AS is included in each vertex of FS and each element of BS is excluded from each vertex of FS. The sets AS and BS are used as tools to encode crucial information on the blocker FS and its root ZS as well as FS′ and ZS′ for preﬁxes S′ of S. If the root ZS is a minimum vertex in FS, we say that S is min-type, otherwise we say that S is max-type. Initial Step. Let S = ∅o. In this case let FS = F. Let ZS be an arbitrarily chosen root of F, i.e. a minimum or maximum of F, which exists due to Theorem 17. Let AS = BS = ∅. Initial Step. Let S = ∅o. In this case let FS = F. Let ZS be an arbitrarily chosen root of F, i.e. a minimum or maximum of F, which exists due to Theorem 17. Let AS = BS = ∅. General Iterative Step. Consider an arbitrary non-empty ordered subset S of Y with \|S\| ≤k −1. Let S′ be the preﬁx of S such that (S′, a) = S for some a ∈Y. Given (FS′, ZS′, AS′, BS′) such that FS′ is a critical (Y −S′)-blocker, ZS′ is a root of FS′, and AS′, BS′ are disjoint sets partitioning AS′ ∪BS′ = S′, we shall construct FS, ZS, AS, and BS. By Lemma 16 and since (Y −S′) \ {a} = Y −S, the sets {F ∈FS′ : a ∈F} and {F ∈FS′ : a /∈F} induce (Y −S)-blockers. General Iterative Step. Consider an arbitrary non-empty ordered subset S of Y with \|S\| ≤k −1. Let S′ be the preﬁx of S such that (S′, a) = S for some a ∈Y. Given (FS′, ZS′, AS′, BS′) such that FS′ is a critical (Y −S′)-blocker, ZS′ is a root of FS′, and AS′, BS′ are disjoint sets partitioning AS′ ∪BS′ = S′, we shall construct FS, ZS, AS, and BS. By Lemma 16 and since (Y −S′) \ {a} = Y −S, the sets {F ∈FS′ : a ∈F} and {F ∈FS′ : a /∈F} induce (Y −S)-blockers. If S′ is min-type, we deﬁne FS to be an arbitrary critical (Y −S)-blocker which is an induced subposet of {F ∈FS′ : a ∈F}. Note that a ∈F for every F ∈FS. 3.4 Construction of the Family {(FS, ZS, AS, BS) : S ∈S} Let AS = AS′ ∪{a} and BS = BS′. If S′ is min-type, we deﬁne FS to be an arbitrary critical (Y −S)-blocker which is an induced subposet of {F ∈FS′ : a ∈F}. Note that a ∈F for every F ∈FS. Let AS = AS′ ∪{a} and BS = BS′. If S′ is max-type, we deﬁne FS to be an arbitrary critical (Y −S)-blocker which is an induced subposet of {F ∈FS′ : a /∈F}. Note that in this case a /∈F for every F ∈FS. Let AS = AS′ and BS = BS′ ∪{a}. If S′ is max-type, we deﬁne FS to be an arbitrary critical (Y −S)-blocker which is an induced subposet of {F ∈FS′ : a /∈F}. Note that in this case a /∈F for every F ∈FS. Let AS = AS′ and BS = BS′ ∪{a}. It remains to select ZS. Theorem 17 provides the existence of a root in FS. If \|S\| ≤k −2, let ZS be an arbitrary root of FS. If \|S\| = k −1, we need to be more careful in choosing ZS. We have that FS is a critical (Y −S)-blocker, for \|Y −S\| = 1. By Lemma 15, FS has exactly two vertices, a minimum and a maximum. If S′ is min-type, let ZS be the minimum of FS, i.e. S is min-type. If S′ is max-type, let ZS be the maximum of FS, here S is max-type. The construction terminates after all ordered subsets of Y of size at most k −1 have been considered. The family {(FS, ZS, AS, BS) : S ∈S} gives a recursive structural decom- position of F into “up” and “down” components, i.e. max-type and min-type blockers, as illustrated in Fig. 5. Note that blockers FS may heavily overlap. Several properties follow immediately from the construction. Lemma 20 Let S be an ordered subset of Y of size at most k −1 and let S′ be a preﬁx of S. Then (i) FS′ ⊆FS, AS′ = AS ∩S′ and BS′ = BS ∩S′. (i) FS′ ⊆FS, AS′ = AS ∩S′ and BS′ = BS ∩S′. (ii) The size of the set AS is equal to the number of min-type strict preﬁxes S′ of S. The size of BS is equal to the number of max-type strict preﬁxes S′ of S. 3.4 Construction of the Family {(FS, ZS, AS, BS) : S ∈S} (iii) If S is min-type, Y ∩ZS = {y ∈Y : y ∈ZS} = AS. If S is max-type, Y \ ZS = {y ∈ Y : y /∈ZS} = BS. (ii) The size of the set AS is equal to the number of min-type strict preﬁxes S′ of S. The size of BS is equal to the number of max-type strict preﬁxes S′ of S. (iii) If S is min-type, Y ∩ZS = {y ∈Y : y ∈ZS} = AS. If S is max-type, Y \ ZS = {y ∈ Y : y /∈ZS} = BS. f S q f yp p ﬁ f (iii) If S is min-type, Y ∩ZS = {y ∈Y : y ∈ZS} = AS. If S is max-type, Y \ ZS = {y ∈ Y : y /∈ZS} = BS. Proof (i) and (ii) are easy to see. For part (iii), recall that FS is a (Y −S)-blocker. If S is min-type, then ZS is a minimum of FS, so ZS ∩(Y −S) = ∅by Lemma 8. Thus Proof (i) and (ii) are easy to see. For part (iii), recall that FS is a (Y −S)-blocker. If S is min-type, then ZS is a minimum of FS, so ZS ∩(Y −S) = ∅by Lemma 8. Thus ZS ∩Y = ZS ∩S = AS. Similarly, if FS is max-type, then Lemma 8 provides ZS ∩(Y −S) = (Y −S). Hence imilarly, if FS is max-type, then Lemma 8 provides ZS ∩(Y −S) = (Y −S). Hence ZS ∩Y = ZS ∩S ∪(Y −S) = AS ∪(Y −S) = Y \ BS, ZS ∩Y = ZS ∩S ∪(Y −S) = AS ∪(Y −S) = Y \ BS, ⊓⊔ 123 Order (2024) 41:401–418 413 Fig. 5 Exemplary construction of F(1,3) and F(1,2) for Y = {1, 2, 3} Fig. 5 Exemplary construction of F(1,3) and F(1,2) for Y = {1, 2, 3} Lemma 21 Let S be an ordered subset of Y of size k −1 and let S′ be a strict preﬁx of S. Then ZS ∩(Y −S′) /∈{∅, Y −S′}. Lemma 21 Let S be an ordered subset of Y of size k −1 and let S′ be a strict preﬁx of S. Then ZS ∩(Y −S′) /∈{∅, Y −S′}. Proof Note that \|Y −S\| = 1, so let Y −S = {b}. 3.4 Construction of the Family {(FS, ZS, AS, BS) : S ∈S} First we consider the case that \|S′\| = k −2, i.e. S = (S′, a) for some a ∈Y. Note that Y −S′ = {a, b}. We shall show that one of the two elements a, b is in ZS while the other is not. If S is min-type, we obtain from the construction that a ∈AS. By Lemma 20 (iii), AS = ZS ∩Y, so in particular a ∈ZS ∩(Y −S′). On the other hand, AS ⊆S, so b /∈AS = ZS ∩Y and thus b /∈ZS ∩(Y −S′). If S is max-type we can argue similarly. Note that a ∈BS. By Lemma 20 (iii), BS = Y \ ZS, so a /∈ZS ∩(Y −S′). Furthermore, BS ⊆S, so b /∈BS = Y \ ZS. Thus b ∈ZS ∩Y and hence b ∈ZS ∩(Y −S′). It remains to consider the case \|S′\| < k −2. Let S′′ = S[k −2] be the preﬁx of S of size k −2, then S′ is a preﬁx of S′′. Observe that Y −S′′ ⊆Y −S′. We already showed that ZS ∩(Y −S′′) /∈{∅, Y −S′′}, so in particular ZS ∩(Y −S′) /∈{∅, Y −S′}. ⊓⊔ 4 Proof of Theorem 4 Proof of Theorem 4 Let k and N be arbitrary integers with N ≥k, let n such that N = n + k. Let Y be a set on \|Y\| = k elements, say without loss of generality Y = {1, . . . , k}. Fix Z with Y ⊆Z and \|Z\| = N. Suppose that there is an N-free, critical Y-blocker F in Q(Z). In other words, suppose that the integer N is sufﬁciently large with respect to k such that there exists an F with these properties in Z. In the following we show that F contains at least k!2−k−1 vertices. Since \|F\| ≤2\|Z\|, it follows that n + k = N = \|Z\| ≥k log k, which implies that k ≤(1 + o(1))n/log n, i.e. N = n + k ≤n + (1 + o(1))n/log n. Then Theorem 9 provides the required bound. Next we argue that there exists a subposet in F with many vertices. Let S be the set of all ordered subsets of Y of size at most k −1. Consider the family {(FS, ZS, AS, BS) : S ∈S} given by Construction 19. Let S1 be the family of all ordered 123 414 Order (2024) 41:401–418 subsets of Y of size exactly k −1. We introduce two kinds of equivalence between elements in S1, type-equivalence and intersection-equivalence. In the following we will show the existence of a large subfamily S3 ⊆S1 such that its elements are pairwise type-equivalent but not intersection-equivalent. We further prove that the vertices {ZS : S ∈S3} induce a large antichain in Q(X ∪Y). Let S1, S2 ∈S1 be two ordered subsets of Y of size k −1. We say that S1 and S2 are type-equivalent if for any preﬁxes S′ 1 of S1 and S′ 2 of S2 of the same size, S′ 1 is min-type if and only if S′ 2 is min-type. Equivalently, S′ 1 is max-type if and only if S′ 2 is max-type. The ordered sets S1 and S2 are intersection-equivalent if for any same-sized preﬁxes S′ 1 of S1 and S′ 2 of S2, ZS′ 1 ∩Y = ZS′ 2 ∩Y. It is obvious that both notions deﬁne equivalence relations on S1. Note that intersection-equivalence of two ordered sets in S1 is a very strong property. It provides a good intuition to think of intersection-equivalent ordered sets as equal. 4 Proof of Theorem 4 Several technical parts of the proof, in particular in Claim 1, arise from the fact that there might be intersection-equivalent ordered sets which are distinct. Claim 1. There exists a subfamily S3 ⊆S1 of size at least 2−k−1k! such that any two distinct ordered sets S1, S2 ∈S3, are type-equivalent but not intersection-equivalent. Proof of Claim 1. Recall that \|S1\| = k!. For every i ∈{0, . . . , k −1} and for every S ∈S1, the preﬁx S[i] of S of size i is either min-type or max-type. By pigeonhole principle for ﬁxed i, there are at least \|S1\|/2 ordered subsets S ∈S1 such that all preﬁxes S[i] are of the same type. Inductively, we ﬁnd a subfamily S2 ⊆S1 of size at least 2−k\|S1\| such that for any ﬁxed i ∈{0, . . . , k −1}, all preﬁxes S[i], S ∈S2 have the same type. Equivalently, the elements of S2 are pairwise type-equivalent. In the following we show that each intersection-equivalence class in S2 has size at most 2. Thus by selecting a representative of each equivalence class we obtain a subfamily S3 as required. Given an arbitrary ﬁxed S1 ∈S2, consider an ordered set S2 ∈S2 such that S1 and S2 are intersection-equivalent, i.e. ZS′ 1 ∩Y = ZS′ 2 ∩Y for every two same-sized preﬁxes S′ 1 of S1 and S′ 2 of S2. Without loss of generality suppose that S1 = (1, 2, . . . , k −1) and Y −S1 = {k}. Let S2 = (y1, . . . , yk−1) and Y −S2 = {yk}. We shall show that yi = i for all but at most two indices i ∈[k], which implies that S2 is either equal to S1 or obtained from S1 by interchanging the two differing members, and therefore the intersection-equivalence class of S1 consists of at most 2 members. Since S1 and S2 are both in S2, i.e. type-equivalent, we know that for every i ∈{0, . . . , k− 1} either both S1[i] and S2[i] are min-type or both S1[i] and S2[i] are max-type. We enumerate the index set {0, . . . , k −1} as follows. Let i1, . . . , i p be the indices i ∈{0, . . . , k −1} such that S1[i] and S2[i] are min-type in increasing order. Similarly, let j1, . . . Similarly for S2 we obtain 2 we obtain (ZS2[i′] ∩Y) \ (ZS2[i] ∩Y) = AS2[i′] \ AS2[i] = {yi+1}. (ZS2[i′] ∩Y) \ (ZS2[i] ∩Y) = AS2[i′] \ AS2[i] = {yi+1}. Since S1 and S2 are intersection-equivalent, we ﬁnd that yi+1 = i + 1. Since S1 and S2 are intersection-equivalent, we ﬁnd that yi+1 = i + 1. We obtain that yiℓ+1 = iℓ+1 for every ℓ∈[p −1]. For j1, . . . , jq a symmetric argument j = jℓand j′ = jℓ+1 considering the set difference We obtain that yiℓ+1 = iℓ+1 for every ℓ∈[p −1]. For j1, . . . , jq a symmetric argument for j = jℓand j′ = jℓ+1 considering the set difference We obtain that yiℓ+1 = iℓ+1 for every ℓ∈[p −1]. Fo for j = jℓand j′ = jℓ+1 considering the set difference (ZS1[ j] ∩Y) \ (ZS1[ j′] ∩Y) = (Y \ BS1[ j]) \ (Y \ BS1[ j′]) = BS1[ j′] \ BS1[ j] = { j + 1 (ZS1[ j] ∩Y) \ (ZS1[ j′] ∩Y) = (Y \ BS1[ j]) \ (Y \ BS1[ j′]) = BS1[ j′] \ BS1[ j] = { j + 1} yields that y jℓ+1 = jℓ+1 for every ℓ∈[q−1]. Thus yi+1 = i+1 for all indices i ∈{0, . . . , k− 1} \ {i p, jq}, so S1 and S2 coincide in all but at most two members. As a consequence, S2 is either equal to S1 or obtained from S1 by interchanging the two differing members. Therefore the intersection-equivalence class of S1 consists of at most 2 ordered sets. As S1 was chosen arbitrary, every intersection-equivalence class of S2 has size at most 2. Select S3 ⊆S2 by choosing an arbitrary representative from each intersection-equivalence class, i.e. let S3 be the largest subfamily of S2 where every two distinct S2, S3 ∈S3 are not intersection-equivalent. Then \|S3\| ≥\|S2\|/2 ≥2−k−1\|S1\| = 2−k−1k!, which concludes the proof of Claim 1. which concludes the proof of Claim 1. Claim 2. The set {ZS : S ∈S3} has size \|S3\| = k!2−k−1. Remark Although not necessary for the proof of Theorem 4, Claim 2 holds in greater general- ity. 4 Proof of Theorem 4 , jq enumerate in increasing order the indices j ∈{0, . . . , k −1} where S1[ j] and S2[ j] are max-type. Note that {i1, . . . , i p} ∪{ j1, . . . , jq} = {0, . . . , k −1}. Now consider any two consecutive indices i = iℓand i′ = iℓ+1 for some ﬁxed ℓ∈[p−1]. Note that i < i′, so in particular i + 1 ≤i′ as well as i < k −1. We know from Lemma 20 (iii) that ZS1[i] ∩Y = AS1[i] and ZS1[i′] ∩Y = AS1[i′]. Our next step is to show that i + 1 is the unique element in the set difference of those two sets. Recall that S1[i] is min-type. In Construction 19 in the iterative step for S1[i + 1] = (S1[i], i + 1), we deﬁned AS1[i+1] = AS1[i] ∪{i + 1}. Observe that i + 1 /∈AS1[i] because by Lemma 20 (i) we know AS1[i] ⊆S1[i] = {1, . . . , i}. Furthermore, the fact that i + 1 ≤i′ implies that AS1[i+1] ⊆AS1[i′], so i + 1 ∈AS1[i′]. Lemma 20 (ii) provides that \|AS1[i]\| = ℓ and \|AS1[i′]\| = ℓ+ 1, thus (ZS1[i′] ∩Y) \ (ZS1[i] ∩Y) = AS1[i′] \ AS1[i] = {i + 1}. 123 123 Order (2024) 41:401–418 415 Similarly for S2 we obtain Analogously to the following proof, one can obtain that for every family S′ of ordered sets such that any two distinct members of S′ are type-equivalent and not intersection-equivalent, the vertices {ZS : S ∈S′} induce an antichain in Q(Z) of size \|S′\|. Proof of Claim 2. Recall that any two distinct, ordered sets in S3 are type-equivalent but not intersection-equivalent. We shall prove that for every two distinct S1, S2 ∈S3, the vertices ZS1 and ZS2 are distinct. We show an even stronger property: In fact, any two vertices ZS1 and ZS2 are incomparable. Assume towards a contradiction that ZS1 ⊆ZS2. Since S1 and S2 are not intersection-equivalent, there are same-sized preﬁxes S′ 1 of S1 and S′ 2 of S2 such that ZS′ 1 ∩Y ̸= ZS′ 2 ∩Y. Since S1 and S2 are type-equivalent, both S′ 1 and S′ 2 have the same type, suppose that they are min-type. First we argue that the sets ZS′ 1 ∩Y and ZS′ 2 ∩Y are not comparable. Lemma 20 (iii) shows that ZS′ 1 ∩Y = AS′ 1 and ZS′ 2 ∩Y = AS′ 2. Type-equivalence implies that pairs of same-sized preﬁxes of S′ 1 and S′ 2 always have the same type. Thus Lemma 20 (ii) yields that \|AS′ 1\| = \|AS′ 2\|. We obtain that ZS′ 1 ∩Y = AS′ 1 and ZS′ 2 ∩Y = AS′ 2 are distinct but of the same size, consequently ZS′ 1 ∩Y and ZS′ 2 ∩Y are not comparable. 1 2 If S′ 1 = S1 and S′ 2 = S2, then ZS1 ∩Y and ZS2 ∩Y are incomparable, and so ZS1\|\|ZS2, a contradiction to our assumption that ZS1 ⊆ZS2. For the remaining proof suppose that the size \|S′ 1\| = \|S′ 2\| is strictly less than k −1. We will show that there is a copy of N in F contradicting the deﬁnition of F to be an N-free poset. Let Y′ = Y −S′ 2, note that FS′ 2 is a Y′-blocker. Since ZS′ 1 ∩Y and ZS′ 2 ∩Y are not comparable, there exists an element a ∈ZS′ 1 ∩Y with a /∈ZS′ 2. Lemma 8 (ii) yields the existence of a vertex U ∈FS′ 2 with U ∩Y′ = Y′ \ {a}. Now we verify that ZS′ 1, ZS2, ZS′ 2 and U form a copy of N in F, see Fig. 6. • ZS′ 2 ⊆ZS2 because ZS′ 2 is a minimum of FS′ 2 and ZS2 ∈FS2 ⊆FS′ 2 by Lemma 20 (i). • ZS′ 1 ⊆ZS1 ⊆ZS2 as ZS′ 1 is a minimum of FS′ 1 and ZS1 ∈FS1 ⊆FS′ 1 by Lemma 20 (i). Similarly for S2 we obtain • ZS′ 2 ⊆ZS2 because ZS′ 2 is a minimum of FS′ 2 and ZS2 ∈FS2 ⊆FS′ 2 by Lemma 20 (i). • ZS′ 1 ⊆ZS1 ⊆ZS2 as ZS′ 1 is a minimum of FS′ 1 and ZS1 ∈FS1 ⊆FS′ 1 by Lemma 20 (i). 123 Order (2024) 41:401–418 416 Fig. 6 Copy of N constructed in the proof of Claim 2 Fig. 6 Copy of N constructed in the proof of Claim 2 • ZS′ 1\|\|ZS′ 2 because ZS′ 1 ∩Y and ZS′ 2 ∩Y are not comparable. • ZS′ 1\|\|ZS′ 2 because ZS′ 1 ∩Y and ZS′ 2 ∩Y are not comparable. 1 2 1 2 • ZS′ 2 ⊆U because ZS′ 2 is a minimum of FS′ 2 and by deﬁnition U ∈FS′ 2. 2 2 2 2 • Note that a /∈U and a ∈ZS′ 1, so ZS′ 1 ⊈U. Since ZS′ 2 ⊈ZS′ 1 but ZS′ 2 ⊆U, transitivity yields U ⊈ZS′ 1. Therefore U and ZS′ 1 are incomparable. 2 2 2 2 • Note that a /∈U and a ∈ZS′ 1, so ZS′ 1 ⊈U. Since ZS′ 2 ⊈ZS′ 1 but ZS′ 2 ⊆U, transitivity yields U ⊈ZS′ 1. Therefore U and ZS′ 1 are incomparable. 1 1 • We know that a /∈U but a ∈ZS2, thus ZS2 ⊈U. Note that ZS2 ∩Y′ ̸= Y′ \ {a} since a ∈ZS2. Furthermore, Lemma 20 provides that ZS2 ∩Y′ ̸= Y′. Thus ZS2 ∩Y′ is not a superset of Y′ \ {a} = U ∩Y′, therefore U ⊈ZS2. We obtain that U\|\|ZS2. • The four vertices are distinct because otherwise we ﬁnd an immediate contradiction to one of the above relations. • The four vertices are distinct because otherwise we ﬁnd an immediate contradiction to one of the above relations. Thus, there is a copy of N in F, which is a contradiction to the fact that F is N-free. Thus, there is a copy of N in F, which is a contradiction to the fact that F is N-free. Similarly for S2 we obtain If S′ 1 and S′ 2 are max-type, a symmetric argument can be applied: As a ﬁrst step, it follows similarly that ZS′ 1 ∩Y and ZS′ 2 ∩Y are incomparable, and then for Y′ = Y −S′ 1 and for a vertex U ∈FS′ 1 with U ∩Y′ = {a} we ﬁnd a copy of N on vertices ZS′ 1, ZS1, ZS′ 2 and U, which is a contradiction as before. This concludes the proof of Claim 2. Claims2impliestheexistenceofasubposetof F ofsizeatleastk!2−k−1.Since F ⊆Q(Z), we know that k!2−k−1 ≤\|F\| ≤2\|Z\|, so \|Z\| ≥log k! 2k+1 ≥log kk 2k+1ek ≥k(log(k) −c), for a ﬁxed constant c > 0. Recall that \|Z\| = n + k, so we obtain that n ≥k log k − 1 + c k, which implies that k ≤ 1 + o(1) n log n . Then Theorem 9 provides that for a ﬁxed constant c > 0. Recall that \|Z\| = n + k, so we obtain that n ≥k log k − 1 + c k, which implies that k ≤ 1 + o(1) n log n . Then Theorem 9 provides that for a ﬁxed constant c > 0. Recall that \|Z\| = n + k, so we obtain that n ≥ which implies that k ≤ 1 + o(1) n log n . Then Theorem 9 provides that R(N, Qn) ≤N = n + k ≤n + 1 + o(1) n log n . The lower bound on R(N, Qn) follows from Theorem 3. ⊓⊔ ⊓⊔ The lower bound on R(N, Qn) follows from Theorem 3. The lower bound on R(N, Qn) follows from Theorem 3. ⊓⊔ Competing of interest The authors declare no competing interests. Competing of interest The authors declare no competing interests. Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/. 5 Concluding Remarks In this paper we showed that R(N, Qn) ≤n + n log n . A key ingredient in our approach is Theorem 9 where we showed a connection between the poset Ramsey number of R(P, Qn) 123 Order (2024) 41:401–418 417 for a poset P and the extremal function m P(n) deﬁned as for a poset P and the extremal function m P(n) deﬁned as m P(n):= min{N : there is no P-freeY-blocker in Q([N]) for some Y ⊆[N], \|Y\|= N−n}. A Y-blocker can be seen as a transversal of a set of speciﬁc Boolean lattices, and is related to other notions of transversals, e.g. clique-transversals in graphs as introduced by Erd˝os, Gallai and Tuza [9]. Seen in this context, research on m P(n) or similar extremal functions on Y-blockers might be of independent interest. Acknowledgements We thank the referee for feedback on the manuscript. Acknowledgements We thank the referee for feedback on the manuscript. Author Contributions M.A. and C.W. wrote and reviewed the manuscript. Author Contributions M.A. and C.W. wrote and reviewed the manuscript. Funding Open Access funding enabled and organized by Projekt DEAL. The research of both authors was partially supported by Deutsche Forschungsgemeinschaft, grant FKZ AX 93/2-1. Funding Open Access funding enabled and organized by Projekt DEAL. The research of both authors was partially supported by Deutsche Forschungsgemeinschaft, grant FKZ AX 93/2-1. Data Availability Data sharing not applicable to this article as no datasets were generated or analysed during the current study. Data Availability Data sharing not applicable to this article as no datasets were generated or analysed during the current study. 11. Grósz, D., Methuku, A., Tompkins, C.: Ramsey numbers of Boolean lattices. Bull. Lond. Math. Soc. 55(2), 914–932 (2023) Declarations Conﬂicts of interest None. Competing of interest The authors declare no competing interests. 10. Falgas-Ravry, V., Markström, K., Treglown, A., Zhao, Y.: Existence thresholds and Ramsey properties of random posets. Random. Struct. Algor. 57(4), 1097–1133 (2020) References 1. Axenovich, M., Walzer, S.: Boolean lattices: Ramsey properties and embeddings. Order 34(2), 287–298 (2017) ( ) 2. Axenovich, M., Winter, C.: Poset Ramsey numbers: large Boolean lattice versus a ﬁxed poset. Combina- torics, Probability and Computing pp. 1–16 (2023) 3. Bohman, T., Peng, F.: A construction for boolean cube ramsey numbers. Order (2022) 4. Carroll, T., Katona, G.: Bounds on maximal families of sets not containing three sets with A ∩B ⊂ A¬ ⊂B. Order 25, 229–236 (2008) 5. Chang, F.-H., Gerbner, D., Li, W.-T., Methuku, A., Nagy, D., Patkós, B., Vizer, M.: Rainbow Ramsey problems for the Boolean lattice. Order 39, 453–463 (2022) p ( ) 6. Chen, H.-B., Chen, W.-H., Cheng, Y.-J., Li, W.-T., Liu, C.-A.: Ramsey properties for V-shaped posets in the boolean lattices. Preprint, available at arXiv:2108.08033v1, (2021) 7. Chen, H.-B., Cheng, Y.-J., Li, W.-T., Liu, C.-A.: The boolean rainbow ramsey number of anticha b l d h i El i J l f C bi i 27(4) (2020) 7. Chen, H.-B., Cheng, Y.-J., Li, W.-T., Liu, C.-A.: The boolean rainbow ramsey number of antichain boolean posets and chains. Electronic Journal of Combinatorics 27(4), (2020) 7. Chen, H.-B., Cheng, Y.-J., Li, W.-T., Liu, C.-A.: The boolean rainbow ramse boolean posets and chains. Electronic Journal of Combinatorics 27(4), (2020) boolean posets and chains. Electronic Journal of Combinatorics 27(4), (2020) 8. Cox, C., Stolee, D.: Ramsey numbers for partially-ordered sets. Order 35(3), 557–579 (2018) 9. Erd˝os, P., Gallai, T., Tuza, Zs.: Covering the cliques of a graph with vertices. Discrete Math.108, 279–289, 1992 10. Falgas-Ravry, V., Markström, K., Treglown, A., Zhao, Y.: Existence thresholds and Ramsey properties of random posets. Random. Struct. Algor. 57(4), 1097–1133 (2020) 11. Grósz, D., Methuku, A., Tompkins, C.: Ramsey numbers of Boolean lattices. Bull. Lond. Math. Soc. 55(2), 914–932 (2023) 123 12 418 Order (2024) 41:401–418 12. Keszegh, B., Lemons, N., Martin, R.R., Pálvölgyi, D., Patkós, B.: Induced and non-induced poset satu tion problems. Journal of Combinatorial Theory, Series A (2021) p y 13. Lu, L., Milans, K.: Set families with forbidden subposets. J. Comb. Theory Ser. 136, 126–142 (2015 14 h C b f l i O d 39 1 1 185 (2022) 13. Lu, L., Milans, K.: Set families with forbidden subposets. J. Comb. Theory Ser. 136, 126–142 (2015) 14. Lu, L., Thompson, C.: Poset Ramsey Numbers for Boolean Lattices. Order 39, 171–185 (2022) 14. References Lu, L., Thompson, C.: Poset Ramsey Numbers for Boolean Lattices. Order 39, 171–185 (2022) Lu, L., Thompson, C.: Poset Ramsey Numbers for Boolean Lattices. Order 39, 171–185 (2022) Méroueh A : A LYM inequality for induced posets J Comb Theory Ser A 155 398 417 (2018) p y ( ) 15. Méroueh, A.: A LYM inequality for induced posets. J. Comb. Theory Ser. A 155, 398–417 (2018) 16. Methuku, A., Pálvölgyi, D.: Forbidden hypermatrices imply general bounds on induced forbidden s poset problems. Comb. Probab. Comput. 26(4), 593–602 (2017) p p p 17. Methuku, A., Tompkins, C.: Exact forbidden subposet results using chain decompositions of the cy Electronic Journal of Combinatorics 22(4), (2015) 18. Tomon, I.: Forbidden induced subposets of given height. J. Comb. Theory Ser. A 161, 537–562 (2019) 18. Tomon, I.: Forbidden induced subposets of given height. J. Comb. Theory Ser. A 161, 537–562 (2019) 19. Valdes, J.: Parsing ﬂowcharts and series-parallel graphs. Technical report STAN-CS-78-682, Computer 18. Tomon, I.: Forbidden induced subposets of given height. J. Comb. Theory Ser. A 161, 537–562 (2019) 19. Valdes, J.: Parsing ﬂowcharts and series-parallel graphs. Technical report STAN-CS-78-682, Computer Science Department, Stanford University, Stanford (1978) p g g y 19. Valdes, J.: Parsing ﬂowcharts and series-parallel graphs. Technical report STAN-CS-78-682, Computer Science Department, Stanford University, Stanford (1978) Parsing ﬂowcharts and series-parallel graphs. Techn 19. Valdes, J.: Parsing ﬂowcharts and series-parallel graphs. Tec Science Department, Stanford University, Stanford (1978) 20. Walzer, S.: Ramsey Variant of the 2-Dimension of Posets. In: Master Thesis, karlsruhe institute of te nology (2015) 21. Winter, C.: Poset Ramsey number R(P, Qn). I. Complete multipartite posets. Order (2023) 21. Winter, C.: Poset Ramsey number R(P, Qn). I. Complete multipartite posets. Order (2023) Publisher’s Note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional afﬁliations. Publisher’s Note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional afﬁliations. 123
https://openalex.org/W2804945208	https://link.springer.com/content/pdf/10.1007%2F978-3-319-90415-3_4.pdf	English	null	Scalable Traffic Quality and System Efficiency Indicators Towards Overall Telecommunication System’s QoE Management	Lecture notes in computer science	2,018	cc-by	9,149	Scalable Traﬃc Quality and System Eﬃciency Indicators Towards Overall Telecommunication System’s QoE Management Stoyan Poryazov1(✉) , Emiliya Saranova1,2 Stoyan Poryazov1(✉) , Emiliya Saranova1,2 , and Ivan Ganchev1,3,4 1 Institute of Mathematics and Informatics, Bulgarian Academy of Sciences, Soﬁa, Bulgaria stoyan@math.bas.bg, emiliya@cc.bas.bg 2 University of Telecommunications and Post, Soﬁa, Bulgaria 3 University of Limerick, Limerick, Ireland ivan.ganchev@ul.ie 4 University of Plovdiv “Paisii Hilendarski”, Plovdiv, Bulgaria 1 Institute of Mathematics and Informatics, Bulgarian Academy of Sciences, Soﬁa, Bulgaria stoyan@math.bas.bg, emiliya@cc.bas.bg 2 University of Telecommunications and Post, Soﬁa, Bulgaria 3 University of Limerick, Limerick, Ireland ivan.ganchev@ul.ie 4 University of Plovdiv “Paisii Hilendarski”, Plovdiv, Bulgaria 4 University of Plovdiv “Paisii Hilendarski”, Plovdiv, Bulgaria Abstract. Conceptual and analytical models of an overall telecommunication system are utilized in this chapter for the definition of scalable indicators towards Quality of Service (QoS) monitoring, prediction, and management. The telecom‐ munication system is considered on different levels – service phase, service stage, network, and overall system. The network itself is presented in seven service stages – A-user, A-terminal, Dialing, Switching, B-terminal Seizure, B-terminal, and B- user, each having its own characteristics and specifics. Traffic quality indicators are proposed on each level. Two network cost/quality ratios are proposed – mean and instantaneous – along with illustrative numerical predictions of the latter, which could be useful for dynamic pricing policy execution, depending on the network load. All defined indicators could be considered as sources for Quality of Experi‐ ence (QoE) prediction. Keywords: Overall telecommunication system · Performance model Dynamic quality of service (QoS) · Telecommunication subservices Diﬀerentiated QoS subservice indicator · QoS prediction · Human factors of QoS Instantaneous Cost/Quality Ratio · Quality of Experience (QoE) © The Author(s) 2018 I. Ganchev et al. (Eds.): Autonomous Control for a Reliable Internet of Services, LNCS 10768, pp. 81–103, 2018. https://doi.org/10.1007/978-3-319-90415-3_4 2 Service Phase Concept and Traﬃc Quality Indicators The conceptual model utilized in this chapter is described in detail in the Chapter “Conceptual and Analytical Models for Predicting the Quality of Service of Overall Telecommunication Systems” of this book. It consists of ﬁve levels: (1) overall tele‐ communication system and its environment; (2) overall telecommunication network; (3) service stages; (4) service phases; and (5) basic virtual devices. In the following subsections, the concepts of ‘service phase’ and ‘service stage’ are elaborated. 1 Introduction Starting from 2010, e.g. [1], a new attitude towards the Quality of Service (QoS) has become dominant, namely to consider QoS and Quality of Experience (QoE) as goods, and the usage of Experience Level Agreement (ELA) [2] has started to be discussed. The importance of the teletraﬃc models, particularly of the overall QoS indicators, for QoE assessment is emphasized by Fiedler [3]. Until now, however, the usage of performance models of overall telecommunication systems was not very popular. This chapter utilizes the models, elaborated in the Chapter “Conceptual and Analytical Models for Predicting the Quality of Service of Overall Telecommunication Systems” of this book, for the deﬁnition of scalable QoS indicators towards overall S. Poryazov et al. 82 telecommunication system’s QoS monitoring, prediction, and management. Some indi‐ cators reﬂect predominantly the users’ experience. All deﬁned indicators depend on human (users’) characteristics and technical characteristics, and may be considered as sources for QoE prediction. For this, in Sect. 2, traﬃc characterization of a service in a real device (service phase) is ﬁrst elaborated. Deﬁnitions of served-, carried-, parasitic-, ousted-, and oﬀered carried traﬃc are proposed, based on the ITU-T deﬁnitions, and eight service phase traﬃc quality indicators are proposed. In Sect. 3, the service stage concept is developed and corresponding traﬃc quality indicators are deﬁned. In Sect. 4, telecommunication system and network eﬃciency indicators are proposed as follows: eight indicators – on the service stage level, ﬁve indicators – on the network level, and three indicators – on the overall system level. The relationship between indi‐ cators on the service stage-, network-, and system level are described. A comparison with classical network eﬃciency indicators is made. The applicability of the approach and results obtained for deﬁning other indicators, as well as for numerical prediction of indicators’ values, is shown. In Sect. 5, two network cost/quality ratios are proposed – mean and instantaneous – and illustrative numerical predictions of the latter are presented, which may be useful for dynamic pricing policy execution, depending on the network load. In the Conclusion, possible directions for future research are brieﬂy discussed. 1 In the expressions, formulas and ﬁgures, the sign (.) is used only as a separator and NOT as a sign of multiplication. The multiplication operation is indicated by a gap between multiplied variables, e.g. X Y. 2.1 Service Phase Based on the ITU-T deﬁnition of a service, provided in [4] (Term 2.14), i.e. “A set of functions oﬀered to a user by an organization constitutes a service”, we propose the following deﬁnition of a service phase. Deﬁnition 1: The Service Phase is a service presentation containing: Deﬁnition 1: The Service Phase is a service presentation containing: • One of the functions, realizing the service, which is considered indivisible; • All modeled reasons for ending/ﬁnishing this function, i.e. the causal structure of the function; • Hypothetic characteristics, related to the causal structure of the function (a well- known example of a hypothetic characteristic is the oﬀered traﬃc concept). Scalable Traﬃc Quality and System Eﬃciency Indicators 83 Following the Structural Normalization and Causal Structure approaches described in the Chapter “Conceptual and Analytical Models for Predicting the Quality of Service of Overall Telecommunication Systems” of this book, we may present the service phase in device s by means of k + 1 basic virtual causal devices, each representing a diﬀerent reason for ending this service phase (Fig. 1). Offered Carried (ofr.crr.Ys) srv.Fs ous.Yk Served (srv.Ys) 2 ous.Y2 ous.Y1 crr.Ys 1 k prs.Yk Parasitic (prs.Ys) Ousted (ost.Ys) prs.Y2 prs.Y1 Fig. 1. Traﬃc characterization of a service phase, represented as device S, by means of k + 1 basic virtual causal devices. Fig. 1. Traﬃc characterization of a service phase, represented as device S, by means of k + 1 basic virtual causal devices. In Fig. 1, only one causal device represents successful completion of the service in device s – with carried traﬃc (crr.Ys1), whereas the remaining causal devices represent k diﬀerent reasons for unsuccessful ending of the service – respectively with traﬃcs prs.Y1, prs.Y2, … , prs.Yk. Generalizing, for more precise traﬃc characterization in a pool of resources, we propose the following deﬁnitions. Deﬁnition 2: The Served Traﬃc in a pool of resources is the traﬃc, occupying (using) resources in the pool. In Fig. 1, the served traﬃc in device s (srv.Ys) is the following sum: srv.Ys = prs.Y1 + rs.Y2 + … + prs.Yk + crr.Ys . (1) (1) Deﬁnition 3: The Carried Traﬃc in a pool of resources is the traﬃc, which was success‐ fully served in the pool (and carried to the next service phase). In Fig. 1, the carried traﬃc in device s is crr.Ys. In Fig. 1, the carried traﬃc in device s is crr.Ys. but the service times are diﬀerent. but the service times are diﬀerent. but the service times are diﬀerent. The hypothetic service time for every ousted device (ous.Ti) equals the carried service time (crr.Ts): ous.Ti = crr.Ts, for i = [1, k] . (3) (3) The ousted traﬃc is a hypothetic one with the following intensity: ous.Yi = prs.Fi crr.Ts, for i = [1, k] . (4) ous.Yi = prs.Fi crr.Ts, for i = [1, k] . (4) (4) 2.1 Service Phase Deﬁnition 4: The Parasitic Traﬃc in a pool of resources is the traﬃc, which was unsuc‐ cessfully served in the pool. Deﬁnition 4: The Parasitic Traﬃc in a pool of resources is the traﬃc, which was unsuc‐ cessfully served in the pool. Deﬁnition 4: The Parasitic Traﬃc in a pool of resources is the traﬃc, which was unsuc‐ cessfully served in the pool. 1 In the expressions, formulas and ﬁgures, the sign (.) is used only as a separator and NOT as a sign of multiplication. The multiplication operation is indicated by a gap between multiplied variables, e.g. X Y. 84 S. Poryazov et al. In Fig. 1, each of traﬃcs prs.Y1, prs.Y2, … , prs.Yk is a parasitic one. Parasitic traﬃc occupies real resources but not for an eﬀective service execution. In Fig. 1, each of traﬃcs prs.Y1, prs.Y2, … , prs.Yk is a parasitic one. Parasitic traﬃc occupies real resources but not for an eﬀective service execution. In Deﬁnitions 2 and 3, the served- and carried traﬃc are diﬀerent terms, despite the ITU-T deﬁnition of the carried traﬃc as “The traﬃc served by a pool of resources” ([5], Term 5.5). We believe that this distinction leads to a better and more detailed traﬃc- and QoS characterization. Deﬁnition 5: The Ousted Traﬃc is the traﬃc that would be carried, if there is no unsuc‐ cessful service ending in the pool of resources. In Fig. 1, each parasitic traﬃc prs.Y1, prs.Y2, … , prs.Yk ousts a corresponding traﬃc that would be carried, if there is no unsuccessful service ending of the corre‐ sponding type: ous.Y1, ous.Y2, … , ous.Yk. The ﬂow intensity to a parasitic device and the corresponding ousted device is the same by deﬁnition, i.e.: ous.Fi = prs.Fi, for i = [1, k], (2) (2) he service times are diﬀerent. 2.3 Oﬀered Carried Traﬃc Deﬁnition 6: The Oﬀered Carried Traﬃc (ofr.crr.Ys) in a pool s of resources is the sum of the carried traﬃc (crr.Ys) and ousted traﬃc (ous.Ys) in the pool: ofr.crr.Ys = ous.Ys + crr.Ys. (10) (10) From (10), (6), (4), (8), prs.Fs = k∑ i=1 prs.Fi and crr.Ys = crr.Fs crr.Ts, the following formula could be obtained: ofr.crr.Ys = srv.Fs crr.Ts. ofr.crr.Ys = srv.Fs crr.Ts. Deﬁnition 6 is analogous to the ITU-T deﬁnition of an Equivalent Oﬀered Traﬃc [7] but considers the traﬃc related to the carried call attempts, whereas the ITU-T deﬁnition considers the traﬃc that would be served. 2.2 Causal Generalization In the Chapter “Conceptual and Analytical Models for Predicting the Quality of Service of Overall Telecommunication Systems” of this book, the causal presentation and causal aggregation are discussed. The causal aggregation is understood as an aggregation of all cases in the model, corresponding to diﬀerent reasons for service ending (referred to as unsuccessful cases further in this chapter). Here a causal generalization is proposed, as an aggregation of all unsuccessful cases (prs.Ys). Besides this, an aggregation of all cases of ousted traﬃc (ous.Ys) could be used: prs.Ys= k∑ i=1 prs.Yi; (5) ous.Ys= k∑ i=1 ous.Yi. (6) (5) (6) By Deﬁnition 2, the served traﬃc is a sum of the parasitic and carried traﬃc (c.f. Fig. 1): Scalable Traﬃc Quality and System Eﬃciency Indicators 85 Scalable Traﬃc Quality and System Eﬃciency Indicators 85 srv.Ys = prs.Ys + crr.Ys; (7) srv.Ys = prs.Ys + crr.Ys; (7) srv.Fs = prs.Fs + crr.Fs . (8) srv.Ys = prs.Ys + crr.Ys; (7) (7) srv.Fs = prs.Fs + crr.Fs . (8) (8) If the system is considered as being in a stationary state, by using the Little’s formula [6] we have: prs.Ys = prs.Fs prs.Ts and crr.Ys = crr.Fs crr.Ts . Hence: srv.Ys = srv.Fs srv.Ts = prs.Fs prs.Ts + crr.Fs crr.Ts . (9) (9) Formulas (7), (8), and (9) illustrate the advantage of the traﬃc qualiﬁers – the nota‐ tion is invariant to the number of cases considered in a service phase. 2.4 Traﬃc Quality Indicators This indicator estimates the aggregated, by all reasons, ratio of missed beneﬁts to the ineﬀective service in a service phase. Indicator 8: Causal Ousted/Parasitic Traﬃc Ratio – this is the ratio of the ousted traﬃc, due to reason i, to the parasitic traﬃc due to the same reason. From Deﬁnition 5 and Formula (2): I8(i) = ous.Yi prs.Yi = ous.Ti prs.Ti . (18) (18) This indicator gives another important estimation of a reason for ineﬀective service in a service phase. 2.4 Traﬃc Quality Indicators Indicator 1: Oﬀered Carried Traﬃc Eﬃciency – the ratio of the carried traﬃc, in a service phase, to the oﬀered carried traﬃc: I1 = crr.Ys ofr.crr.Ys = 1 − ous.Ys ofr.crr.Ys. (11) (11) Indicator 2: Causal Ousted Importance – the ratio of the ousted traﬃc due to reason i (ous.Yi) to the oﬀered carried traﬃc of a service phase (ofr.crr.Ys): Indicator 2: Causal Ousted Importance – the ratio of the ousted traﬃc due to reason i (ous.Yi) to the oﬀered carried traﬃc of a service phase (ofr.crr.Ys): I2(i) = ous.Yi ofr.crr.Ys . (12) (12) This indicator allows the estimation of missed beneﬁts due to reason i and therefore the necessity of countermeasures against this reason. Indicator 3: Ousted Traﬃc Importance – the sum of all causal ousted importance indicators of a service phase. From Fig. 1, and Formulas (6) and (11), it is: S. Poryazov et al. S. Poryazov et al. 86 I3 = k∑ i=1 I2(i) = k∑ i=1 ous.Yi ofr.crr.Ys = ous.Ys ofr.crr.Ys = 1 − crr.Ys ofr.crr.Ys. (13) (13) Indicator 4: Service Eﬃciency – the ratio of the carried traﬃc to the served traﬃc: I4 = crr.Ys srv.Ys = 1 −prs.Ys srv.Ys. (14) (14) Indicator 5: Causal Parasitic Importance – the ratio of the parasitic traﬃc due to reason i (prs.Yi) to the served traﬃc of a service phase (srv.Ys): Indicator 5: Causal Parasitic Importance – the ratio of the parasitic traﬃc due to reason i (prs.Yi) to the served traﬃc of a service phase (srv.Ys): I5(i) = prs.Yi srv.Ys. (15) (15) This indicator allows the estimation of an ineﬀective service due to a reason and therefore the necessity of countermeasures against this reason. y g Indicator 6: Parasitic Traﬃc Importance – the sum of all causal parasitic impor‐ tance indicators of a service phase. From Fig. 1, and Formulas (5) and (14), it is: I6 = k∑ i=1 I5(i) = k∑ i=1 prs.Yi srv.Ys =prs.Ys srv.Ys = 1 −crr.Ys srv.Ys. (16) (16) Indicator 7: Ousted/Parasitic Traﬃc Ratio – this is the ratio of the ousted traﬃc to the parasitic traﬃc: Indicator 7: Ousted/Parasitic Traﬃc Ratio – this is the ratio of the ousted traﬃc to the parasitic traﬃc: I7 = ous.Ys prs.Ys . (17) (17) This indicator estimates the aggregated, by all reasons, ratio of missed beneﬁts to the ineﬀective service in a service phase. 3 Service Stage Concept and Traﬃc Quality Indicators Deﬁnition 7: The Service Stage is a service presentation containing: • One service phase, realizing one function of the service; • One service phase, realizing one function of the service; Scalable Traﬃc Quality and System Eﬃciency Indicators 87 • All auxiliary service phases that directly support this function realization but are not part of the realized function itself. • All auxiliary service phases that directly support this function realization but are not part of the realized function itself. • All auxiliary service phases that directly support this function realization but are not part of the realized function itself. Examples of auxiliary service phases are the entry, exit, buﬀer, and queue virtual devices. The performance of the auxiliary devices depends directly on the service phase, realizing a function of the service. The service stage concept allows the division of the overall telecommunication service into subservices and therefore makes easier the system modeling process. 3.1 Service Stage For simplicity in this subsection, the simplest possible service stage, consisting of only two service phases, is considered (Fig. 2). For more complex service stages with more phases, please refer to the Chapter “Conceptual and Analytical Models for Predicting the Quality of Service of Overall Telecommunication Systems” of this book. srv.Fs blc.Ys ofr.Fg Blocked (blc.Ys) Served (srv.Ye) ous.Ys Served (srv.Ys) prs.Ys crr.Ys ofr.crr.Ys ous.Ye crr.Ye prs.Ye ofr.crr.Ye Entrance Device Service Device crr.Fg prs.Fg Stage g Fig. 2. A service stage g, consisting of Entrance and Service phases. Fig. 2. A service stage g, consisting of Entrance and Service phases. The service stage g, in Fig. 2. consists of Entrance and Service phases (represented by corresponding virtual devices). The Entrance device (e) may check the service request (call) attempt for having the relevant admission rights, whereas the Service device (s) checks for service availability or existence of free service resources, etc. Let ofr.Fg is the ﬂow intensity of the service request attempts oﬀered to this stage, crr.Fg – the inten‐ sity of the outgoing carried ﬂow, prs.Fg – the ﬂow intensity of the parasitic served requests, and prs.Fe – the intensity of the parasitic call attempts ﬂow. Then from Fig. 2 we have: S. Poryazov et al. 88 prs.Fe = ofr.Fg prs.Pe, (19) (19) where prs.Pe is the probability of directing the service request attempts to the general‐ ized parasitic service in device e. By analogy: where prs.Pe is the probability of directing the service request attempts to the general‐ ized parasitic service in device e. By analogy: prs.Fs = ofr.Fg (1 −prs.Pe) prs.Ps. (20) (20) The total parasitic ﬂow in service stage g is: The total parasitic ﬂow in service stage g is: prs.Fg = prs.Fe + prs.Fs. (21) (21) The carried traﬃc (crr.Yg) in service stage g is a sum of the carried traﬃc in devices e and s: crr.Yg = crr.Ye + crr.Ys = crr.Fe crr.Te + crr.Fs crr.Ts, (22) crr.Yg = crr.Ye + crr.Ys = crr.Fe crr.Te + crr.Fs crr.Ts, (22) (22) where: crr.Fe = ofr.Fg (1 −prs.Pe); (23) crr.Fs = ofr.Fg (1 −prs.Pe) (1 −prs.Ps). (24) (23) (24) The total carried traﬃc in service stage g is: The total carried traﬃc in service stage g is: crr.Yg = ofr.Fg (1 −prs.Pe) ( crr.Te + (1 −prs.Ps) crrTs). 3.1 Service Stage (25) (25) The estimation of the carried traﬃc in a service stage could be problematic due to the fact that some of the carried service requests attempts in the ﬁrst device (e) are not carried to the next device (s), i.e. they become parasitic service requests with probability prs.Ps (c.f. Fig. 2). Based on the ITU-T deﬁnition of ‘eﬀective traﬃc’ [5], i.e. as “The traﬃc corre‐ sponding only to the conversational portion of eﬀective call attempts”, we propose here the Eﬀective Carried Traﬃc concept. Deﬁnition 8: The Eﬀective Carried Traﬃc in a service stage is the traﬃc corresponding to the service request attempts leaving the stage with a fully successful (carried) service. In Fig. 2, the eﬀective carried traﬃc (eﬀ.crr.Yg) of service stage g is: eﬀ.crr.Yg = eﬀ.crr.Fg eﬀ.crr.Tg, (26) (26) eﬀ.crr.Yg = eﬀ.crr.Fg eﬀ.crr.Tg, where: Scalable Traﬃc Quality and System Eﬃciency Indicators Note the diﬀerence between (25) and (29), i.e. in general, in a service stage, the eﬀective carried traﬃc is less than the carried traﬃc. The Oﬀered Traﬃc is a fundamental teletraﬃc engineering concept. We use the ITU-T deﬁnition of the Equivalent Oﬀered Traﬃc [7], i.e. “Oﬀered traﬃc, to a pool of resources, is the sum of carried and blocked traﬃc of this pool”. The blocked traﬃc corresponds to the blocked attempts, as per Deﬁnition 2.8 in [5]: “Blocked call attempt: A call attempt that is rejected owing to a lack of resources in the network”. This deﬁnition, however, is too narrow to be applied directly to blocked service request attempts as it does not include most of the reasons for rejection, including access control, service unavailability, called terminal unavailability or busyness, and many others. Thus we propose the following extension of it. Deﬁnition 9: The Blocked Service Request Attempt is a service request attempt with rejected service, in the intended pool of resources, due to any reason. Blocked traﬃc is a service stage concept because it considers blocking of service requests before entering the service phase, or in other words, blocking that occurs in another virtual device before the corresponding service device. In Fig. 2, blocking occurs in the Entrance device. The blocked traﬃc (blc.Ys) corre‐ sponds to the service request attempts oﬀered to service stage g (ofr.Fg), but not belonging to the served attempts (srv.Fs). From Fig. 2 and the Little’s theorem, we obtain: blc.Ys = blc.Fs blc.Ts. (30) (30) The service request attempts that are not carried in phase e, and hence are rejected to the next service phase, are considered parasitic in phase e. For the intensity of the blocked attempts, the following equality holds: blc.Fs = prs.Fe = ous.Fe = ofr.Fg prs.Pe. (31) (31) The oﬀered traﬃc is a hypothetic one “that would be served” if it is not blocked, and therefore: blc.Ts = srv.Ts (32) (32) From (30), (31), and (32), we obtain the following formula: blc.Ys = ofr.Fg prs.Pe srv.Ts, (33) (33) which is valid for the generalized reason for service request attempts rejection in service phase e (i.e. in the Entrance device). which is valid for the generalized reason for service request attempts rejection in service phase e (i.e. in the Entrance device). From the deﬁnition of the equivalent oﬀered traﬃc, Fig. where: eﬀ.crr.Fg = crr.Fs = ofr.Fg (1 −prs.Pe)(1 −prs.Ps); (27) eﬀ.crr.Tg = crr.Te + crr.Ts. (28) (27) (28) From (26), (27), and (28), we obtain: From (26), (27), and (28), we obtain: From (26), (27), and (28), we obtain: From (26), (27), and (28), we obtain: eﬀ.crr.Yg = ofr.Fg (1 −prs.Pe)(1 −prs.Ps)(crr.Te + crr.Ts). (29) (29) Scalable Traﬃc Quality and System Eﬃciency Indicators 89 3.2 Traﬃc Quality Indicators Many of the service-stage traﬃc quality indicators may be reformulated as service-stage performance indicators as done below. Indicator 9: Carried Eﬀectiveness of a Service Stage – the ratio of the eﬀective carried traﬃc to the carried traﬃc: I9 = eﬀ.crr.Y crr.Y . (35) (35) For instance, from (25) and (29), the Carried Eﬀectiveness of service stage g in Fig. 2 is: eﬀ.crr.Yg crr.Yg = (1 −prs.Ps)(crr.Te + crr.Ts) crr.Te + (1 −prs.Ps) crr.Ts . (36) (36) Scalable Traﬃc Quality and System Eﬃciency Indicators 2, and Formulas (23), (33), and srv.Fs = crr.Fe, the traﬃc oﬀered to the service device s is: ofr.Ys = blc.Ys + srv.Ys = ofr.Fg srv.Ts. (34) (34) 90 S. Poryazov et al. 4.1 Telecommunication System QoS Concept Users are shown in “Fig. 1 – Schematic contributions to end-to-end QoS” in [4] but they are not connected to the network. In Fig. 3, schematic contributions to QoS in an overall telecommunication system, including users, is presented in more detail. Fig. 3. Schematic contributions to QoS in an overall telecommunication system, including users. Fig. 3. Schematic contributions to QoS in an overall telecommunication system, including us In Fig. 3, the calling (A) and called (B) users and terminals, as well as the main service stages of the service request (call) attempts in a telecommunication network, are presented. The telecommunication network is usually presented as having ﬁve service stages: A-terminal, Dialing, Switching, B-terminal Seizure, and B-terminal, where each Scalable Traﬃc Quality and System Eﬃciency Indicators 91 service stage has its own characteristics. However, there are two other stages – A-user and B-user –, with their own speciﬁcs. In Fig. 3, the possible paths of service request (call2) attempts are the following: 1. int.Fa: The calling users (A-users) generate intent call attempts, with intensity int.Fa, represented as a Generate device in the A-User block. Call intent is “The desire to establish a connection to a user”. “This would normally be manifested by a call demand. However, demands may be suppressed or delayed by the calling user’s expectation of poor Quality of Service performance at a particular time” [5]. 1. int.Fa: The calling users (A-users) generate intent call attempts, with intensity int.Fa, represented as a Generate device in the A-User block. Call intent is “The desire to establish a connection to a user”. “This would normally be manifested by a call demand. However, demands may be suppressed or delayed by the calling user’s expectation of poor Quality of Service performance at a particular time” [5]. 2. sup.Fa: The intensity of suppressed intent call attempts. Suppressed traﬃc is “The traﬃc that is withheld by users who anticipate a poor quality of service (QoS) performance” [5]. “At present, suitable algorithms for estimating suppressed traﬃc have not been deﬁned” [7]. 3. dem.Fa: The intensity of demand call attempts. Call demand is: “A call intent that results in a ﬁrst call attempt” [5]. 4. rep.Fa: The intensity of repeated call attempts. Repeated call attempt is: “Any of the call attempts subsequent to a ﬁrst call attempt related to a given call demand. NOTE – Repeated call attempts may be manual, i.e. 2 Throughout the rest of this chapter, the term ‘call’ should be interpreted in a broader meaning of a ‘service request’. 4.1 Telecommunication System QoS Concept ofr.Fz = prs.Fz + crr.Fz: The intensity of the oﬀered-, parasitic-, and carried ﬂows of call attempts of the ‘B-terminal seizure’ stage. The intended B-terminal may be busy or unavailable and this will cause blocking of call attempts. 11. ofr.Fb = prs.Fb + crr.Fb: The intensity of the oﬀered-, parasitic-, and carried ﬂows of call attempts of the B-terminal stage. 12. ofr.Fbu = prs.Fbu + crr.Fbu: The intensity of the oﬀered-, parasitic-, and carried ﬂows of call attempts of the B-user stage. The B-user may be absent, busy, tired, etc. 4.1 Telecommunication System QoS Concept generated by humans, or auto‐ matic, i.e. generated by machines” [7]. 5. ofr.Fa: The intensity of all call attempts (demand and repeated) trying to occupy A-terminals. A-terminals are considered as the ﬁrst service stage (c.f. Sect. 3) in the telecommunication network. From Fig. 3: 5. ofr.Fa: The intensity of all call attempts (demand and repeated) trying to occupy A-terminals. A-terminals are considered as the ﬁrst service stage (c.f. Sect. 3) in the telecommunication network. From Fig. 3: ofr.Fa = dem.Fa + rep.Fa. (37) (37) 6. prs.Fa: The intensity of all parasitic (unsuccessfully served, c.f. Sect. 2) call attempts in A-terminals. We are modeling the system in a stationary state and for each considered service stage the intensity of the oﬀered call attempts equals the sum of the outgoing para‐ sitic and carried ﬂows, e.g. ofr.Fa = prs.Fa + crr.Fa. For each service stage, part of the parasitic attempts are terminated by the A-user (c.f. devices of type ‘terminator’ in Fig. 3) and the rest join the repeated attempt’s ﬂow (rep.Fa). 6. prs.Fa: The intensity of all parasitic (unsuccessfully served, c.f. Sect. 2) call attempts in A-terminals. We are modeling the system in a stationary state and for each considered service stage the intensity of the oﬀered call attempts equals the sum of the outgoing para‐ sitic and carried ﬂows, e.g. ofr.Fa = prs.Fa + crr.Fa. For each service stage, part of the parasitic attempts are terminated by the A-user (c.f. devices of type ‘terminator’ in Fig. 3) and the rest join the repeated attempt’s ﬂow (rep.Fa). 7. ofr.Fd = crr.Fa: The intensity of carried (in A-terminals) call attempts (crr.Fa) is equal to the intensity of the oﬀered call attempts (ofr.Fd) to the Dialing stage in the network. 7. ofr.Fd = crr.Fa: The intensity of carried (in A-terminals) call attempts (crr.Fa) is equal to the intensity of the oﬀered call attempts (ofr.Fd) to the Dialing stage in the network. 8. prs.Fd: The intensity of all parasitic (unsuccessfully served, c.f. Sect. 2) call attempts in the Dialing stage. 8. prs.Fd: The intensity of all parasitic (unsuccessfully served, c.f. Sect. 2) call attempts in the Dialing stage. 9. ofr.Fs = prs.Fs + crr.Fs: The intensity of the oﬀered-, parasitic-, and carried ﬂows of call attempts of the Switching stage. 92 S. Poryazov et al. 10. 4.2 Eﬃciency Indicators The eﬃciency indicators, proposed in this chapter, are considered on ﬁve levels: (1) service phase; (2) service stage; (3) part of network; (4) overall network; and (5) overall telecommunication system. Indicator 16: Eﬃciency indicator Qb on the B-terminal stage: I16 = Qb = crr.Fb ofr.Fb . (46) (46) Indicator 17: Eﬃciency indicator Qbu on the B-user stage: I17 = Qbu = crr.Fbu ofr.Fbu . (47) (47) 4.2.1 Proposed Eﬃciency Indicators on Service Stage Level In each service stage, a basic performance indicator is the ratio between intensities of the carried ﬂow and oﬀered ﬂow of call attempts. An exception is the A-User stage because there are two sub-stages in it – Ai (considering the intent call attempts) and Ad (considering the demand call attempts). Indicator 10: Eﬃciency indicator Qai on the Ai sub-stage: I10 = Qai = dem.Fa int.Fa (38) (38) Indicator 11: Eﬃciency indicator Qad on the Ad sub-stage. Indicator 11: Eﬃciency indicator Qad on the Ad sub-stage. Let Pr is the aggregated probability of repetition of the oﬀered (to the A-terminals) call attempts: Pr = rep.Fa ofr.Fa . (39) (39) From (37) and (39), the following formula could be obtained for the eﬃciency indi‐ cator Qad: I11 = (1 −Pr) = dem.Fa ofr.Fa = dem.Fa dem.Fa + rep.Fa = 1 𝛽= Qad, (40) (40) where β is deﬁned in [7] as: 𝛽= All call attempts First call attempts. (41) (41) In (40), Qad is de-facto the probability corresponding to the ratio of the primary (demand) call attempts’ intensity to the oﬀered attempts’ intensity. It may be considered as an aggregated overall network performance indicator (as per the initial attempt in [8]). In (40), Qad is de-facto the probability corresponding to the ratio of the primary (demand) call attempts’ intensity to the oﬀered attempts’ intensity. It may be considered as an aggregated overall network performance indicator (as per the initial attempt in [8]). Scalable Traﬃc Quality and System Eﬃciency Indicators 93 Indicator 12: Eﬃciency indicator Qa on the A-terminal stage: Indicator 13: Eﬃciency indicator Qd on the Dialing stage: I13 = Qd = crr.Fd ofr.Fd . (43) (43) Indicator 14: Eﬃciency indicator Qs on the Switching stage: I14 = Qs = crr.Fs ofr.Fs . (44) (44) Indicator 15: Eﬃciency indicator Qz on the ‘B-terminal Seizure’ stage: I15 = Qz = crr.Fz ofr.Fz . (45) (45) Indicator 16: Eﬃciency indicator Qb on the B-terminal stage: Indicator 12: Eﬃciency indicator Qa on the A-terminal stage: I12 = Qa = crr.Fa ofr.Fa . (42) (42) Indicator 13: Eﬃciency indicator Qd on the Dialing stage: 4.2.2 Proposed Eﬃciency Indicators on Network Level 4.2.2 Proposed Eﬃciency Indicators on Network Level p y Network eﬃciency indicators estimate QoS characteristics of portions of the network comprising more than one service stages, or the overall network. In this subsection, as usually, the indicated network portion begins with the starting points of the network and ends in another network point of interest. All network eﬃciency indicators are fractions with denominators oﬀered to the A-terminals’ ﬂow intensity ofr.Fa. The classic network eﬃciency indicators are the following three, e.g. as deﬁned in [9]: 1. “Answer Seizure Ratio (ASR) = (number of seizures that result in an answer signal)/(the total number of seizures)” … “Measurement of ASR may be made on a route or on a destination code basis” … “A destination can be a mobile network, a country, a city, a service, etc.” [9]. 1. “Answer Seizure Ratio (ASR) = (number of seizures that result in an answer signal)/(the total number of seizures)” … “Measurement of ASR may be made on a route or on a destination code basis” … “A destination can be a mobile network, a country, a city, a service, etc.” [9]. 2. “Answer Bid Ratio (ABR) = (number of bids that result in an answer signal)/(total number of bids); ABR is similar to ASR except that it includes bids that do not result in a seizure” [9]. 2. “Answer Bid Ratio (ABR) = (number of bids that result in an answer signal)/(total number of bids); ABR is similar to ASR except that it includes bids that do not result in a seizure” [9]. 94 S. Poryazov et al. 3. “Network Eﬀectiveness Ratio (NER): NER is designed to express the ability of networks to deliver calls to the far-end terminal. NER expresses the relationship between the number of seizures and the sum of the number of seizures resulting in either an answer message, or a user busy, or a ring no answer, or in the case of ISDN a terminal rejection/unavailability. Unlike ASR, NER excludes the eﬀects of customer behavior and terminal behavior” [9]. 3. “Network Eﬀectiveness Ratio (NER): NER is designed to express the ability of networks to deliver calls to the far-end terminal. NER expresses the relationship between the number of seizures and the sum of the number of seizures resulting in either an answer message, or a user busy, or a ring no answer, or in the case of ISDN a terminal rejection/unavailability. 4.2.2 Proposed Eﬃciency Indicators on Network Level Unlike ASR, NER excludes the eﬀects of customer behavior and terminal behavior” [9]. These classic eﬃciency indicators reﬂect network providers’ attitude but don’t consider the possibilities for initiated but unsuccessful communication as well as the inﬂuence of repeated attempts. Below we propose new network eﬃciency indicators, all having as an index the ﬁrst letter of the last service stage considered. Indicator 18: Network eﬃciency indicator Ea on the A-terminal stage, c.f. also (42): I18 = Ea = crr.Fa ofr.Fa = Qa = I12. (48) (48) Indicator 19: Network eﬃciency indicator Ed on the Dialing stage, c.f. also (43) I19 = Ed = crr.Fd ofr.Fa = Qa Qd = I12 I13. (49) (49) Indicator 20: By taking into account that crr.Fd = ofr.Fs, c.f. Fig. 3 and (44), the network eﬃciency indicator Es on the Switching stage is: Indicator 20: By taking into account that crr.Fd = ofr.Fs, c.f. Fig. 3 and (44), the network eﬃciency indicator Es on the Switching stage is: I20 = Es = crr.Fs ofr.Fa = Qa Qd Qs = I14 I19 . (50) (50) Indicator 21: By taking into account that crr.Fs = ofr.Fz, c.f. Fig. 3 and (45), the network eﬃciency indicator Ez on the ‘B-terminal seizure’ stage is: Indicator 21: By taking into account that crr.Fs = ofr.Fz, c.f. Fig. 3 and (45), the network eﬃciency indicator Ez on the ‘B-terminal seizure’ stage is: I21 = Ez = crr.Fz ofr.Fa = Qa Qd Qs Qz = I15 I20 . (51) (51) Indicator 22: By taking into account that crr.Fz = ofr.Fb, c.f. Fig. 3 and (46), the network eﬃciency indicator Eb on the B-terminal stage is: Indicator 22: By taking into account that crr.Fz = ofr.Fb, c.f. Fig. 3 and (46), the network eﬃciency indicator Eb on the B-terminal stage is: I22 = Eb = crr.Fb ofr.Fa = Qa Qd Qs Qz Qb = I16 I21 . (52) (52) This indicator corresponds to the cases of B-user answers, but does not consider the successfulness of the communication. This indicator corresponds to the cases of B-user answers, but does not consider the successfulness of the communication. 4.2.3 Proposed Eﬃciency Indicators on Overall System Level Indicator 23: By taking into account that crr.Fb = ofr.Fbu, c.f. Fig. 4.2.2 Proposed Eﬃciency Indicators on Network Level 3 and (47), the system eﬃciency indicator Ebu on the B-user stage is: Scalable Traﬃc Quality and System Eﬃciency Indicators Scalable Traﬃc Quality and System Eﬃciency Indicators 95 I23 = Ebu = crr.Fbu ofr.Fa = Qa Qd Qs Qz Qb Qbu = I17 I22 . (53) (53) This indicator corresponds to the cases of fully successful communication, from the users’ point of view, regarding all call attempts oﬀered to the network. p g g p Indicator 24: System eﬃciency indicator Eu on the Ad sub-stage, c.f. also (40): I24 = Eu = Qad Ebu = Qad Qa Qd Qs Qz Qb Qbu = I11 I23. (54) (54) This indicator corresponds to the cases of fully successful communication, from the A-users’ point of view, regarding demand call attempts. It shows what part of the ﬁrst (demand) attempts is fully successful. It may be called ‘Demand Eﬃciency’. It is a user- oriented indicator, compounding explicitly repeated attempts, connection and commu‐ nication parameters. dicator 25: System eﬃciency indicator Ei on the Ai sub-stage, c.f. also (38): I25 = Ei = Qai Eu = Qai Qad Qa Qd Qs Qz Qb Qbu = I10 I24. (55) (55) This indicator corresponds to the cases of fully successful communication, from the A-users’ point of view, regarding intent call attempts. It shows what part of the intent attempts is fully successful. It is very diﬃcult to measure Ei directly because suppressed attempts (forming the demands w.r.t. point 2 in Subsect. 4.1) can’t reach the network and therefore can’t be measured there. 4.3 Approach Applicability and Results Most of the proposed indicators are ﬂow-oriented as they take into account the ﬂow intensities. Flow-oriented indicators are in the core of time- and traﬃc-oriented indica‐ tors. In this subsection, numerical results for some of the proposed ﬂow indicators and other time- and traﬃc-oriented indicators, built on their basis, are presented. An analyt‐ ical model of the overall telecommunication system, corresponding to Fig. 3, is used. Methods of building such models are described in the Chapter “Conceptual and Analyt‐ ical Models for Predicting the Quality of Service of Overall Telecommunication Systems” of this book. y The numerical results are presented for the entire theoretical network-traﬃc-load interval, i.e. the terminal traﬃc of all A- and B-terminals (Yab) is ranging from 0% to 100% of the number Nab of all active terminals in the network. The input parameters are the same, excluding the capacity of the network (the number of the equivalent connection lines), given as a percentage of all terminals in the system. Diﬀerences in the network capacity cause diﬀerent blocking probabilities due to resource insuﬃciency. Three cases have been considered: • Case 1: Without repeated service request attempts and without blocking; • Case 1: Without repeated service request attempts and without blocking; • Case 2: With repeated service request attempts but without blocking; • Case 3: With repeated service request attempts and with blocking. 96 S. Poryazov et al. Time and traﬃc AB-eﬃciency for Case 2 ((Eu Tcc) ∕Tab is sensitive to repeated service request attempts in contrast to (Ebu Tcc) ∕Tab, which is not). Fig. 9. Time and traﬃc AB-eﬃciency for Case 3 (the indicator paid Traﬃc ∕Yab is not sensible enough in the network-load interval without blocking). Fig. 9. Time and traﬃc AB-eﬃciency for Case 3 (the indicator paid Traﬃc ∕Yab is not sensible enough in the network-load interval without blocking). Fig. 9. Time and traﬃc AB-eﬃciency for Case 3 (the indicator paid Traﬃc ∕Yab is not sensible enough in the network-load interval without blocking). S. Poryazov et al. Figures 4, 5, 6 present numerical results obtained for some of the proposed eﬃciency indicators, whereas Figs. 7, 8, 9 present numerical results obtained for some time- and traﬃc-oriented indicators. Fig. 4. Eﬃciency indicators Qad, Ebu and Eu for Case 1 (Qad = 1 and Ebu = Eu because there are no repeated service request attempts in the system) Fig. 4. Eﬃciency indicators Qad, Ebu and Eu for Case 1 (Qad = 1 and Ebu = Eu because there are no repeated service request attempts in the system). Fig. 4. Eﬃciency indicators Qad, Ebu and Eu for Case 1 (Qad = 1 and Ebu = Eu because there are no repeated service request attempts in the system). Fig. 5. Eﬃciency indicators Qad, Ebu and Eu for Case 2 (the network performance is degraded considerably due to repeated service request attempts). Fig. 5. Eﬃciency indicators Qad, Ebu and Eu for Case 2 (the network performance is degraded considerably due to repeated service request attempts). Fig. 5. Eﬃciency indicators Qad, Ebu and Eu for Case 2 (the network performance is degraded considerably due to repeated service request attempts). Scalable Traﬃc Quality and System Eﬃciency Indicators 97 Fig. 6. Eﬃciency indicators Qad, Ebu and Eu for Case 3 (the network performance degrades sharply due to blocking). Fig. 6. Eﬃciency indicators Qad, Ebu and Eu for Case 3 (the network performance degrades sharply due to blocking). Fig. 6. Eﬃciency indicators Qad, Ebu and Eu for Case 3 (the network performance degrades sharply due to blocking). Fig. 7. Time and traﬃc AB-eﬃciency for Case 1 ((Ebu Tcc)∕Tab = (Eu Tcc)∕Tab, because there are no repeated service request attempts). Fig. 7. Time and traﬃc AB-eﬃciency for Case 1 ((Ebu Tcc)∕Tab = (Eu Tcc)∕Tab, because there are no repeated service request attempts). Fig. 7. Time and traﬃc AB-eﬃciency for Case 1 ((Ebu Tcc)∕Tab = (Eu Tcc)∕Tab, because there are no repeated service request attempts). Fig. 8. Time and traﬃc AB-eﬃciency for Case 2 ((Eu Tcc) ∕Tab is sensitive to repeated service request attempts in contrast to (Ebu Tcc) ∕Tab, which is not). 98 S. Poryazov et al. Fig. 8. Time and traﬃc AB-eﬃciency for Case 2 ((Eu Tcc) ∕Tab is sensitive to repeated service request attempts in contrast to (Ebu Tcc) ∕Tab, which is not). 98 S. Poryazov et al. S. Poryazov et al. 98 Fig. 8. 5.1 Mean Cost/Quality Ratio Based on these assumptions and the deﬁnition of the traﬃc volume, i.e. “The traﬃc volume in a given time interval is the time integral of the traﬃc intensity over this time interval” [5], the ‘Cost per Unit’ quantity is: Cost per Unit = Full System′s Costs [Euro] Paid Traﬃc Volume [Erlang × Δt] . (56) (56) By dividing this to the QoS indicator (Q), we obtain the following: Cost per Unit Quality = Full System′s Costs [Euro] Q paid.V [Erlang × Δt] = SC Q paid.V . By dividing this to the QoS indicator (Q), we obtain the following: Cost per Unit Quality = Full System′s Costs [Euro] Q paid.V [Erlang × Δt] = SC Q paid.V . (57) (57) The deﬁnition of the paid traﬃc may depend on the telecommunication service provider. The estimation of the paid traﬃc volume is a routine operation (c.f. ITU-T Recommendations Series D: General Tariﬀ Principles). The deﬁnition of the paid traﬃc may depend on the telecommunication service provider. The estimation of the paid traﬃc volume is a routine operation (c.f. ITU-T Recommendations Series D: General Tariﬀ Principles). The deﬁnition of the QoS indicator (Q) may diﬀer from users’ perspective (i.e. as a generalized QoE parameter) to the telecommunication service provider’s perspective. In general, the best is to include the Q deﬁnition in the Service Level Agreement (SLA). In any case, the value of the QoS indicator (Q) in (57) is the mean value in the time interval considered. The mean cost/quality ratio (57) is suitable for relatively long intervals – days, months, years. 5 Network Cost/Quality Ratios We consider the overall telecommunication system model, presented in Fig. 3, with the following assumptions: Scalable Traﬃc Quality and System Eﬃciency Indicators 99 Assumption 1: The observation time interval Δt is limited; Assumption 2: The full system costs (SC) in the time interval Δt are known; Assumption 3: The cost/quality ratio depends on the paid volume of traﬃc (paid.V) in this time interval and the QoS indicator (Q); Assumption 3: The cost/quality ratio depends on the paid volume of traﬃc (paid.V) in this time interval and the QoS indicator (Q); Assumption 4: The full system costs (SC) don’t depend considerably on the served traﬃc volume in the time interval Δt; Assumption 4: The full system costs (SC) don’t depend considerably on the served traﬃc volume in the time interval Δt; ssumption 5: The QoS indicator (Q) is dimensionless with values from the interval 1] and is proportional to the quality (Q = 1 means ‘the best quality’) Assumption 5: The QoS indicator (Q) is dimensionless with values from the interval (0,1] and is proportional to the quality (Q = 1 means ‘the best quality’). p ,1] and is proportional to the quality (Q = 1 means ‘the best quality’). 5.2 Instantaneous Cost/Quality Ratio The expressions (57) and (58) are similar (the mean value of the instantaneous indi‐ ator, in Δt, gives the value of the Mean Cost/Quality Ratio indicator in Δt), but the The paid traﬃc intensity depends on the network traﬃc load. In any case, the instan‐ taneous values of the QoS indicator (Q) depend on many factors, including the network load. The expressions (57) and (58) are similar (the mean value of the instantaneous indi‐ cator, in Δt, gives the value of the Mean Cost/Quality Ratio indicator in Δt), but the methods for their estimation and usage are diﬀerent. The Instantaneous Cost/Quality Ratio may be useful for dynamic pricing policies, depending on the network load. Related works on this subject were not found in the literature. 5.2 Instantaneous Cost/Quality Ratio We consider the traﬃc intensity (Y) as per the ITU-T deﬁnition, i.e. “The instantaneous traﬃc in a pool of resources is the number of busy resources at a given instant of time” [5]. From assumptions made and (57), the following formula could be obtained: Cost per Unit Quality = Full System′s Costs [Euro] Δt [Time] paid.Y [Erlang] Q = SC Δt Q paid.Y . (58) (58) The paid traﬃc intensity (paid.Y) is an instantaneous quantity but the ratio ‘Cost per Unit/Quality’ (58) depends on the time interval duration. We deﬁne the ‘System’s Costs The paid traﬃc intensity (paid.Y) is an instantaneous quantity but the ratio ‘Cost per Unit/Quality’ (58) depends on the time interval duration. We deﬁne the ‘System’s Costs 100 S. Poryazov et al. Intensity’ (SCI) parameter, independent of the time interval duration (but dependent of the interval position in the service provider’s life time), as per the following formula: System′s Costs Intensity (SCI) = Full System′s Costs [Euro] Δt [Time] = SC Δt . (59) (59) The System’s Costs Intensity (SCI) parameter allows deﬁning a new useful param‐ eter – the Normalized Cost/Quality Ratio (NCQR): The System’s Costs Intensity (SCI) parameter allows deﬁning a new useful param‐ eter – the Normalized Cost/Quality Ratio (NCQR): Normalized Cost∕Quality Ratio (NCQR) = 1 Q paid.Y [Erlang]. (60) (60) The Normalized Cost/Quality Ratio (NCQR) is independent of the absolute system’s costs amount. It is normalized, because it is the cost/quality ratio per 1 Euro cost. From (57), (58), and (59), we obtain: Cost per Unit Quality = Full System′s Costs [Euro] Δt [Time] 1 Q paid.Y [Erlang] = = SCI NCQR (61) Cost per Unit Quality = Full System′s Costs [Euro] Δt [Time] 1 Q paid.Y [Erlang] = = SCI NCQR (61) (61) The proposed quantities SCI and NCQR allow the estimation of the cost/quality ratio for every suitable (paid) time interval with a relatively short duration, e.g. seconds, minutes, hours. The proposed quantities SCI and NCQR allow the estimation of the cost/quality ratio for every suitable (paid) time interval with a relatively short duration, e.g. seconds, minutes, hours. minutes, hours. The paid traﬃc intensity depends on the network traﬃc load. In any case, the instan‐ aneous values of the QoS indicator (Q) depend on many factors, including the network oad. 5.3 Prediction of Instantaneous Cost/Quality Ratio An advantage of the Normalized Cost/Quality Ratio (NCQR) is its independence of the absolute system’s costs amount. This allows separation of the estimations for NCQR and System’ Cost Intensity (SCI). In this subsection, we estimate NCQR using the tele‐ communication system model described in the Chapter “Conceptual and Analytical Models for Predicting the Quality of Service of Overall Telecommunication Systems” of this book. We consider Q as an overall telecommunication system’s QoS indicator. Each of the described indicators on the overall system level (c.f. Subsect. 4.2.3) may be used. Numerical examples below are for the indicator Q = Ebu c.f. (53). This corresponds to cases of fully successful communication, from the users’ point of view, regarding all call attempts, oﬀered to the network. Sometimes it is called “Network Call Eﬃciency”. 101 Scalable Traﬃc Quality and System Eﬃciency Indicators As a paid traﬃc, the successful communication (carried) traﬃc is used: As a paid traﬃc, the successful communication (carried) traﬃc is used: NCQR = 1 Ebu paid.Y . (62) (62) The values of input parameters of human behavior and technical system, to the model, in the presented output numerical results are typical for voice-oriented networks. The values of input parameters of human behavior and technical system, to the model, in the presented output numerical results are typical for voice-oriented networks. Figures 10 and 11 present numerical results for the entire theoretical network traﬃc load interval, i.e. the terminal traﬃc of all A- and B-terminals (Yab) is within the range of 0% to 100% of the number Nab of all active terminals in the system. The input parameters are the same, excluding the capacity of the network (the number of equivalent connection lines), given as a percentage of all terminals in the system. Diﬀerences in the network capacity cause diﬀerent blocking probabilities due to resource insuﬃciency. Two cases have been considered: • Case 1: The network capacity equals 10% of all terminals presented in the system; C 2 Th t k it l 25% f ll t i l t d i th t • Case 1: The network capacity equals 10% of all terminals presented in the system; • Case 2: The network capacity equals 25% of all terminals presented in the system. 5.3 Prediction of Instantaneous Cost/Quality Ratio Numerical prediction of the Normalized Cost/Quality Ratio (NCQR), Network Call Eﬃciency (Ebu), and Paid Traﬃc Intensity in an overall telecommunication system with QoS guarantees (Case 2: Network capacity = 25% of terminals). The results show considerable sensitivity of the Normalized Cost/Quality Ratio (NCQR) from the network capacity and traﬃc load. 5.3 Prediction of Instantaneous Cost/Quality Ratio Case 2: The network capacity equals 25% of all terminals presented in the system Paid Network Traffic 0 10 20 30 40 50 60 70 80 90 100 0 10 20 30 40 50 60 70 80 Normalized Cost / Quality Ratio Network Traffic Load [%] Call Efficiency Network Call Efficiency (Ebu) [%] Normalized Cost / Quality Ratio (NCQR) [%] Network capacity = 10% of Terminals Network Paid Network Traffic 0 10 20 30 40 50 60 70 80 90 100 0 10 20 30 40 50 60 70 80 Normalized Cost / Quality Ratio Network Traffic Load [%] Call Efficiency Network Call Efficiency (Ebu) [%] Normalized Cost / Quality Ratio (NCQR) [%] Network capacity = 10% of Terminals Network Fig. 10. Numerical prediction of the Normalized Cost/Quality Ratio (NCQR), Network Call Eﬃciency (Ebu), and Paid Traﬃc Intensity in an overall telecommunication system with QoS guarantees (Case 1: Network capacity = 10% of terminals). Network capacity = 10% of Terminals 20 30 40 50 60 70 80 Network Traffic Load [%] Fig. 10. Numerical prediction of the Normalized Cost/Quality Ratio (NCQR), Network Call Eﬃciency (Ebu), and Paid Traﬃc Intensity in an overall telecommunication system with QoS guarantees (Case 1: Network capacity = 10% of terminals). 102 S. Poryazov et al. Paid Network Traffic 0 10 20 30 40 50 60 70 80 90 100 0 10 20 30 40 50 60 70 80 Normalized Cost / Quality Ratio Network Traffic Load [%] Network Call Efficiency Network Call Efficiency (Ebu) [%] Normalized Cost / Quality Ratio (NCQR) [%] Network capacity = 25% of Terminals Fig. 11. Numerical prediction of the Normalized Cost/Quality Ratio (NCQR), Network Call Eﬃciency (Ebu), and Paid Traﬃc Intensity in an overall telecommunication system with QoS guarantees (Case 2: Network capacity = 25% of terminals). 102 S. Poryazov et al. Paid Network Traffic 0 10 20 30 40 50 60 70 80 90 100 0 10 20 30 40 50 60 70 80 Normalized Cost / Quality Ratio Network Traffic Load [%] Network Call Efficiency Network Call Efficiency (Ebu) [%] Normalized Cost / Quality Ratio (NCQR) [%] Network capacity = 25% of Terminals Network capacity = 25% of Terminals 0 10 20 30 40 50 60 70 80 90 100 0 Network Traffic Load [%] Fig. 11. References 1. Reichl, P.: From charging for quality of service to charging for quality of experience. Ann. Telecommun. 65(3–4), 189–199 (2010) 1. Reichl, P.: From charging for quality of service to charging for quality of experience. Ann. Telecommun. 65(3–4), 189–199 (2010) 2. Varela, M., Zwickl, P., Reichl, P., Xie, M., Schulzrinne, H.: From service level agreements (SLA) to experience level agreements (ELA): the challenges of selling QoE to the user. In: Proceedings of IEEE ICC QoE-FI, London, June 2015. ISSN: 2164-7038, https://doi.org/ 10.1109/iccw.2015.7247432 2. Varela, M., Zwickl, P., Reichl, P., Xie, M., Schulzrinne, H.: From service level agreements (SLA) to experience level agreements (ELA): the challenges of selling QoE to the user. In: Proceedings of IEEE ICC QoE-FI, London, June 2015. ISSN: 2164-7038, https://doi.org/ 10.1109/iccw.2015.7247432 3. Fiedler, M.: Teletraﬃc models for quality of experience assessment. Tutorial at 23rd International Teletraﬃc Congress (ITC 23), San Francisco, CA, September 2011. http:// iteletraﬁc.org/_Resources/Persistent/9269df1c3dca0bf58ee715c3b9afabbc71d4fb26/ ﬁedler11.pdf. Accessed 20 July 2017 4. ITU T Recommendation E.800 (09/08): Deﬁnitions of terms related to quality of service 5. ITU-T Recommendation E.600 (03/93): Terms and deﬁnitions of traﬃc engineering 5. ITU-T Recommendation E.600 (03/93): Terms and deﬁnitions of traﬃc engineering , Q g p , ( ) 7. ITU-T Recommendation E.501(05/97): Estimation of Traﬃc Oﬀered in The Network 8. Poryazov, S., Saranova, E.: User-oriented, overall traﬃc and time eﬃciency indicators in telecommunications. In: TELFOR 2016 International IEEE Conference #39555, Belgrade, Serbia, 22–23 November 2016, IEEE Catalog Number: CFP1698P-CDR. IEEE (2016). ISBN: 978-1-5090-4086-5/16, INSPEC Accession Number: 16603129, https://doi.org/10.1109/ telfor.2016.7818729 9. ITU-T Rec. E.425 (03/2002): Network management – Internal automatic observations Open Access This chapter is licensed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license and indicate if changes were made. Open Access This chapter is licensed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license and indicate if changes were made. 6 Conclusion In this chapter, a more detailed and precise approach to the traﬃc characterization has been taken, which allows the deﬁnition of new eﬃciency indicators on diﬀerent levels, starting with the service phases and stages, continuing with (part of) the network, and ﬁnishing with the overall telecommunication system. The proposed Instantaneous Cost/ Quality Ratio may be used for the establishment and utilization of dynamic pricing policies, depending on the network load. The use of the instantaneous cost/quality ratios as a source for QoE prediction is a very perspective direction of research. Using similar approaches and speciﬁc QoS indicators for other types of networks, e.g. multimedia- and multiservice networks, seems also very topical. Acknowledgments. This work was coordinated under the EU COST Action IC1304 “Autonomous Control for a Reliable Internet of Services (ACROSS)”. The work was partially funded by the Bulgarian NSF Projects DCOST 01/9 (the work of S. Poryazov) and DCOST 01/20 (the work of E. Saranova). 103 Scalable Traﬃc Quality and System Eﬃciency Indicators 9. ITU-T Rec. E.425 (03/2002): Network management – Internal automatic observations References The images or other third party material in this chapter are included in the chapter’s Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the chapter’s Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder.
https://openalex.org/W1995366282	https://europepmc.org/articles/pmc4284981?pdf=render	English	null	Evidence of Insulin Resistance in Adult Uncomplicated Malaria: Result of a Two-Year Prospective Study	Malaria research and treatment	2,014	cc-by	7,560	Samuel Acquah,1 Johnson Nyarko Boampong,2 Benjamin Ackon Eghan Jnr,3 and Magdalena Eriksson1,4 Benjamin Ackon Eghan Jnr,3 and Magdalena Eriksson1,4 1Department of Medical Biochemistry, School of Medical Sciences, University of Cape Coast, Cape Coast, Ghana 2Department of Biomedical and Forensic Sciences, School of Biological Sciences, University of Cape Coast, Cape Coast, Ghana 3Department of Medicine, School of Medical Sciences, Kwame Nkrumah University of Science and Technology, Kumasi, Ghana 4African Institute for Mathematical Science Ghana (AIMS-Ghana), P.O. Box LG 197, Legon-Accra, Ghana 1Department of Medical Biochemistry, School of Medical Sciences, University of Cape Coast, Cape Coast, Ghana 2Department of Biomedical and Forensic Sciences, School of Biological Sciences, University of Cape Coast, Cape Coast, Ghana 3Department of Medicine, School of Medical Sciences, Kwame Nkrumah University of Science and Technology, Kumasi, Ghana 4African Institute for Mathematical Science Ghana (AIMS-Ghana), P.O. Box LG 197, Legon-Accra, Ghana Correspondence should be addressed to Samuel Acquah; s.acquah@uccsms.edu.gh Received 28 July 2014; Revised 25 November 2014; Accepted 8 December 2014; Published 23 December 2014 Academic Editor: Clara Men´endez Academic Editor: Clara Men´endez Copyright © 2014 Samuel Acquah et al. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. The study aimed at investigating the effects of adult uncomplicated malaria on insulin resistance. Fasting levels of blood glucose (FBG), glycosylated hemoglobin (HbA1c), and serum insulin were measured in 100 diabetics and 100 age-matched controls before and during Plasmodium falciparum malaria. Insulin resistance and beta cell function were computed by homeostatic models assessment of insulin resistance (HOMAIR) and beta cell function (HOMAB) formulae, respectively. Body mass index (BMI) was computed. At baseline, diabetics had significantly (𝑃< 0.05) higher levels of BMI, FBG, HbA1c, and HOMAIR but lower level of HOMAB than controls. Baseline insulin levels were comparable (𝑃> 0.05) between the two study groups. During malaria, diabetics maintained significantly (𝑃< 0.05) higher levels of BMI, FBG, and HbA1c but lower levels of insulin and HOMAB than controls. Malaria-induced HOMAIR levels were comparable (𝑃> 0.05) between the two study groups but higher than baseline levels. Apart from BMI and HOMAB, mean levels of all the remaining parameters increased in malaria-infected controls. In malaria-infected diabetics, significant (𝑃< 0.05) increase was only observed for insulin and HOMAIR but not the other measured parameters. Uncomplicated malaria increased insulin resistance in diabetics and controls independent of BMI. Hindawi Publishing Corporation Malaria Research and Treatment Volume 2014, Article ID 136148, 8 pages http://dx.doi.org/10.1155/2014/136148 Hindawi Publishing Corporation Malaria Research and Treatment Volume 2014, Article ID 136148, 8 pages http://dx.doi.org/10.1155/2014/136148 Hindawi Publishing Corporation Malaria Research and Treatment Volume 2014, Article ID 136148, 8 pages http://dx.doi.org/10.1155/2014/136148 Samuel Acquah,1 Johnson Nyarko Boampong,2 Benjamin Ackon Eghan Jnr,3 and Magdalena Eriksson1,4 This finding may have implications for the evolution of T2DM in malaria-endemic regions. Research Article Evidence of Insulin Resistance in Adult Uncomplicated Malaria: Result of a Two-Year Prospective Study Samuel Acquah,1 Johnson Nyarko Boampong,2 Benjamin Ackon Eghan Jnr,3 and Magdalena Eriksson1,4 1Department of Medical Biochemistry, School of Medical Sciences, University of Cape Coast, Cape Coast, Ghana 2Department of Biomedical and Forensic Sciences, School of Biological Sciences, University of Cape Coast, Cape Coast, Ghana 3Department of Medicine, School of Medical Sciences, Kwame Nkrumah University of Science and Technology, Kumasi, Ghana 4African Institute for Mathematical Science Ghana (AIMS-Ghana), P.O. Box LG 197, Legon-Accra, Ghana 1. Introduction With Ghana being a sub-Saharan African country experiencing increased incidence of diabetes, especially, T2DM, while maintaining her endemic status with respect to malaria infection, the interaction between malaria and diabetes in the same individ- ual cannot be overlooked and this may pose additional health challenge that ought to be investigated.i type 2 diabetics and nondiabetic controls over a two-year period for malaria infection. The inclusion criteria used in selecting study participants were being 40 years or older, non- alcoholic, nonsmoker, not suffering from any cardiovascular disease, hepatitis B, C, and HIV, not pregnant, not nursing a baby, with no evidence of hepatic dysfunction, with absence of severe arthritis, lupus, or inflammatory bowel disease, and not undergoing hormonal therapy or having any medical condition that can significantly influence the levels of any of the biomarkers measured. Insulin resistance, which signifies reduced sensitivity of insulin-sensitive cells to insulin-mediated actions, has long been recognized as a key factor in the pathogenesis of T2DM [6]. In a prospective Italian study, Bonora et al. [7] reported insulin resistance as an independent risk factor for cardiovascular disease (CVD) in diabetics. Since then, a number of subsequent studies [8, 9] have linked insulin resistance to various CVD or CVD risk factors. Diabetics were randomly selected from a database of registered diabetes patients attending the diabetic clinic at CCTH and invited to participate in the study. Controls from the general inhabitants of the Cape Coast metropolis were age-matched with the diabetic group. In all, 200 respondents consisting of 100 diabetics and 100 nondiabetic controls were enrolled for the study in 2011. Before and during malaria infection, 10 mL of venous blood sample was obtained after overnight fast. Blood sample was separated into plasma and serum for determination of various biochemical indices. Whereas plasma samples were analyzed the same day, serum samples were aliquoted appropriately and stored at −80∘C till further analysis. Weight and height were measured for computation of body mass index (BMI) using standard protocol. Apart from a study in Sudan that investigated insulin resistance in the context of childhood malaria [10], all other investigations looked at insulin resistance in terms of lipid accumulation particularly in the development of T2DM. To the best of our knowledge, no study has considered the effect of malaria on insulin resistance in adult diabetics and nondiabetic controls. 1. Introduction occurred in 2010, with associated 660 000 deaths worldwide [3], suggesting a global decline in malaria-related morbidity and mortality. In spite of the decline, the contribution of Africa to the global malaria burden remains unchanged at 80% and 90%, respectively, for malaria-related morbidity and mortality. Malaria and type 2 diabetes mellitus (T2DM) continue to affect millions of people globally. Whereas T2DM can be considered a truly global phenomenon, malaria seems to be concentrated in relatively less developed economies of the globe. This implies that the economically less developed continents of the globe have to battle with the negative impact of malaria and T2DM separately or synergistically on the health of the populace. Ironically, current report by the International Diabetes Federation (IDF) indicates that the incidence of diabetes is increasing globally [4]. It is projected that, by the year 2030, an estimated 366 million people affected by the disease in 2011 will reach 552 million with 80% of them living in low and middle income countries [4]. These statistics indicate that, in developing countries, the probability of coexistence of diabetes and malaria in the same individual is quite high making the need for studying the interaction between the two health conditions long overdue. Indeed, Danquah et al. Malaria is an intricate disease viewed as the most detri- mental and pervasive parasitic infection worldwide [1]. It impinges on the health and wellbeing of more than 500 million people and results in 1-2 million deaths yearly [1, 2]. According to the current estimates of the World Health Organization (WHO), 219 million cases of the disease Malaria Research and Treatment 2 [5] established that diabetics have increased risk of getting malaria compared to their nondiabetic controls. With Ghana being a sub-Saharan African country experiencing increased incidence of diabetes, especially, T2DM, while maintaining her endemic status with respect to malaria infection, the interaction between malaria and diabetes in the same individ- ual cannot be overlooked and this may pose additional health challenge that ought to be investigated.i [5] established that diabetics have increased risk of getting malaria compared to their nondiabetic controls. 1. Introduction In view of the predicted increased incidence of T2DM in sub-Saharan Africa, a continent that is still grappling with malaria-imposed health challenges, coupled with a recent finding that diabetics are more prone to getting malaria [5], we sought to investigate whether uncomplicated malaria in adult could alter insulin resistance and beta cell function in diabetics receiving treatment at the Cape Coast Teaching Hospital (CCTH) compared with age- matched nondiabetic controls. Fasting blood glucose (plasma) and glycosylated hemo- globin were determined the same day by the BT3000 autoan- alyzer with reagents from JAS diagnostics (JAS Diagnostics Inc., USA). Blood glucose determination followed the enzy- matic protocol with that of glycosylated hemoglobin relying on the agglutination procedure. gg p Serum insulin was measured by a commercially available quantitative enzyme-linked immunosorbent assay (ELISA) kit from Calbiotech (Calbiotech Inc., USA) with reagents from the same manufacturer. The specific ELISA is the direct sandwich type. Briefly, samples, standards, and reagents were brought to room temperature. Exactly 25 𝜇L each of the various concentrations of insulin standards (6.25, 12.5, 25, 50, and 100 𝜇IU/mL) was placed in appropriate wells. The same volume of sample was also placed in each well, followed by the addition of 100 𝜇L of 1X enzyme conjugate prepared by diluting 500 𝜇L of 20X enzyme conjugate concentrate with 9500 𝜇L of assay diluent. After thorough mixing, the setup was then incubated for 1 hour. Using an automated microplate washer (Thermo Electron Co-Operation, Fin- land), the microplate wells were washed 6 times each with 300 𝜇L of 1X wash buffer prepared by mixing 25 mL of 20X wash buffer concentrate with 475 mL of distilled water, after the incubation period. Exactly 100 𝜇L of TMB substrate was then added to each well followed by incubation at room temperature for 15 minutes and addition of 50 𝜇L stop solution. Absorbance was then read immediately at 450 nm using a Multiscan Microplate Reader (Thermo Scientific, Finland). Appropriate standard curve was prepared from absorbance values of the insulin standards. The concentration of insulin in sample was subsequently determined from the standard curve. 2. Materials and Methods [13] Malaria Research and Treatment 3 5.792 2.55 7.08 5.02 5.79 1.43 3.833 4.879 2.464 4.83 3.93 5.62 1.02 9.141 0 2 4 6 8 10 12 FBG Insulin HOMAIR Diabetic Age × 10 BMI × 10 HOMAB × 10 P < 0.001 P < 0.001 P < 0.001 P < 0.001 P = 0.021 P = 0.838 P = 0.488 Nondiabetic HbA1c Figure 1: Age, BMI, glycemic indicators, insulin, HOMAIR, and HOMAB levels of male respondents by study group. 5.792 2.55 7.08 5.02 5.79 1.43 3.833 4.879 2.464 4.83 3.93 5.62 1.02 9.141 0 2 4 6 8 10 12 FBG Insulin HOMAIR Diabetic Age × 10 BMI × 10 HOMAB × 10 P < 0.001 P < 0.001 P < 0.001 P < 0.001 P = 0.021 P = 0.838 P = 0.488 Nondiabetic HbA1c Figure 1: Age, BMI, glycemic indicators, insulin, HOMAIR, and HOMAB levels of male respondents by study group. 5.792 2.55 7.08 5.02 5.79 1.43 3.833 4.879 2.464 4.83 3.93 5.62 1.02 9.141 0 2 4 6 8 10 12 FBG Insulin HOMAIR Age × 10 BMI × 10 HOMAB × 10 P < 0.001 P < 0.001 P < 0.001 P < 0.001 P = 0.021 P = 0.838 P = 0.488 HbA1c as HOMAIR = fasting insulin (𝜇U) × fasting glucose (mmol/ L)/22.5: as HOMAIR = fasting insulin (𝜇U) × fasting glucose (mmol/ L)/22.5: HOMAB = 20 × fasting insulin (𝜇U/mL) fasting glucose (mmol/L) −3.5. (1) (1) Malaria was diagnosed by the CareStart Malaria HRP2Pf rapid diagnostic test kit manufactured by Access Bio (Access Bio Inc., USA). The CareStart Malaria HRP2Pf has been found to exhibit high sensitivity and specificity and correlates strongly with microscopy [14–16] in diagnosing falciparum malaria. Diabetics were already diagnosed type 2 diabetic patients receiving appropriate treatment at CCTH excluding exoge- nous insulin. 2.3. Statistical Analysis. Data obtained were analyzed by SPSS software version 17. Data are presented as mean ± standard deviation except the HOMAIR values that were log-transformed and presented as geometric mean ± stan- dard deviation. Mean levels of measured parameters were compared between study groups and between genders with independent sample 𝑡-test. Mean levels of parameters before and during malaria were compared across groups by one- way ANOVA with Tukey post hoc test. Pearson correlation test was used to assess linear relationship between any pair of measured parameters. A stepwise linear regression model was then applied to identify independent predictors of malaria-induced HOMAIR. A 𝑃value < 0.05 was considered statistically significant. Figure 1: Age, BMI, glycemic indicators, insulin, HOMAIR, and HOMAB levels of male respondents by study group. Figure 1: Age, BMI, glycemic indicators, insulin, HOMAIR, and HOMAB levels of male respondents by study group. Table 1: Age, BMI, glycemic indicators, HOMAIR, and HOMAB levels of respondents by study group. Table 1: Age, BMI, glycemic indicators, HOMAIR, and HOMAB levels of respondents by study group. Table 1: Age, BMI, glycemic indicators, HOMAIR, and HOMAB levels of respondents by study group. Parameter Diabetic 𝑁= 100 Nondiabetic 𝑁= 100 𝑃value Age (years) 53.43 ± 6.57 52.88 ± 10.29 0.778 BMI (kg/m2) 28.50 ± 4.65 26.09 ± 5.34 0.025∗ FBG (mmol/L) 6.68 ± 1.46 4.54 ± 1.17 <0.001∗ HbA1c (%) 5.12 ± 1.03 4.07 ± 1.02 <0.001∗ Insulin (𝜇IU/mL) 5.70 ± 1.69 5.33 ± 1.72 0.838 HOMAIR 1.59 ± 0.19 1.08 ± 0.18 <0.001∗ HOMAB 43.90 ± 25.31 116.90 ± 21.72 <0.001∗ Figures represent mean ± standard deviation; BMI: body mass index; FBG: fasting blood glucose; HbA1c: glycosylated hemoglobin; HOMAIR: homeostatic model assessment of insulin resistance; HOMAB: homeostatic model assessment of beta cell secretion; ∗significant 𝑃value; 𝑁: number of respondents. 3. Results Of the 200 respondents, 100 (70 diabetics and 30 controls) had malaria after a two-year follow-up. Age and baseline serum insulin level did not differ (𝑃> 0.05) between the defined study groups (Table 1). However, diabetics had significantly (<0.001) higher levels of FBG, HbA1c, HOMAIR, and BMI but lower level of HOMAB than their nondiabetic coun- terpart. Stratification of baseline data by intragroup gender revealed a comparable (𝑃> 0.05) mean level of all the meas- ured parameters between the sexes in both study groups except FBG and BMI. FBG was higher (𝑃= 0.023) in control males compared with their female counterpart with BMI being higher (𝑃< 0.001) in diabetics females in comparison with their male counterpart (data not shown). Figures represent mean ± standard deviation; BMI: body mass index; FBG: fasting blood glucose; HbA1c: glycosylated hemoglobin; HOMAIR: homeostatic model assessment of insulin resistance; HOMAB: homeostatic model assessment of beta cell secretion; ∗significant 𝑃value; 𝑁: number of respondents. Table 2: BMI, FBG, HbA1c, insulin, HOMAIR, and HOMAB levels of respondents during malaria by study group. 2. Materials and Methods 2.1. Study Area. The study was carried out at the Diabetic Clinic of CCTH in the Cape Coast Metropolis. CCTH serves as the referral hospital for the various health facilities in the region. It has a well-structured and active Diabetic Clinic with patients from various parts of the region. The Central Region shares boundaries with Ashanti and Eastern regions to the north, Western region to the west, Greater Accra region to the east, and the Atlantic Ocean to the south. It covers an area of 9,826 km2 with Cape Coast as its capital [11]. The region is considered the educational hub of Ghana as the majority of the country’s top senior high schools are located in the region. According to the 2010 Population and Housing Census, the regional population stood at 2,201,863 with that of the Cape Coast metropolis being 169,894 [12]. Ecologically, the region is divided into the forest and coastal zones. In terms of climate, Central region is mainly tropical but experiences South West Monsoon and North East Trade winds with average temperature of 21–31∘C and annual rainfall figure of 750–1000 mm [11]. The people are mainly farmers and fishermen in the informal sector with a relatively small proportion of the working population in the formal sector [11]. y p g g According to the 2010 Population and Housing Census, the regional population stood at 2,201,863 with that of the Cape Coast metropolis being 169,894 [12]. Ecologically, the region is divided into the forest and coastal zones. In terms of climate, Central region is mainly tropical but experiences South West Monsoon and North East Trade winds with average temperature of 21–31∘C and annual rainfall figure of 750–1000 mm [11]. The people are mainly farmers and fishermen in the informal sector with a relatively small proportion of the working population in the formal sector [11]. 2.2. Study Participants and Laboratory Determinations. The study was an outpatient-based comparative type that followed HOMAIR and HOMAB were estimated from glucose and insulin levels by the formulae of Matthews et al. 4. Discussion This study investigated malaria-induced insulin resistance in adult diabetics and controls in the Cape Coast metropolis of the Central region of Ghana. The comparable mean age of respondents between the two study groups in the current study appears contrary to a study in Kumasi that found higher age for diabetics than nondiabetic controls [5] though the higher age of diabetic males compared with their control counterpart in the present study supports the Kumasi study [5]. This seeming age-specific contradictory observation in the current study compared to that of Danquah et al. [5] could be ascribed to differences in sample size, study design, and characteristics of respondents in the two studies. Whereas the current study was limited to 200 adults aged 40 years and above, that of Danquah et al. [5] included a much larger number of adults 18 years and beyond. Indeed, gender- specific consensus on age at diagnosis as a substitute for age of onset of T2DM is yet to be reached by the scientific community. Kolb et al. [18] observed lower age of onset for men contrary to earlier one [19] that reported rather higher age for men. Recently, a number of reports [20, 21] have found comparable age of onset for males and female diabetics in line with the current study. The varied conclusions in literature regarding age of onset of T2DM between the sexes could be attributed to possible differences in racial, genetic, and environmental factors related to the evolution of the disease in various populations [22–24]. Figure 2: Age, BMI, glycemic indicators, insulin, HOMAIR, and HOMAB levels of female respondents by study group. Analysis of variance comparison of means revealed that mean levels of all the measured parameters differed signifi- cantly (𝑃< 0.001) across groups with and without malaria (Table 3). Tukey’s post hoc HSD test rather showed that, apart from the mean HOMAIR and serum insulin that increased significantly (𝑃 < 0.05) in diabetic respondents during malaria, the mean levels of all the other measured parameters were comparable (𝑃 > 0.05) before and during malaria (Table 4). However, in the control group, significantly (𝑃< 0.05) higher mean levels were observed for all the measured parameters except BMI and HOMAB (Table 4). 4. Discussion As expected, diabetics in the current study recorded higher baseline mean levels of FBG and HbA1c than controls except that the observed mean levels for diabetic respondents were lower than those reported by other researchers in China [21], India [20], Nigeria [25], and United States of America [26]. The rather low levels of baseline glycemic indicators in the current study as opposed to earlier reports [20, 21, 25] could be due to differences in characteristics of enrolled respondents with respect to duration of disease, adherence to prescribed treatment regimen, and analytical method employed in determination of the glycemic parameters. In the presence of malaria, levels of the glycemic indicators in both study groups increased with diabetics maintaining their higher levels compared with controls as observed at baseline. However, the extent of increase due to malaria was only appreciable in the nondiabetic group suggesting that diabetics in the current study did not respond to malaria by increasing blood glucose level. The appreciable increase in FBG due to malaria in the control arm of the current work supports earlier finding in Netherlands [27] and has been attributed to increased gluconeogenesis. Considering the fact that HbA1c is a marker for relative long-term blood glucose control, its appreciable increase due to Plasmodium falciparum malaria in the control group of the current study is an indication that the increased FBG was not transient. Interestingly, the comparable levels of FBG and HbA1c Malaria-induced HOMAIR correlated positively and moderately (𝑅= 0.435; 𝑃= 0.001) but strongly (𝑅= 0.901; 𝑃 < 0.001) with malaria-induced HOMAB and insulin, respectively, in the diabetic group. In a subsequent stepwise linear regression model, malaria-induced HOMAB and insulin were significant (𝑃< 0.05; 𝑅2 = 0.954; adjusted 𝑅2 = 0.915) independent predictors of HOMAIR, jointly explaining over 91% of the observed variation. Similar analyses in the control group showed positive correlation of malaria-induced HOMAIR level with malaria- induced HOMAB (𝑅= 0.479; 𝑃= 0.013), malaria-induced insulin (𝑅= 0.978; 𝑃< 0.001), and baseline insulin level (𝑅= 0.634; 𝑃= 0.011). Also, malaria-induced HOMAB correlated positively with baseline HOMAB (𝑅= 0.66; 𝑃= 0.002) and malaria-induced insulin level (𝑅= 0.686; 𝑃< 0.001). Above all, baseline HOMAB correlated positively with malaria- induced insulin level (𝑅= 0.498; 𝑃= 0.025). Table 2: BMI, FBG, HbA1c, insulin, HOMAIR, and HOMAB levels of respondents during malaria by study group. Parameter Diabetic 𝑁= 70 Nondiabetic 𝑁= 30 𝑃value FBG (mmol/L) 7.12 ± 0.60 5.27 ± 0.60 <0.001∗ BMI (kg/m2) 27.98 ± 4.53 26.01 ± 5.75 0.026∗ HbA1c (%) 5.67 ± 1.03 4.63 ± 1.07 <0.001∗ Insulin (𝜇IU/mL) 10.57 ± 5.27 15.45 ± 8.60 0.02∗ HOMAIR 3.58 ± 0.34 3.32 ± 0.55 0.563 HOMAB 69.74 ± 25.52 189.15 ± 21.40 <0.001∗ Figures represent mean ± standard deviation; BMI: body mass index; FBG: fasting blood glucose; HbA1c: glycosylated hemoglobin; HOMAIR: homeostatic model assessment of insulin resistance; HOMAB: homeostatic model assessment of beta cell secretion; ∗significant 𝑃value. Intergroup gender comparison of baseline data showed that diabetic males were significantly (𝑃< 0.05) older and had higher levels of FBG, HbA1c, and HOMAIR but lower level of HOMAB than control males (Figure 1). With respect to the females, a similar trend was observed with exception of age and BMI which were comparable (𝑃= 0.182) and higher (𝑃< 0.001), respectively, in diabetic respondents (Figure 2). Baseline insulin level did not (𝑃> 0.05) show either intergroup or intragroup gender variation (Figures 1 and 2). Figures represent mean ± standard deviation; BMI: body mass index; FBG: fasting blood glucose; HbA1c: glycosylated hemoglobin; HOMAIR: homeostatic model assessment of insulin resistance; HOMAB: homeostatic model assessment of beta cell secretion; ∗significant 𝑃value. than controls (Table 2). Interestingly, the two groups did not differ (𝑃= 0.563) in their malaria-induced HOMAIR levels (Table 2). than controls (Table 2). Interestingly, the two groups did not differ (𝑃= 0.563) in their malaria-induced HOMAIR levels (Table 2). In the presence of malaria, diabetics maintained sig- nificantly (𝑃< 0.001) higher mean levels of BMI, FBG, and HbA1c but lower levels of serum insulin and HOMAB 4 Malaria Research and Treatment Malaria Research and Treatment Malaria Research and Treatment 5.537 3.093 6.5 5.17 6.65 1.66 4.817 5.309 2.594 4.45 4.12 5.22 1.12 13.044 0 2 4 6 8 10 12 14 16 FBG Insulin HOMAIR Age × 10 BMI × 10 HOMAB × 10 P = 0.182 P < 0.001 P < 0.001 P < 0.001 P = 0.003 P = 0.451 P < 0.001 Diabetic Nondiabetic HbA1c Figure 2: Age, BMI, glycemic indicators, insulin, HOMAIR, and HOMAB levels of female respondents by study group. Conspicuously, no significant (𝑃> 0.05) association was observed between BMI and any of the measured indices of insulin sensitivity or glycemia. 4. Discussion As expected, a stepwise linear regression model revealed that malaria- induced HOMAB, malaria-induced insulin, and baseline insulin levels were independent (𝑅2 = 0.996; adjusted 𝑅2 = 0.994; 𝑃< 0.001) predictors of the observed HOMAIR level during malaria infection. The model explained over 99% of the observed variation in HOMAIR levels in controls who had malaria. Malaria Research and Treatment 5 Table 3: FBG, HbA1c, HOMAIR, and HOMAB levels of respondents with and without malaria. Parameter Diabetic Nondiabetic 𝐹value 𝑃value DM (𝑁= 70) BM (𝑁= 100) DM (𝑁= 30) BM (𝑁= 100) BMI (kg/m2) 27.98 ± 4.53 28.50 ± 4.65 26.01 ± 5.75 26.09 ± 5.34 10.314 0.01∗ FBG 7.12 ± 0.06 6.68 ± 1.46 5.27 ± 0.06 4.54 ± 1.17 35.564 <0.001∗ HbA1c 5.67 ± 1.03 5.12 ± 1.03 4.63 ± 1.07 4.07 ± 1.02 16.978 <0.001∗ Insulin 10.57 ± 5.27 5.79 ± 1.78 15.45 ± 8.60 5.33 ± 1.72 11.616 <0.001∗ HOMAIR 3.58 ± 0.34 1.59 ± 0.19 3.32 ± 0.55 1.08 ± 0.18 8.473 <0.001∗ HOMAB 69.74 ± 6.09 43.90 ± 25.30 189.15 ± 21.40 116.90 ± 21.70 6.677 <0.001∗ Figures represent mean ± standard deviation; BMI: body mass index; FBG: fasting blood glucose (mmol/L); HbA1c: glycosylated hemoglobin (%); HOMAIR: homeostatic model assessment of insulin resistance; HOMAB: homeostatic model assessment of beta cell secretion; DM: during malaria; BM: before malaria; ∗significant 𝑃value; 𝑁: number of respondents. Table 3: FBG, HbA1c, HOMAIR, and HOMAB levels of respondents with and without malaria. Figures represent mean ± standard deviation; BMI: body mass index; FBG: fasting blood glucose (mmol/L); HbA1c: glycosylated hemoglobin (%); HOMAIR: homeostatic model assessment of insulin resistance; HOMAB: homeostatic model assessment of beta cell secretion; DM: during malaria; BM: before malaria; ∗significant 𝑃value; 𝑁: number of respondents. Table 4: Tukey’s HSD comparison of mean levels of parameters in respondents with and without malaria. 4. Discussion Parameter Malaria No malaria 𝑃value Diabetic 𝑁= 70 𝑁= 100 BMI (kg/m2) 27.98 ± 4.53 28.50 ± 4.65 0.74 FBG (mmol/L) 7.12 ± 0.06 6.68 ± 1.46 0.377 HbA1c (%) 5.67 ± 1.03 5.12 ± 1.03 0.139 Insulin (𝜇IU/mL) 10.57 ± 0.73 5.79 ± 1.78 0.011∗ HOMAIR 3.58 ± 0.34 1.59 ± 0.19 0.009∗ HOMAB 69.74 ± 6.09 43.90 ± 2.53 0.364 Nondiabetic 𝑁= 30 𝑁= 100 BMI (kg/m2) 26.01 ± 5.75 26.09 ± 5.34 0.81 FBG (mmol/L) 5.27 ± 0.06 4.54 ± 1.17 0.02∗ HbA1c (%) 4.63 ± 1.07 4.07 ± 1.02 0.032∗ Insulin (𝜇IU/mL) 15.45 ± 2.04 5.33 ± 1.72 0.006∗ HOMAIR 3.32 ± 0.55 1.08 ± 0.18 0.009∗ HOMAB 189.15 ± 21.40 116.90 ± 21.70 0.169 Figures represent mean ± standard deviation; BMI: body mass index; FBG: fasting blood glucose; HbA1c: glycosylated haemoglobin, HOMAIR: homeostatic model assessment of insulin resistance, HOMAB: homeostatic model assessment of beta cell secretion; ∗significant 𝑃value; 𝑁: number of respondents. Table 4: Tukey’s HSD comparison of mean levels of parameters in respondents with and without malaria. Parameter Malaria No malaria 𝑃valu Table 4: Tukey’s HSD comparison of mean levels of parameters in respondents with and without malaria. dard deviation; BMI: body mass index; FBG: fasting blood glucose; HbA1c: glycosylated haemoglobin, HOMAIR: homeostati sistance, HOMAB: homeostatic model assessment of beta cell secretion; ∗significant 𝑃value; 𝑁: number of respondents. Figures represent mean ± standard deviation; BMI: body mass index; FBG: fasting blood glucose; HbA1c: glycosylated hae model assessment of insulin resistance, HOMAB: homeostatic model assessment of beta cell secretion; ∗significant 𝑃value of diabetics before and during malaria could be due to treatment effect since most diabetics were on metformin, an antidiabetic drug with acknowledged suppression effect on hepatic gluconeogenesis [28, 29]. On the other hand, the slight increase could be a pointer to a probable attenuated inhibition of gluconeogenesis by metformin, suggesting that increased intensity of infection may eliminate the metformin- induced inhibition of hepatic gluconeogenesis and thus result in significant increase in FBG and HbA1c. This view requires further studies to be proven in the light of a recent finding that associates HbA1c with intima-media thickness, a surrogate marker for CVD risk [30], and a larger study [31] suggesting that every 1% rise in HbA1c is associated with 1.67 risk of developing coronary heart disease in apparently healthy adults. Acknowledgments The authors are grateful to the management of UCC for funding this study. They are also grateful to the study participants for their critical collaboration and for donating blood samples necessary for the study. Above all, they would like to acknowledge the contribution of staff and management of CCTH for granting them permission to use their facility for the study. References [1] K. Haldar, S. C. Murphy, D. A. Milner, and T. E. Taylor, “Malaria: mechanisms of erythrocytic infection and pathological corre- lates of severe disease,” Annual Review of Pathology, vol. 2, pp. 217–249, 2007. Interestingly, the various measures of glycemia and insulin sensitivity did not associate with BMI, suggesting that general fat accumulation played no significant role in the observed insulin resistance in the current study. [2] D. A. Milner Jr., J. Montgomery, K. B. Seydel, and S. J. Rogerson, “Severe malaria in children and pregnancy: an update and perspective,” Trends in Parasitology, vol. 24, no. 12, pp. 590–595, 2008. Based on the above we propose that malaria-induced insulin resistance may contribute to the predicted higher increased incidence of T2DM for malaria-endemic sub- Saharan African countries [4]. Hence, control measures against malaria may contribute to reduced incidence of T2DM. [3] World Health Organization, World Malaria Report 2012, World Health Organization, Geneva, Switzerland, 2013. [4] International Diabetes Federation, “Diabetes at a glance, 2012 Africa,” in IDF Diabetes Atlas, Brussels, Belgium, 5th edition, 2012. [5] I. Danquah, G. Bedu-Addo, and F. P. Mockenhaupt, “Type 2 diabetes mellitus and increased risk for malaria infection,” Emerging Infectious Diseases, vol. 16, no. 10, pp. 1601–1604, 2010. Conflict of Interests The authors declare that there is no conflict of interests regarding the publication of this paper. Ethical Approval The study was approved by the Committee on Human Research, Publications and Ethics of the Kwame Nkrumah University of Science and Technology, School of Medical Sciences, and Komfo Anokye Teaching Hospital, Kumasi. All protocols followed were in accordance with the ethical requirements of the Ghana Health Service, CCTH, and the World Medical Association declaration of Helsinki [17]. In addition, written informed consent was obtained from all study participants. 4. Discussion Although the use of HOMAIR-dependent diagnosis of insulin resistance in different populations varies with cutoffs [38–40], a value of 2.6 appears generally acceptable as the threshold above which insulin resistance is diagnosed [41]. Generally, HOMAIR values in diabetics have been reported to be higher than nondiabetic controls [38, 38, 40] as observed at baseline in the current study. However, malaria-induced increased mean HOMAIR values above 3.2 in diabetics and controls observed in the current work have never been reported in the adult populations to the best of our knowledge. Only one study in Sudan [10] reported insulin resistance in children with severe complicated malaria, the form of the disease normally absent in semi-immune adults living in malaria-endemic countries like Ghana. As such, our finding that uncomplicated malaria in semi-immune adults could raise HOMAIR levels by over 120% and 200% in diabetic and nondiabetic controls, respectively, to levels above the accepted 2.6 value is novel and signifies overt insulin resistance due to falciparum malaria. This observa- tion, notably, in the nondiabetic group may have implications for the development of T2DM in malaria-endemic countries. With the possibility of multiple episodes of clinical malaria in the same individual, the corresponding insulin resistance per episode could predispose the affected people to the devel- opment of overt T2DM with time since insulin resistance precedes development of overt T2DM [32]. This is because the significantly increased mean HOMAIR values found in malaria-infected respondents in the current work did not appear to have resulted in a significantly increased secretory function of their beta cells as determined by HOMAB. Thus, although insulin level increased significantly due to malaria, probably to compensate for the increased resistance, the level was not high enough to overcome the observed resistance. More so, insulin secretion may not be perpetually increased to respond to malaria-induced glucose increase, envisaged in multiple malaria episodes as the beta cells can theoretically be apoptosed by persistent malaria-induced hyperglycemia, a scenario that can result in explicit T2DM in nondiabetic individuals. In the case of diabetics, malaria could attenuate the expected pharmacologic effects of the hypoglycemic drugs, increasing their risk of CVD as their insulin resistance increases, as suggested by the observed HOMAIR in the current report. Authors’ Contribution Samuel Acquah contributed to the design, conduct, data collection, analysis, interpretation, and writing of paper. Johnson Nyarko Boampong, Benjamin Ackon Eghan Jnr, and Magdalena Eriksson were involved in the design, analysis, interpretation, critical review, and writing of paper. 4. Discussion Thus, with diabetics being already at risk for CVD, further investigations ought to be carried out to ascertain the clinical relevance of the slightly elevated FBG and HbA1c levels as a result of malaria found in the current study. Insulin resistance, defined as the reduced response of insulin sensitive cells to the action of insulin, has been amply demonstrated to be central to the development of T2DM [32] and CVD [33]. The condition which coexists with reduced beta cell secretory function has been observed in the Ghanaian population [34]. It is ideally determined by the hyperinsulinemic-euglycemic clamp technique developed by DeFronzo, Tobin, and Andres over three decades ago [35]. However, this technique, which directly measures insulin resistance and is widely accepted as the gold standard, is time and labor intensive and costly and requires advanced techni- cal expertise [36], discouraging its use in a routine manner in clinical practice and research with insulin resistance as secondary objective. To this end, a number of indirect math- ematical models have been developed with the HOMAIR and HOMAB models, developed by Matthews and colleagues [13] gaining broad acceptance by the scientific community. The model, which depends on blood levels of insulin and glucose, of diabetics before and during malaria could be due to treatment effect since most diabetics were on metformin, an antidiabetic drug with acknowledged suppression effect on hepatic gluconeogenesis [28, 29]. On the other hand, the slight increase could be a pointer to a probable attenuated inhibition of gluconeogenesis by metformin, suggesting that increased intensity of infection may eliminate the metformin- induced inhibition of hepatic gluconeogenesis and thus result in significant increase in FBG and HbA1c. This view requires further studies to be proven in the light of a recent finding that associates HbA1c with intima-media thickness, a surrogate marker for CVD risk [30], and a larger study [31] suggesting that every 1% rise in HbA1c is associated with 1.67 risk of developing coronary heart disease in apparently healthy adults. Thus, with diabetics being already at risk for CVD, further investigations ought to be carried out to ascertain the clinical relevance of the slightly elevated FBG and HbA1c levels as a result of malaria found in the current study. 6 Malaria Research and Treatment 6 independent of BMI. This finding has implications for the evolution of T2DM in malaria-endemic regions of the globe. 4. Discussion has good correlation with the gold standard technique [37]. Although the use of HOMAIR-dependent diagnosis of insulin resistance in different populations varies with cutoffs [38–40], a value of 2.6 appears generally acceptable as the threshold above which insulin resistance is diagnosed [41]. Generally, HOMAIR values in diabetics have been reported to be higher than nondiabetic controls [38, 38, 40] as observed at baseline in the current study. However, malaria-induced increased mean HOMAIR values above 3.2 in diabetics and controls observed in the current work have never been reported in the adult populations to the best of our knowledge. Only one study in Sudan [10] reported insulin resistance in children with severe complicated malaria, the form of the disease normally absent in semi-immune adults living in malaria-endemic countries like Ghana. As such, our finding that uncomplicated malaria in semi-immune adults could raise HOMAIR levels by over 120% and 200% in diabetic and nondiabetic controls, respectively, to levels above the accepted 2.6 value is novel and signifies overt insulin resistance due to falciparum malaria. This observa- tion, notably, in the nondiabetic group may have implications for the development of T2DM in malaria-endemic countries. With the possibility of multiple episodes of clinical malaria in the same individual, the corresponding insulin resistance per episode could predispose the affected people to the devel- opment of overt T2DM with time since insulin resistance precedes development of overt T2DM [32]. This is because the significantly increased mean HOMAIR values found in malaria-infected respondents in the current work did not appear to have resulted in a significantly increased secretory function of their beta cells as determined by HOMAB. Thus, although insulin level increased significantly due to malaria, probably to compensate for the increased resistance, the level was not high enough to overcome the observed resistance. More so, insulin secretion may not be perpetually increased to respond to malaria-induced glucose increase, envisaged in multiple malaria episodes as the beta cells can theoretically be apoptosed by persistent malaria-induced hyperglycemia, a scenario that can result in explicit T2DM in nondiabetic individuals. In the case of diabetics, malaria could attenuate the expected pharmacologic effects of the hypoglycemic drugs, increasing their risk of CVD as their insulin resistance increases, as suggested by the observed HOMAIR in the current report. Interestingly, the various measures of glycemia and has good correlation with the gold standard technique [37]. 5. Conclusion 5, pp. 214–217, 2010. [11] A publication of Ministry of Local Government and Rural Development on Cape Coast Metropolitan demographic char- acteristics, 2013, http://www.ghanadistricts.com/districts/?r=3 & =50&sa=2657. [26] H. E. Bays, R. B. Goldberg, K. E. Truitt, and M. R. Jones, “Colesevelam hydrochloride therapy in patients with type 2 diabetes mellitus treated with metformin: glucose and lipid effects,” Archives of Internal Medicine, vol. 168, no. 18, pp. 1975– 1983, 2008. [12] Ghana Statistical Service, 2010 Population and Housing Census, Summary Results of Final Report, Sakoa Press, Accra, Ghana, 2012. [27] R. M. E. Bl¨umer, H. van Thien, A. F. C. Ruiter et al., “Adiponectin and glucose production in patients infected with Plasmodium falciparum,” Metabolism: Clinical and Experimental, vol. 54, no. 1, pp. 60–66, 2005. [13] D. R. Matthews, J. P. Hosker, A. S. Rudenski, B. A. Naylor, D. F. Treacher, and R. C. Turner, “Homeostasis model assessment: insulin resistance and 𝛽-cell function from fasting plasma glucose and insulin concentrations in man,” Diabetologia, vol. 28, no. 7, pp. 412–419, 1985. [28] M. Foretz, S. H´ebrard, J. Leclerc et al., “Metformin inhibits hep- atic gluconeogenesis in mice independently of the LKB1/AMPK pathway via a decrease in hepatic energy state,” The Journal of Clinical Investigation, vol. 120, no. 7, pp. 2355–2369, 2010. [14] J. Maltha, P. Gillet, E. Bottieau, L. Cnops, M. van Esbroeck, and J. Jacobs, “Evaluation of a rapid diagnostic test (CareStart Malaria HRP-2/pLDH (Pf/pan) Combo Test) for the diagnosis of malaria in a reference setting,” Malaria Journal, vol. 9, article 171, 2010. [29] Y. D. Kim, K.-G. Park, Y.-S. Lee et al., “Metformin inhibits hep- atic gluconeogenesis through AMP-activated protein kinase- dependent regulation of the orphan nuclear receptor SHP,” Diabetes, vol. 57, no. 2, pp. 306–314, 2008. [15] A. Woyessa, W. Deressa, A. Ali, and B. Lindtjørn, “Evaluation of CareStart malaria Pf/Pv combo test for Plasmodium falciparum and Plasmodium vivax malaria diagnosis in Butajira area, south-central Ethiopia,” Malaria Journal, vol. 12, article 218, 2013. [30] T. Bobbert, K. Mai, A. Fischer-Rosinsk´y, A. F. H. Pfeiffer, and J. Spranger, “A1𝐶is associated with intima-media thickness in individuals with normal glucose tolerance,” Diabetes Care, vol. 33, no. 1, pp. 203–204, 2010. [16] B. Moges, B. Amare, Y. Belyhun et al., “Comparison of CareStart HRP2/pLDH COMBO rapid malaria test with light microscopy in north-west Ethiopia,” Malaria Journal, vol. 11, article 234, 2012. [31] J. K. Pai, L. E. Cahill, F. B. Hu, K. M. 5. Conclusion [6] M. K. Badman and J. S. Flier, “The adipocyte as an active partic- ipant in energy balance and metabolism,” Gastroenterology, vol. 132, no. 6, pp. 2103–2115, 2007. Uncomplicated falciparum malaria caused insulin resis- tance in diabetic and nondiabetic semi-immune adults, Malaria Research and Treatment 7 [7] E. Bonora, G. Formentini, F. Calcaterra et al., “HOMA-es- timated insulin resistance is an independent predictor of car- diovascular disease in type 2 diabetic subjects: prospective data from the verona diabetes complications study,” Diabetes Care, vol. 25, no. 7, pp. 1135–1141, 2002. [21] B. Zhao, Y. Liu, Y. Zhang et al., “Gender difference in carotid intima-media thickness in type 2 diabetic patients: a 4-year follow-up study,” Cardiovascular Diabetology, vol. 11, article 51, 2012. [22] S. Ali, R. Chopra, S. Manvati et al., “Replication of type 2 diabetes candidate genes variations in three geographically unrelated Indian population groups,” PLoS ONE, vol. 8, no. 3, Article ID e58881, 2013. [8] H. Knobler, F. Abbasi, C. Lamendola, and G. M. Reaven, “Insulin resistance and cardiovascular disease risk factors in subjects with prehypertension,” Diabetes and Vascular Disease Research, vol. 8, no. 1, pp. 43–46, 2011. [23] S. R. de Rooij, G. Nijpels, P. M. Nilsson et al., “Low-grade chronic inflammation in the relationship between insulin sen- sitivity and cardiovascular disease (RISC) population: associa- tions with insulin resistance and cardiometabolic risk profile,” Diabetes Care, vol. 32, no. 7, pp. 1295–1301, 2009. [9] P. Xun, Y. Wu, Q. He, and K. He, “Fasting insulin concentrations and incidence of hypertension, stroke, and coronary heart disease: a meta-analysis of prospective cohort studies,” The American Journal of Clinical Nutrition, vol. 98, no. 6, pp. 1543– 1554, 2013. [24] J. C. Florez, J. Hirschhorn, and D. Altshuler, “The inherited basis of diabetes mellitus: implications for the genetic analysis of c omplex traits,” Annual Review of Genomics and Human Genetics, vol. 4, pp. 257–291, 2003. [10] E. M. Eltahir, G. ElGhazali, T. M. E. A-Elgadir, I. E. A-Elbasit, M. I. Elbashir, and H. A. Giha, “Raised plasma insulin level and homeostasis model assessment (HOMA) score in cerebral malaria: evidence for insulin resistance and marker of viru- lence,” Acta Biochimica Polonica, vol. 57, no. 4, pp. 513–520, 2010. [25] E. J. Ikekpeazu, E. E. Neboh, I. C. Maduka, I. J. Nwagbara, and M. W. Nwobodo, “Type-2 diabetes mellitus and malaria parasitaemia: effect on liver function tests,” Asian Journal of Medical Science, vol. 2, no. 5. Conclusion Rexrode, J. E. Manson, and E. B. Rimm, “Hemoglobin A1𝐶is associated with increased risk of incident coronary heart disease among apparently healthy, nondiabetic men and women,” Journal of the American Heart Association, vol. 2, no. 2, 2013. [17] World Health Organization, “World Medical Association Dec- laration of Helsinki. Ethical principles for medical research involving human subjects,” Bulletin of the World Health Orga- nization, vol. 79, no. 4, pp. 373–374, 2001. [32] Y. Song, J. E. Manson, L. Tinker et al., “Insulin sensitivity and insulin secretion determined by homeostasis model assessment and risk of diabetes in a multiethnic cohort of women: the women’s health initiative observational study,” Diabetes Care, vol. 30, no. 7, pp. 1747–1752, 2007. [18] H. Kolb, B. Schneider, L. Heinemann et al., “Type 2 diabetes phenotype and progression is significantly different if diag- nosed before versus after 65 years of age,” Journal of Diabetes Science and Technology, vol. 2, no. 1, pp. 82–90, 2008. [33] E. Bonora, S. Kiechl, J. Willeit et al., “Insulin resistance as estimated by homeostasis model assessment predicts incident symptomatic cardiovascular disease in caucasian subjects from the general population: the Bruneck study,” Diabetes Care, vol. 30, no. 2, pp. 318–324, 2007. [19] R. A. Hegele, H. Cao, S. B. Harris, B. Zinman, A. J. G. Hanley, and C. M. Anderson, “Gender, obesity, hepatic nuclear factor-1𝛼 G319S and the age-of-onset of type 2 diabetes in Canadian Oji- Cree,” International Journal of Obesity, vol. 24, no. 8, pp. 1062– 1064, 2000. [34] A. G. B. Amoah, D. P. Schuster, T. Gaillard, and K. Osei, “Insulin resistance, beta cell function and cardiovascular risk factors in Ghanaians with varying degrees of glucose tolerance,” Ethnicity & Disease, vol. 12, no. 4, supplement 3, pp. 10–17, 2002. [20] M. Nakhjavani, A. Morteza, Y. Jenab et al., “Gender difference in albuminuria and ischemic heart disease in type 2 diabetes,” Clinical Medicine and Research, vol. 10, no. 2, pp. 51–56, 2012. Malaria Research and Treatment 8 [35] R. A. DeFronzo, J. D. Tobin, and R. Andres, “Glucose clamp technique: a method for quantifying insulin secretion and resistance,” The American Journal of Physiology, vol. 237, no. 3, pp. E214–E223, 1979. [36] R. Muniyappa, S. Lee, H. Chen, and M. J. Quon, “Current approaches for assessing insulin sensitivity and resistance in vivo: advantages, limitations, and appropriate usage,” The Amer- ican Journal of Physiology—Endocrinology and Metabolism, vol. 294, no. 1, pp. E15–E26, 2008. [37] K.-H. 5. Conclusion Lin, T.-L. Liou, L.-C. Hsiao, and C.-M. Hwu, “Clinical and biochemical indicators of homeostasis model assessment- estimated insulin resistance in postmenopausal women,” Jour- nal of the Chinese Medical Association, vol. 74, no. 10, pp. 442– 447, 2011. [38] A. Esteghamati, H. Ashraf, O. Khalilzadeh et al., “Optimal cut-off of homeostasis model assessment of insulin resistance (HOMA-IR) for the diagnosis of metabolic syndrome: third national surveillance of risk factors of non-communicable diseases in Iran (SuRFNCD-2007),” Nutrition & Metabolism, vol. 7, article 26, 2010. [39] H.-Q. Qu, Q. Li, A. R. Rentfro, S. P. Fisher-Hoch, and J. B. McCormick, “The definition of insulin resistance using HOMA- IR for americans of mexican descent using machine learning,” PLoS ONE, vol. 6, no. 6, Article ID e21041, 2011. [40] S. Lee, S. Choi, H. J. Kim et al., “Cutoff values of surrogate measures of insulin resistance for metabolic syndrome in Korean non-diabetic adults,” Journal of Korean Medical Science, vol. 21, no. 4, pp. 695–700, 2006. [41] J. F. Ascaso, S. Pardo, J. T. Real, R. I. Lorente, A. Priego, and R. Carmena, “Diagnosing insulin resistance by simple quantitative methods in subjects with normal glucose metabolism,” Diabetes Care, vol. 26, no. 12, pp. 3320–3325, 2003.
https://openalex.org/W2982245552	https://europepmc.org/articles/pmc6837212?pdf=render	English	null	Attractiveness of thermally different, uniformly black targets to horseflies:<i>Tabanus tergestinus</i>prefers sunlit warm shiny dark targets	Royal Society open science	2,019	cc-by	12,576	royalsocietypublishing.org/journal/rsos Research Cite this article: Horváth G, Pereszlényi Á, Tóth T, Polgár S, Jánosi IM. 2019 Attractiveness of thermally different, uniformly black targets to horseflies: Tabanus tergestinus prefers sunlit warm shiny dark targets. R. Soc. open sci. 6: 191119. Research Cite this article: Horváth G, Pereszlényi Á, Tóth T, Polgár S, Jánosi IM. 2019 Attractiveness of thermally different, uniformly black targets to horseflies: Tabanus tergestinus prefers sunlit warm shiny dark targets. R. Soc. open sci. 6: 191119. Gábor Horváth1, Ádám Pereszlényi1,3, Tímea Tóth1, Szabolcs Polgár1 and Imre M. Jánosi2,4 GH, 0000-0002-9008-2411 From a large distance tabanid flies may find their host animal by means of its shape, size, motion, odour, radiance and degree of polarization of host-reflected light. After alighting on the host, tabanids may use their mechano-, thermo-, hygro- and chemoreceptors to sense the substrate characteristics. Female tabanids prefer to attack sunlit against shady dark host animals, or dark against bright hosts for a blood meal, the exact reasons for which are unknown. Since sunlit darker surfaces are warmer than shady ones or sunlit/shady brighter surfaces, the differences in surface temperatures of dark and bright as well as sunlit and shady hosts may partly explain their different attractiveness to tabanids. We tested this observed warmth preference in field experiments, where we compared the attractiveness to tabanids (Tabanus tergestinus) of a warm and a cold shiny black barrel imitating dark hosts with the same optical characteristics. Using imaging polarimetry, thermography and Schlieren imaging, we measured the optical and thermal characteristics of both barrels and their small-scale models. We recorded the number of landings on these targets and measured the time periods spent on them. Our study revealed that T. tergestinus tabanid flies prefer sunlit warm shiny black targets against sunlit or shady cold ones with the same optical characteristics. These results support our new hypothesis that a blood-seeking female tabanid prefers elevated temperatures, partly because Subject Category: Biology (whole organism) Subject Areas: physiology/ecology/behaviour Keywords: horsefly, host choice, blood-seeking, thermoreception, polarization vision, parasite–host interaction Subject Category: Biology (whole organism) Subject Areas: physiology/ecology/behaviour Keywords: horsefly, host choice, blood-seeking, thermoreception, polarization vision, parasite–host interaction 1Environmental Optics Laboratory, Department of Biological Physics, and 2Department of Physics of Complex Systems, ELTE Eötvös Loránd University, 1117 Budapest, Pázmány sétány 1, Hungary 3Hungarian Natural History Museum, Department of Zoology, Bird Collection, 1083 Budapest, Ludovika tér 2-6, Hungary 4Max Planck Institute for the Physics of Complex Systems, Nöthnitzer Strasse 38, 01187 Dresden, Germany Author for correspondence: Gábor Horváth e-mail: gh@arago.elte.hu Author for correspondence: Gábor Horváth e-mail: gh@arago.elte.hu Electronic supplementary material is available online at https://doi.org/10.6084/m9.figshare.c. 4692362. Electronic supplementary material is available online at https://doi.org/10.6084/m9.figshare.c. 4692362. © 2019 The Authors. Published by the Royal Society under the terms of the Creative Commons Attribution License http://creativecommons.org/licenses/by/4.0/, which permits unrestricted use, provided the original author and source are credited. Gábor Horváth1, Ádám Pereszlényi1,3, Tímea Tóth1, Szabolcs Polgár1 and Imre M. Jánosi2,4 http://dx.doi.org/10.1098/rsos.191119 Received: 25 June 2019 Accepted: 23 September 2019 Received: 25 June 2019 Accepted: 23 September 2019 1. Introduction Cattle annoyed by tabanids grow thinner, thus their meat and milk production decreases [1]. Tabanid attack and annoyance result in health problems of livestock and economical losses for horse- and cattle-keepers [2,3]. Thus, there is a great need for effective tabanid traps for tabanid control. In order to design such traps, it is important to know the reasons for host attractiveness to tabanids. The attractiveness of host animals to tabanids has several different reasons, such as feeding on vertebrate secretions (e.g. sweat, tears and blood as nutrition sources), or the use of the host’s body surface as a basking or swarming site [4]. From a large distance, tabanids find their host animals by means of different cues, such as the radiance (brightness/darkness and/or colour and/or contrast) [5–11] and the degree of polarization [12–18] of host-reflected light, shape and size [5,8,9,19,20] and odour [21–27]. The role of the host’s movement in host finding by tabanids is equivocal, because depending on the target’s velocity, some species prefer moving targets, others are not influenced by the target’s motion [5,9,28,29]. After alighting on the body surface of a host, female tabanids decide whether they remain on it to (i) look for vessels for blood-sucking or (ii) rest (bask, for instance). Being on the host’s body surface, tabanids use their mechano-, thermo-, hygro- and chemoreceptors to test the physical and physiological characteristics of the substrate [4,30]. These receptors are most effective if the fly is in contact with the host’s body surface. Therefore, mainly odour and vision are used to locate hosts from a distance. However, tabanids may decide to land or not to land already in close vicinity of the target without contact. The thermoreceptors in tabanids (as usually in Diptera) are in the legs, antennae and mouthparts [4,31,32]. Female tabanids prefer to attack sunlit against shady dark (brown, black) host animals, or dark against bright (grey, white) hosts for a blood meal [16,33–35]. A qualitative observation by Tashiro & Schwardt [36] showed that tabanids attack dark cattle more frequently than white. In a quantitative field experiment, the attractiveness of sunlit brown horses to tabanids was about four times larger than that of sunlit white ones [33]. That is why the most effective tabanid traps use black decoy targets [8,19,37–41]. Field experiments and observations have shown that tabanids prefer sunlit dark hosts/targets [16,34,42]. 2 her wing muscles are more rapid and her nervous system functions better (due to faster conduction velocities and synaptic transmission of signals) in a warmer microclimate, and thus, she can avoid the parasite-repelling reactions of host animals by a prompt take-off. her wing muscles are more rapid and her nervous system functions better (due to faster conduction velocities and synaptic transmission of signals) in a warmer microclimate, and thus, she can avoid the parasite-repelling reactions of host animals by a prompt take-off. royalsocietypublishing.org/journal/rsos R. Soc. open sci. 6: 191119 3 In this work, we will test the hypothesis that the thermal characteristics of (sunlit/shady) dark hosts play an important role in the host choice by tabanids. In sunshine, the surface features of dark and bright hosts of the same species differ mainly in temperature: sunlit darker surfaces are warmer than brighter ones, because the former absorb more sunlight than the latter. Therefore, our working hypothesis is that the differences in the surface temperature between dark and bright hosts can explain the differences in their attractiveness to tabanids landed (because from a distance, a tabanid cannot sense the host’s surface temperature). Tabanids prefer to congregate in enclosures that accumulate heat during daylight, notably motor vehicles [46]. Most tabanid species feed during full daylight and are often most in evidence on the hottest, sunniest days. Most species are exophilic and exophagic, an exception being the loiasis vector, Chrysops silaceus, which will enter houses to feed [4]. However, Bracken & Thorsteinson [28] found that black or white decoys simulating warmth of mammalian hosts were not significantly more attractive to tabanids than similar objects at ambient temperature. On the other hand, when a female tabanid has landed on a host, heat is also required to stimulate her blood-sucking [47]. Thus, the role of the surface temperature of the hosts in the host choice of tabanids is equivocal. In previous studies, the attractiveness of thermally different but optically uniform targets to tabanids has not been compared. To fill this gap, we performed field experiments, in which the attractiveness to tabanids of a warm and a cold shiny black barrel (imitating dark hosts) with the same optical characteristics was compared. Using imaging polarimetry, thermography and Schlieren imaging, we measured the physical (optical and thermal) characteristics of both targets and their small-scale models. We recorded the number of landings on these targets and measured the time periods spent on them. Our study revealed the tabanid attractiveness of optically uniform dark host-mimicking targets versus the target’s surface temperature. We showed that tabanids prefer to land on sunlit surfaces of elevated temperatures and they stay there much longer than on colder targets of equivalent optical properties. We formulate a hypothesis that such preference might be related to a thermal strategy where an optimal temperature of wing muscles can be a selective advantage. 1. Introduction Horse- and cattle-keepers know well that tabanids practically do not follow their host animals to the shade of stables or forests, thus hosts can escape from tabanid attacks into shady regions. Horváth et al. [33] found that in comparison with a white horse, a brown horse spent 2.2 times longer period in a tabanid-free shady forest than in the sunny field with intense tabanid attacks. Using video recordings of tabanids in the field, Caro et al. [43] could study the behaviour of horseflies only around sunlit zebras and horses, because tabanids practically do not attack shady host animals. Otártics et al. [44] found that H-traps in sunny places caught significantly more female tabanids than at shady sites. The tabanid preference of sunlit dark hosts/targets is rather enigmatic, because in the case of a given host species, the blood quality (with respect to the egg production in female tabanids) of darker hosts should practically be the same as that of brighter ones. Similarly, the mechanical (skin, hair) and chemical (carbon dioxide, odour) characteristics of the body surface of darker and brighter hosts of the same species cannot be very different. A relevant difference can be, however, that sunlit animals could sweat more than shady ones (especially if a sunlit dark animal is compared with a shady bright one). A more sweating host animal may have a bacteria population growing differently compared to a less sweaty host, which can thus influence chemicals released by the host’s skin. According to a widespread explanation, the responsiveness of blood-sucking insects (including tabanids) to targets showing strong radiance contrast with the background is not surprising as large homeothermal host animals as blood sources are low-radiance objects that will appear dark in contrast with high-radiance vegetation [45]. Therefore, host-seeking insects (together with tabanids) prefer to land on dark, low-radiance colours similar to those of many host animals [4]. However, the explanation of the attraction of tabanids by dark hosts is more complex, because bright brown and grey ungulates (e.g. horses, cattle, antelopes) belong also to the hosts of tabanids. Apart from darkness, the degree of polarization of target-reflected light plays also an important role: a dark target is more attractive if it is shiny, thus strongly polarizing, because matt black surfaces, with only a weak polarizing ability, are much less attractive to tabanids than shiny ones [15,33]. 1. Introduction Furthermore, high degrees of polarization of reflected light help tabanids to select sunlit dark host animals from the dark patches of their visual environment [13]. is more attractive if it is shiny, thus strongly polarizing, because matt black surfaces, with only a weak polarizing ability, are much less attractive to tabanids than shiny ones [15,33]. Furthermore, high degrees of polarization of reflected light help tabanids to select sunlit dark host animals from the dark patches of their visual environment [13]. 3 2.1. Choice experiments Our choice experiments 1–5 were performed on a horse farm in Göd (47°430 N, 19°090 E) in July 2018 from early forenoon to late afternoon of warm, sunny and windless days (experiment 1: 10 July, 09.50–17.30 = UTC + 2 h; experiment 2: 14 July, 10.30–18.00; experiment 3: 15 July, 10.30–18.00; experiment 4: 21 July, 10.00–17.30; experiment 5: 27 July, 10.00–14.00) where tabanid flies (Tabanus tergestinus Egger 1859) are abundant in summer, as we have experienced in our earlier field experiments [14,16,33,48,49]. Two shiny black plastic cylindrical barrels (height = 42 cm, diameter = 30 cm, wall thickness = 5 mm) with black caps were placed on white tetrapodal plastic stools (height = 46 cm) 1 m apart at a site where there was not any shade cast by an object (vegetation, building) during the experiments. We used barrels with a shiny black outer surface, because due to the rule of Umow [50], smooth black surfaces reflect light with the highest degree of polarization, and blood-seeking female tabanids are the strongest attracted to such polarized light. The warm barrel contained only air which became and remained warm in sunshine, because its closed cap hindered the emanation of warm/hot air. The cold barrel was filled with tap water, into which 10 iceakkus (Aspico G40, 0.25 l, 0.76 kg plastic container filled with a gel of low temperature of congelation) frozen in a common kitchen-refrigerator were submerged. Its closed cap hindered the outflow of the cold water or the emanation of water vapour. Since in sunshine, the outer surface of the air-filled barrel became warm, while that of the water-iceakku-filled barrel remained relatively cold, the former and latter barrel are henceforward called ‘warm’ and ‘cold’, respectively. The positions of the warm and cold barrels were hourly inverted during each experiment. Since in experiment 1 the iceakkus were not refreshed, the outer surface temperature of the cold barrel gradually increased in time, and near the end of this experiment it approximated that of the warm barrel. 2.1. Choice experiments To eliminate this 15 20 25 30 35 40 45 50 55 60 65 cloudy cloudy cloudy experiment 1 (10 July 2018) sw aw sc ac 15 20 25 30 35 40 45 50 55 60 65 sw aw sc ac 10.00 11.00 12.00 13.00 14.00 15.00 16.00 17.00 18.00 time t (UTC + 2 h) 10.00 11.00 12.00 13.00 14.00 15.00 16.00 17.00 18.00 time t (UTC + 2 h) surface temperature T (°C) surface temperature T (°C) cloudy cloudy iceakku refreshment experiment 2 (14 July 2018) (b) (a) e 1. (a) Average (different symbols) ± s.d. (vertical bars) of the surface temperature T of the solar and antisolar sides of the and cold barrels in experiment 1 (on 10 July 2018) measured by thermography as a function of time t (local solar time = + 2 h). At a given point of time, T was averaged in the rectangles on the barrel’s shells shown in figure 2c,d. sw, solar side of arm barrel; aw, antisolar side of the warm barrel; sc, solar side of the cold barrel; ac, antisolar side of the cold barrel. (b) As (a) periment 2 (on 14 July 2018) in which the iceakkus were refreshed in the cold barrel at 13.30 represented by a vertical orange royalsocietypublishing.org/journal/rsos R. Soc. open sci. 6: 191119 4 royalsocietypublishing.org/journal/rsos R. Soc. open sci. 6: 191119 4 15 20 25 30 35 40 45 50 55 60 65 cloudy cloudy cloudy experiment 1 (10 July 2018) sw aw sc ac 10.00 11.00 12.00 13.00 14.00 15.00 16.00 17.00 18.00 time t (UTC + 2 h) surface temperature T (°C) (a) royalsocietypublishing.org/journal/rsos R. Soc. open sci. 6: 191119 4 4 15 20 25 30 35 40 45 50 55 60 65 sw aw sc ac 10.00 11.00 12.00 13.00 14.00 15.00 16.00 17.00 18.00 time t (UTC + 2 h) ( ) surface temperature T (°C) cloudy cloudy iceakku refreshment experiment 2 (14 July 2018) (b) Figure 1. (a) Average (different symbols) ± s.d. (vertical bars) of the surface temperature T of the solar and antisolar sides of the warm and cold barrels in experiment 1 (on 10 July 2018) measured by thermography as a function of time t (local solar time = UTC + 2 h). At a given point of time, T was averaged in the rectangles on the barrel’s shells shown in figure 2c,d. 2.1. Choice experiments sw, solar side of the warm barrel; aw, antisolar side of the warm barrel; sc, solar side of the cold barrel; ac, antisolar side of the cold barrel. (b) As (a) for experiment 2 (on 14 July 2018) in which the iceakkus were refreshed in the cold barrel at 13.30 represented by a vertical orange line. Grey vertical bands show the intervals when the sun was occluded by clouds. 15 20 25 30 35 40 45 50 55 60 65 sw aw sc ac 10.00 11.00 12.00 13.00 14.00 15.00 16.00 17.00 18.00 time t (UTC + 2 h) surface temperature T (°C) cloudy cloudy iceakku refreshment experiment 2 (14 July 2018) (b) Figure 1. (a) Average (different symbols) ± s.d. (vertical bars) of the surface temperature T of the solar and antisolar sides of the warm and cold barrels in experiment 1 (on 10 July 2018) measured by thermography as a function of time t (local solar time = UTC + 2 h). At a given point of time, T was averaged in the rectangles on the barrel’s shells shown in figure 2c,d. sw, solar side of the warm barrel; aw, antisolar side of the warm barrel; sc, solar side of the cold barrel; ac, antisolar side of the cold barrel. (b) As (a) for experiment 2 (on 14 July 2018) in which the iceakkus were refreshed in the cold barrel at 13.30 represented by a vertical orange line. Grey vertical bands show the intervals when the sun was occluded by clouds. problem, in experiments 2–5, the iceakkus were refreshed in the cold barrel at half-time. Thus, we could ensure relatively great temperature differences between the warm and cold barrels (figure 1). Since we did not want to lose the results of experiment 1, we used these data, although the temporal temperature variation of the water-filled cold barrel was different in the first experiment (without refreshing the iceakkus) compared to the other four experiments (with refreshing the iceakkus). The only irrelevant consequence of this was that in the second half of experiment 1, higher surface 40 60 65 temperature (°C) 70 45 50 55 20 25 30 35 15 sunshine, solar side warm cold photograph thermogram sunshine, antisolar side warm cold warm cold warm cold (a) (c) (d) (b) Figure 2. 2.1. Choice experiments Photographs (a,b) and thermograms (c,d) of the solar (a,c) and antisolar (b,d) sides of the warm and cold black barrels used in the choice experiments measured by a thermocamera in sunshine at 09.50 (=UTC + 2 h) on 10 July 2018. The rectangles on the barrel’s shells show the areas where the surface temperature T of the barrels was averaged. royalsocietypublishing.org/journal/rsos R. Soc. open sci. 6: 191119 5 sunshine, solar side warm cold photograph (a) 5 sunshine, antisolar side warm cold (b) royalsocietypublishing.org/journal/rsos R. Soc. open sci. 6: 191119 5 (a) (b) photograph warm cold (d) 40 60 65 temperature (°C) 70 45 50 55 20 25 30 35 15 thermogram warm cold warm cold (c) (d) (d) thermogram Figure 2. Photographs (a,b) and thermograms (c,d) of the solar (a,c) and antisolar (b,d) sides of the warm and cold black barrels used in the choice experiments measured by a thermocamera in sunshine at 09.50 (=UTC + 2 h) on 10 July 2018. The rectangles on the barrel’s shells show the areas where the surface temperature T of the barrels was averaged. temperatures of the water-filled barrel occurred than in experiments 2–5. This, however, did not influence our results and conclusions: our main aim was to determine the number of landings and the average time period spent by tabanids on target surfaces as a function of the surface temperature T. For this aim, many different T-values of the barrel’s surface were necessary. It was all the same that these data are collected by one experimenter observing only the solar sides of the warm and cold barrels, or by two experimenters observing simultaneously both solar and antisolar sides. We used plastic barrels, the wall (thickness = 5 mm) of which has much lower thermal conductivity than that of metal barrels. Thus, there were significant temperature differences between our warmer and colder barrels during the whole experiment. This temperature difference was further enhanced with the use of iceakkus instead of frozen water (i.e. ice). The thermal capacity of the former is much larger than that of the latter, thus the submerged iceakkus kept the water in our plastic barrels colder for a longer period than submerged ice. The tabanids landing on the barrels were observed by one (experiments 1–2) or two (experiments 3–5) persons at a distance of 2 m symmetrically from the barrels. 2.1. Choice experiments These observers wore white clothes and hats against direct sunshine and to minimize their visual attractiveness to tabanids. They were sitting on a chair during the observations. Since we wanted to determine on the spot the species of tabanids landed on our barrels, we observed them with the naked eye 2 m from the barrels, rather than tallying them from a greater distance with binoculars as, for example, Vale [51] performed with tsetse flies. In experiments 1–2, only the solar sides (facing towards the sun) of the barrels were continuously observed by a person (because then only one researcher was available for this task), while in experiments 3–5, both the solar and antisolar (facing towards the antisolar point) sides were observed by two researchers (because then two researchers were available). The direction of observation was perpendicular to the connecting line passing through the centre of the barrels. This connecting line was hourly changed in such a way that its normal vector pointed towards the sun. By this, we ensured that the sunlight hitting the barrels was always nearly parallel to the direction of observation. During observations, we counted with a 5 min temporal resolution how many landings were on the solar/antisolar sides of the warm/cold barrels, and how many seconds the landed tabanids spent on them. Counting happened with the naked eye, and time periods were measured with a stopwatch. The distribution of the outer surface temperature of the barrels was periodically (experiment 1: half hourly, experiments 2–5: hourly) measured by a thermocamera on their solar and antisolar sides from the direction of observation. We used dry (non-sticky) barrels, because we wanted to measure the time period spent by alighted tabanids on the barrels. Since our barrels were non-sticky, a given tabanid individual could have landed more than once on them. Such pseudoreplication could be eliminated only if the barrels were sticky. However, from a sticky barrel, a landed tabanid cannot fly away, thus the time period spent by the insect on the barrel cannot be measured. Therefore, we decided not to cover the surface of our barrels with an insect-monitoring glue. This method is frequently used in field experiments studying the attractiveness of certain (e.g. matt) test surfaces to insects (e.g. [15,17,18,52]). 2.3. Thermography The distributions of the surface temperature (thermograms) of two (one sunlit and one shady) white, two brown and two black horses as well as the two black barrels used in our field experiments were measured with an infrared camera (VarioCAM®, Jenoptik Laser Optik Systeme GmbH, Jena, Germany) with a nominal precision of ±1.5°C. Horses were measured under sunny (illuminated with direct sunlight) and shady (when the sun was behind a cloud) conditions. At a given (sunlit or shady) horse, we recorded several thermograms and selected a typical thermogram where the optical axis of the thermocamera was nearly perpendicular to the long axis of the horse. The thermograms of the sunlit and shady sides of barrels were measured periodically (half hourly or hourly) in full sunshine under cloudless sky from the beginning to the end of each experiment. To determine the surface temperature distribution on horses and barrels along straight lines and in arbitrary areas, we used a custom- developed software. On the thermograms of horses, tilted and nearly horizontal lines were selected, along which the average 〈T〉and standard deviation ΔT of the surface temperature T were calculated in order to probe typical T-values of sunlit and shady horses. On the barrel’s sides, a uniform rectangular area was selected, where 〈T〉and ΔT were determined. The air temperature was not continuously monitored during our experiments, but it changed between about 28 and 36°C from the beginning to the end of the experiments. The correctness of the temperature measurement of our thermocamera was validated by a calibration procedure with a contact thermometer (GAO Digital Multitester EM392B 06554H, EverFlourish Europe Gmbh, Friedrichsthal, Germany) with a nominal precision of ±1°C. Further details about this calibration can be read in electronic supplementary material, figures S1–S3. 2.1. Choice experiments Finally, the insect- trapping efficiencies of warm and cold sticky surfaces could be different due to the different viscosities of the warm and cold glue covering the surfaces. This is a further reason why sticky barrels cannot be used in our choice experiments. royalsocietypublishing.org/jou 6 6 royalsocietypublishing.org/journal/rsos R. Soc. open sci. 6: 191119 2.2. Imaging polarimetry Tabanids are polarotactic insects attracted to (among other cues) linearly polarized light when females and males look for water by horizontal polarization and females seek host animals by high degrees of linear polarization [52]. Therefore, we measured the reflection–polarization characteristics of the two (warm, cold) black barrels used in our field experiments with imaging polarimetry in the red (650 nm), green (550 nm) and blue (450 nm) parts of the spectrum [53,54]. Note, however, that our thermophysiological results on tabanid thermal preference are not influenced by the reflection– polarization patterns of the two optically identical shiny black barrels of different temperatures. The reason for this is the physical fact that the reflection–polarization characteristics of the barrels’ surface are independent of its temperature in the visible spectral range. Therefore, all optical parameters (radiance, degree of linear polarization and angle of polarization) of our warm and black barrels were the same. Only tiny water droplets condensed from air humidity on the outer surface of the cold barrel could have resulted in some differences in reflection–polarization, but such a condensation did not occur in our field experiments. The temperature of the antisolar outer surface of the cold barrel was between 20 and 30°C (figure 1), because the plastic barrel’s wall with 5 mm thickness was a good thermal insulator. Therefore, the outer surface temperature was never so cold that the air humidity could have been condensed on the wall. 2.4. Schlieren imagery In order to demonstrate the warm/cold air flow in the vicinity of the warm/cold barrels used in our field experiments, we performed Schlieren imagery [55]. With this technique, the spatio-temporal variation of the temperature-dependent refractive index of the air can be visualized. The barrels were modelled by a small hot windless windy warm cold w w w w w w w w w c c c c c c c c 1 cm w w w w w w w c c c c c c 1 cm w w w w w w w w w w w c c c c c c c c c c c 1 cm 1 cm 1 cm 1 cm (a) (c) (d) (b) (e) ( f ) Figure 3. Schlieren images of the air next to the hot (a,b; T = 70°C), warm (c,d; T = 45°C) and cold (e,f; T = 8°C) plastic flask modelling the warm/cold barrels used in the choice experiments. In the upper row (a,c,e), the air was windless (calm), while in the lower row (b,d,f ), the air was windy (turbulent), because the flask was very weakly blown by mouth from 1 m. Different shades of grey represent different gradients of the temperature-dependent refractive index of air. w, warmer air; c, colder air. cold w w w w w w w c c c c c c 1 cm (e) hot windless w w w w w w w w w w w c c c c c c c c c c c 1 cm (a) warm w w w w w w w w w c c c c c c c c 1 cm (c) 7 (e) (a) (c) royalsocietypublishing.org/journal/rsos R. Soc. open sci. 6: 191119 windless 1 cm 1 cm ( f ) windy 1 cm (b) 1 cm (d) ( f ) (b) (d) windy 1 cm Figure 3. Schlieren images of the air next to the hot (a,b; T = 70°C), warm (c,d; T = 45°C) and cold (e,f; T = 8°C) plastic flask modelling the warm/cold barrels used in the choice experiments. In the upper row (a,c,e), the air was windless (calm), while in the lower row (b,d,f ), the air was windy (turbulent), because the flask was very weakly blown by mouth from 1 m. 3.1. Reflection–polarization and thermal characteristics of the warm and cold barrels Electronic supplementary material, figure S6 and Results, show that the optical characteristics of the warm and cold barrels used in our field experiments were the same. In figures 1 and 2 and electronic supplementary material, Results, it is clearly seen that in sunshine, the relation among the surface temperatures of the solar and antisolar sides of the warm and cold barrels was: Twarm(solar) > Twarm(antisolar) > Tcold(solar) > Tcold(antisolar). Thus, during the experiments, always four test surfaces (sides of the barrels) with different temperatures and uniform optical characteristics were offered to flying tabanids. 2.5. Statistical analysis We applied Wilcoxon signed-rank test (R Statistics 3.2.3) to find differences in tabanid attractiveness between the warm and cold barrels (in experiments 1–2) and between the warm sunlit, warm shady, cold sunlit and cold shady barrel’s sides (in experiments 3–5). In the dataset, the number N of landings and the time spent on the barrels were calculated with a 5 min temporal resolution. The statistical analysis compared the same time periods of the solar/antisolar sides of the warm/cold barrels. 2.4. Schlieren imagery Different shades of grey represent different gradients of the temperature-dependent refractive index of air. w, warmer air; c, colder air. cylindrical black plastic flask (height = 16 cm, diameter = 7 cm). In its cold state, the flask was filled with tap water and ice cubes, so that the inside water temperature was T = 0°C until the ice melted away. In its warm and hot state, the flask was filled with warm (T = 47°C) and hot water (T = 72°C), respectively. The surface temperature of the flask was measured with thermography. The cold/warm flask was placed into the light beam of the Schlieren equipment (electronic supplementary material, figures S4 and S5). 3.2. Schlieren imagery of the warm/cold air next to the warm/cold barrels According to figure 3 and electronic supplementary material, Results, a warm target heats up the boundary layer (i.e. the air next to the flask’s wall) and this warm air streams upward due to its lower density than photograph photograph thermogram thermogram s h a d y s u n l i t black horse brown horse white horse 40 60 65 temperature (°30) 70 45 50 55 20 25 30 35 15 Figure 4. Photographs and thermograms of sunlit and shady black, brown and white horses. On the thermograms, the closed curve(s) as well as the tilted and nearly horizontal straight lines represent the regions on/along which the surface temperature was averaged. royalsocietypublishing.org/journal/rsos R. Soc. open sci. 6: 191119 8 8 royalsocietypublishing.org/journal/rsos R. Soc. open sci. 6: 191119 8 thermogram photograph brown horse l/rsos R. Soc. open sci. 6: 191119 Figure 4. Photographs and thermograms of sunlit and shady black, brown and white horses. On the thermograms, the closed curve(s) as well as the tilted and nearly horizontal straight lines represent the regions on/along which the surface temperature was averaged. that of the surrounding air of 20°C room temperature. On the other hand, a cold target cools down the boundary layer and this cold air streams downward due to its higher density. In calmness (under windless conditions), the thickness of the upward flowing warm and downward streaming cold boundary layer is around 1–2 cm over the surface. Under turbulent (windy) conditions, the warm and cold air layers mix turbulently, thus an elongated warm/cold turbulent train of several centimetres long forms beyond a warm/cold target along the wind direction. Note that tabanids also fly in weaker winds, when the air flow is turbulent around any target (host animals or our barrels). 3.3. Surface temperatures of sunlit and shady horses of different colours Figure 4 shows the thermograms of sunlit and shady white, brown and black horses, and electronic supplementary material, table S1, contains the surface temperatures of a white, a brown and a black horse measured by thermography on an area and along a nearly horizontal and vertical line under sunlit (illuminated by direct sunlight) and shady (sun behind a cloud) conditions. In sunshine, the warmest (44.4–43.1°C) and coldest (36.4°C) were the sunlit body surfaces of the black and white horses, respectively, while the surface temperature of the brown horse was in between (42°C). On the other hand, the mean temperatures of the shady body surfaces of these horses were very similar (35.4–37.3°C). These data demonstrate that in sunshine, the surface temperature T of animals is mainly determined by the albedo a (light absorptance) of the body surface: the lower is a, the higher is T [56–59]. 3.4. Attractiveness of the warm and cold barrels to tabanids In our field experiment, we observed T. tergestinus Egger 1859 tabanid flies. During our experiment, on the horse farm’s ground, there was a rectangular black plastic tray (30 × 30 cm) filled with common transparent, yellowish salad oil functioning as an efficient tabanid trap [41]. This trap captured only T. tergestinus horseflies. They were identified on the basis of the following morphological characteristics [60, pp. 362–363]: T. tergestinus is a medium-sized (length 15–18 mm) species with green eyes and three red bands, the frons in females are narrow, the median callus is linear and connected with the lower callus, the thorax is dark grey, notopleural lobes are dark, the abdomen is reddish-brown at sides from one to four tergites and with dark median stripe with indistinct paler median triangles, while posterior antisolar antisolar solar solar solar solar side * n.s. * *** * * warm warm barrel experiments 3 + 4 + 5 cold cold 1 + 2 54.3% (400) 45.7% (337) 35.6% (299) 30.2% (254) 21.3% (179) 12.9% (109) percentage (%) of landings 0 20 40 60 80 100 Figure 5. Percentage (%) of the total number N (in parentheses) of landings on the solar and antisolar sides of the warm and cold barrels in experiments 1–2 and 3–5 (electronic supplementary material, table S2) with the results of the Wilcoxon signed-rank test. n.s., not significant, p > 0.05; 0.01 < p < 0.05; 0.001 < p < 0.01; *p < 0.001 (electronic supplementary material, table S3). royalsocietypublishing.org/journal/rsos R. Soc. open sci. 9 9 3 + 4 + 5 Figure 5. Percentage (%) of the total number N (in parentheses) of landings on the solar and antisolar sides of the warm and cold barrels in experiments 1–2 and 3–5 (electronic supplementary material, table S2) with the results of the Wilcoxon signed-rank test. n.s., not significant, p > 0.05; 0.01 < p < 0.05; 0.001 < p < 0.01; p < 0.001 (electronic supplementary material, table S3). 22 ± 46 61 ± 78 65 ± 83 51 ± 115 22 ± 48 4 ± 9 antisolar antisolar solar solar solar solar side * * * * * ** warm warm barrel experiments 3 + 4 + 5 cold cold 0 100 200 220 1 + 2 average time ·tÒ ± s.d. (s) spent by tabanids on the barrel Figure 6. Average time 〈t〉± s.d. 4. Discussion The maximum coat temperature of sunlit dark (brown or black) horses (even if they sweat, when the evaporation of sweat decreases the surface temperature) can easily be 50.9–51.5°C (electronic supplementary material, table S1). This is because sunlit surface parts of the dark coat can locally warm up very strongly if the surface is nearly perpendicular to the incident sunlight. Such surface temperatures of sunlit dark horses are fully realistic, because the interior body temperature of adult horses is 37.2–38.3°C (https://thehorse.com/133159/normal-horse-vitals-signs-and-health-indicators/), and the thermal conductivity of horse hair is low. Depending on the orientation of the local body surface relative to direct sunlight, the albedo (low for black stripes, high for white stripes) of the striped coat and the air temperature, the surface temperature of sunlit black zebra stripes, for example, can change between 35 and 55°C [61] or 44 and 56°C [62]. In the top-left (sunlit black horse) and bottom-right (shady white horse) thermograms of figure 4, the temperature of ground with heterogeneous albedo and thermophysical characteristics was approximately 60–65°C. These hot patches correspond to dry soil regions or dung piles, which warmed up very much in direct sunshine. One of the plausible explanations of the heat preference of tabanids may be based on the fact that an elevated skin temperature of potential hosts is related to either internal (heavy muscular work) or external (strong insolation) heat stresses of the host, the usual thermoregulatory response of which is an increased skin blood flow (and sweating). 3.4. Attractiveness of the warm and cold barrels to tabanids As we have seen above, the mean surface temperature of the barrel’s sides decreased also in the same order: Twarm(solar) > Twarm(antisolar) > Tcold(solar) > Tcold(antisolar) (figures 1 and 2). 254 > Ncold(solar) = 179 > Ncold(antisolar) = 109 (figure 5). These differences are also significant (electronic supplementary material, table S3A), except the Nwarm(solar) versus Nwarm(antisolar) where p = 0.0917. As we have seen above, the mean surface temperature of the barrel’s sides decreased also in the same order: Twarm(solar) > Twarm(antisolar) > Tcold(solar) > Tcold(antisolar) (figures 1 and 2). 10 Considering the average time 〈t〉spent by the alighted individual tabanids on the different barrel’s surface, the following trends were obtained (figure 6; electronic supplementary material, tables S2 and S3): (i) in experiments 1–2, the tabanids spent highly significantly ( p < 0.0001) more time on the solar side of the warm barrel (〈t〉= 61 ± 78 s) than on the solar side of the cold barrel (22 ± 46 s). (ii) In experiments 3–5 again, tabanids spent highly significantly (p < 0.0001) more time on the warm barrel (〈t〉= 58 ± 99 s) than on the cold one (15 ± 39 s). (iii) The average time 〈t〉spent by tabanids on the different barrel’s sides decreased in the following order: 〈t〉warm(solar) = 65 ± 83 s > 〈t〉warm(antisolar) = 51 ± 115 s > 〈t〉cold(solar) = 22 ± 48 s > 〈t〉cold(antisolar) = 4 ± 9 s. These differences are highly significant (electronic supplementary material, table S3B). Figure 7a shows the total numbers N of landings on the barrels as a function of the barrel’s surface temperature T. If T was below 18°C (T < 18°C), tabanids did not land on the barrels. Although the dispersion (standard deviation) of the N(T ) data points is relatively large, the regression line and its 95% confidence area clearly show the tendency that N increases with increasing T. Figure 7b represents the average time 〈t〉spent by tabanids on the barrels versus T averaged for the periods (30 or 60 min) between two sequential thermographic measurements of T. 〈t〉increased tendentiously with increasing T, and its maximum was 136 s. R. Soc. open sci. 6: 191119 3.4. Attractiveness of the warm and cold barrels to tabanids (s) per individual spent on the solar and antisolar sides of the warm and cold barrels in experiments 1–2 and 3–5 (electronic supplementary material, table S2) with the results of the Wilcoxon signed-rank test. n.s., not significant, p > 0.05; 0.01 < p < 0.05; 0.001 < p < 0.01; *p < 0.001 (electronic supplementary material, table S3). 3 + 4 + 5 Figure 6. Average time 〈t〉± s.d. (s) per individual spent on the solar and antisolar sides of the warm and cold barrels in experiments 1–2 and 3–5 (electronic supplementary material, table S2) with the results of the Wilcoxon signed-rank test. n.s., not significant, p > 0.05; 0.01 < p < 0.05; 0.001 < p < 0.01; *p < 0.001 (electronic supplementary material, table S3). tergites are uniformly greyish. Hence, during our experiment, only this tabanid species swarmed in the study area. With the naked eye, we could not detect the sex (male or female) of tabanids landed on the barrels, but they might have been females in all probability, because black targets above the ground level are attacked only by female tabanids [4]. Figures 5 and 6 as well as electronic supplementary material, tables S2 and S3, show the total numbers N of landings on the solar and antisolar sides of the warm and cold barrels used in our field experiments and the average time 〈t〉± s.d. (second) per individual spent on the different barrel’s surface together with the results of the statistical analyses of these data. (i) In experiments 1 and 2, only the solar side of the barrels was observed. On the warm barrel (N = 400), significantly ( p = 0.0139) more tabanids landed than on the cold barrel (N = 337). (ii) In experiments 3–5, both the solar and antisolar sides of the barrels were monitored. Then again, the warm barrel (N = 553) was highly significantly ( p < 0.0001) more attractive to tabanids than the cold barrel (N = 288). The attractiveness of the different barrel’s sides decreased in the following order: Nwarm(solar) = 299 > Nwarm(antisolar) = 254 > Ncold(solar) = 179 > Ncold(antisolar) = 109 (figure 5). These differences are also significant (electronic supplementary material, table S3A), except the Nwarm(solar) versus Nwarm(antisolar) where p = 0.0917. 4.1. A new hypothesis We propose here the novel hypothesis that the warmer microclimate above the body surface of sunlit dark hosts is a selective advantage for blood-sucking female tabanids: tabanid bites are very painful, which therefore elicit intense defensive behaviour of the host, such as tail flicking, skin twitching, foot stomping and head shaking, for example [63]. Due to these defensive movements, a blood-sucking tabanid can easily get hurt. To avoid such a damage, the biting tabanid should fly away promptly before the host’s tail hits it. For such a rapid escape, the tabanid wing muscles should function quickly. Only high enough temperature of these muscles can ensure their quick enough functioning. Thus, we suggest the following new hypothesis: — Blood-seeking tabanids choose preferentially sunlit darker hosts, partly because their wing muscles and neurons function rapidly enough in the warmer microclimate on the body surface of such hosts in order to escape by flying away from the parasite-repelling reactions of the hosts. (Electrophysiological recordings from flight muscle and associated nerves showed that in certain insects, the number of nerve impulses impinging on a wing muscle is much fewer than that of — Blood-seeking tabanids choose preferentially sunlit darker hosts, partly because their wing muscles and neurons function rapidly enough in the warmer microclimate on the body surface of such hosts in order to escape by flying away from the parasite-repelling reactions of the hosts. (Electrophysiological recordings from flight muscle and associated nerves showed that in certain insects, the number of nerve impulses impinging on a wing muscle is much fewer than that of muscle contractions. In this case, high wingbeat frequencies could not be explained on the basis of a one-to-one relationship between nerve impulse and muscle contraction. On the other hand, in other 0 5 10 15 20 25 30 35 (a) (b) number N of landings surface temperature T (°C) average time ·tÒ (s) spent by tabanids on the test surface 15 20 25 30 35 40 45 50 55 60 15 20 25 30 35 40 45 50 55 60 0 20 40 60 80 100 120 140 ure 7. (a) Total number N of landings on the barrels as a function of the barrel’s surface temperature T (°C). (b) Average time = t/N (s) spent by tabanids on the barrels versus T averaged for the periods (30 or 60 min) between two sequential mographic measurements of T. 4.1. A new hypothesis The straight lines show the regression line fitted to the data points represented by symbol The grey area around the regression line shows 95% confidence interval. All temperature data originate from the mographically measured, temporally changing surface temperatures of the solar and antisolar sides of the warm and cold els shown in figure 1. royalsocietypublishing.org/journal/rsos R. Soc. open sci. 6: 191119 11 0 5 10 15 20 25 30 35 (a) number N of landings 15 20 25 30 35 40 45 50 55 60 royalsocietypublishing.org/journal/rsos R. Soc. open sci. 6: 191119 11 11 number N of landings number N of landings (b) surface temperature T (°C) average time ·tÒ (s) spent by tabanids on the test surface 15 20 25 30 35 40 45 50 55 60 0 20 40 60 80 100 120 140 surface temperature T (°C) Figure 7. (a) Total number N of landings on the barrels as a function of the barrel’s surface temperature T (°C). (b) Average time 〈t〉= t/N (s) spent by tabanids on the barrels versus T averaged for the periods (30 or 60 min) between two sequential thermographic measurements of T. The straight lines show the regression line fitted to the data points represented by symbol +. The grey area around the regression line shows 95% confidence interval. All temperature data originate from the thermographically measured, temporally changing surface temperatures of the solar and antisolar sides of the warm and cold barrels shown in figure 1. muscle contractions. In this case, high wingbeat frequencies could not be explained on the basis of a one-to-one relationship between nerve impulse and muscle contraction. On the other hand, in other insects, there is a one-to-one relationship between nerve impulses and muscle contractions [64, p. 218]. In the colder microclimate on shady darker host animals or sunlit/shady brighter hosts, the wing muscles of tabanids may not function rapidly enough for a successful escape. The prerequisite of this hypothesis is to prove that tabanids indeed prefer warmer hosts against colder ones. In this work, we tested this prerequisite by studying the thermal preference of host-seeking female tabanids. A possible explanation of the heat preference of tabanids might be related to an association between elevated skin temperature and increased skin blood flow of a potential host animal. royalsocietypublishing.org/journal/rsos R. Soc. open sci. 6: 191119 This seeking lasts a minimum of a few seconds, but it is generally much longer. During this time, the temperature of her wing muscles can reach that of the host’s boundary layer. According to our hypothesis, if this temperature is high enough (sunlit dark host), then after the start of blood-sucking with a painful bite, the tabanid is able to avoid the host’s parasite-repelling reactions by a rapid flying away, which reactions are very dangerous for a blood-sucking fly. If the temperature of the boundary layer is not high enough, then the wing muscles cooled down in it are not able to function quickly enough; therefore, after the start of blood-sucking, the tabanid is not able to avoid the host’s parasite- repelling reactions by flying away. It is an established fact that muscles of ectothermic animals function less well and contract less rapidly and less forcefully under cold conditions, and the temperature differences measured by us in this work between horses in the sun versus shade, or black versus white horses, are physiologically clearly relevant. Our idea to link the activity and readiness of the flight musculature with ambient temperature is therefore justified. It has been shown for numerous ectothermic animals from earthworms [66] to Antarctic fish [67] that neuronal function is strongly dependent on environmental temperature, so that conduction velocities are linearly related to temperature. Not only are conduction velocities faster at higher temperatures, synaptic transmission (together with conduction crucial for initiating a muscular contraction) would also be faster at higher temperature. Since poikilothermic (cold-blooded) animals can react more swiftly to the approach of a threat, be this a swishing tail or another animal, under a warm condition, tabanids do not need to depart too quickly, but can rely on their rapid reaction to take off faster (when necessary) than if they were colder. This extra advantage of being able to stay a little longer on the host when warm allows the tabanid to ingest a little more blood. And this could be one additional (if not the real reason) for tabanids to seek out warmer hosts. Feeding on a warmer host may have the advantage that tabanids could feed faster (e.g. because of the faster functioning of the musculature of their mouthparts [32]) with which they could minimize the time spent on the host, and thus minimize the hosts’ anti-parasitic behaviours [4]. royalsocietypublishing.org/journal/rsos R. Soc. open sci. 6: 191119 In our opinion, both the warm/cold boundary layer of a few centimetres and the warm/cold turbulent air train of a few decimetres next to the target’s wall (figure 3) are important to explain the behaviour of tabanids flying around the barrels in our field experiments: within a few to several centimetres, flying tabanids may sense the temperature of the boundary layer (microclimate) and thus could estimate the temperature of the barrel’s surface and decide whether they do or do not alight on it. On the basis of the possibly promptly sensed temperature of the boundary layer, a female tabanid could decide whether the host’s body surface is or is not warm enough for alighting. When she lands on the host, her non-contracting wing muscles begin to cool, the speed of which depends on the temperature of the boundary layer over hosts as follows: — In a warm boundary layer (sunlit dark host), the cooling process of wing muscles is slow, or it is even negligible, thus a blood-sucking female tabanid can fly away quickly when the host wants to swat her (by tail, for example). According to our new hypothesis, this can be one of the reasons why tabanids prefer sunlit dark host animals. — In a warm boundary layer (sunlit dark host), the cooling process of wing muscles is slow, or it is even negligible, thus a blood-sucking female tabanid can fly away quickly when the host wants to swat her (by tail, for example). According to our new hypothesis, this can be one of the reasons why tabanids prefer sunlit dark host animals. — In a cooler boundary layer (shady dark or sunlit bright host), the cooling process of wing muscles is rapid, thus a blood-sucking tabanid cannot fly away quickly when the host wants to swat her. In our opinion, this can be one of the reasons why tabanids dislike bright host animals and shady dark hosts. — In a cooler boundary layer (shady dark or sunlit bright host), the cooling process of wing muscles is rapid, thus a blood-sucking tabanid cannot fly away quickly when the host wants to swat her. In our opinion, this can be one of the reasons why tabanids dislike bright host animals and shady dark hosts. When a female tabanid is seeking the veins directly below the skin of the host, she does not immediately bite the host. 4.1. A new hypothesis However, it is also possible that the clearly observed warmth preference is an indication of the thermal strategy of tabanids [65], where they prefer environments that keep an optimal body temperature. When a female tabanid lands on a host animal, her wing muscles are still sufficiently warm for flying, because prior to alighting, the muscles were kept warm by the heat produced by muscle contractions. According to our hypothesis, during flight, she may sense with her thermoreceptors placed on her body surface that the boundary air layer of the picked host is warm enough. For this rapid thermoreception, the short time period could be sufficient during which she overflies through the host’s boundary layer. 1 12 royalsocietypublishing.org/journal/rsos R. Soc. open sci. 6: 191119 The prerequisite of the above hypothesis is that tabanids prefer warmer hosts against colder ones. The results of our field experiments show that this prerequisite is fulfilled in the case of T. tergestinus which horsefly species evidently preferred the warmer black barrel against the colder one, and both targets possessed the same optical characteristics. Since a remotely thermosensitive organ detecting infrared radiation by tabanids has not been discovered, the only explanation of our finding is that prior to alighting, the tabanids flying close around the barrels sense the colder air (boundary) layer next to the vertical wall of the colder barrel. This boundary layer was not warm enough, thus tabanids landed only rarely on the colder barrel, and if they looked for blood, they remained only for a very short periods on the cold substratum. Although we could not determine the sex of tabanids (T. tergestinus) in our field experiments, most of them were females in all probability, because such targets attract only females. On the one hand, it is well known that the shiny black sphere of the Manitoba trap attracts exclusively female tabanids; thus, this and similar traps capture only females [4,8,9,10,19,21,22,23,27,38,68]. On the other hand, in our earlier field experiments, the shiny black sticky or dry (non-sticky) vertical traps caught only female tabanids [12,15,16,33,35,40,48,49]. Therefore, our results are valid for female T. tergestinus horseflies. Finally, our hypothesis is also supported by the fact that there is an optimal interval of the air temperature for the flight (muscles) of tabanids. This has been found by Herczeg et al. [49], for example: the six day-active tabanid species (Atylotus loewianus Villeneuve 1920, T. tergestinus Egger 1859, Tabanus bovinus Linnaeus 1758, Tabanus maculicornis Zetterstedt 1842, Tabanus bromius Linnaeus 1758, Haematopota pluvialis Linnaeus 1758) studied by Herczeg et al. [49] need an ambient air temperature of at least 18°C to fly. This temperature threshold corresponds to the minimum temperature required to activate the enzyme function and flight muscle fibres of tabanids [69]. Herczeg et al. [49] found that 31–35°C is the optimal temperature range for the flight of the mentioned tabanids. This result confirms the similar findings of Amano [70]. Kohane & Watt [71] and Josephson et al. [72] have observed that the optimal frequency and power of isolated flight muscles increase as the temperature rises, peaking at about 39°C. royalsocietypublishing.org/journal/rsos R. Soc. open sci. 6: 191119 As far as we know, quantitative measurements of the rate of heat transfer from the wing muscles of flies to the ambient air have not been performed. Thus, there is no information about the relevant time periods during which the wing muscles of tabanids can warm up from the local environmental (air or substratum) temperature to the minimum temperature necessary for efficient flight. Only general knowledge is available in this subject: tabanids (as other flies) are ectothermic, that is, their body temperature changes with the environment [32]. As flying insects in general, tabanids have to maintain their flight muscles above a minimum temperature to gain power enough to fly. Since smaller tabanids (as animals in general) have larger surface per volume ratio, smaller tabanids cool down more quickly than larger ones. Flying tabanids (as other insects) raise their body temperature through the action of flight, above environmental temperatures [73]. Vibrating or shivering the wing muscles and basking, for example, allow tabanids (as other insects) to actively increase the temperature of their flight muscles, enabling flight [73]. From this point of view, warmth preference is a trivial thermal strategy to minimize active energy consumption for maintaining an optimal body temperature. royalsocietypublishing.org/journal/rsos R. Soc. open sci. 6: 191119 However, this has the potential cost of thermal stress [65]. Another advantage could be that feeding faster minimizes the decrease in the temperature of the insect’s body, including wing muscles [64]. Note that both advantages are also related to the faster functioning of tabanid muscles. 4.2. Fulfilment of the prerequisite of the new hypothesis 13 4.3. A previous study on the thermal preference of tabanids Using two shiny black and two shiny white metal barrels filled with warm or cold water as decoys in a pair of Manitoba canopy traps, Bracken & Thorsteinson [28] studied the attractiveness to tabanids of thermally different targets. Their canopy traps with a warmer (31.1°C) and a colder (11°C) black barrel captured 308 ± 60.5 and 227 ± 39 tabanids (species not published), while those with a warmer (34.6°C) and a colder (22°C) white barrel caught 17 ± 8 and 9 ± 3 tabanids, respectively. They concluded that decoys with surface temperature approximating to body temperature of mammals are not significantly more attractive to tabanids than decoys at air temperature. Our results and conclusions (see later) obtained for T. tergestinus partly contradict those of Bracken & Thorsteinson [28]. The reason for this is that we used a different method. Bracken & Thorsteinson [28] captured tabanids with Manitoba canopy traps having black or white decoys with different temperatures. This trap type cannot capture all tabanids landed on the decoy [68]. While Bracken & Thorsteinson [28] determined the attractiveness to tabanids only for two surface temperatures (31.1°C and 11°C) of their black decoys, measuring at 98 different temperatures, we determined the thermal preference of tabanids as a function of the target’s surface temperature. 5. Conclusion This finding corroborates our hypothesis that biting tabanids should fly away promptly before the host’s tail hits them and that for such rapid escapes, the tabanid wing muscles should function optimally, i.e. with as little delay as possible. We suggest that biting tabanids seek out warmer hosts, because only sufficiently high temperatures can ensure the rapid and good functioning of their flight muscles and neurons. Data accessibility. Our paper has the following electronic supporting materials: Supplementary Materials and methods, Calibration of the thermocamera, Supplementary results, Reflection–polarization characteristics of the warm and cold barrels, Thermal characteristics of the warm and cold barrels, Schlieren imagery of the warm/cold air next to the warm/cold barrels, Attractiveness of the warm and cold barrels to tabanids; electronic supplementary material, tables S1–S3 and figures S1–S7. Authors’ contributions. G.H. and Á.P.: substantial contributions to conception and design; G.H., Á.P., S.P., T.T. and I.M.J.: performing experiments and data acquisition; G.H., Á.P., S.P. and T.T.: data analysis and interpretation; G.H., Á.P. and I.M.J.: drafting the article or revising it critically for important intellectual content. mpeting interests. The authors have no competing interests. Competing interests. The authors have no competing interests. Funding. This work was supported by the grant no. NKFIH K-123930 (Experimental Study of the Functions of Zebra Stripes: A New Thermophysiological Explanation) received by G.H. from the Hungarian National Research, Development and Innovation Office. Acknowledgements. We are grateful to István Simon, who permitted the performance of our field experiment and thermographic measurements in his Szálender horse farm in Göd, Hungary. We thank the logistic and technical help of Dr András Barta and Péter Takács (Estrato Research and Development Ltd). We thank Prof. Tim Caro (Department of Wildlife, Fish and Conservation Biology, University of California, Davis, CA, USA) for reading and commenting an earlier version of our manuscript. We are grateful to two anonymous reviewers for their constructive and valuable comments. 5. Conclusion 14 From our results, we conclude the following: (1) The warmer the shiny black barrel’s surface, the more T. tergestinus tabanid flies alighted on it (figures 5 and 7a). In spite of the fact that the optical characteristics of the warm and cold barrels were the same, significantly more tabanids landed on the warm barrel than on the cold barrel. Hence, these tabanids prefer warmer dark targets with uniform optical characteristics against colder ones. (2) From conclusion (1), it follows that tabanids decided to land on the barrels not only by means of optical cues, but also thermal cues governed their choice. (2) From conclusion (1), it follows that tabanids decided to land on the barrels not only by means of optical cues, but also thermal cues governed their choice. (3) We have frequently observed that tabanids flew so close to the barrel’s surface that their wings almost touched it, and often, we also heard a typical knock sound produced by their body when they hit the barrel during flight. From this observation and on the basis of the Schlieren measurements and (1) and (2), we conclude that tabanids flying in the immediate vicinity of the barrels can sense the temperature of the air (boundary) layer next (a few centimetres in calmness and a few decimetres in wind) to the barrel’s surface, and the warmer is it, the higher is the chance of their landing. g (4) The warmer was the barrel’s surface, tabanids spent the longer time on it (figures 6 and 7b). (4) The warmer was the barrel’s surface, tabanids spent the longer time on it (figures 6 and 7 All these support our main conclusion that T. tergestinus tabanid horseflies prefer warmer alighting surfaces (host animals) against colder ones with the same optical properties (remote visual attractiveness). This finding corroborates our hypothesis that biting tabanids should fly away promptly before the host’s tail hits them and that for such rapid escapes, the tabanid wing muscles should function optimally, i.e. with as little delay as possible. We suggest that biting tabanids seek out warmer hosts, because only sufficiently high temperatures can ensure the rapid and good functioning of their flight muscles and neurons. All these support our main conclusion that T. tergestinus tabanid horseflies prefer warmer alighting surfaces (host animals) against colder ones with the same optical properties (remote visual attractiveness). by the use of Fenvalerate ear tags during the dry period. J. Dairy Res. 54, 165–171. (doi:10.1017/S0022029900 025309) Lehane MJ. 2005 The biology of blood-sucking in insects, 2nd edn. Cambridge, UK: Cambridge University Press, Bracken GK, Hanec W, Thorsteinson AJ. 1962 The orientation behaviour of horse flies and deer flies (Tabanidae: Diptera). II. The role of some visual factors in the attractiveness of decoy royalsocietypublishing.org/journal/rsos R. Soc. open sci. 6: 191119 35. 35. Blahó M, Egri Á, Száz D, Kriska G, Åkesson S, Horváth G. 2013 Stripes disrupt odour attractiveness to biting horseflies: battle between ammonia, CO2, and colour pattern for dominance in the sensory systems of host- seeking tabanids. Physiol. Behav. 119, 168–174. (doi:10.1016/j.physbeh.2013.06.013) 21. 21. Hall MJR, Farkas R, Chainey JE. 1998 Use of odour-baited sticky boards to trap tabanid flies and investigate repellents. Med. Vet. Entomol. 12, 241–245. (doi:10.1046/j.1365-2915.1998. 00107.x) 10. Allan SA, Stoffolano JG. 1986 The effects of hue and intensity on visual attraction of adult Tabanus nigrovittatus (Diptera: Tabanidae). J. Med. Entomol. 23, 83–91. (doi:10.1093/ jmedent/23.1.83) 22. Krcmar S. 2005 Response of horse flies (Diptera, Tabanidae) to different olfactory attractants. Biol. Brat. 60, 611–613. 36. Tashiro H, Schwardt HH. 1953 Biological studies of horse flies in New York. J. Econ. Entomol. 46, 813–822. (doi:10.1093/jee/46.5.813) 23. Krcmar S. 2013 Comparison of the efficiency of the olfactory and visual traps in the collection of horseflies (Diptera: Tabanidae). Entomol. Gen. 34, 261–267. (doi:10.1127/entom.gen/34/2013/ 261) 11. Sasaki H. 2001 Comparison of capturing tabanid flies (Diptera: Tabanidae) by five different color traps in the fields. Appl. Entomol. Zool. 36, 515–519. (doi:10.1303/aez.2001.515) 37. Catts EP. 1970 A canopy trap for collecting Tabanidae. Mosq. News 30, 472–474. 38. Wall WJ, Doane OW. 1980 Large scale use of box traps to study and control saltmarsh greenhead flies (Diptera: Tabanidae) on Cape Cod, Massachusetts. Environ. Entomol. 9, 371–375. (doi:10.1093/ee/9.4.371) 12. Horváth G, Majer J, Horváth L, Szivák I, Kriska G. 2008 Ventral polarization vision in tabanids: horseflies and deerflies (Diptera: Tabanidae) are attracted to horizontally polarized light. Naturwissenschaften 95, 1093–1100. (doi:10. 1007/s00114-008-0425-5) 24. Mihok S, Mulye H. 2010 Responses of tabanids to Nzi traps baited with octenol, cow urine and phenols in Canada. Med. Vet. Entomol. 24, 266–272. (doi:10.1111/j.1365-2915.2010.00889. x) 39. Blahó M, Egri Á, Barta A, Antoni G, Kriska G, Horváth G. 2012 How can horseflies be captured by solar panels? A new concept of tabanid traps using light polarization and electricity produced by photovoltaics. Vet. Parasitol. 189, 353–365. (doi:10.1016/j.vetpar.2012.04.016) 25. Mihok S, Lange K. 2012 Synergism between ammonia and phenols for Hybomitra tabanids in northern and temperate Canada. Med. Vet. Entomol. 26, 282–290. (doi:10.1111/j.1365- 2915.2011.00999.x) 13. Horváth G, Szörényi T, Pereszlényi Á, Gerics B, Hegedüs R, Barta A, Åkesson S. 2017 Why do horseflies need polarization vision for host detection? royalsocietypublishing.org/journal/rsos R. Soc. open sci. 6: 191119 PLoS ONE 14, e0210831 (doi:10.1371/ journal.pone.0210831) 29. 29. Mizell III RF, Mizell IV RF, Mizell RA. 2002 Trolling: a novel trapping method for Chrysops spp. (Diptera: Tabanidae). Fla Entomol. 85, 356–366. (doi:10.1653/0015- 4040(2002)085[0356:TANTMF]2.0.CO;2) 17. Blahó M et al. 2014 Unexpected attraction of polarotactic water-leaving insects to matt black car surfaces: mattness of paintwork cannot eliminate the polarized light pollution of black cars. PLoS ONE 9, e103339 (doi:10.1371/journal. pone.0103339) 44. Otártics MZ, Altbäcker V, Solymosi K, Mátics R, Romvári R, Farkas S. 2019 Efficacy of H-traps is affected by exposure to sunshine. Nat. Croat. 28, 257–269. 30. Dethier VG. 1957 Parasitological reviews: the sensory physiology of blood-sucking arthropods. Exp. Parasitol. 6, 68–122. (doi:10.1016/0014- 4894(57)90009-7) 45. Allan SA, Day JF, Edman JD. 1987 Visual ecology of biting flies. Annu. Rev. Entomol. 32, 297–316. (doi:10.1146/annurev.en.32.010187. 001501) 18. Száz D, Mihályi D, Farkas A, Egri Á, Barta A, Kriska G, Robertson B, Horváth G. 2016 Polarized light pollution of matte solar panels: anti-reflective photovoltaics reduce polarized light pollution but benefit only some aquatic insects. J. Insect Conserv. 20, 663–675. (doi:10. 1007/s10841-016-9897-3) 31. McGavin GC. 2001 Essential entomology: an order-by-order introduction. Oxford, UK: Oxford University Press. 46. Thorsteinson AJ. 1958 The orientation behaviour of horse flies and deer flies (Tabanidae: Diptera). I. The attractance of heat to tabanids. Entomol. Exp. Appl. 1, 191–196. (doi:10.1111/j. 1570-7458.1958.tb00023.x) 32. Elzinga RJ. 2003 Fundamentals of entomology, 6th edn. Upper Saddle River, NJ: Prentice Hall. 33. Horváth G, Blahó M, Kriska G, Hegedüs R, Gerics B, Farkas R, Åkesson S. 2010 An unexpected advantage of whiteness in horses: the most horsefly-proof horse has a depolarizing white coat. Proc. R. Soc. B 277, 1643–1650. (doi:10. 1098/rspb.2009.2202) 19. Thorsteinson AJ, Bracken GK, Tostawaryk W. 1966 The orientation behaviour of horse flies and deer flies (Tabanidae: Diptera). V. The influence of the number of reflecting surfaces on attractiveness to tabanids of glossy black polyhedra. Can. J. Zool. 44, 275–279. (doi:10. 1139/z66-026) 47. Lall SB. 1969 Phagostimulants of haematophagous tabanids. Entomol. Exp. Appl. 12, 325–336. (doi:10. 1111/j.1570-7458.1969.tb02528.x) 48. Herczeg T, Blahó M, Száz D, Kriska G, Gyurkovszky M, Farkas R, Horváth G. 2014 Seasonality and daily activity of male and female tabanid flies monitored in a Hungarian hill-country pasture by new polarization traps 48. Herczeg T, Blahó M, Száz D, Kriska G, Gyurkovszky M, Farkas R, Horváth G. References by the use of Fenvalerate ear tags during the dry period. J. Dairy Res. 54, 165–171. (doi:10.1017/S0022029900 025309) . Lehane MJ. 2005 The biology of blood-sucking in insects, 2nd edn. Cambridge, UK: Cambridge University Press, . Bracken GK, Hanec W, Thorsteinson AJ. 1962 The orientation behaviour of horse flies and deer flies (Tabanidae: Diptera). II. The role of some visual factors in the attractiveness of decoy silhouettes. Can. J. Zool. 40, 685–695. (doi:10. 1139/z62-064) 6. Granger CA. 1970 Trap design and color as factors in trapping the salt marsh greenhead fly. J. Econ. Entomol. 63, 1670–1673. (doi:10.1093/ jee/63.5.1670) 7. Roberts RH. 1970 Color of malaise trap and collection of Tabanidae. Mosq. News 29, 236–238. 8. Roberts RH. 1977 Attractancy of two black decoys and CO2 to tabanids (Diptera: Tabanidae). Mosq. News 37, 169–172. by the use of Fenvalerate ear tags during the dry period. J. Dairy Res. 54, 165–171. (doi:10.1017/S0022029900 025309) 4. Lehane MJ. 2005 The biology of blood-sucking in insects, 2nd edn. Cambridge, UK: Cambridge University Press, 5. Bracken GK, Hanec W, Thorsteinson AJ. 1962 The orientation behaviour of horse flies and deer flies (Tabanidae: Diptera). II. The role of some visual factors in the attractiveness of decoy silhouettes. Can. J. Zool. 40, 685–695. (doi:10. 1139/z62-064) 6. Granger CA. 1970 Trap design and color as factors in trapping the salt marsh greenhead fly. J. Econ. Entomol. 63, 1670–1673. (doi:10.1093/ jee/63.5.1670) 7. Roberts RH. 1970 Color of malaise trap and collection of Tabanidae. Mosq. News 29, 236–238. 8. Roberts RH. 1977 Attractancy of two black decoys and CO2 to tabanids (Diptera: Tabanidae). Mosq. News 37, 169–172. 1. Perich MJ, Wright RE, Lusby KS. 1986 Impact of horse flies (Diptera: Tabanidae) on beef cattle. J. Econ. Entomol. 79, 128–131. (doi:10.1093/ jee/79.1.128) 2. Hunter DM, Moorhouse DW. 1976 The effects of Austrosimulium pestilens on the milk production of dairy cattle. Aust. Vet. J. 52, 97–99. (doi:10. 1111/j.1751-0813.1976.tb13868.x) 3. Harris JA, Hillerton JE, Morant SV. 1987 Effect on milk production of controlling muscid flies, and reducing fly-avoidance behaviour ONE 7, e41138 (doi:10.1371/journal.pone. 0041138) traps of selected targets as attractants for tabanid species (Diptera: Tabanidae). J. Am. Mosq. Control Assoc. 9, 182–188. 9. Browne SM, Bennett GF. 1980 Colour and shape as mediators of host-seeking responses of simuliids and tabanids (Diptera) in the Tantramar marshes, New Brunswick, Canada. J. Med. Entomol. 17, 58–62. (doi:10.1093/ jmedent/17.1.58) 15 royalsocietypublishing.org/journal/rsos R. Soc. open sci. 6: 191119 Polarization helps tabanid flies to select sunlit dark host animals from the dark patches of the visual environment. R. Soc. open sci. 4, 170735 (doi:10.1098/rsos.170735) 40. Egri Á, Blahó M, Száz D, Barta A, Kriska G, Antoni G, Horváth G. 2013 A new tabanid trap applying a modified concept of the old flypaper: linearly polarising sticky black surfaces as an effective tool to catch polarotactic horseflies. Int. J. Parasitol. 43, 555–563. (doi:10.1016/j.ijpara.2013.02.002) 40. 26. Baldacchino F, Carier J, Porciani A, Buatois B, Dormont L, Jay-Robert P. 2013 Behavioural and electrophysiological responses of females of two species of tabanid to volatiles in urine of different mammals. Med. Vet. Entomol. 27, 77–85. (doi:10.1111/j.1365-2915.2012.01022.x) 14. Kriska G, Bernáth B, Farkas R, Horváth G. 2009 Degrees of polarization of reflected light eliciting polarotaxis in dragonflies (Odonata), mayflies (Ephemeroptera) and tabanid flies (Tabanidae). J. Insect Physiol. 55, 1167–1173. (doi:10.1016/j.jinsphys.2009.08.013) 27. 27. Baldacchino F, Manon S, Puech L, Buatois B, Dormont L, Jay-Robert P. 2014 Olfactory and behavioural responses of tabanids to octenol, phenols and aged horse urine. Med. Vet. Entomol. 28, 201–209. (doi:10.1111/mve. 12038) 41. Egri Á, Blahó M, Száz D, Kriska G, Majer J, Herczeg T, Gyurkovszky M, Farkas R, Horváth G. 2013 A horizontally polarizing liquid trap enhances the tabanid-capturing efficiency of the classic canopy trap. Bull. Entomol. Res. 103, 665–674. (doi:10.1017/S0007485313000357) 15. Egri Á, Blahó M, Sándor A, Kriska G, Gyurkovszky M, Farkas R, Horváth G. 2012 New kind of polarotaxis governed by degree of polarization: attraction of tabanid flies to differently polarizing host animals and water surfaces. Naturwissenschaften 99, 407–416. (doi:10.1007/s00114-012-0916-2) 28. Bracken GK, Thorsteinson AJ. 1965 The orientation behaviour of horse flies and deer flies (Tabanidae: Diptera). IV. The influence of some physical modifications of visual decoys on orientation of horse flies. Entomol. Exp. Appl. 8, 314–318. (doi:10.1111/j.1570-7458.1965. tb00868.x) 42. Krcmar S, Radolic V, Lajos P, Lukacevic I. 2014 Efficiency of colored modified box traps for sampling tabanids. Parasite 21, 67 (doi:10. 1051/parasite/2014068) 16. Egri Á, Blahó M, Kriska G, Farkas R, Gyurkovszky M, Åkesson S, Horváth G. 2012 Polarotactic tabanids find striped patterns with brightness and/or polarization modulation least attractive: an advantage of zebra stripes. J. Exp. Biol. 215, 736–745. (doi:10.1242/jeb.065540) 43. Caro T, Argueta Y, Briolat ES, Bruggink J, Kasprowsky M, Lake J, Mitchell MJ, Richardson S, How M. 2019 Benefits of zebra stripes: behaviour of tabanid flies around zebras and horses. royalsocietypublishing.org/journal/rsos R. Soc. open sci. 6: 191119 67. 49. Herczeg T, Száz D, Blahó M, Barta A, Gyurkovszky M, Farkas R, Horváth G. 2015 The effect of weather variables on the flight activity of horseflies (Diptera: Tabanidae) in the continental climate of Hungary. Parasitol. Res. 114, 1087–1097. (doi:10. 1007/s00436-014-4280-3) 67. Macdonald JA. 1981 Temperature compensation in the peripheral nervous system: Antarctic versus temperate poikilotherms. J. Comparative Physiol. 142, 411–418. (doi:10.1007/ BF00605453) 58. Benesch AR, Hilsberg S. 2003 Infrarot- thermographische Untersuchungen der Oberflächentemperatur bei Zebras. Zoologischer Garten NF 2, 74–82. 59. Benesch AR, Hilsberg-Merz S. 2006 Oberflächentemperaturen bei Zebrastreifen. Nat. Mus. 136, 49–56. 68. Thomson RCK, Saunders DS. 1986 Relative efficiency of Manitoba traps and adhesive panels for the capture of the common cleg, Haematopota pluvialis (Meigen) (Diptera: Tabanidae). Ann. Trop. Med. Parasitol. 80, 345–349. (doi:10.1080/ 00034983.1986.11812027) 50. Umow N. 1905 Chromatische Depolarisation durch Lichtzerstreuung. Phys. Z. 6, 674–676. 60. Chvala M, Lyneborg L, Moucha J. 1972 The horse flies of Europe (Diptera, Tabanidae). Copenhagen, Denmark: Entomological Society of Copenhagen. for the capture of the common cleg, Haematopota pluvialis (Meigen) (Diptera: Tabanidae). Ann. Trop. Med. Parasitol. 80, 345–349. (doi:10.1080/ 00034983.1986.11812027) 69. Pybus J, Tragear RT. 1975 The relationship of adenosine triphosphatase activity to tension and power output of insect flight muscle. J. Physiol. 247, 71–89. (doi:10.1113/jphysiol.1975. sp010921) 70. Amano K. 1985 Statistical analysis of the influence of meteorological factors on flight activity of female tabanids. Kontyu (Tokyo) 53, 161–172. 51. Vale GA. 1998 Responses of tsetse flies (Diptera: Glossinidae) to vegetation in Zimbabwe: implications for population distribution and bait siting. Bull. Entomol. Res. 88 (Suppl. 1), S7–S59. (doi:10.1017/s0007485300041961) 69. 9. Pybus J, Tragear RT. 1975 The relationship of adenosine triphosphatase activity to tension and power output of insect flight muscle. J. Physiol. 247, 71–89. (doi:10.1113/jphysiol.1975. sp010921) 61. Horváth G, Pereszlényi Á, Száz D, Barta A, Jánosi IM, Gerics B, Åkesson S. 2018 Experimental evidence that stripes do not cool zebras. Sci. Rep. 8, 9351 (doi:10.1038/s41598-018- 27637-1) 52. Horváth G (ed.) 2014 Polarized light and polarization vision in animal sciences. Berlin, Germany: Springer. 62. Cobb A, Cobb S. 2019 Do zebra stripes influence thermoregulation? J. Nat. Hist. 53, 863–879. (doi:10.1080/00222933.2019.1607600) 53. Horváth G, Varjú D. 1997 Polarization pattern of freshwater habitats recorded by video polarimetry in red, green and blue spectral ranges and its relevance for water detection by aquatic insects. J. Exp. Biol. 200, 1155–1163. 70. Amano K. royalsocietypublishing.org/journal/rsos R. Soc. open sci. 6: 191119 2014 Seasonality and daily activity of male and female tabanid flies monitored in a Hungarian hill-country pasture by new polarization traps 34. Blahó M, Egri Á, Báhidszki L, Kriska G, Hegedüs R, Åkesson S, Horváth G. 2012 Spottier targets are less attractive to tabanid flies: on the tabanid-repellency of spotty fur patterns. PLoS 20. Schreck CE, Kline DL, Williams DC, Tidwell MA. 1993 Field evaluations in Malaise and canopy giant nerve fibres of Lumbricus terrestris L. J. Exp. Biol. 47, 471–480. and traditional canopy traps. Parasitol. Res. 113, 4251–4260. (doi:10.1007/s00436-014-4103-6) and traditional canopy traps. Parasitol. Res. 113, 4251–4260. (doi:10.1007/s00436-014-4103-6) animals. J. Mammal. 54, 1003–1007. (doi:10. 2307/1379105) 16 royalsocietypublishing.org/journal/rsos R. Soc. open sci. 6: 191119 1985 Statistical analysis of the influence of meteorological factors on flight activity of female tabanids. Kontyu (Tokyo) 53, 161–172. 63. Muzari MO, Skerratt LF, Jones RE, Duran TL. 2010 Alighting and feeding behaviour of tabanid flies on horses, kangaroos and pigs. Vet. Parasitol. 170, 104–111. (doi:10.1016/j.vetpar. 2010.01.028) 71. Kohane MJ, Watt WB. 1999 Flight-muscle adenylate pool responses to flight demands and thermal constraints in individual Colias eurytheme (Lepidoptera, Pieridae). J. Exp. Biol. 202, 3145–3154. 54. Horváth G, Varjú D. 2004 Polarized light in animal vision—polarization patterns in nature. Berlin, Germany: Springer. 64. Romoser WS, Stoffolano Jr. JG. 1998 The science of entomology. New York, NY: WCB McGraw-Hill Companies, Inc. 55. Settles GS. 2012 Schlieren and shadowgraph techniques: visualizing phenomena in transparent media. Berlin, Germany: Springer. 72. Josephson RK, Malamud JG, Stokes DR. 2000 Power output by an asynchronous flight muscle from a beetle. J. Exp. Biol. 203, 2667–2689. 65. Moyes CD, Schulte PM. 2014 Principles of animal physiology, 2nd edn. Harlow, Essex, UK: Pearson Education Ltd. 56. Cena K. 1966 Investigations of absorption of solar radiation by cattle and horses with various coat colours. Acta Agraria et Silvestria 6, 93–138. 73. Triplehorn CA, Johnson NF. 2005 Borror and DeLong’s introduction to the study of insects, 7th edn. Australia: Thomson, Brooks/Cole. 66. Lagerspetz KYH, Talo A. 1967 Temperature acclimation of the functional parameters of the 57. Cena K, Clark JA. 1973 Thermographic measurements of the surface temperatures of
https://openalex.org/W157389418	https://cfjournal.hse.ru/article/download/1536/2253	English	null	Role of Corporate Governance in Banking Sector: Evidence from All Over the World	Korporativnye finansy	2,013	cc-by	4,504	20Corporate Finance Department, Corporate Finance Center, National Research University Higher School of Economics, Moscow, RUSSIA: Assistant Professor, Researcher, anastasianstepanova@gmail.com 21Corporate Finance Department, Corporate Finance Center, National Research University Higher School of Economics, Moscow, RUSSIA, Junior Researcher, olgamivantsova@gmail.com ЖУРНАЛ "КОРПОРАТИВНЫЕ ФИНАНСЫ" ЖУРНАЛ "КОРПОРАТИВНЫЕ ФИНАНСЫ" №4(24) 2012 80 Keywords: corporate governance, bank performance, commercial banks, independent directors, ownership structure, emerging markets Keywords: corporate governance, bank performance, commercial banks, independent directors, ownership structure, emerging markets Anastasia Stepanova20, Olga Ivantsova21 This paper is focused on the recent research in the area of performance effect of corporate governance in banking sector. We review the results of studies devoted to two key nonfinancial characteristics of a commercial bank influencing its performance. In the first part of the paper we analyze the evidence on board of directors structure impact on bank performance. We focus on the performance effect of board size, independent directors and gender diversity of the board. In the second part we discuss the issue of bank ownership structure. In this paper we pay special attention to the difference between performance drivers in developed and emerging markets as well as to the performance drivers changes in times of financial crisis. Role of Corporate Governance in Banking Sector: Evidence from All Over the World Anastasia Stepanova20, Olga Ivantsova21 JEL: G32, G34. JEL: G32, G34. © Электронный журнал Корпоративные Финансы, 2012 Выпуск #4(24), 2012 © Электронный журнал Корпоративные Финансы, 2012 2.1 Board Size and its impact on Bank Performance From a theoretical point of view, larger boards of directors gather more human capital, knowledge and experience and it helps the board provide the management with better monitoring and advising. On the other hand, if there are too many members on the board it creates additional problems with coordination and communications between directors in comparison to smaller boards. Within larger boards more compromises should be reached to make a decision, which makes this process less flexible and more time consuming. In general, the effect of board size on bank value is a trade-off between advantages (human capital) and disadvantages (coordination problems). p ) Therefore, such a trade-off should create a nonlinear relation between board size and bank performance, however very little proof of this relation can be found in the existing literature. Andres and Vallelado report a reverse nonlinear relationship between number of directors in the board and bank performance measured as return on assets and Tobin’s Q for Western European countries and USA (Andres, Vallelado, 2008). This type of a relationship implies that a board size optimal for performance can be determined. Grove et al. find a concave relationship between financial performance and board size for US banks (Grove et al., 2011). However, the majority of authors find this relation between board size and performance to be negative showing that the disadvantages of large boards are generally stronger than the advantages. Sakawa and Watanabel, analyzing the Japanese banking sector, find that banks with larger boards “underperform their peers in terms of Tobin’s Q” (Sakawa, Watanabel, 2011). Staikouras et al. show negative correlation between board size and performance as ROA, ROE and Tobin’s Q for European banks (Staikouras et al., 2007). For non-developed countries the same negative relationship is reported by Adusei for banks’ return on equity in Ghana and by Pathan et al. for return on assets and equity and Sharp ratio in the Thai banking sector (Adusei, 2011; Pathan et al., 2007). Nevertheless, some authors show the positive relation arguing that the banking sector differs from the other sectors very much and additional knowledge and experience provided by larger boards contributes to better bank performance. For the US banks Adams and Mehran identify the significant positive correlation between board size and Tobin’s Q during the period from 1959 till 1995 and Aebi et al. 2.1 Board Size and its impact on Bank Performance demonstrate the same for bank returns during a very special period – the financial crisis of 2007-2008 (Adams, Mehran, 2008; Aebi et al., 2012). Belkhir for the sample of 174 bank and savings and loans holding companies reports that there is a positive relationship between the size of the board of directors and performance measured by Tobin’s Q and return on assets (Belkhir, 2009). For non-developed countries only Kyereboah-Coleman and Biekpe show a positive relation between ROA and board size for banks in Ghana (Kyereboah-Coleman, Biekpe, 2006). ) However, such authors like Adams and Mehran and Grove et al. for US banks, Bino and Tomar for Jordanian public banks, Love and Rachinsky for Russia and Ukraine find that board size has no effect on bank's return on assets (Adams, Mehran, 2008; Grove et al., 2009; Bino, Tomar, 2012; Love, Rachinsky, 2008). Tandelilin et al. demonstrate that the relationship between board size and bank performance in insignificant for Indonesia and Zulkafli and Samad show it for Malaysia, Thailand, Philippines, Indonesia, Korea, Hong Kong, Taiwan and India (Tandelilin et al., 2007; Zulkafli, Samad, 2007). 1. Introduction The financial crisis of 2007-2009 showed how little we know about the governance of banks and how crucial the sustainability of banking sector is. Corporate governance of banks has special relevance due to the specifics of the banking sector and its very special function in the economy. Hence, we need better understanding of corporate governance as a driver for bank performance. The Basel Committee on Banking Supervision has called attention to the need to study and improve the corporate governance of financial institutions even before the financial crisis by stating that good corporate governance is necessary to guarantee a sound financial system. Improving corporate board structure, in respect of size and composition, has been one of the main issues in corporate governance initiatives undertaken by international authorities in last decade (EU Commission Communication, 2003; Basel Committee, 2006). Additionally, corporate governance is regulated by national institutions and the framework of such regulation is forming and changing over time: UK, Germany and Norway, ones of the most developed European countries, updated their codes of corporate governance in 2012. p g On the other hand ownership structure is not regulated and only a shareholder structure is a subject of mandatory disclosure for some companies. In the last two decades many mergers and acquisitions in the banking sector took place all over the world and changed banks’ ownership type from government to private or from domestic to foreign and increased the average level of ownership concentration. These changes raised important research questions, for example, what type of ownership and level of concentration are better for bank performance. The main purpose of this paper is to review a number of existing research papers on the impact of corporate governance mechanisms over performance of commercial banks. In Section 2 we discuss the empirical evidence on the board of directors’ influence on bank performance. Section 3 reviews the performance effect of ownership concentration and types of owners. Section 4 concludes. Выпуск #4(24), 2012 ЖУРНАЛ "КОРПОРАТИВНЫЕ ФИНАНСЫ" ЖУРНАЛ "КОРПОРАТИВНЫЕ ФИНАНСЫ" 81 №4(24) 2012 81 © Электронный журнал Корпоративные Финансы, 2012 2.2 Independent directors: do they contribute to value or not? The second most popular research question is an effect of board independence. However the previous literature does not provide us with a conclusion regarding the effect of outside directors in the board. On the one hand, independent directors have fewer conflicts of interest when monitoring managers; they are less dependent on the CEO’s opinion and they have reputational incentive to perform their functions better. If the poor performance of the board results in poor performance of the firm it can damage a reputation of an independent director, which is important as many of them © Электронный журнал Корпоративные Финансы, 2012 Выпуск #4(24), 2012 №4(24) 2012 82 ЖУРНАЛ "КОРПОРАТИВНЫЕ ФИНАНСЫ" serve in more than one board of directors. On the other hand, an excessive proportion of non- executive directors could damage the advisory role of boards since they may not have some information about the company, which can be easily accessible for insiders. So there is also a tradeoff between the advantages and disadvantages of non-executive directors. Moreover, some authors point out that the effectiveness of outside directors depends on the cost of acquiring information about the firm: when the cost of acquiring information is low, outsiders in board of directors increase performance (Duchin et al., 2010). Andres and Vallelado support the hypothesis about the tradeoff in the situation with independent directors and show the reverse nonlinear relation between them and US bank performance, which also implies the existence of an optimal percentage of outsiders in the board (Andres, Vallelado, 2008). Both the codes of corporate governance in many countries and the Basel Committee recommend having a substantial proportion of outsiders in the board taking into account the advantages of their independence. Grove et al. provide empirical evidence regarding positive effect of outsiders for return on assets and Tobin’s Q for US banks, Busta does it for Tobin’s Q and returns on invested capital for banks in continental Europe and Pathan et al. prove it for returns of banks in Thailand, though Busta reports an opposite relationship for banks in the United Kingdom (Grove et al., 2009; Busta, 2008; Pathan et al., 2007). Rowe et al. demonstrate that lower percentage of insiders on the board is associated with better bank performance in China (Rowe et al., 2011). 2.2 Independent directors: do they contribute to value or not? p ( ) Still a few authors show a negative effect of outside directors, for example Kyereboah- Coleman and Biekpe for Ghanaian banks and Bino and Tomar, who report that the majority of affiliated directors on the board is correlated with higher performance (Kyereboah-Coleman, Biekpe, 2006; Bino, Tomar, 2012). Pi and Timme, Adams and Mehran and Aebi et al. find no significant relation between the degree of board independence and performance of US banks (Pi, Timme, 1993; Adams, Mehran, 2008; Aebi et al., 2011). For the European banks Staikouras et al. report the same for returns and Tobin’s Q and Busta shows it for return on assets (Staikouras et al., 2007; Busta, 2008). Sakawa and Watanabel find no significant relation between the proportion of outside directors on the board and Tobin’s Q for banks in Japan (Sakawa, Watanabel, 2011). For emerging countries this finding is common for Tandelilin et al. for Indonesia, Choi and Hasan for Korea and Love and Rachinsky for Russia and Ukraine (Tandelilin et al., 2007; Choi, Hasan, 2005; Love, Rachinsky, 2008). © Электронный журнал Корпоративные Финансы, 2012 2.3 Gender diversity efficiency The next characteristic of board of directors that receives more and more attention in the research literature and regulatory discussions in the last decade is gender diversity. The opinion that gender diversity can be beneficial to business is widely accepted, however the rate at which women are being integrated into senior positions remains extremely slow. Amongst the largest public companies across Europe, men account for 89% of board members and women just 11% (European Commission, 2010). Norway stands out as it is the only country that is relatively close to gender equality with 42% women on the boards of the largest listed companies. This is a direct result of the regulation implemented in 2006. Motivation that is behind such regulation is of a great interest. Some studies provide empirical evidence that companies with the highest share of women in executive committees outperform companies with no women, for example, in terms of return on equity by 41 percent (McKinsey&Company, 2010). Farrel and Hersh also find that companies with more women in their boards show better performance (Farrel, Hersh, 2005). Though this does not mean strict causality, it does provide a basis for discussions about benefits of greater gender diversity in board rooms. p g g y One of the explanations of the positive effect of diversity among board members is that it increases creativity and innovation by adding complementary knowledge, skills and experience. More diverse boards evaluate more alternatives compared to homogenous boards during the decision making process and that leads to better corporate performance. Some researches argue that a gender-balanced board is more likely to pay attention to managing and controlling risk (European © Электронный журнал Корпоративные Финансы, 2012 Выпуск #4(24), 2012 ЖУРНАЛ "КОРПОРАТИВНЫЕ ФИНАНСЫ" 83 №4(24) 2012 83 Commission, 2012). Adams and Ferreira show for the sample of US firms that more diverse boards allocate more human resources to monitoring, however on average the impact of gender diversity on firm performance is negative (Adams, Ferreira, 2009). Commission, 2012). Adams and Ferreira show for the sample of US firms that more diverse boards allocate more human resources to monitoring, however on average the impact of gender diversity on firm performance is negative (Adams, Ferreira, 2009). p g ( ) There are other characteristics of the board of directors that receive much less attention in the existing literature. 2.3 Gender diversity efficiency For example, CEO duality which means the positions of the CEO and Chairman of the Board of Directors are taken by the same person. Pi and Timme find that banks with a dual CEO underperform in terms of ROA (Pi, Timme, 1993). However Simpson and Gleason report that CEO duality is related to lower probability of financial distress. Their argument is that such a CEO will make less risky decisions to keep his position. Grove et al. demonstrate for US banks that CEO duality is negatively associated with financial performance (Grove et al., 2011). 3. Ownership structure and Bank Performance An ownership structure can be defined along two main dimensions. First is the degree of ownership concentration. Second is the type of the owners: state, foreign, institutional, management, etc. 3.1 Performance Effect of Ownership concentration A high ownership concentration may have a positive effect on a firm’s values as large shareholders have more incentives to monitor the bank’s management. On the other side, large shareholders may have too much influence on the board and the management and if they have other goals besides share value maximization (like government), it may not be effective for the firm itself and also for the minority shareholders. The quality of their monitoring may be also not very high if they do not have relevant knowledge and experience. Grove et al. show that a presence of a majority shareholder is positively associated with performance of US banks in terms of Tobin’s Q and ROA (Grove et al., 2009). Chang et al., studying effects of diversification and reporting its negative impact of bank’s performance, show that “banks with a dominant owner suffer less from diversification than those without a dominant owner” (Chang et al., 2009). Rowe et al. for Chinese banks and Riewsathirathorn et al. for East- Asian banks demonstrate that lower block ownership is associated with better performance (Rowe et al., 2011; Riewsathirathorn et al., 2011). Pi and Timme for US banks and Ianotta et al. for large European banks show that ownership concentration is unrelated to bank’s profitability (Pi, Timme, 1993; Ianotta et al., 2007). Some authors find more complicated nonlinear relationship between ownership concentration and bank’s performance. For example, Antoniadis et al. find a statistically significant quadratic relationship between ownership and performance for Greek banks (Antoniadis et al., 2010). Furthermore, Magalhaes et al. for 795 banks from 47 countries report a cubic relationship between ownership concentration and bank performance (Magalhaes et al., 2010). Busta confirms that the effect may differ for the different institutional settings (Busta, 2008). The author finds that the effect of an increase in ownership concentration on Tobin’s Q may be beneficial in France and Scandinavian countries but may be punished by investors in the Germany and UK. © Электронный журнал Корпоративные Финансы, 2012 3.2 Type of owner: does it matter for bank performance? yp f f p f Among other characteristics of bank ownership that are being studied are types of shareholders: institutions, management, foreigners and government. Many authors show a negative influence of state ownership (e.g. see Berger et al., 2005; Farazi et al., 2011 for Middle East and North Africa; Berger et al., 2009 and Lin & Zhang, 2009 for China; Micco et al., 2007 for developing countries). At the same time in times of global financial crisis we should take into account the potential positive influence of state as a bank owner on bank financial stability and performance. Although we failed to find significant evidence for this relationship during recent financial global crisis, we could support this statement with strong evidence from Asian crisis of © Электронный журнал Корпоративные Финансы, 2012 © Электронный журнал Корпоративные Финансы, 2012 Выпуск #4(24), 2012 ЖУРНАЛ "КОРПОРАТИВНЫЕ ФИНАНСЫ" №4(24) 2012 84 early 2000s (Cornett et al., 2010). y ( ) Stable governance traditions, deeper experience, higher level of discipline, better access to global capital market generally allow banks with foreign owners outperforming domestically-owned banks. A number of empirical studies find a positive performance effect of foreign ownership in emerging markets (e.g. Bonin et al., 2005; Kim, Rasiah, 2010). But results regarding foreign ownership change from country to country. For example, Berger et al. show that in China foreign banks are most efficient, but Lensink et al. report for the worldwide sample that increase in foreign ownership negatively affects bank performance (Berger et al., 2009; Lensink et al., 2007). 4. Conclusion The financial crisis demonstrated the great importance of commercial banks in the economic system, which strengthens the need to study the corporate governance of banks. We analyzed the existing empirical evidence regarding the relationship between corporate governance mechanisms and performance and we observe that there is no common opinion about such relationship. The researchers obtain the mixed results on the impact of board of directors characteristics - its size and independence - over bank performance. Despite the arguments about positive effect of additional human capital provided by larger boards, the most widespread opinion is that the relationship between bank performance and board size is negative. This view is supported by the majority of the papers and shows that the coordination problems in larger boards outweigh its advantages of bigger collection of knowledge. The similar trade-off may be seen for the board independence. Here we observe that if independent directors have a significant effect on bank performance than it is most likely positive due to outside knowledge and experience they bring to the board. The research of the performance effect of gender diversity is becoming popular among researchers only in the past few years. Most existing empirical evidence shows that female directors provide banks with a better monitoring which leads to a positive relationship between the percentage of women in the board of directors and bank performance. The results of previous research regarding the ownership concentration effect vary significantly. Nevertheless, some authors agree that having high ownership concentration is not effective for banks and therefore the its relationship with bank performance is negative. The findings about the impact of state ownership are more homogenous and they state that the government as a shareholder may have other goals besides the increasing value of the company which may hurt the bank performance. The diversity of obtained evidence might be explained by countries’ peculiarities, which means that there are some important dissimilarities in how corporate governance influences bank performance in various countries and during different stages of economic cycle. Therefore, when corporate governance principles are being established there has to be taken an individual approach to every country. ЖУРНАЛ "КОРПОРАТИВНЫЕ ФИНАНСЫ" 85 №4(24) 2012 of directors, Journal of Banking and Finance, 32 (2008), 2570-2580. 6. Antoniadis, I., Lazarides, T., Sarrianides, N. (2010), Ownership and performance in the Greek banking sector, International Conference on Applied Economics – ICOAE 2010, Working Paper. g p 7. Basel Committee on Banking Supervision (2006), Enhancing Corporate Governance for Banking Organizations. 8. Belkhir, M. (2009), Board of directors’ size and performance in the banking industry, International Journal of Managerial Finance, Vol. 5, No. 1. 9. Berger, A. N., Clarke, G.R.G., Cull, R., Klapper, L., Udell, G. F. (2005), Corporate governance and bank performance: A joint analysis of the static, selection, and dynamic effects of domestic, foreign, and state ownership, Journal of Banking and Finance, 29 (2005), 2179-2221. 10. Berger, A.N., Hasan, I., Zhou, M. (2009), Bank ownership and efficiency in China: What will happen in the world’s largest nation?, Journal of Banking & Finance, 33 (2009), 113- 130. 11. Bino, A., Tomar, S. (2012), Corporate governance and bank performance: evidence from Jordanian banking industry, Journal of Business Administration, forthcoming. 12. Bonin, J. P., Hasan, I., Wachtel, P. (2005), Bank performance, efficiency and ownership in transition countries, Journal of Banking and Finance, 29 (2005), 31-53. 13. Busta, I. (2008), Corporate governance in banking: a European study, Copenhagen Business School, The PhD School in Economics and Business Administration, PhD Series 15.2008. 14. Chang, Y., Tsai, H., Lai, K. (2009), Ownership structure, supervisory regulation and the diversification effects on bank performance, Journal of Financial Studies, Vol. 17, No 4. 15. Choi, S., Hasan, I., (2005), Ownership, governance, and bank performance: Korean experience, Financial Markets, Institutions & Instruments, Vol. 14, No. 4. 16. Cornett, M., Guo, L., Khaksari, S., Tehranian, H. (2010), The Impact of State Ownership on Performance Differences in Privately-Owned versus State-Owned Banks: An International Comparison, Journal Of Financial Intermediation, 19 (2010), 74-94. 17. Duchin, R., Matsusaka, J.G., Ozbas, O. (2010), When Are Outside Directors Effective?, Journal of Financial Economics, 96 (2010), 195-214. 18. European Commission Recommendation, 2005/162/EC. 19. European Commission, 2010, More women in senior positions – Key to economic stability and growth, Directorate-General for Employment, Social Affairs and Equal Opportunities. 20. European Commission, 2012, Women in economic decision-making in the EU: Progress report. A Europe 2020 initiative. 21. Farazi, S., Feyen, E., Rocha, R. (2011), Bank ownership and performance in the Middle East and North Africa region, The World Bank, Policy Research Working Paper 5620. 22. © Электронный журнал Корпоративные Финансы, 2012 References 1. Adams, R.B., Ferreira, D. (2009), Women in the Boardroom and their impact on governance and performance, Journal of Financial Economics, 94 (2009), 291-309. 2. Adams, R.B., Mehran, H. (2008), Corporate performance, board structure, and their determinants in the banking industry, Federal Reserve Bank of New York Staff Reports, no. 330. 3. Adusei, M. (2011), Board structure and bank performance in Ghana, Journal of Money, Investment and Banking 19. 3. Adusei, M. (2011), Board structure and bank performance in Ghana, Journal of Money, Investment and Banking 19. 4. Aebi, V., Sabato, G., Schmid, M. (2012), Risk management, corporate governance, and bank performance in the financial crisis, Journal of Banking & Finance, 36 (2012), 3213-3226. 4. Aebi, V., Sabato, G., Schmid, M. (2012), Risk management, corporate governance, and bank performance in the financial crisis, Journal of Banking & Finance, 36 (2012), 3213-3226. p g ( ) 5. Andres, P. de, Vallelado, E. (2008), Corporate governance in banking: The role of the board p g ( ) 5. Andres, P. de, Vallelado, E. (2008), Corporate governance in banking: The role of the board © Электронный журнал Корпоративные Финансы, 2012 Выпуск #4(24), 2012 ЖУРНАЛ "КОРПОРАТИВНЫЕ ФИНАНСЫ" 28. Kyereboah-Coleman, A., Biekpe, N. (2006), Do boards and CEOs matter f ЖУРНАЛ "КОРПОРАТИВНЫЕ ФИНАНСЫ" Farrel, K.A., Hersh, P.L. (2005), Additions to corporate boards: the effect of gender, Journal of Corporate Finance, 11 (2005), 85– 106. 23. Grove, H., Patelli, L., Victoravich, L.M., Xu, P. (2009), Corporate governance and performance: evidence from U.S. Commercial Banks, Working paper. 24. Grove, H., Patelli, L., Victoravich, L.M., Xu, P. (2011), Corporate Governance and Performance in the Wake of the Financial Crisis: Evidence from US Commercial Banks, Corporate Governance: An International Review, Vol. 19, 5, 418-436. p 25. Hermalin, B.E., Weisbach, M.S. (2003), Board of directors as an endogenously determined institution: A survey of the economic literature, FRBNY Economic Policy Review, 7–26. 26. Iannotta, G., Nocera, G., Sironi, A. (2007), Ownership structure, risk and performance in the European banking industry, Journal of Banking & Finance, 31 (2007), 2127–2149. 27. Kim, P.K., Rasiah, D. (2010), Relationship between corporate governance and bank performance in Malaysia during the pre and post Asian financial crisis, European Journal of Economics, Finance and Administrative Sciences, 21 (2010), 39-63. ( ) 28. Kyereboah-Coleman, A., Biekpe, N. (2006), Do boards and CEOs matter for bank 28. Kyereboah-Coleman, A., Biekpe, N. (2006), Do boards and CEOs matter for bank 28. Kyereboah-Coleman, A., Biekpe, N. (2006), Do boards and CEOs matter for bank Выпуск #4(24), 2012 ЖУРНАЛ "КОРПОРАТИВНЫЕ ФИНАНСЫ" 86 №4(24) 2012 86 performance? A comparative analysis of banks in Ghana, Corporate Ownership and Control, Vol. 4, Issue 1. performance? A comparative analysis of banks in Ghana, Corporate Ownership and Control, Vol. 4, Issue 1. 29. Laeven, L., Levine, R. (2009), Bank governance, regulation and risk-taking, Journal of Financial Economics, 93 (2009), 259-275. 30. Lensink, R., Naaborg, I. (2007), Does foreign ownership foster bank performance?, Applied Financial Economics, 17 (2007), 881–885. 31. Lin, X., Zhang, Y. (2009). Bank ownership reform and bank performance in China, Journal of Banking & Finance, 33(1), 20-29. 32. Love, I., Rachinsky, A. (2007), Corporate Governance, ownership and bank performance in emerging markets: Evidence from Russia and Ukraine, Working paper. 33. Magalhaes, R., Urtiaga, M.G., Tribo, J.A. (2010), Banks’ ownership structure, risk and performance, Working Paper. 34. McKinsey&Company, 2010, Women Matter 2010, Women at the top of corporations: Making it happen. 35. Mehran, H., Morrison, A., Shapiro, J. (2011), Corporate Governance and Banks: What Have We Learned from the Financial, Federal Reserve Bank of New York Staff Reports, no. 502. 36. Micco, A., Panizza, U., Yanez, M. (2007), Bank ownership and performance. ЖУРНАЛ "КОРПОРАТИВНЫЕ ФИНАНСЫ" Does politics matter?, Journal of Banking & Finance, 31 (2007), 219–241. 37. Minton, B.A., Taillard, J.P.A., Williamson, R. (2010), Do independence and financial expertise of the board matter for risk taking and performance?, Charles A. Dice Center Working Paper No. 2010-14. g p 38. Pathan, S., Skully, M., Wickramanayake, J. (2007), Board size, independence and performance: an analysis of Thai banks, Asia-Pacific Finan Markets, 14 (2007), 211-227. 39. Pi, L., Timme, S.G. (1993), Corporate control and bank efficiency, Journal of Banking and Finance, 17 (1993), 515-530. 40. Riewsathirathorn, P., Jumroenvong, S., Jiraporn, P. (2011), The impact of ownership concentration on bank performance and risk-taking: evidence from East Asia, The Journal of Emerging Markets, forthcoming. 41. Rowe, W., Shi, W., Wang, C. (2011), Board governance and performance of Chinese banks, Banks and Bank Systems, No. 1, 2011. 42. Sakawa H., Watanabel, N. (2011), Corporate board structures and performance in the banking industry: evidence from Japan, Working Paper. 43. Simpson, W.G., Gleason, A.E. (1999), Board structure, ownership, and financial distress in banking firms, International Review of Economics and Finance, 8 (1999), 281–292. 44. Staikouras, P.K., Staikouras, C.K., Agoraki, M.E. (2007), The effect of board size and composition on European bank performance, European Journal of Law and Economics, 23 (2007), 1-27. 45. Tandelilin, E., Kaaro, H., Mahadwartha, P.A. (2007), Corporate governance, risk management and bank performance: Does type of ownership matter?, EADN individual research grant project No. 34. 46. Zulkafli, A.H., Samad, F.A. (2007), Corporate governance and performance of banking firms: evidence from Asian emerging markets, Advances in Financial Economics, 12 (2007), 49-74. Выпуск #4(24), 2012 © Электронный журнал Корпоративные Финансы, 2012
https://openalex.org/W2898599855	https://europepmc.org/articles/pmc6206822?pdf=render	English	null	Hematopoietic stem cell transplantation and quality of life during the first year of treatment	Revista latino-americana de enfermagem	2,018	cc-by	7,938	p p 2 Hospital Vita Curitiba, Supervisão de Enfermagem, Curitiba, PR, Brazil. 1 Universidade Federal do Paraná, Complexo Hospital de Clínicas, Curitiba, PR, Brazil. Hematopoietic stem cell transplantation and quality of life during the first year of treatment Angela da Costa Barcellos Marques1 Ana Paula Szczepanik1 Celina Angélica Mattos Machado1 Pâmella Naiana Dias Santos2 Paulo Ricardo Bittencourt Guimarães3 Luciana Puchalski Kalinke4 Objective: to evaluate the quality of life of adult patients with hematological cancer comparing Hematopoietic Stem Cell Transplantation modalities during the first year of treatment. Method: this is an observational and longitudinal research with 55 participants. Data collection was performed in six steps: before transplantation, pancytopenia, before hospital discharge, after 100, 180 and 360 days, in a reference hospital in Brazil for this treatment. The international instruments Quality of Life Questionnaire - Core 30 and Functional Assessment Cancer Therapy - Bone Marrow Transplantation were validated and translated into Portuguese (Brazil). Results: the mean age of participants was 36 years, 65% (n = 36) had leukemia diagnosis and 71% (n = 39) had undergone allogenic transplantation. In the Quality of Life Questionnarie - Core30 instrument, the pain symptom was significant between the first and second stages, and loss of appetite between the third and fourth stages, both in the allogenic group. In the Functional Assessment Cancer Therapy - Bone Marrow Transplantation, the functional well-being domain was significant between the third and fourth stages, also in the allogenic group. Conclusions: although the aggressiveness of treatment affects quality of life, patients consider it satisfactory after the first year. There are few significant differences between autologous and allogenic patients, and both groups have recovered in the course of the process. Descriptors: Quality of Life; Hematopoietic Stem Cell Transplantation; Bone Marrow Transplantation; Hematologic Neoplasms; Oncology Nursing; Drug Therapy. How to cite this article Marques ACB, Szczepanik AP, Machado CAM, Santos PND, Guimarães PRB, Kalinke LP. Hematopoietic stem cell transplantation and quality of life during the first year of treatment. Rev. Latino-Am. Enfermagem. 2018;26:e3065. [Access ___ __ ____]; Available in: ___________________. DOI: http://dx.doi.org/10.1590/1518-8345.2474.3065. month day year URL 4 Universidade Federal do Paraná, Departamento de Enfermagem, Curitiba, PR, Brazil. p , p g , , , 3 Universidade Federal do Paraná, Departamento de Estatística, Curitiba, PR, Brazil. Rev. Latino-Am. Enfermagem 2018;26:e3065 DOI: 10.1590/1518-8345.2474.3065 www.eerp.usp.br/rlae Hematopoietic stem cell transplantation and quality of life during the first year of treatment Original Article Rev. Latino-Am. Enfermagem 2018;26:e3065 DOI: 10.1590/1518-8345.2474.3065 www.eerp.usp.br/rlae Hematopoietic stem cell transplantation and quality of life during the first year of treatment Original Article Original Article Original Article Rev. Latino-Am. Enfermagem 2018;26:e3065 DOI: 10.1590/1518-8345.2474.3065 www.eerp.usp.br/rlae Hematopoietic stem cell transplantation and quality of life during the first year of treatment Universidade Federal do Paraná, Complexo Hospital de Clínicas, Curitiba, PR, Brazil. 2 Hospital Vita Curitiba, Supervisão de Enfermagem, Curitiba, PR, Brazil. 3 Universidade Federal do Paraná, Departamento de Estatística, Curitiba, PR, Brazil. 4 Universidade Federal do Paraná Departamento de Enfermagem Curitiba PR Brazil Introduction 516 HSCTs, of which 316 were autologous and 200 allogenic(12). Approximately 30,000 transplantations have been performed in our country since 1979(13). Some malignant and non-malignant diseases can be treated with hematopoietic stem cell transplantation (HSCT). Since it is a complex and aggressive procedure, it demands demand specific care with professionals from different areas inserted in the same therapeutic context. The treatment is relatively long and involves risks that predispose the patient to a broad spectrum of complications that need to be managed in order not to threaten his/her life or affect his/her survival and quality of life (QoL). The modalities of HSCT are denominated as autologous or autogenic and allogenic. They are determined according to the type of donor of hematopoietic stem cells (HSCs). In autologous HSCT, the HTCs are collected from the patient him/herself before the conditioning phase; they are basically stored and reinfused later. In allogenic HSCT, the HTCs come from a donor that may be related or not to the patient(14-16). The hematological cancers are among the malignancies, with considerable incidence in Brazil and in the world. According to the National Cancer Institute José Alencar Gomes da Silva (INCA), estimates for the biennium 2016/2017 indicate the occurrence of 22,780 cases, 12,210 among men and 10,570 among women(1). For the International Agency for Research on Cancer (IARC), from the World Health Organization (WHO), 948,942 cases are estimated in 2020(2). Each HSCT modality has specificities, with pre- established protocols and chemotherapeutic regimens according to the disease, besides requiring care at different levels of complexity. The allogenic HSCT has some particularities in terms of variables to be controlled, as there is concern about the necessary compatibilities between donor and recipient. In addition, the fact that HSCs come from another person, relative or not, may increase the risk of complications, such as Graft versus Host Disease (GVHD), thus affecting multiple organ systems and implying changes in the domains of QoL(16-17). Cancer has the potential to negatively impact patients’ QoL, as it creates suffering because of uncertainty about the future and how their body will react to treatment(3). It entails physical, psychological and emotional changes, with consequent loss of self- esteem interfering in patients’ survival and QoL(4-5). Introduction However, both modalities have their own demands, and there are several factors, such as the existence of previous comorbidities associated or not with hematologic cancer, older age or not, diagnosis time, prognosis and even social conditions, among others that may interfere in the treatment, regardless of the type of transplantation performed. In recent decades, QoL has been the focus of studies, mainly in the area of oncology. The development and technological improvement of the therapeutics for hematological cancer has increased the possibility of prolonging the life of patients, reflecting in a greater attention to the QoL, that started to be as important as the survival(6) and aroused the interest of researchers in knowing how patients’ lives are affected by the diseases(7-8). During the therapeutic process, the patient undergoes some critical stages in which complications, besides putting their life at risk, can negatively affect their QoL, as symptoms that have a disabling potential may appear(16,18). In addition to the physical complications, the patient may suffer with emotional and social changes during treatment. They may feel fear and anguish, and miss the family and friends, since social isolation is necessary in the early stages of treatment. There are varied definitions and conceptions for the term “Quality of Life”. Its concept is broad and multidimensional, with social, health or economic parameters(9-10). The World Health Organization (WHO) defines it as individual’s perception of their position in life in the context of the culture and value systems in which they live in relation to their goals, expectations, standards and concerns”(11). For this research, we decided to use the WHO definition because we understood that its concept is more comprehensive and possibly contemplates the entire therapeutic path that the patient with hematological cancer submitted to HSCT can go through. The QoL of patients worsens as the severity of the symptoms increases(19). Knowing the specific changes in the QoL of the patient at each stage of the treatment enables the professionals involved in this context, especially the nurse, to establish an individualized and effective care plan, assisting the patient in facing their clinical condition, as well as aiming at a better survival. In this sense, it is relevant to carry out studies focused on this theme. How to cite this article Marques ACB, Szczepanik AP, Machado CAM, Santos PND, Guimarães PRB, Kalinke LP. Hematopoietic stem cell transplantation and quality of life during the first year of treatment. Rev. Latino-Am. Enfermagem. 2018;26:e3065. [Access ___ __ ____]; Available in: ___________________. DOI: http://dx.doi.org/10.1590/1518-8345.2474.3065. month day year URL 2 Rev. Latino-Am. Enfermagem 2018;26:e3065. www.eerp.usp.br/rlae Results The sociodemographic characterization of the sample showed that the mean age was 36 years, 53% (n = 29) were male, 55% (n = 30) were married or declared stable union. Regarding education, 44% (n = 24) declared having completed high school and 64% (n = 35) declared themselves to be economically active. Regarding the clinical characterization, 65% (n = 36) presented some type of leukemia and 71% (n = 39) had undergone allogenic HSCT. This is a longitudinal and observational research developed in the Bone Marrow Transplantation Service (STMO in Portuguese) of a federal teaching hospital in Curitiba, national reference in HSCT, from September 2013 to November 2016. The non-probabilistic sample, however, was based on the number of patients attended in the years 2010 to 2012 plus 50%, due to the possibility of losses because of the treatment characteristics, was composed of 55 participants. Inclusion criteria were age equal to or greater than 18 years, diagnosis of hematologic cancer and having been submitted to HSCT. Participants who did not have the physical conditions to complete the instruments were excluded from the survey. Three patients were excluded due to loss of follow-up; 12 died before completing 100 days of HSCT, two before completing 180 days and six days before completing 360 days. The results expressed in Table 1 demonstrate the means obtained between the autologous and allogenic groups, measured by the QLQ-C30 in the six stages of the research. The functional scale shows that in both groups there was recovery of physical function and social function after 360 days of transplantation when compared to the pre-HSCT stage; in this scale, the higher the mean, the better the performance. In the symptom scale, the means demonstrate increased fatigue in the autologous group after 360 days in relation to the pre- HSCT stage; in this scale, the higher the mean, the more intense the symptomatology. Data were collected in the inpatient and outpatient wards of HSCT in six stages, namely before HSCT, pancytopenia period, before discharge, after 100 days, after 180 days, and after 360 days of HSCT. Sociodemographic and clinical data were collected with a specific instrument in pre-HSCT stage. Introduction Thus, the objective of the present study was to evaluate the QoL of adult patients with hematological cancer submitted to HSCT in the different stages of treatment during the first year and to compare the autologous and allogenic transplantation modalities. Despite the complexity and aggressiveness of HSCT, its performance has increased every year. In 2016, 2,270 HSCTs were performed in Brazil by 43 transplantation centers. In the first quarter of 2017, there were www.eerp.usp.br/rlae 3 Marques ACB, Szczepanik AP, Machado CAM, Santos PND, Guimarães PRB, Kalinke LP. Marques ACB, Szczepanik AP, Machado CAM, Santos PND, Guimarães PRB, Kalinke LP. Method organized into tables and analyzed according to EORTC and FACIT guidelines, expressed as mean (M). This research is part of the thematic project “Quality of Life Assessment of Patients with Hematologic Neoplasm Submitted to Hematopoietic Stem Cell Transplantation”, approved by the Research Ethics Committee of the Health Sciences Sector of the Federal University of Paraná under the opinion number 411,548 whose objective was to evaluate QoL of patients up to five years after HSCT. The Mann Whitney test was used to compare the types of transplantation and the Friedman test was applied for the comparison between the stages, complemented by the Test of Minimum Important Difference, for multiple comparisons (p-value), in which the level of significance of 5% for results with p-value below 0.05 were considered significant (p <0.05). The calculations were performed by a statistician. The Statistica 7.0 software was used for analysis. The use of the instruments was authorized by the European Organization Research Treatment of Cancer (EORTC) and the Functional Assessment of Chronic Illness Therapy (FACIT), which made the questionnaires available via download directly to the researcher after registering the research project. www.eerp.usp.br/rlae Results Curitiba, PR, Brazil, 2013-2016 (n=55) Quality of Life Questionnaire - Core 30 SCORES Before HSTC* (n=55) Pancytopenia (n=50) Before discharge (n=49) After 100 days (n=41) After 180 days (n=38) After 360 days (n=32) Means Means Means Means Means Means Aut† Alo‡ Aut† Alo‡ Aut† Alo‡ Aut† Alo‡ Aut† Alo‡ Aut† Alo‡ Overall QoL§ 70.8 79.2 59.3 55.3 73.9 66.7 80.7 71.4 75.6 77.5 72.7 70.6 Functional Scale Physical function 72.9 77 50.8 57.6 57.5 68.2 82.5 74.7 79.4 80.5 83.6 83.8 Personal Performance 79 79 50 41.1 52 61.6 82 72.6 88.2 77.5 90 .9 87.3 Social role 62.5 52.1 46.8 33.3 46.8 35.8 71.7 52.9 86.1 67.9 81.8 75.4 Scale Symptoms/Items Fatigue 22.2 21.3 47.9 55.8 35.4 45.4 26.5 28.9 27.7 24.7 29.8 22.2 Nausea and vomiting 8.3 9.4 43.7 47.5 35.4 35.8 6.4 13.1 8.3 5.7 7.5 11.1 Pain 30.2 14.1 38.5 68.1 18.7 28.7 20.5 17.2 18 13.4 25.7 12.7 Loss of appetite 20.8 16.2 56.2 68.6 47.9 61.6 5.1 27.3 16.6 16.6 6 12.7 Diarrhea 4.1 5.9 50 54.9 33.3 26.2 15.3 13.1 13.8 7.6 12.1 1.5 HSCT: Hematopoietic stem cell transplantation; † Aut: Autologous; ‡ Alo: Allogenic; §QoL: Quality of Life Table 2 – Significant scores of the Functional Assessment of Cancer Therapy - Bone Marrow Transplantation of patients submitted to autologous and allogenic transplantation obtained in the six stages of the research. Curitiba, PR, Brazil, 2013-2016 (n=55) Table 2 – Significant scores of the Functional Assessment of Cancer Therapy - Bone Marrow Transplantation of patients submitted to autologous and allogenic transplantation obtained in the six stages of the research. Results In all stages, we applied the instruments Quality of Life Questionnaire-Core 30 (QLQ C-30) - version 3.0 Brazilian Portuguese, developed by the European Organization for Research and Treatment of Cancer (EORTC) and the Functional Assessment Cancer Therapy - Bone Marrow transplantation (FACT-BMT) - version 4.0 Brazilian Portuguese, prepared by the Functional Assessment of Chronic Illness Therapy (FACIT). The QLQ C-30 is divided into functional and symptom scales and its results are calculated according to the EORTC Scoring Manual(20). The FACT-BMT is divided into domains and its results are calculated as described in the FACIT Scoring Manual(21). Table 2 shows the means obtained by the group of patients submitted to autologous and allogenic HSCT, measured by FACT-BMT in the six stages of the study. It is observed that only in the pre-HSCT stage the allogeneic group obtained averages higher than the autologous group in all domains. However, in all other stages, the means of the allogeneic group were lower than those of the autologous group, but with no significant difference. The means of overall QoL values measured by the QLQ-C30 and by the FACT-BMT, expressed in Tables 1 and 2, respectively, showed similar performance among patients submitted to autologous and allogenic HSCT in the six stages of the research. Sociodemographic and clinical data were analyzed descriptively and expressed in absolute and relative frequency. The data of the QoL instruments were The comparison between the modalities of HSCT, as demonstrated in Table 3, showed that in both the overall QoL measured by QLQ-C30 and in the general 4 Rev. Latino-Am. Enfermagem 2018;26:e3065. Rev. Latino-Am. Enfermagem 2018;26:e3065. period, in both groups, with gradual reduction during the treatment. It is noteworthy that the pain symptom was more intense in the allogenic group in this period. QoL measured by FACT-BMT, there was no statistically significant difference in any stage of the study. However, in relation to the symptom of pain and loss of appetite measured by the QLQ-C30, as well as the functional well-being evaluated with the FACT-BMT, there was a significant difference between autologous and allogenic transplantation in the pancytopenia period and after 100 days, respectively. Figure 2 shows that, among the six stages of the study, in the domains measured by the FACT-BMT, there was a similar performance between the autologous and allogenic groups between pre-HSCT and the period of pancytopenia. Results In physical well-being, both groups present gradual recovery of the means between the period of pancytopenia and after 360 days. In functional well-being, the autologous group presented better performance than the allogenic group during the same period. Concerning the domain additional concerns and the FACTG (overall evaluation encompassing physical well-being/social and family welfare/emotional well- being/functional well-being), both groups presented recovery during treatment, however the autologous group presented slightly higher means. The comparison of the autologous and allogenic transplantations, expressed in Figure 1, of the domains measured by the QLQ-C30, shows that in the personal performance there is a significant reduction of the means in the periods of pancytopenia and before discharge in relation to the pre-HSCT stage, with gradual recovery up to 360 days. In the social function, there was a similar behavior of the means in both groups, and in after 360 days the means exceed those presented in the pre-HSCT stage. Regarding fatigue and pain symptoms, they presented the highest averages in the pancytopenia Table 1 - Significant scores of the Quality of Life Questionnaire - Core 30 of the patients submitted to autologous and allogenic transplantation obtained in the six stages of the research. www.eerp.usp.br/rlae Results Curitiba, PR, Brazil, 2013-2016 (n=55) , , ( ) Functional Assessment of Cancer Therapy Bone Marrow Transplantation Domain Before HSCT (n=55) Pancytopenia (n=50) Before discharge (n=49) After 100 days (n=41) After 180 days (n=38) After 360 days (n=32) Means Means Means Means Means Means Aut† Alo‡ Aut† Alo‡ Aut† Alo‡ Aut† Alo‡ Aut† Alo‡ Aut† Alo‡ Physical well-being 21 22.3 16.4 14.5 19.9 18.7 23 21.2 23.9 22.7 24.2 23.2 Functional well-being 18.9 19.5 15.1 14.6 16.9 15.4 20.2 16.2 18.6 18.6 19.9 18.2 Additional concerns 27.5 27.9 22.6 23 24.7 23.4 30.3 27.4 31.1 29 29.9 29.1 FACTG§ 80.2 80.7 68.4 67.3 74.2 71.5 84.3 77.5 84.1 81.3 83.9 80.3 Overall QoL 107 108 91.4 90.3 98.9 95 114 105 115 110 113 109 HSCT: Hematopoietic stem cell transplantation; †Aut: Autologous; ‡Alo: Allogenic; §FACTG: Overall assessment (physical well-being/social and family well-being/emotional well-being/functional well-being) Functional Assessment of Cancer Therapy Bone Marrow Transplantation Functional Assessment of Cancer Therapy Bone Marrow Transplantation www.eerp.usp.br/rlae 5 Marques ACB, Szczepanik AP, Machado CAM, Santos PND, Guimarães PRB, Kalinke LP. Table 3- Comparison between autologous and allogeneic transplantation measured by the Quality of Life Questionnaire - Core 30 questionnaire and the Functional Assessment of Cancer Therapy - Bone Marrow Transplantation in the six stages of the research. Curitiba, PR, Brazil, 2013-2016 (n=55) Table 3- Comparison between autologous and allogeneic transplantation measured by the Quality of Life Questionnaire - Core 30 questionnaire and the Functional Assessment of Cancer Therapy - Bone Marrow Transplantation in the six stages of the research. Results In Iran, a longitudinal study of 587 patients submitted to HSCT revealed that the mean survival time of patients was 517 days(23). literature. In Brazil, a study conducted in Campinas-SP, Brazil, with 62 patients submitted to HSCT, showed a death rate of 21% in the period before one year after transplantation(22). In Iran, a longitudinal study of 587 patients submitted to HSCT revealed that the mean survival time of patients was 517 days(23). www.eerp.usp.br/rlae Results Curitiba, PR, Brazil, 2013-2016, (n=55) Physical well-being 26 24 22 20 18 16 14 1 2 3 4 5 6 Additional concerns 32 31 30 29 28 27 26 25 24 23 22 1 2 3 4 5 6 21 20 19 18 17 16 15 14 1 2 3 4 5 6 Functional well-being 86 84 82 80 78 76 74 72 70 68 66 1 2 3 4 5 6 FACTG autologous allogenic Notes: Friedman test; FACTG: Overall assessment (physical well-being/social and family well-being/emotional well-being/functional well-being) Figure 2 - Comparison of the six stages of the research using the domains measured by the Functional Assessment of Cancer Therapy - Bone Marrow Transplantation (FACT-BMT) of patients submitted to autologous and allogenic HSCT. Curitiba, PR, Brazil, 2013-2016, (n=55). 21 20 19 18 17 16 15 14 1 2 3 4 5 6 Functional well-being Physical well-being 26 24 22 20 8 6 4 1 2 3 4 5 6 21 20 19 18 17 16 15 14 1 2 3 4 5 6 Functional well-being Additional concerns 32 31 30 29 28 27 26 25 24 23 22 1 2 3 4 5 6 ocial and family well-being/emotional well-being/functional well-being Notes: Friedman test; FACTG: Overall assessment (physical well-being/social and family well-being/emotional well-being/functional well-being) Figure 2 - Comparison of the six stages of the research using the domains measured by the Functional Assessment of Cancer Therapy - Bone Marrow Transplantation (FACT-BMT) of patients submitted to autologous and allogenic HSCT. Curitiba, PR, Brazil, 2013-2016, (n=55). Notes: Friedman test; FACTG: Overall assessment (physical well-being/social and family well-being/emotional well-being/functional well-being) Figure 2 - Comparison of the six stages of the research using the domains measured by the Functional Assessment of Cancer Therapy - Bone Marrow Transplantation (FACT-BMT) of patients submitted to autologous and allogenic HSCT. Curitiba, PR, Brazil, 2013-2016, (n=55). Notes: Friedman test; FACTG: Overall assessment (physical well-being/social and family well-being/emotional well-being/functional well-being) Figure 2 - Comparison of the six stages of the research using the domains measured by the Functional Assessment of Cancer Therapy - Bone Marrow Transplantation (FACT-BMT) of patients submitted to autologous and allogenic HSCT. Curitiba, PR, Brazil, 2013-2016, (n=55). literature. In Brazil, a study conducted in Campinas-SP, Brazil, with 62 patients submitted to HSCT, showed a death rate of 21% in the period before one year after transplantation(22). Results Curitiba, PR, Brazil, 2013-2016 (n=55) Domain Before HSCT* (n=55) Pancytopenia (n=50) Before discharge (n=49) After 100 days (n=41) After 180 days (n=38) After 360 days (n=32) p-value p-value p-value p-value p-value p-value Quality of Life Questionnaire - Core 30 Overall QoL† 0.094 0.571 0.266 0.298 0.841 0.785 Physical function 0.503 0.337 0.075 0.353 0.792 0.969 Personal Performance 0.993 0.251 0.428 0.249 0.312 0.611 Social role 0.213 0.099 0.223 0.186 0.207 0.639 Fatigue 0.905 0.369 0.325 0.836 0.525 0.289 Nausea and vomiting 0.847 0.750 0.924 0.151 0.841 0.815 Pain 0.052 0.004‡ 0.285 0.480 0.359 0.180 Loss of appetite 0.734 0.297 0.240 0.031‡ 0.938 0.755 Diarrhea 0.862 0.688 0.576 0.688 0.653 0.155 Functional Assessment of Cancer Therapy Bone Marrow Transplantation Physical well-being 0.250 0.278 0.620 0.382 0.792 0.852 Functional well-being 0.790 0.658 0.240 0.031‡ 0.699 0.348 Additional concerns 0.491 0.734 0.428 0.151 0.327 0.884 FACTG§ 0.804 0.297 0.506 0.186 0.676 0.574 Overall QoL† 0.666 0.688 0.506 0.205 0.545 0.519 Notes: Mann-Whitney test; *HSCT: Hematopoietic stem cell transplantation; †QoL: Quality of life; ‡Statistically significant data; §FACTG: Overall assessment (physical well-being/social and family well-being/emotional well-being/functional well-being) Personal performance 100 90 80 70 60 50 40 30 1 2 3 4 5 6 Fatigue 60 55 50 45 40 35 30 25 20 15 1 2 3 4 5 6 90 80 70 60 50 40 30 1 2 3 4 5 6 Social role 80 70 60 50 40 30 20 10 1 2 3 4 5 6 Pain autologous allogenic Note: Friedman test Figure 1- Comparison of the six stages of the research using the domains measured by Quality of Life Questionnaire - Core 30 (QLQ-C30) of patients submitted to autologous and allogenic HSCT. Curitiba, PR, Brazil, 2013-2016, (n=55) 80 70 60 50 40 30 20 10 1 2 3 4 5 6 Pain autologous allogenic 80 70 60 50 40 30 20 10 1 2 3 4 5 6 Pain autologous allogenic 1 2 3 4 5 6 Note: Friedman test Figure 1- Comparison of the six stages of the research using the domains measured by Quality of Life Questionnaire - Core 30 (QLQ-C30) of patients submitted to autologous and allogenic HSCT. Curitiba, PR, Brazil, 2013-2016, (n=55) Figure 1- Comparison of the six stages of the research using the domains measured by Quality of Life Questionnaire - Core 30 (QLQ-C30) of patients submitted to autologous and allogenic HSCT. Discussion HSCT is a complex and aggressive treatment, with a large number of variables that must be controlled so that patients’ lives are not jeopardized or their QoL is compromised. During the study, 20 patients (36.36% of the sample) died. The mortality rate associated with treatment is expressed in national and international Despite the significant mortality rate, a significant proportion of patients survives after HSCT and demands 6 Rev. Latino-Am. Enfermagem 2018;26:e3065. 6 the changes that occur during the first year after the patient undergoes autologous and allogenic HSCT at the different stages of treatment, highlighting the most intense symptoms. The results of this research showed that the overall QOL measured by the QLQ-C30 presented lower mean values in the pancytopenia stage when compared to the pre-HSCT stage; however, there is recovery of these means during the treatment, exceeding the baseline parameters after one year. With the exception of the pre-HSCT stage, in all others, the autologous group presented slightly higher means than the allogenic ones. After one year of HSCT, the means presented above 70 points, indicating a satisfactory performance. support, as well as care at different stages of treatment. In this context, health professionals must know the profile of these patients and the changes they suffer in their QoL during the course of the therapeutic process. With regard to sociodemographic analysis, the mean age of the study was 36 years, with extremes between 18 and 69 years. This finding coincides with those found in a study conducted in São Paulo, Brazil, with 71 adult patients submitted to HSCT, whose mean age was 37 years, ranging from 18 to 65 years(24). It is important to emphasize that the mean age of the present study belongs to the age group that would normally be at the peak of one’s productive life, inserted in the labor market. Added to this is the fact that 64% of the patients declared themselves to be economically active. This makes treatment an additional concern, since many of them are family providers and, because of their health condition, need to interrupt their professional activities, reducing their family income in a moment of fragility, carrying with them the uncertainty of the future, translated into feelings of fear and anguish. Discussion Likewise, the general QoL verified by the FACT- BMT presented means above 90 points in all stages of the research, in both groups, exceeding the pre- HSCT parameters after one year, suggesting that the patients consider their general health status as good. It is emphasized that the gradual recovery of the means throughout the therapeutic process and the exceedance of the baseline parameters demonstrate that the overall QoL improves after one year of HSCT. In relation to gender, the results of the research showed a small prevalence of males, 53% (n = 29). Corroborating with this result, the statistics of the occurrence of hematological cancer in the Brazilian population in the biannual report 2016-2017 published by INCA estimated that 54% of the cases may occur among men(1). Men and women suffer from the impact of the diagnosis and the fear of undergoing aggressive and life-threatening treatment. They share the same feelings and concerns about their children or the possibility of infertility. They are uncomfortable with the frailty, both physical and emotional, arising from the treatment and dependence caused by the disease. These factors undoubtedly interfere negatively with their QoL. These results are corroborated by similar studies conducted at transplantation centers in Germany, Denmark, Finland, Norway, Sweden, Canada, Taiwan and the United States, using QLQ-C30 before and after transplantation, which found that the mean overall QoL becomes reduced during hospitalization; however, they recover at baseline parameter up to one year after transplantation(30). A study carried out in Spain evaluated the QoL of patients submitted to autologous and allogenic HSCT and revealed that the overall QoL is worse at two months compared to baseline; however, it improves at nine months after HSCT, confirming a positive evolution throughout the therapeutic process. This leads to conclude that the type of HSCT did not influence the QoL in the evaluated period. Nevertheless, a significant difference between the groups was observed in the physical scale(28). In addition, the survey revealed that 55% (n = 30) are married or declared to be in a stable union, data confirmed by similar studies conducted in Brazil and abroad(8,25-27). The presence of a spouse or partner can serve as an emotional support to the patient who undergoes transplantation from the time of diagnosis and at different stages of treatment, especially during hospitalization, when social isolation is relatively long. www.eerp.usp.br/rlae Discussion The physical function, personal performance and social role domains measured by the QLQ-C30, as well as the physical well-being, functional well-being and additional concerns domains assessed by FACT- BMT, presented lower averages during hospitalization, especially in pancytopenia stage, but there was recovery of baseline values after one year. The comparison of autologous and allogenic groups reveals that the means of the autologous group are slightly higher to the allogenic, but without statistically significant differences. Regarding the type of transplantation, the results of the present study differ from results found in studies conducted in Spain and India, in which 59% and 70% of HSCTs, respectively, were autologous(28-29). Importantly, the place where the research was carried out is considered a world reference transplantation center also due to the amount of allogenic HSCTs that it performs annually. The improvement in these domains is an expected result, since in the hospitalization period, mainly in The domains of QoL measured by the QLQ-C30 and FACT-BMT instruments made it possible to identify 7 Marques ACB, Szczepanik AP, Machado CAM, Santos PND, Guimarães PRB, Kalinke LP. pancytopenia, the patient experiences critical moments of the treatment, when complications can occur that put their life at risk or negatively interfere with their QoL. However, after the hospitalization period, it is expected that they will recover and resume their life and the social life they had before starting treatment. For authors(30), the pre-HSCT means are reached between seven and 12 months after transplantation. A study conducted in Turkey with 82 adult patients submitted to HSCT showed that fatigue, pain and loss of appetite become exacerbated after transplantation and emphasized that patients should be evaluated individually in all domains of QoL, as the results of this evaluation will serve as a basis for nursing to put in place effective interventions that will assist patients in coping with these symptoms(19). In the symptom scale, all items presented the highest mean values in the pancytopenia stage, both in the autologous and in the allogenic groups, evidencing a significant increase of the symptoms evaluated in this critical stage of the treatment. In relation to fatigue, there was an increase of more than 25 points in the means of this period when compared with pre-HSCT stage. Discussion These means decreased during the first year after HSCT, but remained higher than those at the beginning of treatment, suggesting that this symptom remains, although less pronounced, after one year. It is noteworthy that in the first months the patient suffers from the residual effects of drugs received, since the conditioning phase and many of these have side effects or adverse effects with potential to increase the symptoms, thus impacting the QoL. The results of the symptoms scale offer indications of changes in the domains and in which stages of treatment these changes occur, enabling the nursing team to carry out a plan of care with actions aimed at the reduction of the symptoms, minimizing the negative effects on QoL. The general FACTG assessment encompassing the domains of physical well-being, functional well-being, social and family well-being and emotional well-being presented high means, except in the pancytopenia stage, ranging from 74.2 to 84.3 in the autologous group and 71.5 to 81.3 in the allogenic group. These results reflect the good performance and satisfaction of patients during the first year after transplantation. The analysis of the results did not confirm significant differences between the groups. As for the loss of appetite symptom, it was more intense in the autologous group only in the pre-HSCT stage; in all other stages, the means of the allogenic group were higher, with a significant difference between the groups after100 days (p = 0.031). National and international studies highlight the loss of appetite as one of the most present symptoms during HSCT; it has a substantial increase during hospitalization and remains a problem for up to six months after HSCT(4,8,19,30). This symptom must be early detected by the care team in order to implement actions that prevent a possible malnutrition, which may compromise the health and the QoL of this patient. The domains of QoL measured in the present study made it possible to identify the changes that occur throughout the therapeutic process. The importance of this evaluation is emphasized by authors(32), when they state that in order to support these patients, one should look closely at their satisfaction with life through the domains of QoL. They also add that the evaluation of QoL has been increasingly common in the studies, since it allows the professionals to understand and identify patients’ needs more clearly and to approach them comprehensively. Discussion The authors(33) reinforce that changes in questionnaire scores may indicate changes in care needs. Despite clinical differences, patients submitted to autologous and allogenic HSCT present similar changes in the domains of QoL(34). This reveals that, despite the fact that HSCT is a rigorous and aggressive treatment that predisposes the patient to a broad spectrum of complications, surviving patients eventually consider their quality of life to be good, gradually regaining their life routine. The nursing team should be as close as possible to the patient, elucidating their doubts, conducting guidelines and also trying to involve the family. They must keep a close eye on the signs and symptoms that impact QoL and try to reverse the picture. Another recurrent symptom in patients undergoing HSCT is pain. In this study, the symptom intensified in the pancytopenia stage in both groups, but the allogenic group presented a significant difference (p = 0.004) in relation to the autologous group. This differs from a study carried out in the United States in which there was no significant difference between the types of HSCT(31). Throughout the therapeutic process, some clinical conditions are frequent, such as mucositis, and can provoke this symptom causing discomfort and suffering, thus harming QoL. Pain in HSCT should be early detected and managed, both with pharmacological and non-pharmacological measures. The care team must keep in mind the factors that may be contributing or potentiating this symptom and offer comfort measures, thus contributing to a better QoL. www.eerp.usp.br/rlae Conclusion Quality of life in hematologic oncology patients undergoing chemotherapy. Rev Esc Enferm USP. [Internet]. 2013 Apr [cited 2015 May 15]; 47(2):355-61. Available from: http://www.scielo.br/pdf/reeusp/v47n2/en_12.pdf 9. Almeida MAB de, Gutierrez GL, Marques R. Qualidade de vida: definições, conceitos e interfaces com outras áreas de pesquisa. São Paulo: Escola de Artes, Ciências e Humanidades – EACH/USP, 2012. [Acesso 15 jun 2015]. 142p. Disponível em: http://each.uspnet.usp.br/ edicoes-each/qualidade_vida.pdf However, the lack of both national and international studies on this subject, mainly comparing QoL between autologous and allogenic transplantations, became a limiting condition of this study, since it made it difficult to compare with different results in the literature. However, this research may direct new studies on QoL in oncology, indicating other aspects of the theme. 10. Mollevi-Bascoul C, Castan F, Azria D, Gourgou- Bourgade S. EORTC QLQ-C30 descriptive analysis with the qlqc30 command. Stata J. [Internet]. 2015 [cited 2015 Mar 13]; 15(4):1060-74. Available from: https:// ideas.repec.org/a/tsj/stataj/v15y2015i4p1060-1074. html Conclusion The diagnosis of hematologic cancer alone is already a stressful factor for the patient and his/her Rev. Latino-Am. Enfermagem 2018;26:e3065. 8 8 3. Guimarães RCR, Gonçalves RPF, Limas CA, Torres MR, Silva CSO e. Nursing actions facing reactions to chemotherapy in oncological patients. J Res Fundam Care Online. [Internet]. 2015 Abr/Jun [cited 2016 Jun 27];7(2):2440-52. Available from: http://dx.doi. org/10.9789/2175-5361.2015.v7i2.2440-2552 4. Públio GB, Silva KO, Viana GFS. Qualidade de vida de pacientes oncológicos submetidos à quimioterapia. C&D Rev Eletr Fainor. [Internet]. 2014 Jul/Dec [Acesso 30 jul 2017]; 7(2):244-57. Available from: http://srv02. fainor.com.br/revista237/index.php/memorias/article/ view/329/209 3. Guimarães RCR, Gonçalves RPF, Limas CA, Torres MR, Silva CSO e. Nursing actions facing reactions to chemotherapy in oncological patients. J Res Fundam Care Online. [Internet]. 2015 Abr/Jun [cited 2016 Jun 27];7(2):2440-52. Available from: http://dx.doi. org/10.9789/2175-5361.2015.v7i2.2440-2552 3. Guimarães RCR, Gonçalves RPF, Limas CA, Torres MR, Silva CSO e. Nursing actions facing reactions to chemotherapy in oncological patients. J Res Fundam Care Online. [Internet]. 2015 Abr/Jun [cited 2016 Jun 27];7(2):2440-52. Available from: http://dx.doi. org/10.9789/2175-5361.2015.v7i2.2440-2552 family. Besides this, the treatment to which they will be submitted has numerous risks, including death. The feelings involved are very intense for both the patient and the family members who, in the early stages of treatment, are still shocked due to the load of worries. This research made it possible to highlight the changes in the domains of QoL and some symptoms of patients with hematological cancer submitted to HSCT during the first year of treatment. These changes can serve as indicators of the overall satisfaction of patients with their own life, as well as guide the nursing actions in a more specific and individualized manner. 4. Públio GB, Silva KO, Viana GFS. Qualidade de vida de pacientes oncológicos submetidos à quimioterapia. C&D Rev Eletr Fainor. [Internet]. 2014 Jul/Dec [Acesso 30 jul 2017]; 7(2):244-57. Available from: http://srv02. fainor.com.br/revista237/index.php/memorias/article/ view/329/209 fainor.com.br/revista237/index.php/memorias/article/ view/329/209 5. European Organization For Research And Treatment Of Cancer (EORTC). [Internet]. Bélgica: EORTC; 2015 [cited 2015 Jun 23]. Available from:http://www.eortc.be 6. Matias AB, de Oliveira-Cardoso EA, Mastropietro AP, Voltarelli JC, dos Santos MA. Quality of life following allogeneic, hematopoietic stem cell transplantation: a longitudinal study. Estud Psicol. [Internet]. 2011 Apr/ June [cited 2016 Jun 27];28(2):187-97. Available from: http://dx.doi.org/10.1590/S0103-166X2011000200006 7. Sawada NO, Nicolussi AC, Okino L, Cardozo FMC, Zago MMF. Quality of life evaluation in cancer patients to submitted to chemotherapy. Rev Esc Enferm USP. [Internet]. Conclusion 2009 Sept [cited 2015 Jun 23]; 43(3):578- 84. Available from:http://www.scielo.br/pdf/reeusp/ v43n3/en_a12v43n3.pdf 8. Andrade V, Sawada NO, Barichello E. Quality of life in hematologic oncology patients undergoing chemotherapy. Rev Esc Enferm USP. [Internet]. 2013 Apr [cited 2015 May 15]; 47(2):355-61. Available from: http://www.scielo.br/pdf/reeusp/v47n2/en_12.pdf 9. Almeida MAB de, Gutierrez GL, Marques R. Qualidade de vida: definições, conceitos e interfaces com outras áreas de pesquisa. São Paulo: Escola de Artes, Ciências e Humanidades – EACH/USP, 2012. [Acesso 15 jun 2015]. 142p. Disponível em: http://each.uspnet.usp.br/ edicoes-each/qualidade_vida.pdf 10. Mollevi-Bascoul C, Castan F, Azria D, Gourgou- Bourgade S. EORTC QLQ-C30 descriptive analysis with the qlqc30 command. Stata J. [Internet]. 2015 [cited 2015 Mar 13]; 15(4):1060-74. Available from: https:// ideas.repec.org/a/tsj/stataj/v15y2015i4p1060-1074. html 11. World Health Organization (WHO). WHOQOL 5. European Organization For Research And Treatment Of Cancer (EORTC). [Internet]. Bélgica: EORTC; 2015 [cited 2015 Jun 23]. Available from:http://www.eortc.be 6. Matias AB, de Oliveira-Cardoso EA, Mastropietro AP, Voltarelli JC, dos Santos MA. Quality of life following allogeneic, hematopoietic stem cell transplantation: a longitudinal study. Estud Psicol. [Internet]. 2011 Apr/ June [cited 2016 Jun 27];28(2):187-97. Available from: http://dx.doi.org/10.1590/S0103-166X2011000200006 7. Sawada NO, Nicolussi AC, Okino L, Cardozo FMC, Zago MMF. Quality of life evaluation in cancer patients to submitted to chemotherapy. Rev Esc Enferm USP. [Internet]. 2009 Sept [cited 2015 Jun 23]; 43(3):578- 84. Available from:http://www.scielo.br/pdf/reeusp/ v43n3/en_a12v43n3.pdf Among the affected domains measured by QLQ-C30, the symptoms of pain in the pancytopenia stage and loss of appetite after 100 days, both in the group of patients submitted to allogenic HSCT, were statistically significant. Regarding FACT-BMT, the statistical analysis showed a significant result in the functional well-being domain after100 days, in the group of patients submitted to autologous HSCT. These results express changes in the QoL of these patients, revealing impairment in these domains. Despite the aggressiveness of the treatment, these findings demonstrate that patients undergoing both modalities of HSCT consider their overall quality of life to be relatively good throughout the first year of the therapeutic process. The findings of the present research corroborate studies conducted in Brazil and abroad, serving as indicators on changes in the domains of QoL, since they signal to the care team, especially to nurses, which aspects are bothering the patient and in which stage of the treatment this discomfort occurs. Thus, it allows for more specific and individualized interventions, minimizing impairments in QoL and contributing to a better survival. 8. Andrade V, Sawada NO, Barichello E. References 1. Ministério da Saúde (BR), Instituto Nacional de Câncer José de Alencar Gomes da Silva (INCA). Estimativa 2016: incidência de câncer no Brasil. [Internet]. Rio de Janeiro, RJ; 2016. [Acesso 30 maio 2016]. Available from: http://www.inca.gov.br/bvscontrolecancer/ publicacoes/edicao/Estimativa_2016.pdf 2. WHO. International Agency for Research on Cancer (IARC). [Internet]. [cited 2017 May 18]. Available from: http://globocan.iarc.fr/Pages/burden_sel.aspx 11. World Health Organization (WHO). WHOQOL Measuring quality of life. The world health organization quality of life instruments (The WHOQOL 100 and the WHOQOL Bref). Geneva, 1997. [cited 2015 May 29]. Available from: http://www.who.int/mental_health/ media/68.pdf 12. Rede Brasileira de Transplantes (RBT). Ano XXII nº 12. Rede Brasileira de Transplantes (RBT). Ano XXII nº 1. Dados numéricos da doação de órgãos e transplantes realizados por estado e instituição no período Janeiro- 12. Rede Brasileira de Transplantes (RBT). Ano XXII nº 1. Dados numéricos da doação de órgãos e transplantes realizados por estado e instituição no período Janeiro- 1. Dados numéricos da doação de órgãos e transplantes realizados por estado e instituição no período Janeiro- www.eerp.usp.br/rlae 9 9 Marques ACB, Szczepanik AP, Machado CAM, Santos PND, Guimarães PRB, Kalinke LP. from: http://ehp.sagepub.com/content/28/2/212.full. pdf março de 2017. São Paulo, 2017. [Acesso 18 maio 2017]. Disponível em: http://www.abto.org.br/abtov03/ Upload/file/RBT/2017/2017_leitura_1T.pdf março de 2017. São Paulo, 2017. [Acesso 18 maio 2017]. Disponível em: http://www.abto.org.br/abtov03/ Upload/file/RBT/2017/2017_leitura_1T.pdf 22. Silva JB, Póvoa VCO, Lima MHM, Oliveira HC, Padilha KG, Secoli SR. Nursing workload in hematopoietic stem cell transplantation: a cohort study. Rev Esc Enferm USP. [Internet]. 2015 Jul/Aug [cited 2016 Jul 08]; 49(Esp):93-100. Available from: http://dx.doi. org/10.1590/S0080-623420150000700014 13. Sociedade Brasileira de Transplante de Medula Óssea (SBTMO). Notícias. [Internet] [Acesso 18 maio 2017]. Disponível em: http://www.sbtmo.org.br/noticia. php?id=417 14. Wingard JR, Majhail NS, Brazauskas R, Wang Z, Sobocinski KA, Jacobsohn D, et al. Long-term survival and late deaths after allogeneic hematopoietic cell transplantation. J Clin Oncol. [Internet] 2011 [cited 2015 Aug 9];29(16):2230-9. Available from:https:// www.ncbi.nlm.nih.gov/pubmed/21464398 23. Shokouhi S, Brays S, Bakhtiyari S, Sayehmiri K, Alimoghadam K, Ghavamzadeh A. Effects of aGVHD and cGVHD on survival rate in patients with acute myeloid leukemia after allogeneic stem cell transplantation. Int J Hematol Oncol Stem Cell Res. [Internet]. 2015 Jul [cited 2016 Set 12]; 9(3):112–21. Available from: https:// www.ncbi.nlm.nih.gov/pmc/articles/PMC4529677/pdf/ IJHOSCR-9-112.pdf 15. Li WH, Sykes M. Emerging concepts in haematopoietic cell transplantation. Nat Rev. Immunology. [Internet]. 2012 May [cited 2015 Sept 23];12(6):403-16. Available from: http://www.ncbi.nlm.nih.gov/pmc/articles/ PMC4006975/pdf/nihms-574294.pdf 24. Rodrigues HF, Garbin LM, Castanhol LEC, Simões BP, Curcioli ACJV, Silveira RCCP. www.eerp.usp.br/rlae References Cateter de Hickman no transplante de células-tronco hematopoéticas: implante cirúrgico, retirada e assistência de enfermagem. Rev Enferm UERJ. [Internet]. 2015 My/Jun [Acesso 12 maio 2017]; 23(3):304-9. Disponível em: http://www.e- publicacoes.uerj.br/index.php/enfermagemuerj/article/ view/4995/13766 16. Sureda A, Bader P, Cesaro S, Dreger P, Duarte RF, Dufour C, et al. Indications for allo- and auto-SCT for hematological diseases, solid tumours and immune disorders: current practice in Europe, 2015. Bone Marrow Transplant. [Internet]. 2015 Aug [cited 2016 Sept 2]; 50(8):1037-56. Available from: http://www. nature com/bmt/journal/v50/n8/pdf/bmt20156a pdf 16. Sureda A, Bader P, Cesaro S, Dreger P, Duarte RF, Dufour C, et al. Indications for allo- and auto-SCT for hematological diseases, solid tumours and immune disorders: current practice in Europe, 2015. Bone Marrow Transplant. [Internet]. 2015 Aug [cited 2016 Sept 2]; 50(8):1037-56. Available from: http://www. nature.com/bmt/journal/v50/n8/pdf/bmt20156a.pdf 17. DePalo J, Chai X, Lee SJ, Cutler CS, Treister N. Assessing the relationship between oral chronic graft-versus-host disease and global measures of quality of life. Oral Oncol. [Internet]. 2015 Oct [cited 2016 Sept 5];51(10):944-9. Available from: http://www.sciencedirect.com/science/article/pii/ S1368837515002882 25. Cohen MZ, Rozmus CL, Mendoza TR, Padhye NS, Neumann J, Gning I, et al. Symptoms and quality of life in diverse patients undergoing hematopoietic stem cell transplantation. J Pain Symptom Manages. [Internet]. 2012 Aug [cited 2015 Oct 11];44(2):168- 80. Available from: https://dx.doi.org/10.1016/j. jpainsymman.2011.08.011 nature.com/bmt/journal/v50/n8/pdf/bmt20156a.pdf 17. DePalo J, Chai X, Lee SJ, Cutler CS, Treister N. Assessing the relationship between oral chronic graft-versus-host disease and global measures of quality of life. Oral Oncol. [Internet]. 2015 Oct [cited 2016 Sept 5];51(10):944-9. Available from: http://www.sciencedirect.com/science/article/pii/ S1368837515002882 26. Ferreira Júnior MA, Ivo ML, Pontes ERJC, Carvalho DPSRP, Vitor AF. Perfil sociodemográfico e clínico de portadores de síndromes mielodisplásicas. Biosci J. [Internet]. 2014 Jul/Aug [Acesso 21 ago 2015];30(4):1269-77. Disponível em: http://www. seer.ufu.br/index.php/biosciencejournal/article/ view/23749/14779 18. Ikeda ALC, Rosa LM da, Cruz FJB da. Care nursing in the collection of hematopoietic stem cells by apheresis. J Nurs UFPE. on line [Internet]. 2015 Mar [cited 2017 May 18]; 9(2):7039-45. Available from: http://www.revista. ufpe.br/revistaenfermagem/index.php/revista/article/ view/7104/pdf_7367 27. Braamse AM, van Meijel B, Visser O, Huijgens PC, Beekman AT, Dekker J. Distress, problems and supportive care needs of patients treated with auto- orallo-SCT. Bone Marrow Transplant. [Internet]. 2014 Feb [cited 2016 jul 21]; 49(2):292-8. Available from: https://dx.doi.org/10.1038/bmt.2013 19. Ovayolu O, Ovayolu N, Kaplan E, Pehlivan M, Karadag G. Symptoms and quality of life before and after stem cell transplantation in Cancer. Pak J Med Sci. [Internet]. Corresponding Author: Luciana Puchalski Kalinke E-mail: lucianakalinke@yahoo.com.br https://orcid.org/0000-0003-4868-8193 Copyright © 2018 Revista Latino-Americana de Enfermagem This is an Open Access article distributed under the terms of the Creative Commons (CC BY). This license lets others distribute, remix, tweak, and build upon your work, even commercially, as long as they credit you for the original creation. This is the most accommodating of licenses offered. Recommended for maximum dissemination and use of licensed materials. References 2014 May/Jun [cited 2016 Aug 29]; 29(3):803-8. Available from: http://dx.doi.org/10.12669/pjms.293.3290 20. Fayers PM, Aaronson NK, Bjordal K, Groenvold M, Curan D, Bottoomley. On behalfofthe EORTC Qualityof Life Group. The EORTC QLQ-C30 Scoring Manual. 3ªed. Brussels (BE): European Organization for Research and Treatment of Cancer; 2001. 28. Seixas MR, Rodríguez LL, Fernández JMP, Moncada MV, Quijano-Campos JC. Calidad de vida relacionada com La salud en pacientes com trasplante de progenitores hematopoyéticos. Index Enferm. [Internet]. 2014 Oct/Dec [Acceso 2 ago 2015]; 23(4). Disponible en: http://scielo.isciii.es/scielo.php?script=sci_arttext&pid =S1132-12962014000300004 21. Eremenco SL, Cella D, Arnold BJ. A comprehensive method for the translation and cross-cultural validation of health status questionnaires. Eval Health Prof. [Internet]. 2005 Jun [cited 2015 Jan 23]; 28(2):212-32. Available 21. Eremenco SL, Cella D, Arnold BJ. A comprehensive method for the translation and cross-cultural validation of health status questionnaires. Eval Health Prof. [Internet]. 2005 Jun [cited 2015 Jan 23]; 28(2):212-32. Available 29. Shah CA, Karanwal A, Desai M, Pandya M, Shah R, Shah R. Hematopoietic stem-cell transplantation in the 29. Shah CA, Karanwal A, Desai M, Pandya M, Shah R, Shah R. Hematopoietic stem-cell transplantation in the 10 Rev. Latino-Am. Enfermagem 2018;26:e3065. developing world: experience from a center in western Índia. J Oncol. [Internet]. 2015 [cited 2016 May 4]; 2015. Available from: http://dx.doi.org/10.1155/2015/710543 30. Grulke N, Albani C, Bailer H. Quality of life in patients before and after haematopoietic stem cell transplantation measured with the European Organization for Research and Treatment of Cancer (EORTC) Quality of Life Core Questionnaire QLQ-C30. Bone Marrow Transplant. [Internet]. 2012 [cited 2015 Sep 1];47(4):473- 82. Available from: https://dx.doi.org/10.1038/ bmt.2011.107 31. El-Jawahri A.; Traeger LN, Kuzmuk K, Eusebio JR, Vandusen HB, Shin JA, et al. Quality of life and mood of patients and family caregivers during hospitalization for hematopoietic stem cell transplantation. Cancer. [Internet]. 2015 Mar [cited 2016 Aug 8]; 121(6):951-59. Available from: https://dx.doi.org/10.1002/cncr.29149 32. Amler S, Sauerland MC, Deiters C, Büchner T, Schumacher A. Factors influencing life satisfaction in acute myeloid leukemia survivors following allogeneic stem cell transplantation: a cross-sectional study. Health Qual Life Outcomes. [Internet]. 2015 Feb [2016 Feb 4];13(28). Available from: https://www.ncbi.nlm.nih. gov/pmc/articles/PMC4349480/ 33. Snyder CF, Blackford AL, Sussman J, Bainbridge D, Howell D, Seow HY, et al. Identifying Changes in Scores on the EORTC-QLQ-C30 Representing a Change in Patients’ Supportive Care Needs. Qual Life Res. [Internet]. 2015 May [cited 2016 Sep 20]; 24(5):1207– 16. Available from: https://www.ncbi.nlm.nih.gov/ pubmed/25398495 34. www.eerp.usp.br/rlae References García CM, Mumby PB, Thilges S, Stiff PJ. Comparison of early quality of life outcomes in autologous and allogeneic transplant patients. Bone marrow transplant. [Internet]. 2012 Dec [cited 2015 Sep 1]; 47(12):1577- 82. Available from: https://dx.doi.org/10.1038/ bmt.2012.77 Received: Sep 29th 2017 Accepted: Ago 13th 2018 Received: Sep 29th 2017 Accepted: Ago 13th 2018 Received: Sep 29th 2017 Accepted: Ago 13th 2018 www.eerp.usp.br/rlae
https://openalex.org/W4253072262	https://peerj.com/articles/9020v0.5/submission	English	null	Peer Review #2 of "Characterization of biofilm production in different strains of Acinetobacter baumannii and the effects of chemical compounds on biofilm formation (v0.1)"	null	2,020	cc-by	14,144	Manuscript to be reviewed Manuscript to be reviewed Characterization of bioﬁlm production in diﬀerent strains of Acinetobacter baumannii and the eﬀects of chemical compounds on bioﬁlm formation Ming-Feng Lin Corresp., 1 , Yun-You Lin 2 , Chung-Yu Lan 2, 3 Ming-Feng Lin Corresp., 1 , Yun-You Lin 2 , Chung-Yu Lan 2, 3 52, Chih Shan Road, Chutung Town, Hsinchu Count 18 Email address: allenlmf2007@gmail.com 19 20 Abstract 21 Acinetobacter baumannii, an important emerging pathog 22 for its ability to form biofilms. Biofilm formation increas 23 dry surfaces and may contribute to its persistence in the h 24 the probability of nosocomial infections and outbreaks. T 25 characterize the biofilm production of different strains of 26 chemical compounds, especially antibiotics, on biofilm fo 27 significant relationship was observed between the ability 28 susceptibility of the A. baumannii clinical isolates. Biofi 29 ATCC 17978 after gene knockout of two-component regu 30 genes emrA/emrB and outer membrane coding gene ompA 31 less biofilm formation than the wild-type strain, which w 32 scanning electron microscopy and confocal laser scannin 33 amikacin, colistin, LL-37 or tannic acid decreased the bio 34 In contrast, the addition of lower subinhibitory concentra 35 formation ability of A. baumannii. Minimum biofilm erad 36 imipenem, colistin, and tigecycline were increased obvio 37 resistant clinical strain A. baumannii VGH2. In conclusio 1 2 Characterization of biofilm production in different 3 strains of Acinetobacter baumannii and the effects 4 chemical compounds on biofilm formation 5 6 7 Ming-Feng Lin1, Yun-You Lin2, Chung-Yu Lan2,3 8 9 10 1 Department of Medicine, National Taiwan University Hospital Chu-Tung Branch, Hsin 11 County, Taiwan 12 2 Institute of Molecular and Cellular Biology, National Tsing Hua University, Hsinchu, T 13 3 Department of Life Science, National Tsing Hua University, Hsinchu, Taiwan 14 15 Corresponding Author: 16 Ming-Feng Lin1 17 No. 52, Chih Shan Road, Chutung Town, Hsinchu County 31064, Taiwan. 18 Email address: allenlmf2007@gmail.com 19 20 Abstract 21 Acinetobacter baumannii, an important emerging pathogen of nosocomial infections, is k 22 for its ability to form biofilms. Biofilm formation increases the survival rate of A. bauma 23 dry surfaces and may contribute to its persistence in the hospital environment, which inc 24 the probability of nosocomial infections and outbreaks. This study was undertaken to 25 characterize the biofilm production of different strains of A. baumannii and the effects of 26 chemical compounds, especially antibiotics, on biofilm formation. In this study, no statis 27 significant relationship was observed between the ability to form a biofilm and the antim 28 susceptibility of the A. baumannii clinical isolates. Biofilm formation caused by A. Ming-Feng Lin Corresp., 1 , Yun-You Lin 2 , Chung-Yu Lan 2, 3 baum 29 ATCC 17978 after gene knockout of two-component regulatory system gene baeR, efflu 30 genes emrA/emrB and outer membrane coding gene ompA revealed that all mutant strain 31 less biofilm formation than the wild-type strain, which was further supported by the imag 32 scanning electron microscopy and confocal laser scanning microscopy. The addition of 33 amikacin, colistin, LL-37 or tannic acid decreased the biofilm formation ability of A. bau 34 In contrast, the addition of lower subinhibitory concentration tigecycline increased the bi 35 formation ability of A. baumannii. Minimum biofilm eradication concentrations of amika 36 imipenem, colistin, and tigecycline were increased obviously for both wild type and mul 37 resistant clinical strain A. baumannii VGH2. In conclusion, the biofilm formation ability 1 2 Characterization of biofilm production in different 3 strains of Acinetobacter baumannii and the effects of 4 chemical compounds on biofilm formation 5 6 7 Ming-Feng Lin1, Yun-You Lin2, Chung-Yu Lan2,3 8 9 10 1 Department of Medicine, National Taiwan University Hospital Chu-Tung Branch, Hsinchu 11 County, Taiwan 12 2 Institute of Molecular and Cellular Biology, National Tsing Hua University, Hsinchu, Taiwan 13 3 Department of Life Science, National Tsing Hua University, Hsinchu, Taiwan 14 15 Corresponding Author: 16 Ming-Feng Lin1 17 No. 52, Chih Shan Road, Chutung Town, Hsinchu County 31064, Taiwan. 18 Email address: allenlmf2007@gmail.com 19 20 Abstract 21 Acinetobacter baumannii, an important emerging pathogen of nosocomial infections, is known 22 for its ability to form biofilms. Biofilm formation increases the survival rate of A. baumannii on 23 dry surfaces and may contribute to its persistence in the hospital environment, which increases 24 the probability of nosocomial infections and outbreaks. This study was undertaken to 25 characterize the biofilm production of different strains of A. baumannii and the effects of 26 chemical compounds, especially antibiotics, on biofilm formation. In this study, no statistically 27 significant relationship was observed between the ability to form a biofilm and the antimicrobial 28 susceptibility of the A. baumannii clinical isolates. Biofilm formation caused by A. baumannii Ming-Feng Lin Corresp., 1 , Yun-You Lin 2 , Chung-Yu Lan 2, 3 1 Department of Medicine, National Taiwan University Hospital Chu-Tung Branch, Hsinchu County, Taiwan 2 Institute of Molecular and Cellular Biology, National Tsing Hua University, Hsinchu, Taiwan 3 Department of Life Science, National Tsing Hua University, Hsinchu, Taiwan Corresponding Author: Ming-Feng Lin Email address: allenlmf2007@gmail.com 1 Department of Medicine, National Taiwan University Hospital Chu-Tung Branch, Hsinchu County, Taiwan 2 Institute of Molecular and Cellular Biology, National Tsing Hua University, Hsinchu, Taiwan 3 Department of Life Science, National Tsing Hua University, Hsinchu, Taiwan Acinetobacter baumannii, an important emerging pathogen of nosocomial infections, is known for its ability to form bioﬁlms. Bioﬁlm formation increases the survival rate of A. baumannii on dry surfaces and may contribute to its persistence in the hospital environment, which increases the probability of nosocomial infections and outbreaks. This study was undertaken to characterize the bioﬁlm production of diﬀerent strains of A. baumannii and the eﬀects of chemical compounds, especially antibiotics, on bioﬁlm formation. In this study, no statistically signiﬁcant relationship was observed between the ability to form a bioﬁlm and the antimicrobial susceptibility of the A. baumannii clinical isolates. Bioﬁlm formation caused by A. baumannii ATCC 17978 after gene knockout of two-component regulatory system gene baeR, eﬄux pump genes emrA/emrB and outer membrane coding gene ompA revealed that all mutant strains had less bioﬁlm formation than the wild-type strain, which was further supported by the images from scanning electron microscopy and confocal laser scanning microscopy. The addition of amikacin, colistin, LL-37 or tannic acid decreased the bioﬁlm formation ability of A. baumannii. In contrast, the addition of lower subinhibitory concentration tigecycline increased the bioﬁlm formation ability of A. baumannii. Minimum bioﬁlm eradication concentrations of amikacin, imipenem, colistin, and tigecycline were increased obviously for both wild type and multidrug resistant clinical strain A. baumannii VGH2. In conclusion, the bioﬁlm formation ability of A. baumannii varied in diﬀerent strains, involved many genes and could be inﬂuenced by many chemical compounds. PeerJ reviewing PDF \| (2019:10:42050:4:0:NEW 20 Mar 2020) Manuscript to be reviewed 1 2 Characterization of biofilm prod 3 strains of Acinetobacter bauman 4 chemical compounds on biofilm 5 6 7 Ming-Feng Lin1, Yun-You Lin2, C 8 9 10 1 Department of Medicine, National Taiwan University H 11 County, Taiwan 12 2 Institute of Molecular and Cellular Biology, National T 13 3 Department of Life Science, National Tsing Hua Unive 14 15 Corresponding Author: 16 Ming-Feng Lin1 17 No. Manuscript to be reviewed 38 baumannii varied in different strains, involved many genes and could be influenced by many 39 chemical compounds. 40 41 Introduction 42 A. baumannii, as an important emerging pathogen of nosocomial infection, is known for its 43 ability to form biofilms (Longo, Vuotto & Donelli 2014). Biofilm formation increases the 44 survival rate of A. baumannii on dry surfaces and may contribute to its persistence in the hospital 45 environment, increasing the probability of causing nosocomial infections and outbreaks (Espinal, 46 Marti & Vila 2012). 47 The mechanisms to explain the increased drug resistance of bacteria related with biofilms are 48 various and at least included delayed penetration of the antimicrobial agents into the biofilm and 49 reduced growth rate of the microorganisms within the biofilm (Donlan 2000). A positive 50 correlation between biofilm formation and antimicrobial resistance in Acinetobacter baumannii 51 has been confirmed (Badave & Kulkarni 2015), although one study suggested an inverse 52 relationship between biofilm production and meropenem resistance in nosocomial A. baumannii 53 isolates (Perez 2015). The ability to form a biofilm may affect antibiotic susceptibility and 54 clinical failure, even when the dose administered is in the susceptible range (Kim et al. 2015). 55 Modulation of biofilm formation is as diverse as the surface on which the bacteria reside and 56 the cellular components involved in the programmed multi-step process (Gaddy & Actis 2009). 57 The regulation of biofilm formation is influenced by sensing bacterial cell density, the presence 58 of different nutrients and the concentration of cations, which was through the two-component 59 regulatory system BfmRS (Luo et al. 2015a). This transcriptional regulatory system activates the 60 expression of the CsuA/BABCDE usher-chaperone assembly system responsible for the 61 production of pili, which are needed for cell attachment and biofilm formation on polystyrene 62 surfaces (Tomaras et al. 2008). Both biofilm-associated proteins and OmpA protein play a role in 63 biofilm formation in A. baumannii (Loehfelm, Luke & Campagnari 2008; Gaddy, Tomaras & 64 Actis 2009). Together, the regulation of biofilm formation is multifactorial, and many molecular 65 mechanisms remain unidentified. 66 Although biofilm formation in bacteria has been studied extensively (Kalkanci & TunÇCan 67 2019) and described in A. baumannii previously (Kim et al. 2015), the characteristics of biofilm 68 production in clinical isolates from hospitals in Taiwan have seldom been well characterized. 69 Therefore, we used the A. 20 Abstract 21 Acinetobacter baumannii, an important emerging pathogen of nosocomial infections, is known 22 for its ability to form biofilms. Biofilm formation increases the survival rate of A. baumannii on 23 dry surfaces and may contribute to its persistence in the hospital environment, which increases 24 the probability of nosocomial infections and outbreaks. This study was undertaken to 25 characterize the biofilm production of different strains of A. baumannii and the effects of 26 chemical compounds, especially antibiotics, on biofilm formation. In this study, no statistically 27 significant relationship was observed between the ability to form a biofilm and the antimicrobial 28 susceptibility of the A. baumannii clinical isolates. Biofilm formation caused by A. baumannii 29 ATCC 17978 after gene knockout of two-component regulatory system gene baeR, efflux pump 30 genes emrA/emrB and outer membrane coding gene ompA revealed that all mutant strains had 31 less biofilm formation than the wild-type strain, which was further supported by the images from 32 scanning electron microscopy and confocal laser scanning microscopy. The addition of 33 amikacin, colistin, LL-37 or tannic acid decreased the biofilm formation ability of A. baumannii. 34 In contrast, the addition of lower subinhibitory concentration tigecycline increased the biofilm 35 formation ability of A. baumannii. Minimum biofilm eradication concentrations of amikacin, 36 imipenem, colistin, and tigecycline were increased obviously for both wild type and multidrug 37 resistant clinical strain A. baumannii VGH2. In conclusion, the biofilm formation ability of A. PeerJ reviewing PDF \| (2019:10:42050:4:0:NEW 20 Mar 2020) 41 Introduction 42 A. baumannii, as an important emerging pathogen of nosocomial infection, is known for its 43 ability to form biofilms (Longo, Vuotto & Donelli 2014). Biofilm formation increases the 44 survival rate of A. baumannii on dry surfaces and may contribute to its persistence in the hospital 45 environment, increasing the probability of causing nosocomial infections and outbreaks (Espinal, 46 Marti & Vila 2012). ) 47 The mechanisms to explain the increased drug resistance of bacteria related with biofilms are 48 various and at least included delayed penetration of the antimicrobial agents into the biofilm and 49 reduced growth rate of the microorganisms within the biofilm (Donlan 2000). A positive 50 correlation between biofilm formation and antimicrobial resistance in Acinetobacter baumannii 51 has been confirmed (Badave & Kulkarni 2015), although one study suggested an inverse 52 relationship between biofilm production and meropenem resistance in nosocomial A. baumannii 53 isolates (Perez 2015). The ability to form a biofilm may affect antibiotic susceptibility and 54 clinical failure, even when the dose administered is in the susceptible range (Kim et al. 2015). 55 Modulation of biofilm formation is as diverse as the surface on which the bacteria reside and 56 the cellular components involved in the programmed multi-step process (Gaddy & Actis 2009). 57 The regulation of biofilm formation is influenced by sensing bacterial cell density, the presence 58 of different nutrients and the concentration of cations, which was through the two-component 59 regulatory system BfmRS (Luo et al. 2015a). This transcriptional regulatory system activates the 60 expression of the CsuA/BABCDE usher-chaperone assembly system responsible for the 61 production of pili, which are needed for cell attachment and biofilm formation on polystyrene 62 surfaces (Tomaras et al. 2008). Both biofilm-associated proteins and OmpA protein play a role in 63 biofilm formation in A. baumannii (Loehfelm, Luke & Campagnari 2008; Gaddy, Tomaras & 64 Actis 2009). Together, the regulation of biofilm formation is multifactorial, and many molecular 65 mechanisms remain unidentified. 66 Although biofilm formation in bacteria has been studied extensively (Kalkanci & TunÇCan 67 2019) and described in A. baumannii previously (Kim et al. 2015), the characteristics of biofilm 68 production in clinical isolates from hospitals in Taiwan have seldom been well characterized. 69 Therefore, we used the A. baumannii strains from our previous studies to examine the difference 70 in biofilm formation among these strains with a modified XTT assay. Manuscript to be reviewed baumannii strains from our previous studies to examine the difference 70 in biofilm formation among these strains with a modified XTT assay. In addition, we also closely 71 observed biofilm formation using scanning electron microscopy (SEM) and confocal laser 72 scanning microscopy (CLSM). These experiments were designed to understand the nature of 73 biofilm production among the wild type strain mutant strains and clinical isolates of A 38 baumannii varied in different strains, involve 39 chemical compounds. 40 41 Introduction 42 A. baumannii, as an important emerging patho 43 ability to form biofilms (Longo, Vuotto & Do 44 survival rate of A. baumannii on dry surfaces 45 environment, increasing the probability of cau 46 Marti & Vila 2012). 47 The mechanisms to explain the increased d 48 various and at least included delayed penetrat 49 reduced growth rate of the microorganisms w 50 correlation between biofilm formation and an 51 has been confirmed (Badave & Kulkarni 2015 52 relationship between biofilm production and m 53 isolates (Perez 2015). The ability to form a bi 54 clinical failure, even when the dose administe 55 Modulation of biofilm formation is as dive 56 the cellular components involved in the progr 57 The regulation of biofilm formation is influen 58 of different nutrients and the concentration of 59 regulatory system BfmRS (Luo et al. 2015a). 60 expression of the CsuA/BABCDE usher-chap 61 production of pili, which are needed for cell a 62 surfaces (Tomaras et al. 2008). Both biofilm- 63 biofilm formation in A. baumannii (Loehfelm 64 Actis 2009). Together, the regulation of biofil 65 mechanisms remain unidentified. 66 Although biofilm formation in bacteria has 67 2019) and described in A. baumannii previou 68 production in clinical isolates from hospitals i 69 Therefore, we used the A. baumannii strains f 70 in biofilm formation among these strains with 71 observed biofilm formation using scanning el 72 scanning microscopy (CLSM). These experim 38 baumannii varied in different strains, involved many genes and could be influenced by many 39 chemical compounds. 40 41 Introduction 42 A. baumannii, as an important emerging pathogen of nosocomial infection, is known for its 43 ability to form biofilms (Longo, Vuotto & Donelli 2014). Biofilm formation increases the 44 survival rate of A. Manuscript to be reviewed baumannii on dry surfaces and may contribute to its persistence in the hospital 45 environment, increasing the probability of causing nosocomial infections and outbreaks (Espinal, 46 Marti & Vila 2012). 47 The mechanisms to explain the increased drug resistance of bacteria related with biofilms are 48 various and at least included delayed penetration of the antimicrobial agents into the biofilm and 49 reduced growth rate of the microorganisms within the biofilm (Donlan 2000). A positive 50 correlation between biofilm formation and antimicrobial resistance in Acinetobacter baumannii 51 has been confirmed (Badave & Kulkarni 2015), although one study suggested an inverse 52 relationship between biofilm production and meropenem resistance in nosocomial A. baumannii 53 isolates (Perez 2015). The ability to form a biofilm may affect antibiotic susceptibility and 54 clinical failure, even when the dose administered is in the susceptible range (Kim et al. 2015). 55 Modulation of biofilm formation is as diverse as the surface on which the bacteria reside and 56 the cellular components involved in the programmed multi-step process (Gaddy & Actis 2009). 57 The regulation of biofilm formation is influenced by sensing bacterial cell density, the presence 58 of different nutrients and the concentration of cations, which was through the two-component 59 regulatory system BfmRS (Luo et al. 2015a). This transcriptional regulatory system activates the 60 expression of the CsuA/BABCDE usher-chaperone assembly system responsible for the 61 production of pili, which are needed for cell attachment and biofilm formation on polystyrene 62 surfaces (Tomaras et al. 2008). Both biofilm-associated proteins and OmpA protein play a role in 63 biofilm formation in A. baumannii (Loehfelm, Luke & Campagnari 2008; Gaddy, Tomaras & 64 Actis 2009). Together, the regulation of biofilm formation is multifactorial, and many molecular 65 mechanisms remain unidentified. 66 Although biofilm formation in bacteria has been studied extensively (Kalkanci & TunÇCan 67 2019) and described in A. baumannii previously (Kim et al. 2015), the characteristics of biofilm 68 production in clinical isolates from hospitals in Taiwan have seldom been well characterized. 69 Therefore, we used the A. baumannii strains from our previous studies to examine the difference 70 in biofilm formation among these strains with a modified XTT assay. In addition, we also closely 71 observed biofilm formation using scanning electron microscopy (SEM) and confocal laser 72 scanning microscopy (CLSM). Manuscript to be reviewed These experiments were designed to understand the nature of 73 biofilm production among the wild-type strain, mutant strains and clinical isolates of A. 38 baumannii varied in different strains, involved many genes and could be influenced by many 39 chemical compounds. 84 Bacterial strains 84 Bacterial strains 85 All of the bacterial strains used in this study were isolates from our previous studies (Lin et al. 86 2014; Lin, Lin & Lan 2015; Lin et al. 2015; Lin, Lin & Lan 2017a; Lin et al. 2017b) or 87 American Type Culture Collection (ATCC), as shown in Table 1. The studied strains contained 88 quality control strain ATCC 19606, ATCC 17978 and 21 clinical isolates. Five antibiotic- 89 induced resistant strains from ATCC 17978 or tigecycline resistant clinical isolate ABhl1 were 90 also included. Since biofilm formation is associated with antibiotic resistance and baeR as well 91 as emrA/emrB have been shown related with tigecycline and colistin resistance respectively in 92 our previous studies (Lin et al. 2014; Lin, Lin & Lan 2017a), so we also enrolled the mutant 93 strains for these genes. 94 95 Biofilm quantification 96 Biofilm formation was performed in 24-well polystyrene plates as previously described with 97 some modifications (Marti et al. 2011; Orsinger-Jacobsen et al. 2013; Kim et al. 2015). Biofilm 98 formation was determined in Mueller Hinton Broth (MHB) using an initial OD600 of 0.01, and 99 the plates were incubated at 37°C for 24 h without shaking (24-well, 1 mL). Two wells were left 100 uninoculated and used as negative controls. The culture media was removed by inversion, and 101 the wells were washed twice with phosphate buffered saline (PBS). 102 Both crystal violet and XTT assays show an excellent application for quantification of 103 biofilms. The Crystal violet assay is cheap, easy and is usually used for the quantification of 104 biofilms formed by microorganisms but XTT is more reliable and repeatable (Hendiani, Abdi- 105 Ali & Mohammadi 2014). In this study, XTT assay was used to measure the metabolic activity in 106 order to estimate the burden of viable cells. The electron transport system in the cellular 107 membrane of live bacteria reduces the XTT tetrazolium salt to XTT formazan, which results in a 108 colorimetric change measured at 492 nm (Orsinger-Jacobsen et al. 2013). 250 μL XTT solution 109 (0.0625 mg/mL XTT, 0.5 mM menadione in PBS) was added to each well of 24-well microtiter 110 plate for 30 min at 37°C. Manuscript to be reviewed 78 presence of an A. baumannii biofilm influenced antimicrobial therapy, we determ 79 planktonic minimum inhibitory concentrations and minimum biofilm eradication 80 The result might help clinicians decide the optimal therapy while treating the infe 81 by A. baumannii with a high biofilm-producing ability. 82 83 Materials & Methods 84 Bacterial strains 85 All of the bacterial strains used in this study were isolates from our previous studi 86 2014; Lin, Lin & Lan 2015; Lin et al. 2015; Lin, Lin & Lan 2017a; Lin et al. 2017 87 American Type Culture Collection (ATCC), as shown in Table 1. The studied stra 88 quality control strain ATCC 19606, ATCC 17978 and 21 clinical isolates. Five a 89 induced resistant strains from ATCC 17978 or tigecycline resistant clinical isolate 90 also included. Since biofilm formation is associated with antibiotic resistance and 91 as emrA/emrB have been shown related with tigecycline and colistin resistance re 92 our previous studies (Lin et al. 2014; Lin, Lin & Lan 2017a), so we also enrolled 93 strains for these genes. 94 95 Biofilm quantification 96 Biofilm formation was performed in 24-well polystyrene plates as previously des 97 some modifications (Marti et al. 2011; Orsinger-Jacobsen et al. 2013; Kim et al. 2 98 formation was determined in Mueller Hinton Broth (MHB) using an initial OD600 99 the plates were incubated at 37°C for 24 h without shaking (24-well, 1 mL). Two 100 uninoculated and used as negative controls. The culture media was removed by in 101 the wells were washed twice with phosphate buffered saline (PBS). 102 Both crystal violet and XTT assays show an excellent application for quantific 103 biofilms. The Crystal violet assay is cheap, easy and is usually used for the quanti 104 biofilms formed by microorganisms but XTT is more reliable and repeatable (Hen 105 Ali & Mohammadi 2014). In this study, XTT assay was used to measure the meta 106 order to estimate the burden of viable cells. The electron transport system in the c 107 membrane of live bacteria reduces the XTT tetrazolium salt to XTT formazan, wh 108 colorimetric change measured at 492 nm (Orsinger-Jacobsen et al. 2013). 250 μL 109 (0.0625 mg/mL XTT, 0.5 mM menadione in PBS) was added to each well of 24-w 110 plate for 30 min at 37°C. 41 Introduction In addition, we also closely 71 observed biofilm formation using scanning electron microscopy (SEM) and confocal laser 72 scanning microscopy (CLSM). These experiments were designed to understand the nature of 73 biofilm production among the wild-type strain, mutant strains and clinical isolates of A. 74 baumannii. To examine the impact of different compounds on biofilm formation and find 75 potential candidates for anti-biofilm agents, different kinds of antibiotics, the antimicrobial 76 peptide LL-37 and the iron chelating agent tannic acid, were chosen to test the relationship 77 between their presence and biofilm production in A. baumannii. Finally, to understand how the PeerJ reviewing PDF \| (2019:10:42050:4:0:NEW 20 Mar 2020) 84 Bacterial strains After incubation, 200 µL of the XTT supernatant was transferred to a 111 fresh 96-well plate, and the calorimetric absorbance was measured (OD490nm (XTT) biofilm) at 490 112 nm using iMarkTM Microplate Absorbance Reader (Bio-Rad laboratories, Inc. USA). The 113 negative control (OD490nm (XTT) control) was used to reduce the background absorbance optic 114 density values in the XTT assay. Since different bacterial strains had different growth rates, the 115 total cell numbers would be different at the end of incubation, thus influencing the total biofilm 116 mass or metabolic activities. To correct the bias due to cell numbers, the OD595nm (OD595nm (cells)) 117 was determined to estimate the total cells in each corresponding well of 24-well microtiter plate 85 All of the bacterial strains used in this study were isolates from our previous studies (Lin et al. 86 2014; Lin, Lin & Lan 2015; Lin et al. 2015; Lin, Lin & Lan 2017a; Lin et al. 2017b) or 87 American Type Culture Collection (ATCC), as shown in Table 1. The studied strains contained 88 quality control strain ATCC 19606, ATCC 17978 and 21 clinical isolates. Five antibiotic- 89 induced resistant strains from ATCC 17978 or tigecycline resistant clinical isolate ABhl1 were 90 also included. Since biofilm formation is associated with antibiotic resistance and baeR as well 91 as emrA/emrB have been shown related with tigecycline and colistin resistance respectively in 92 our previous studies (Lin et al. 2014; Lin, Lin & Lan 2017a), so we also enrolled the mutant 93 strains for these genes. 85 All of the bacterial strains used in this study were isolates from our previous studies (Lin et al. 86 2014; Lin, Lin & Lan 2015; Lin et al. 2015; Lin, Lin & Lan 2017a; Lin et al. 2017b) or 87 American Type Culture Collection (ATCC), as shown in Table 1. The studied strains contained 88 quality control strain ATCC 19606, ATCC 17978 and 21 clinical isolates. Five antibiotic- 89 induced resistant strains from ATCC 17978 or tigecycline resistant clinical isolate ABhl1 were 90 also included. Since biofilm formation is associated with antibiotic resistance and baeR as well 91 as emrA/emrB have been shown related with tigecycline and colistin resistance respectively in 92 our previous studies (Lin et al. 2014; Lin, Lin & Lan 2017a), so we also enrolled the mutant 93 strains for these genes. Manuscript to be reviewed The studied strains contained 88 quality control strain ATCC 19606, ATCC 17978 and 21 clinical isolates. Five antibiotic- 89 induced resistant strains from ATCC 17978 or tigecycline resistant clinical isolate ABhl1 were 90 also included. Since biofilm formation is associated with antibiotic resistance and baeR as well 91 as emrA/emrB have been shown related with tigecycline and colistin resistance respectively in 92 our previous studies (Lin et al. 2014; Lin, Lin & Lan 2017a), so we also enrolled the mutant 93 strains for these genes. Manuscript to be reviewed After incubation, 200 µL of the XTT supernatant was tr 111 fresh 96-well plate, and the calorimetric absorbance was measured (OD490nm (XTT) 112 nm using iMarkTM Microplate Absorbance Reader (Bio-Rad laboratories, Inc. U 113 negative control (OD490nm (XTT) control) was used to reduce the background absorban 114 density values in the XTT assay. Since different bacterial strains had different gro 78 presence of an A. baumannii bio 79 planktonic minimum inhibitory 80 The result might help clinicians 81 by A. baumannii with a high bio 82 83 Materials & Methods 84 Bacterial strains 85 All of the bacterial strains used 86 2014; Lin, Lin & Lan 2015; Lin 87 American Type Culture Collecti 88 quality control strain ATCC 196 89 induced resistant strains from A 90 also included. Since biofilm form 91 as emrA/emrB have been shown 92 our previous studies (Lin et al. 2 93 strains for these genes. 94 95 Biofilm quantification 96 Biofilm formation was performe 97 some modifications (Marti et al 98 formation was determined in Mu 99 the plates were incubated at 37° 100 uninoculated and used as negati 101 the wells were washed twice wi 102 Both crystal violet and XTT 103 biofilms. The Crystal violet assa 104 biofilms formed by microorgani 105 Ali & Mohammadi 2014). In thi 106 order to estimate the burden of v 107 membrane of live bacteria reduc 108 colorimetric change measured a 109 (0.0625 mg/mL XTT, 0.5 mM m 110 plate for 30 min at 37°C. After i 111 fresh 96-well plate, and the calo 112 nm using iMarkTM Microplate 113 negative control (OD490nm (XTT) c 114 density values in the XTT assay 78 presence of an A. baumannii biofilm influenced antimicrobial therapy, we determined the 79 planktonic minimum inhibitory concentrations and minimum biofilm eradication concentrations. 80 The result might help clinicians decide the optimal therapy while treating the infections caused 81 by A. baumannii with a high biofilm-producing ability. 81 by A. baumannii with a high biofilm producing ability. 82 83 Materials & Methods 84 Bacterial strains 85 All of the bacterial strains used in this study were isolates from our previous studies (Lin et al. 86 2014; Lin, Lin & Lan 2015; Lin et al. 2015; Lin, Lin & Lan 2017a; Lin et al. 2017b) or 87 American Type Culture Collection (ATCC), as shown in Table 1. Manuscript to be reviewed 118 after transferred to a new 96-well plate for optic density measurement. The ability to form a 119 biofilm was expressed using a biofilm formation index: [BFI = (OD490nm (XTT) biofilm－OD490nm 120 (XTT) control)/OD595nm (cells)]. The ODc was defined as three standard deviations above the mean OD 121 of the negative control. Each isolate was classified as follows: non-biofilm producer: OD ≤ ODc; 122 weak biofilm producer: ODc < OD ≤ 2 × ODc; moderate biofilm producer: 2 × ODc < OD ≤ 4 × 123 ODc; or strong biofilm producer: OD > 4 × ODc (Hu et al. 2016). 124 125 Biofilm imaging 126 Biofilms were examined by SEM and CLSM (Djeribi et al. 2012). For SEM examination, 127 biofilms were developed at 37°C for 24 h in polystyrene coverslips placed into a 24-well plate 128 with MHB containing bacterial cells (OD600nm 0.01) added. After fixed in 3.7% formaldehyde for 129 40 min at room temperature and then rinsed twice with PBS, further fixation was performed with 130 1% osmium tetroxide for 15 min at room temperature and then rinsed twice with PBS. The 131 samples were then dehydrated by passing them through different concentrations of ethanol: 35%, 132 50%, 70%, 80%, and 95%, each for 5 min, followed by 100% ethanol twice for 10 min each 133 time. The samples were then dried at 60°C for 24 h. After being coated with gold-palladium (via 134 sputter coating), the samples were examined under a bioscanning electron microscope (Hitachi, 135 S-4700, Type II). 136 For CLSM examination biofilms formation in polystyrene coverslips and fixation in 3 7% 136 For CLSM examination, biofilms formation in polystyrene coverslips and fixation in 3.7% 137 formaldehyde were the same as mentioned above. Then the samples were stained with a prepared 138 100 nM solution of Syto-9 stain (Invitrogen, Carlsbad, CA) for 30 min (Luo et al. 2015b). Syto-9 139 stained biofilms were excited with a 488 nm solid-state laser, and fluorescence was captured 140 between 500–550 nm. The images of biofilms were rendered and assembled using appropriate 141 computational software (ZEISS LSM780). 136 For CLSM examination, biofilms formation in polystyrene coverslips and fixation in 3.7% 137 formaldehyde were the same as mentioned above. Then the samples were stained with a prepared 138 100 nM solution of Syto-9 stain (Invitrogen, Carlsbad, CA) for 30 min (Luo et al. 2015b). Manuscript to be reviewed Syto-9 139 stained biofilms were excited with a 488 nm solid-state laser, and fluorescence was captured 140 between 500–550 nm. The images of biofilms were rendered and assembled using appropriate 141 computational software (ZEISS LSM780). 84 Bacterial strains 96 Biofilm formation was performed in 24-well polystyrene plates as previously described with 97 some modifications (Marti et al. 2011; Orsinger-Jacobsen et al. 2013; Kim et al. 2015). Biofilm 98 formation was determined in Mueller Hinton Broth (MHB) using an initial OD600 of 0.01, and 99 the plates were incubated at 37°C for 24 h without shaking (24-well, 1 mL). Two wells were left 00 uninoculated and used as negative controls. The culture media was removed by inversion, and 01 the wells were washed twice with phosphate buffered saline (PBS). p p ( ) 102 Both crystal violet and XTT assays show an excellent application for quantification of 103 biofilms. The Crystal violet assay is cheap, easy and is usually used for the quantification of 104 biofilms formed by microorganisms but XTT is more reliable and repeatable (Hendiani, Abdi- 105 Ali & Mohammadi 2014). In this study, XTT assay was used to measure the metabolic activity in 106 order to estimate the burden of viable cells. The electron transport system in the cellular 107 membrane of live bacteria reduces the XTT tetrazolium salt to XTT formazan, which results in a 108 colorimetric change measured at 492 nm (Orsinger-Jacobsen et al. 2013). 250 μL XTT solution 109 (0.0625 mg/mL XTT, 0.5 mM menadione in PBS) was added to each well of 24-well microtiter 110 plate for 30 min at 37°C. After incubation, 200 µL of the XTT supernatant was transferred to a 111 fresh 96-well plate, and the calorimetric absorbance was measured (OD490nm (XTT) biofilm) at 490 112 nm using iMarkTM Microplate Absorbance Reader (Bio-Rad laboratories, Inc. USA). The 113 negative control (OD490nm (XTT) control) was used to reduce the background absorbance optic 114 density values in the XTT assay. Since different bacterial strains had different growth rates, the 115 total cell numbers would be different at the end of incubation, thus influencing the total biofilm 116 mass or metabolic activities. To correct the bias due to cell numbers, the OD595nm (OD595nm (cells)) 117 was determined to estimate the total cells in each corresponding well of 24-well microtiter plate PeerJ reviewing PDF \| (2019:10:42050:4:0:NEW 20 Mar 2020) Manuscript to be reviewed 158 Louis, MO) containing different concentrations of antibiotics, including amikacin, imipenem, 159 colistin and tigecycline, to reach ≈ 5 × 105 CFU/mL, and the cultures were incubated at 37°C for 160 24 h. The lowest antibiotic concentration that completely inhibited bacterial growth was defined 161 as the minimum inhibitory concentration, and growth was determined by unaided eyes and by 162 measuring optical densities using a spectrophotometer. 163 164 Biofilm susceptibility test 165 The biofilm susceptibility test was performed as previously described with some modifications 166 (Kim et al. 2015). In brief, preparation of the antibiotic stock solution and preparation of the 167 inoculum were the same as the method used in checking the MIC. Biofilm susceptibility tests 168 were performed in flat-bottom, polystyrene cell-culture microtiter plates containing 5 × 105 169 CFU/mL in MHB with final well volumes of 100 μL. The plates were incubated at 37°C without 170 shaking. After 24 h, all the suspension fluid was removed from the plates and then washed with 171 PBS. Equal amounts of appropriate antibiotic dilutions were added. Twofold dilutions of the 172 tested antibiotics in CAMHB were diluted from 256 to 0.125 μg/mL. The tested antibiotic 173 concentrations would be increased if necessary. Then, the plates were incubated at 37°C for 24 h 174 without shaking. After 24 h, inhibition of bacteria growth was determined by no turbidity 175 observed by unaided eyes. Regrowth of bacteria was further determined by spot assay. The 176 minimum biofilm eradication concentration (MBEC) represents the lowest drug concentration at 177 which bacteria failed to regrow (Olson et al. 2002). 178 179 Statistical analysis 180 The comparison of biofilm formation, including BFI and RBFI, was analyzed using Student’s t- 181 test. All of the data were from three independent experiments and analyzed using Statistical 182 Package for the Social Sciences version 16 (SPSS Inc., Chicago, IL, USA). p < 0.05 was 183 considered statistically significant. 184 185 Results 186 Biofilm quantification and its relationship with the antimicrobial susceptibility 187 Biofilm quantification of the 33 different strains of A. baumannii (Table 1), including 21 clinical 188 isolates, by the biofilm formation index is presented in Fig. 1. There are obvious differences in 189 biofilm formation among the clinical isolates of A. baumannii. Of the 22 clinical isolates 190 including A. 143 Influence of antibiotics, LL-37 and tannic acid on biofilm formation 143 Influence of antibiotics, LL-37 and tannic acid on biofilm formation 143 Influence of antibiotics, LL-37 and tannic acid on biofilm formation 144 The activity of the various antibiotics, including amikacin, imipenem, colistin and tigecycline, 145 LL-37 and tannic acid, on bacterial growth was tested in 24-well microtiter plates containing the 146 above compounds at different concentrations. The concentrations chosen for different 147 compounds were based on individual MIC with two-fold serial dilution. However, if biofilm 148 formation was not observed at the higher concentrations of compounds, then the lower 149 concentration was adopted. Biofilm production was checked by the same method as described 150 above. Finally, A. baumannii ATCC 17978 without any compound added was used as a control 151 for biofilm formation to calculate the relative biofilm formation capacity (relative biofilm 152 formation index [RBFI] = BFI/BFIA. baumannii ATCC 17978). 153 154 Planktonic susceptibility test 155 A planktonic susceptibility test was performed as previously described to determine the 156 minimum inhibitory concentration (MIC) of A. baumannii (Lin et al. 2014). In brief, bacteria 157 were inoculated into 1 mL cation-adjusted Mueller-Hinton Broth (CAMHB) (Sigma-Aldrich, St. 144 The activity of the various antibiotics, including amikacin, imipenem, colistin and tigecycline, 145 LL-37 and tannic acid, on bacterial growth was tested in 24-well microtiter plates containing the 146 above compounds at different concentrations. The concentrations chosen for different 147 compounds were based on individual MIC with two-fold serial dilution. However, if biofilm 148 formation was not observed at the higher concentrations of compounds, then the lower 149 concentration was adopted. Biofilm production was checked by the same method as described 150 above. Finally, A. baumannii ATCC 17978 without any compound added was used as a control 151 for biofilm formation to calculate the relative biofilm formation capacity (relative biofilm 152 formation index [RBFI] = BFI/BFIA. baumannii ATCC 17978). 153 154 Planktonic susceptibility test 155 A planktonic susceptibility test was performed as previously described to determine the 156 i i i hibit t ti (MIC) f A b ii (Li t l 2014) I b i f b t i PeerJ reviewing PDF \| (2019:10:42050:4:0:NEW 20 Mar 2020) Manuscript to be reviewed baumannii ATCC 17978, twenty isolates were strong biofilm producers (4 × 191 ODc < OD490nm) except VGH1 and CT11 (2 × ODc < OD490nm ≤ 4 × ODc) were moderate biofilm 192 producers (data shown in the raw data set). 193 The antimicrobial susceptibility of the clinical isolates in this study has been published 194 previously (Lin et al. 2017b) as shown in Table S1. Susceptibility breakpoints to antimicrobial 195 agents were determined by the MIC interpretive standards for Acinetobacter spp. of Clinical and 196 Laboratory Standards Institute (CLSI. 2012). The MIC interpretive criteria for each individual 197 antimicrobial agents except cefazolin, cefmetazole and tigecycline (MIC µg/mL) (S=susceptible, 158 Louis, MO) containing different concentrations of antibiotics, including amikacin, imipenem, 159 colistin and tigecycline, to reach ≈ 5 × 105 CFU/mL, and the cultures were incubated at 37°C for 160 24 h. The lowest antibiotic concentration that completely inhibited bacterial growth was defined 161 as the minimum inhibitory concentration, and growth was determined by unaided eyes and by 162 measuring optical densities using a spectrophotometer. 163 164 Biofilm susceptibility test 165 The biofilm susceptibility test was performed as previously described with some modifications 166 (Kim et al. 2015). In brief, preparation of the antibiotic stock solution and preparation of the 167 inoculum were the same as the method used in checking the MIC. Biofilm susceptibility tests 168 were performed in flat-bottom, polystyrene cell-culture microtiter plates containing 5 × 105 169 CFU/mL in MHB with final well volumes of 100 μL. The plates were incubated at 37°C without 170 shaking. After 24 h, all the suspension fluid was removed from the plates and then washed with 171 PBS. Equal amounts of appropriate antibiotic dilutions were added. Twofold dilutions of the 172 tested antibiotics in CAMHB were diluted from 256 to 0.125 μg/mL. The tested antibiotic 173 concentrations would be increased if necessary. Then, the plates were incubated at 37°C for 24 h 174 without shaking. After 24 h, inhibition of bacteria growth was determined by no turbidity 175 observed by unaided eyes. Regrowth of bacteria was further determined by spot assay. The 176 minimum biofilm eradication concentration (MBEC) represents the lowest drug concentration at 177 which bacteria failed to regrow (Olson et al. 2002). 178 179 Statistical analysis 180 The comparison of biofilm formation, including BFI and RBFI, was analyzed using Student’s t- 181 test. Manuscript to be reviewed All of the data were from three independent experiments and analyzed using Statistical 182 Package for the Social Sciences version 16 (SPSS Inc., Chicago, IL, USA). p < 0.05 was 183 considered statistically significant. 184 185 Results 186 Biofilm quantification and its relationship with the antimicrobial susceptibility 158 Louis, MO) containing different concentrations of antibiotics, including amikacin, imipenem, 159 colistin and tigecycline, to reach ≈ 5 × 105 CFU/mL, and the cultures were incubated at 37°C for 160 24 h. The lowest antibiotic concentration that completely inhibited bacterial growth was defined 161 as the minimum inhibitory concentration, and growth was determined by unaided eyes and by 162 measuring optical densities using a spectrophotometer. 163 164 Biofilm susceptibility test p y 165 The biofilm susceptibility test was performed as previously described with some modifications 166 (Kim et al. 2015). In brief, preparation of the antibiotic stock solution and preparation of the 167 inoculum were the same as the method used in checking the MIC. Biofilm susceptibility tests 168 were performed in flat-bottom, polystyrene cell-culture microtiter plates containing 5 × 105 169 CFU/mL in MHB with final well volumes of 100 μL. The plates were incubated at 37°C without 170 shaking. After 24 h, all the suspension fluid was removed from the plates and then washed with 171 PBS. Equal amounts of appropriate antibiotic dilutions were added. Twofold dilutions of the 172 tested antibiotics in CAMHB were diluted from 256 to 0.125 μg/mL. The tested antibiotic 173 concentrations would be increased if necessary. Then, the plates were incubated at 37°C for 24 h 174 without shaking. After 24 h, inhibition of bacteria growth was determined by no turbidity 175 observed by unaided eyes. Regrowth of bacteria was further determined by spot assay. The 176 minimum biofilm eradication concentration (MBEC) represents the lowest drug concentration at 177 which bacteria failed to regrow (Olson et al. 2002). 178 176 minimum biofilm eradication concentration (MBEC) represents the lowest drug concentration at 177 which bacteria failed to regrow (Olson et al. 2002). 178 179 Statistical analysis 180 The comparison of biofilm formation, including BFI and RBFI, was analyzed using Student’s t- 181 test. All of the data were from three independent experiments and analyzed using Statistical 182 Package for the Social Sciences version 16 (SPSS Inc., Chicago, IL, USA). p < 0.05 was 183 considered statistically significant. 184 185 Results 186 Biofilm quantification and its relationship with the antimicrobial susceptibility 187 Biofilm quantification of the 33 different strains of A. baumannii (Table 1), including 21 clinical 188 isolates, by the biofilm formation index is presented in Fig. 1. There are obvious differences in 189 biofilm formation among the clinical isolates of A. baumannii. Of the 22 clinical isolates 190 including A. baumannii ATCC 17978, twenty isolates were strong biofilm producers (4 × 191 ODc < OD490nm) except VGH1 and CT11 (2 × ODc < OD490nm ≤ 4 × ODc) were moderate biofilm 192 producers (data shown in the raw data set). 164 Biofilm susceptibility test 193 The antimicrobial susceptibility of the clinical isolates in this study has been published 194 previously (Lin et al 2017b) as shown in Table S1 Susceptibility breakpoints to antimicrobial Manuscript to be reviewed 198 I= intermediate and R=resistant) are listed respectively as the following : Ampicillin/sulbactam 199 (S ≤ 8/4, I 16/8 and R ≥ 32/16); piperacillin/tazobactam (S ≤ 16/4, I 32/4-64/4 and R ≥ 128/4) ; 200 cefotaxime (S ≤ 8, I 16-32 and R ≥ 64); ceftazidime (S ≤ 8, I 16 and R ≥ 32); cefepime (S ≤ 8, I 201 16 and R ≥ 32); imipenem (S ≤ 4, I 8 and R ≥ 16); meropenem (S ≤ 4, I 8 and R ≥ 16); amikacin 202 (S ≤ 16, I 32 and R ≥ 64); gentamicin (S ≤ 4, I 8 and R ≥ 16); ciprofloxacin (S ≤ 1, I 2 and R ≥ 4); 203 levofloxacin (S ≤ 2, I 4 and R ≥ 8); trimethoprim/sulfamethoxazole (S ≤ 2/38 and R ≥ 4/76). A. 204 baumannii is naturally resistant to cefazolin, whereas the isolates of A. baumannii with 205 cefmetazole MICs of ≥ 32µg/mL are considered resistant (Jones et al. 1986). The provisional 206 MIC breakpoints for tigecycline are ≤2, 4 and ≥8 µg/mL to designate susceptible, intermediate 207 and resistant strains, respectively (Pachon-Ibanez et al. 2004). All of the 21 A. baumannii clinical 208 isolates are not susceptible to many of the tested antimicrobial agents. Although ATCC 17978 is 209 the least resistant strain among the isolates studied, its BFI ranked second. Two isolates, VGH2 210 and VGH7, are pan-drug resistant. However, their biofilm formation ability BFI differed with 211 VGH2 ranking first and VGH7 ranking sixteenth. It seems the capacity of biofilm production 212 does not relate to antimicrobial susceptibility as shown in Table S1. To validate the inference, 213 the correlation between bioﬁlm formation ability and resistance to the 15 antimicrobial agents in 214 A. baumannii was analyzed using the Wilcoxon rank-sum test. Fig. 2 depicts the box plots with 215 the comparison of the biofilm formation indices of the susceptible and resistant strains to the 216 individual antibiotics. It reveals no relationship with statistical significance (p>0.05) between the 217 ability to form biofilm and the antimicrobial susceptibility of the A. baumannii clinical isolates. 218 185 Results 186 Biofilm quantification and its relationship with the antimicrobial susceptibility 187 Biofilm quantification of the 33 different strains of A. baumannii (Table 1), including 21 clinical 188 isolates, by the biofilm formation index is presented in Fig. 1. There are obvious differences in 189 biofilm formation among the clinical isolates of A. baumannii. Of the 22 clinical isolates 190 including A. baumannii ATCC 17978, twenty isolates were strong biofilm producers (4 × 191 ODc < OD490nm) except VGH1 and CT11 (2 × ODc < OD490nm ≤ 4 × ODc) were moderate biofilm 192 producers (data shown in the raw data set). 193 The antimicrobial susceptibility of the clinical isolates in this study has been published 194 previously (Lin et al. 2017b) as shown in Table S1. Susceptibility breakpoints to antimicrobial 195 agents were determined by the MIC interpretive standards for Acinetobacter spp. of Clinical and 196 Laboratory Standards Institute (CLSI. 2012). The MIC interpretive criteria for each individual 197 antimicrobial agents except cefazolin, cefmetazole and tigecycline (MIC µg/mL) (S=susceptible, PeerJ reviewing PDF \| (2019:10:42050:4:0:NEW 20 Mar 2020) 219 Biofilm formation by A. baumannii ATCC 17978 after baeR, emrA, emrB and ompA gene 220 knockout 220 knockout 221 Many proteins, including two-component regulatory system BfmRS (Luo et al. 2015a), 222 CsuA/BABCDE usher-chaperone assembly system (Tomaras et al. 2008), biofilm-associated 223 proteins (Loehfelm, Luke & Campagnari 2008), OmpA protein (Gaddy, Tomaras & Actis 2009), 224 AdeFGH (He et al. 2015) and blaPER (Lee et al. 2008), have been demonstrated to be related 225 with biofilm formation in A. baumannii. To understand the scope of involvement further, biofilm 226 formation caused by A. baumannii ATCC 17978 after gene knockout of another two-component 227 regulatory system gene baeR, efflux pump genes emrA/emrB and ompA is compared as shown in 228 Fig. 3A. The ΔbaeR, ΔemrA, ΔemrB and ΔompA mutant strains all had less biofilm formation 229 than the wild type strain. This result is further confirmed by the SEM and CLSM images shown 230 in Fig. 3B and 3C, respectively. Of these three categories of genes, ompA knockout led to the 231 most marked decrease biofilm formation in A. baumannii. The SEM figure of ΔompA strain 232 shows that only few cells are clustered and its images of CLSM exhibit the thinnest biofilm. 233 Although the influence on biofilm formation reduction after baeR, emrA, or emrB gene knockout 234 was not so obvious as the ΔompA strain, the degree of decrease was still statistically significant 235 while being compared with the ATCC 17978 strain (p<0.05 for ΔbaeR, p<0.01 for ΔemrA, and 236 p<0.001 for ΔemrB). 237 proteins, including two-component regulatory system BfmRS (Luo et al. 2015a), PeerJ reviewing PDF \| (2019:10:42050:4:0:NEW 20 Mar 2020) 238 Influence of different compounds on biofilm formation The ATCC 17978 strain had low MICs to all 263 of the four tested antimicrobial agents, whereas the VGH2 strain had higher MICS for amikacin 264 (1024 μg/mL) and imipenem (16 μg/mL). When biofilms cells of the same strains were tested, 265 MBECs rose astonishingly with at least fourfold higher relative to the planktonic MICs. Of all 266 the four antibiotics tested, colistin had the most obvious MBEC increase from 0.5 to >8192 267 μg/mL for VGH2. This result suggested that biofilm cells of bacteria would become more 268 resistant to tested antibiotics according to MBEC. 269 270 Discussion 271 High variability in biofilm formation exists among the clinical isolates of A. baumannii 272 (Wroblewska et al. 2008). In a previous study that examined 86 clinical isolates of A. baumannii, 273 clinical isolates exhibited an enhanced biofilm formation ability relative to a standard A. 274 baumannii strain (ATCC 19606) (Kim et al. 2015). Interestingly, only A. baumannii VGH2 has 275 better biofilm formation ability than the standard A. baumannii strain (ATCC 17978) in this 276 study. However, Kim et al. used crystal violet to measure the biomass of biofilm whereas we 277 measured the metabolic activity in order to estimate the burden of viable biofilm cells by 238 Influence of different compounds on biofilm formation 238 Influence of different compounds on biofilm formation 238 Influence of different compounds on biofilm formation 239 Table 2 presents minimum inhibitory concentrations (MICs) of A. baumannii ATCC 17978 and 240 VGH2 to amikacin, imipenem, colistin, tigecycline, LL-37 and tannic acid. The influence of 241 different compounds, including amikacin, imipenem, colistin, tigecycline, LL-37 and tannic acid, 242 on biofilm formation in A. baumannii ATCC 17978 and VGH2 is shown in Fig. 4 and Fig. 5. 243 Amikacin, imipenem, colistin and LL-37 were tested at 1/4, 1/2 and 1 x MIC whereas tigecycline 244 was tested at 1/16, 1/8, 1/4 x MIC. Since tannic acid had very high MIC (>300μg/mL), the 245 concentrations of 200, 100 and 50 μg/mL were chosen arbitrarily for testing. The addition of 246 amikacin (> 0.5 μg/mL), colistin (> 0.5 μg/mL), LL-37 (> 4 μg/mL), or tannic acid (> 50 μg/mL) 247 decreased the biofilm formation ability of A. baumannii ATCC 17978, whereas the addition of 248 amikacin (> 256 μg/mL), imipenem (> 4 μg/mL), colistin (> 0.125 μg/mL), LL-37 (> 2 μg/mL), 249 or tannic acid (> 50 μg/mL) decreased the biofilm formation ability of VGH2. In contrast, the 250 addition of tigecycline increased the biofilm formation ability of A. baumannii ATCC 17978 (> 251 62.5 ng/mL) and VGH2 (> 0.5 μg/mL). Figure 6 presents the SEM and CLSM images of biofilm 252 formation by A. baumannii ATCC 17978 in the presence of 1 μg/mL colistin or 125 ng/mL 253 tigecycline. With 1 μg/mL colistin added to the ATCC 17978 strain, the SEM image shows the 254 cells become scattered and the CLSM images exhibit decreased thickness of biofilm. In contrast, 255 125 ng/mL tigecycline exposure has the opposite impact on biofilm formation of A. baumannii 256 ATCC 17978. Both of the SEM and CLSM images validated the findings of biofilm 257 quantification by RBFI analysis (Fig. 4 and Fig. 5). 2 8 257 quantification by RBFI analysis (Fig. 4 and Fig. 5). 258 259 MICs and MBECs of A. baumannii ATCC 17978 and VGH2 260 Besides MICs, MBECs of A. baumannii ATCC 17978 and VGH2 to amikacin, imipenem, 261 colistin and tigecycline is also shown in Table 2. In this study, the MIC values for antimicrobials 262 were determined using a broth microdilution assay. 259 MICs and MBECs of A. baumannii ATCC 17978 and VGH2 260 Besides MICs, MBECs of A. baumannii ATCC 17978 and VGH2 to amikacin, imipenem, 261 colistin and tigecycline is also shown in Table 2. In this study, the MIC values for antimicrobials 262 were determined using a broth microdilution assay. The ATCC 17978 strain had low MICs to all 263 of the four tested antimicrobial agents, whereas the VGH2 strain had higher MICS for amikacin 264 (1024 μg/mL) and imipenem (16 μg/mL). When biofilms cells of the same strains were tested, 265 MBECs rose astonishingly with at least fourfold higher relative to the planktonic MICs. Of all 266 the four antibiotics tested, colistin had the most obvious MBEC increase from 0.5 to >8192 267 μg/mL for VGH2. This result suggested that biofilm cells of bacteria would become more 268 resistant to tested antibiotics according to MBEC. Manuscript to be reviewed 278 tetrazolium salt XTT reduction assay instead because the XTT assay is the test most commonly 279 used to estimate viable biofilm growth and to examine the impact of biofilm therapies (Nett et al. 280 2011). One previous study about candidal biofilm showed the better reproducibility of the XTT 281 assay compared with that of the crystal violet method led the authors to conclude that the former 282 is more reliable (Jin et al. 2003). However, using a metabolic activity assay for measuring 283 antibiotic activity against biofilm suffers from the limitation that usually the absorbance or 284 fluorescence signal decreases in proportion to the number of bacteria only on a limited range of 285 colony forming units (CFUs). Statistical analyses showed that XTT activity was linearly 286 associated with the log of the cell concentration over the cell concentration range tested (from 2 287 X 105 to 108 cells/ml), implying XTT activity can be used as an indicator of cell numbers in a 288 cell suspension (Jin et al. 2003; Orsinger-Jacobsen et al. 2013). Although we didn’t determine 289 the zone of linearity in this study, all of the absorbance OD595 values, even in the presence of 290 subinhibitory antibiotic concentrations, were more than 0.1 (data shown in the raw data set), 291 which indicated that cells were present more than 105. Therefore, the XTT assay in this study 292 was presumed to be applicable in measuring the biofilm formation of A. baumannii. On the other 293 hand, major differences in metabolic activity can be observed when comparing several bacterial 294 isolates from the same species (Zimmermann et al. 2015), so that differences in XTT signal 295 could rely simply on difference in metabolic activity. In this study, the influence of chemical 296 compounds on biofilm formation of ATCC 17978 or VGH2 was compared to itself. Therefore, 297 this phenomenon would not change the main findings of this study. 298 In this study, all of the 21 A. baumannii clinical isolates could form biofilms. However, the 299 clone relatedness and the ability of biofilm formation were not further explored. One previous 300 study reported higher ability of biofilm formation was found for the strains assigning to 301 multilocus sequence typing (MLST) sequence type 2 (ST2), ST25 and ST78 (Giannouli et al. 302 2013). 270 Discussion 271 High variability in biofilm formation exists among the clinical isolates of A. baumannii 272 (Wroblewska et al. 2008). In a previous study that examined 86 clinical isolates of A. baumannii, 273 clinical isolates exhibited an enhanced biofilm formation ability relative to a standard A. 274 baumannii strain (ATCC 19606) (Kim et al. 2015). Interestingly, only A. baumannii VGH2 has 275 better biofilm formation ability than the standard A. baumannii strain (ATCC 17978) in this 276 study. However, Kim et al. used crystal violet to measure the biomass of biofilm whereas we 277 measured the metabolic activity in order to estimate the burden of viable biofilm cells by PeerJ reviewing PDF \| (2019:10:42050:4:0:NEW 20 Mar 2020) Manuscript to be reviewed But in another study the sporadic strains showed significantly higher biofilm-forming 303 capacity than the epidemic isolates assigning to ST2 (Hu et al. 2016). It seems the biofilm- 304 forming capacity in various sequence types of A. baumannii needs further experiments to clarify. 305 A few studies have investigated the role of RND efflux pumps, including AdeABC, AdeIJK 306 and AdeFGH, in A. baumannii biofilm formation (He et al. 2015; Yoon et al. 2015). The role of 307 efflux pumps in A. baumannii biofilm formation has been suggested in whole transcriptome 308 analysis of biofilm and planktonic cells (Rumbo-Feal et al. 2013), although EmrAB was not 309 implicated in gene expression change of biofilm cells. However, emrA and emrB genes have 310 been demonstrated related with biofilm formation in other bacteria. Escherichia coli mutant 311 strains lacking the efflux emrA and emrB were reported to display enhanced biofilm growth 312 compared with control (Bay et al. 2017),where the deletion of the efflux genes emrAB in 313 Salmonella enterica serovars resulted in decreased biofilm formation compared with wild type 314 strain (Baugh et al. 2012). In this study, the efflux pump EmrAB was found to be associated with 315 biofilm formation in A. baumannii for the first time. Although we have demonstrated that 316 EmrAB contributes to adaptation to osmotic stress and resistance to colistin in A. baumannii (Lin 317 Lin & Lan 2017a), its role in biofilm formation still needs to be clarified. PeerJ reviewing PDF \| (2019:10:42050:4:0:NEW 20 Mar 2020) Manuscript to be reviewed 318 Besides emrA, emrB and ompA genes implicated in biofilm formation of A. baumannii in this 319 study, deletion of two-component system gene baeR led to decreased biofilm formation. Deleting 320 baeR of A. baumannii results in vulnerable to certain chemicals, especially tigecycline and tannic 321 acid (Lin et al. 2014; Lin, Lin & Lan 2015). In addition to involvement of disposing chemicals, 322 two-component signal transduction system is also regarded as a major strategy for connecting 323 input stimuli to biofilm formation (Liu et al. 2018). The current proposed model associates low 324 intracellular levels of c-di-GMP with a planktonic lifestyle, whereas high c-di-GMP levels are 325 associated with biofilm formation. However, it has been demonstrated that the BaeSR response 326 in E. coli does not influence biofilm formation, nor is it involved in indole-mediated inhibition of 327 biofilm formation (Leblanc, Oates & Raivio 2011). To investigate the mechanism for biofilm 328 formation decrease after baeR deletion and clarify the true role of BaeSR in biofilm formation of 329 A. baumannii, a secondary messenger c-di-GMP concentration in A. baumannii needs to be 330 measured in the future research since c-di-GMP is a well-known intracellular messenger 331 molecule that affects biofilm formation. 332 Exposure to aminoglycoside and subinhibitory concentrations of imipenem is associated with 333 biofilm-forming A. baumannii isolates (Rodríguez-Baño et al. 2008; Nucleo et al. 2009). In two 334 pairs of clinical colistin-susceptible/colistin-resistant (Csts/Cstr) A. baumannii strains, the Cstr 335 strains showed significantly decreased biofilm formation in static and dynamic assays (p<0.001) 336 and lower relative fitness (p<0.05) compared with those of the Csts counterparts (Dafopoulou et 337 al. 2015). We found that A. baumannii ATCC 17978 decreased biofilm formation upon exposure 338 to subinhibitory concentrations of amikacin, imipenem and colistin. In this study, tannic acid and 339 LL-37 at a concentration below its MIC can decrease biofilm formation, which might occur by 340 causing structural damage to the A. baumannii biofilm (Shi et al. 2014). One previous study has 341 demonstrated increased biofilm formation of Staphylococcus epidermidis after sub-MIC 342 tigecycline treatment (both at 0.25 and 0.5 MIC) by producing increased expression of the icaA 343 (production of transmembrane protein), altE (encoding autolysin related with adhesion) and sigB 344 (biofilm stability) genes and by affecting biofilm architecture in the isolates (Szczuka, Jablonska 345 & Kaznowski 2017). Manuscript to be reviewed 394 Bay DC, Stremick CA, Slipski CJ, and Turner RJ. 2017. Secondary multidrug efflux pump 395 mutants alter Escherichia coli biofilm growth in the presence of cationic antimicrobial 396 compounds. Research in microbiology 168:208-221 DOI: 10.1016/j.resmic.2016.11.003. 397 CLSI. 2012. Performance Standards for Antimicrobial Susceptibility Testing; Twenty-Second 398 Informational Supplement. CLSI document M100-S22. Wayne, PA: Clinical and Laboratory 358 baumannii strains (Song & Cheong 2015). However, the influence of tigecycline on biofilm 359 formation of A. baumannii is contrary to the results of this study. Meropenem plus sulbactam 360 exhibited synergism against biofilm-embedded carbapenem-resistant A. baumannii and caused 361 significantly more damage to biofilm architecture than any of the agents used alone (Wang et al. 362 2016). 358 baumannii strains (Song & Cheong 2015). However, the influence of tigecycline on biofilm 359 formation of A. baumannii is contrary to the results of this study. Meropenem plus sulbactam 360 exhibited synergism against biofilm-embedded carbapenem-resistant A. baumannii and caused 361 significantly more damage to biofilm architecture than any of the agents used alone (Wang et al. 362 2016). 363 The images of SEM and CLSM were used in this study to support the findings by XTT assay 364 about the impact of gene knockout and adding antibiotics on biofilm formation, but no live-dead 365 stain was performed. Live-dead stain consists of SYTO9, which stains live cells green, and 366 propidium iodide (PI), which stains dead cells red (Richmond et al. 2016). This allows live 367 versus dead cells to be detected and is more informative. Although we made an effort to 368 characterize biofilm production of A. baumannii in different conditions, there were still some 369 limitations in this study. As previously described, the interplay of these related genes in biofilm 370 formation of A. baumannii and key secondary messenger c-di-GMP determination were not 371 explored further. Besides, because the blood or tissue concentrations produced by infusions of 372 the studied antimicrobials were not determined, it is hard to infer that the in vitro results will 373 apply to the in vivo condition. 387 Badave GK, and Kulkarni D. 2015. Biofilm producing multidrug resistant Acinetobacter 388 baumannii: an emerging challenge. Journal of clinical and diagnostic research 9:DC08-10 389 DOI: 10.7860/jcdr/2015/11014.5398. 390 Baugh S, Ekanayaka AS, Piddock LJ, and Webber MA. 2012. Manuscript to be reviewed 358 baumannii strains (Song & Cheong 2015). However, the influence of tigecycline on biofilm 359 formation of A. baumannii is contrary to the results of this study. Meropenem plus sulbactam 360 exhibited synergism against biofilm-embedded carbapenem-resistant A. baumannii and caused 361 significantly more damage to biofilm architecture than any of the agents used alone (Wang et al. 362 2016). 363 The images of SEM and CLSM were used in this study to support the findings by XTT assay 364 about the impact of gene knockout and adding antibiotics on biofilm formation, but no live-dead 365 stain was performed. Live-dead stain consists of SYTO9, which stains live cells green, and 366 propidium iodide (PI), which stains dead cells red (Richmond et al. 2016). This allows live 367 versus dead cells to be detected and is more informative. Although we made an effort to 368 characterize biofilm production of A. baumannii in different conditions, there were still some 369 limitations in this study. As previously described, the interplay of these related genes in biofilm 370 formation of A. baumannii and key secondary messenger c-di-GMP determination were not 371 explored further. Besides, because the blood or tissue concentrations produced by infusions of 372 the studied antimicrobials were not determined, it is hard to infer that the in vitro results will 373 apply to the in vivo condition. 374 375 Conclusions 376 In this study, the efflux pump gene emrA/emrB and two-component system gene baeR were 377 found to be associated with biofilm formation and sub-MIC tigecycline led to increased biofilm 378 formation in A. baumannii for the first time. In conclusion, the biofilm formation ability of A. 379 baumannii was diverse in different strains, involved many genes and is influenced by many 380 chemical compounds. The minimum biofilm eradication concentrations of all the tested 381 antibiotics were increased for both the wild-type and clinical isolate of multidrug resistant A. 382 baumannii VGH2. 383 384 385 References 386 387 Badave GK, and Kulkarni D. 2015. Biofilm producing multidrug resistant Acinetobacter 388 baumannii: an emerging challenge. Journal of clinical and diagnostic research 9:DC08-10 389 DOI: 10.7860/jcdr/2015/11014.5398. 390 Baugh S, Ekanayaka AS, Piddock LJ, and Webber MA. 2012. Loss of or inhibition of all 391 multidrug resistance efflux pumps of Salmonella enterica serovar typhimurium results in 392 impaired ability to form a biofilm. The Journal of antimicrobial chemotherapy 67:2409-2417 393 DOI: 10.1093/jac/dks228. Manuscript to be reviewed To explore the mechanism by which sub-MIC tigecycline increased the 346 biofilm formation of A. baumannii, more future relevant studies are needed. 347 Wang et al. have shown that the minimal bactericidal concentrations for biofilm-embedded 348 cells of the tested isolates were more than 50-fold higher than those for their planktonic cells 349 (Wang et al. 2016). The lowest biofilm inhibitory concentration (BIC) value of clinical multidrug 350 resistant A. baumannii (MDRAB) isolates for colistin was 32-fold the MIC90 (minimum 351 inhibitory concentration for 90% of the isolates) value, and BIC90 values were found to be 512- 352 fold the MIC90 value (Milletli Sezgin, Coban & Gunaydin 2013). This result is compatible with 353 those in this study. Because treatment for biofilm-forming MDRAB infection has become more 354 difficult, several regimens have been proposed as anti-biofilm treatment for A. baumannii 355 infection. Treatment with imipenem and rifampicin individually or in combination has obvious 356 anti-biofilm effects (Wang et al. 2014). Song et al. demonstrated that tigecycline and colistin- 357 rifampicin are effective for the prevention of or reduction in biofilm formation caused by A. PeerJ reviewing PDF \| (2019:10:42050:4:0:NEW 20 Mar 2020) Manuscript to be reviewed 400 Dafopoulou K, Xavier BB, Hotterbeekx A, Janssens L, Lammens C, De E, Goossens H, 401 Tsakris A, Malhotra-Kumar S, and Pournaras S. 2015. Colistin-resistant Acinetobacter 402 baumannii clinical strains with deficient biofilm formation. Antimicrobial agents and 403 chemotherapy 60:1892-1895 DOI: 10.1128/aac.02518-15. 404 Djeribi R, Bouchloukh W, Jouenne T, and Menaa B. 2012. Characterization of bacterial 405 biofilms formed on urinary catheters. American journal of infection control 40:854-859 DOI: 406 10.1016/j.ajic.2011.10.009. 407 Donlan RM. 2000. Role of biofilms in antimicrobial resistance. ASAIO Journal 46:S47-52 DOI: 408 10.1097/00002480-200011000-00037. 409 Espinal P, Marti S, and Vila J. 2012. Effect of biofilm formation on the survival of 410 Acinetobacter baumannii on dry surfaces. The Journal of hospital infection 80:56-60 DOI: 411 10.1016/j.jhin.2011.08.013. 412 Gaddy JA, and Actis LA. 2009. Regulation of Acinetobacter baumannii biofilm formation. 413 Future microbiology 4:273-278 DOI: 10.2217/fmb.09.5. 414 Gaddy JA, Tomaras AP, and Actis LA. 2009. The Acinetobacter baumannii 19606 OmpA 415 protein plays a role in biofilm formation on abiotic surfaces and in the interaction of this 416 pathogen with eukaryotic cells. Infection and immunity 77:3150-3160 DOI: 417 10.1128/IAI.00096-09. 418 Giannouli M, Antunes LC, Marchetti V, Triassi M, Visca P, and Zarrilli R. 2013. 419 Virulence-related traits of epidemic Acinetobacter baumannii strains belonging to the 420 international clonal lineages I-III and to the emerging genotypes ST25 and ST78. BMC 421 infectious diseases 13:282 DOI: 10.1186/1471-2334-13-282. 422 He X, Lu F, Yuan F, Jiang D, Zhao P, Zhu J, Cheng H, Cao J, and Lu G. 2015. Biofilm 423 formation caused by clinical Acinetobacter baumannii isolates is associated with 424 overexpression of the AdeFGH efflux pump. Antimicrobial agents and chemotherapy 425 59:4817-4825 DOI: 10.1128/aac.00877-15. 426 Hendiani S, Abdi-Ali A, and Mohammadi P. 2014. Comparison of two methods for 427 quantification of Acinetobacter baumannii biofilm formation. Biological journal of 428 microorganism 2:51-56. g 429 Hu Y, He L, Tao X, Meng F, and Zhang J. 2016. Biofilm may not be necessary for the 430 epidemic spread of Acinetobacter baumannii. Scientific reports 6:32066 DOI: 431 10.1038/srep32066. 432 Jin Y, Yip HK, Samaranayake YH, Yau JY, and Samaranayake LP. 2003. Biofilm-forming 433 ability of Candida albicans is unlikely to contribute to high levels of oral yeast carriage in 434 cases of human immunodeficiency virus infection. Journal of clinical microbiology 41:2961- 435 2967 DOI: 10.1128/jcm.41.7.2961-2967.2003. 436 Jones RN, Barry AL, Fuchs PC, and Thornsberry C. 1986. Manuscript to be reviewed Loss of or inhibition of all 391 multidrug resistance efflux pumps of Salmonella enterica serovar typhimurium results in 392 impaired ability to form a biofilm. The Journal of antimicrobial chemotherapy 67:2409-2417 393 DOI: 10.1093/jac/dks228. 394 Bay DC, Stremick CA, Slipski CJ, and Turner RJ. 2017. Secondary multidrug efflux pump 395 mutants alter Escherichia coli biofilm growth in the presence of cationic antimicrobial 396 compounds. Research in microbiology 168:208-221 DOI: 10.1016/j.resmic.2016.11.003. 397 CLSI. 2012. Performance Standards for Antimicrobial Susceptibility Testing; Twenty-Second 398 Informational Supplement. CLSI document M100-S22. Wayne, PA: Clinical and Laboratory 399 Standards Institute; 2012. 397 CLSI. 2012. Performance Standards for Antimicrobial Susceptibility Testing; Twenty-Second 398 Informational Supplement. CLSI document M100-S22. Wayne, PA: Clinical and Laboratory pp 399 Standards Institute; 2012. PeerJ reviewing PDF \| (2019:10:42050:4:0:NEW 20 Mar 2020) Manuscript to be reviewed 445 Leblanc SK, Oates CW, and Raivio TL. 2011. Characterization of the induction and cellular 446 role of the BaeSR two-component envelope stress response of Escherichia coli. Journal of 447 bacteriology 193:3367-3375 DOI: 10.1128/JB.01534-10. 448 Lee HW, Koh YM, Kim J, Lee JC, Lee YC, Seol SY, Cho DT, and Kim J. 2008. Capacity of 449 multidrug-resistant clinical isolates of Acinetobacter baumannii to form biofilm and adhere to 450 epithelial cell surfaces. Clinical microbiology and infection 14:49-54 DOI: 10.1111/j.1469- 451 0691.2007.01842.x. 452 Lin MF, Lin YY, and Lan CY. 2015. The role of the two-component system baesr in disposing 453 chemicals through regulating transporter systems in Acinetobacter baumannii. PloS one 454 10:e0132843 DOI: 10.1371/journal.pone.0132843. 455 Lin MF, Lin YY, and Lan CY. 2017a. Contribution of EmrAB efflux pumps to colistin 456 resistance in Acinetobacter baumannii. Journal of microbiology 55:130-136 DOI: 457 10.1007/s12275-017-6408-5. 458 Lin MF, Lin YY, Tu CC, and Lan CY. 2017b. Distribution of different efflux pump genes in 459 clinical isolates of multidrug-resistant Acinetobacter baumannii and their correlation with 460 antimicrobial resistance. Journal of microbiology, immunology, and infection 50:224-231 461 DOI: 10.1016/j.jmii.2015.04.004. 461 DOI: 10.1016/j.jmii.2015.04.004. 462 Lin MF, Lin YY, Yeh HW, and Lan CY. 2014. Role of the BaeSR two-component system in 463 the regulation of Acinetobacter baumannii adeAB genes and its correlation with tigecycline 464 susceptibility. BMC microbiology 14:119 DOI: 10.1186/1471-2180-14-119. 465 Lin MF, Tsai PW, Chen JY, Lin YY, and Lan CY. 2015. OmpA binding mediates the effect 466 of antimicrobial peptide LL-37 on Acinetobacter baumannii. PloS one 10:e0141107 DOI: 467 10.1371/journal.pone.0141107. 468 Liu C, Sun D, Zhu J, and Liu W. 2018. Two-component signal transduction systems: a major 469 strategy for connecting input stimuli to biofilm formation. Frontiers in microbiology 9:3279 470 DOI: 10.3389/fmicb.2018.03279. 471 Loehfelm TW, Luke NR, and Campagnari AA. 2008. Identification and characterization of an 472 Acinetobacter baumannii biofilm-associated protein. Journal of bacteriology 190:1036-1044 473 DOI: 10.1128/jb.01416-07. 473 DOI: 10.1128/jb.01416-07. 474 Longo F, Vuotto C, and Donelli G. 2014. Biofilm formation in Acinetobacter baumannii. The 475 new microbiologica 37:119-127. 476 Luo LM, Wu LJ, Xiao YL, Zhao D, Chen ZX, Kang M, Zhang Q, and Xie Y. 2015a. 477 Enhancing pili assembly and biofilm formation in Acinetobacter baumannii ATCC19606 478 using non-native acyl-homoserine lactones. BMC microbiology 15:62 DOI: 10.1186/s12866- 479 015-0397-5. 480 Luo TL, Rickard AH, Srinivasan U, Kaye KS, and Foxman B. 2015b. Manuscript to be reviewed Antimicrobial activity of 437 cefmetazole (CS-1170) and recommendations for susceptibility testing by disk diffusion, 438 dilution, and anaerobic methods. Journal of clinical microbiology 24:1055-1059. Ö 439 Kalkanci A, and TunÇCan Ö. 2019. Biofilm related infection: diagnosis, treatment and 440 prevention. Mediterranean journal of infection microbes and antimicrobials (in press) DOI: 441 10.4274/mjima.galenos.2019.2019.18. 442 Kim HA, Ryu SY, Seo I, Suh SI, Suh MH, and Baek WK. 2015. Biofilm formation and 443 colistin susceptibility of Acinetobacter baumannii isolated from Korean nosocomial samples. 444 Microbial drug resistance 21:452-457 DOI: 10.1089/mdr.2014.0236. PeerJ reviewing PDF \| (2019:10:42050:4:0:NEW 20 Mar 2020) Manuscript to be reviewed 489 Nett JE, Cain MT, Crawford K, and Andes DR. 2011. Optimizing a Candida biofilm 490 microtiter plate model for measurement of antifungal susceptibility by tetrazolium salt assay. 491 Journal of clinical microbiology 49:1426-1433 DOI: 10.1128/jcm.02273-10. 492 Nucleo E, Steffanoni L, Fugazza G, Migliavacca R, Giacobone E, Navarra A, Pagani L, and 493 Landini P. 2009. Growth in glucose-based medium and exposure to subinhibitory 494 concentrations of imipenem induce biofilm formation in a multidrug-resistant clinical isolate 495 of Acinetobacter baumannii. BMC microbiology 9:270 DOI: 10.1186/1471-2180-9-270. 496 Olson ME, Ceri H, Morck DW, Buret AG, and Read RR. 2002. Biofilm bacteria: formation 497 and comparative susceptibility to antibiotics. Canadian journal of veterinary research 66:86- 498 92. 499 Orsinger-Jacobsen SJ, Patel SS, Vellozzi EM, Gialanella P, Nimrichter L, Miranda K, and 500 Martinez LR. 2013. Use of a stainless steel washer platform to study Acinetobacter 501 baumannii adhesion and biofilm formation on abiotic surfaces. Microbiology 159:2594-2604 502 DOI: 10.1099/mic.0.068825-0. 503 Pachon-Ibanez ME, Jimenez-Mejias ME, Pichardo C, Llanos AC, and Pachon J. 2004. 504 Activity of tigecycline (GAR-936) against Acinetobacter baumannii strains, including those 505 resistant to imipenem. Antimicrobial agents and chemotherapy 48:4479-4481 DOI: 506 10.1128/aac.48.11.4479-4481.2004. 507 Perez LR. 2015. Acinetobacter baumannii displays inverse relationship between meropenem 508 resistance and biofilm production. Journal of chemotherapy 27:13-16 DOI: 509 10.1179/1973947813y.0000000159. 510 Richmond GE, Evans LP, Anderson MJ, Wand ME, Bonney LC, Ivens A, Chua KL, 511 Webber MA, Sutton JM, Peterson ML, and Piddock LJ. 2016. The Acinetobacter 512 baumannii two-component system aders regulates genes required for multidrug efflux, 513 biofilm formation, and virulence in a strain-specific manner. mBio 7:e00430-00416 DOI: 514 10.1128/mBio.00430-16. 515 Rodríguez-Baño J, Martí S, Soto S, Fernández-Cuenca F, Cisneros JM, Pachón J, Pascual 516 A, Martínez-Martínez L, McQueary C, Actis LA, Vila J, and Infections. SGftSoN. 2008. 517 Biofilm formation in Acinetobacter baumannii: associated features and clinical implications. 518 Clinical microbiology and infection 14:276-278 DOI: 10.1111/j.1469-0691.2007.01916.x. gy f j 519 Rumbo-Feal S, Gomez MJ, Gayoso C, Alvarez-Fraga L, Cabral MP, Aransay AM, 520 Rodriguez-Ezpeleta N, Fullaondo A, Valle J, Tomas M, Bou G, and Poza M. 2013. Whole 521 transcriptome analysis of Acinetobacter baumannii assessed by RNA-sequencing reveals 522 different mRNA expression profiles in biofilm compared to planktonic cells. PloS one 523 8:e72968 DOI: 10.1371/journal.pone.0072968. 524 Shi P, Gao Y, Lu Z, and Yang L. 2014. Effect of antibacterial peptide LL-37 on the integrity of 525 Acinetobacter baumannii biofilm. Manuscript to be reviewed Association of 481 blaOXA-23 and bap with the persistence of Acinetobacter baumannii within a major 482 healthcare system. Frontiers in microbiology 6:182 DOI: 10.3389/fmicb.2015.00182. 483 Marti S, Rodriguez-Bano J, Catel-Ferreira M, Jouenne T, Vila J, Seifert H, and De E. 484 2011. Biofilm formation at the solid-liquid and air-liquid interfaces by Acinetobacter species. 485 BMC research notes 4:5 DOI: 10.1186/1756-0500-4-5. 486 Milletli Sezgin F, Coban AY, and Gunaydin M. 2013. Investigation of biofilm formation in 487 Acinetobacter baumannii isolates and their colistin susceptibilities in biofilm. International 488 journal of antimicrobial agents 41:199 DOI: 10.1016/j.ijantimicag.2012.10.003. 487 Acinetobacter baumannii isolates and their colistin susceptibilities in biofilm. International 488 journal of antimicrobial agents 41:199 DOI: 10.1016/j.ijantimicag.2012.10.003. PeerJ reviewing PDF \| (2019:10:42050:4:0:NEW 20 Mar 2020) Manuscript to be reviewed Nan fang yi ke da xue xue bao 34:426-429. 526 Song JY, and Cheong HJ. 2015. In vitro comparison of anti-biofilm effects against 527 carbapenem-resistant Acinetobacter baumannii: imipenem, colistin, tigecycline, rifampicin 528 and combinations. Infection & chemotherapy 47:27-32 DOI: 10.3947/ic.2015.47.1.27. 529 Szczuka E, Jablonska L, and Kaznowski A. 2017. Effect of subinhibitory concentrations of 530 tigecycline and ciprofloxacin on the expression of biofilm-associated genes and biofilm 531 structure of Staphylococcus epidermidis. Microbiology 163:712-718 DOI: 532 10.1099/mic.0.000453. 533 Tomaras AP, Flagler MJ, Dorsey CW, Gaddy JA, and Actis LA. 2008. Characterization of a 534 two-component regulatory system from Acinetobacter baumannii that controls biofilm PeerJ reviewing PDF \| (2019:10:42050:4:0:NEW 20 Mar 2020) Manuscript to be reviewed Manuscript to be reviewed Manuscript to be reviewed Manuscript to be reviewed 535 formation and cellular morphology. Microbiology 154:3398-3409 DOI: 536 10.1099/mic.0.2008/019471-0. 537 Wang Y, Bao W, Guo N, Chen H, Cheng W, Jin K, Shen F, Xu J, Zhang Q, Wang C, An Y, 538 Zhang K, Wang F, and Yu L. 2014. Antimicrobial activity of the imipenem/rifampicin 539 combination against clinical isolates of Acinetobacter baumannii grown in planktonic and 540 biofilm cultures. World journal of microbiology & biotechnology 30:3015-3025 DOI: 541 10.1007/s11274-014-1728-7. 542 Wang YC, Kuo SC, Yang YS, Lee YT, Chiu CH, Chuang MF, Lin JC, Chang FY, and 543 Chen TL. 2016. Individual or combined effects of meropenem, imipenem, sulbactam, 544 colistin, and tigecycline on biofilm-embedded Acinetobacter baumannii and biofilm 545 architecture. Antimicrobial agents and chemotherapy 60:4670-4676 DOI: 10.1128/aac.00551- 546 16. 547 Wroblewska MM, Sawicka-Grzelak A, Marchel H, Luczak M, and Sivan A. 2008. Biofilm 548 production by clinical strains of Acinetobacter baumannii isolated from patients hospitalized 549 in two tertiary care hospitals. FEMS immunology and medical microbiology 53:140-144 DOI: 550 10 1111/j 1574-695X 2008 00403 x y p 550 10.1111/j.1574-695X.2008.00403.x. 551 Yoon EJ, Chabane YN, Goussard S, Snesrud E, Courvalin P, De E, and Grillot-Courvalin 552 C. 2015. Contribution of resistance-nodulation-cell division efflux systems to antibiotic 553 resistance and biofilm formation in Acinetobacter baumannii. mBio 6:e00309-00315. DOI: 554 10.1128/mBio.00309-15. 555 Zimmermann M, Escrig S, Hubschmann T, Kirf MK, Brand A, Inglis RF, Musat N, Muller 556 S, Meibom A, Ackermann M, and Schreiber F. 2015. Phenotypic heterogeneity in 557 metabolic traits among single cells of a rare bacterial species in its natural environment 558 quantified with a combination of flow cell sorting and NanoSIMS. Frontiers in microbiology 559 6:243 DOI: 10.3389/fmicb.2015.00243. PeerJ reviewing PDF \| (2019:10:42050:4:0:NEW 20 Mar 2020) PeerJ reviewing PDF \| (2019:10:42050:4:0:NEW 20 Mar 2020) Manuscript to be reviewed Figure 1 Bioﬁlm quantiﬁcation of the 33 diﬀerent strains, including 21 clinical isolates of A. baumannii (highlighted by asterisk markers). There are obvious diﬀerences in bioﬁlm formation among the clinical isolates of A. baumannii. PeerJ reviewing PDF \| (2019:10:42050:4:0:NEW 20 Mar 2020) Manuscript to be reviewed Figure 2 Comparison of the bioﬁlm formation indices of the susceptible and resistant strains to the individual antibiotics among A. baumannii ATCC 17978 and 21 clinical isolates. (A-O) The antimicrobial susceptibility of A. baumannii isolates was mainly determined according to the MIC interpretive criteria published by the CLSI in 2012 (M100-S22). The MICs of individual antibiotics against the clinical isolates were shown in Table S1. No relationship with statistical signiﬁcance was demonstrated between the ability to form bioﬁlm and the antimicrobial susceptibility of the A. baumannii clinical isolates. The Student’s t-test (p <0.05) was used to determine the statistical signiﬁcance of the experimental data. Abbreviations: SAM=ampicillin/sulbactam; TZP=piperacillin/tazobactam; CFZ=cefazolin; CMZ=cefmetazole; CTX=cefotaxime; CAZ=ceftazidime; FEP=cefepime; IPM=imipenem; MEM=meropenem; AMK=amikacin; GEN=gentamicin; CIP=ciproﬂoxacin; LVX=levoﬂoxacin; SXT=trimethoprim/sulfamethoxazole; TGC=tigecycline; NE= not evaluated; S=susceptible; R=resistant; BFI=bioﬁlm formation index. (A-O) The antimicrobial susceptibility of A. baumannii isolates was mainly determined according to the MIC interpretive criteria published by the CLSI in 2012 (M100-S22). The MICs of individual antibiotics against the clinical isolates were shown in Table S1. No relationship with statistical signiﬁcance was demonstrated between the ability to form bioﬁlm and the antimicrobial susceptibility of the A. baumannii clinical isolates. The Student’s t-test (p <0.05) was used to determine the statistical signiﬁcance of the experimental data. Abbreviations: SAM=ampicillin/sulbactam; TZP=piperacillin/tazobactam; CFZ=cefazolin; CMZ=cefmetazole; CTX=cefotaxime; CAZ=ceftazidime; FEP=cefepime; IPM=imipenem; MEM=meropenem; AMK=amikacin; GEN=gentamicin; CIP=ciproﬂoxacin; LVX=levoﬂoxacin; SXT=trimethoprim/sulfamethoxazole; TGC=tigecycline; NE= not evaluated; S=susceptible; R=resistant; BFI=bioﬁlm formation index. PeerJ reviewing PDF \| (2019:10:42050:4:0:NEW 20 Mar 2020) PeerJ reviewing PDF \| (2019:10:42050:4:0:NEW 20 Mar 2020) Manuscript to be reviewed PeerJ reviewing PDF \| (2019:10:42050:4:0:NEW 20 Mar 2020) Manuscript to be reviewed Figure 3 Bioﬁlm formation by A. baumannii ATCC 17978 after gene knockout; (A) comparison between the wild-type stain and its mutants (B-E) SEM and (F-Q) CLSM images. ΔbaeR, ΔemrA, ΔemrB and ΔompA all had less bioﬁlm formation than the wild type. The results are shown as the means ± SD from three independent experiments. The Student’s t- test was used to determine the statistical signiﬁcance of the experimental data. , p < 0.05 and , p < 0.01 and *, p < 0.001 between ATCC 17978 and the gene knockout mutant strains. PeerJ reviewing PDF \| (2019:10:42050:4:0:NEW 20 Mar 2020) Manuscript to be reviewed Manuscript to be reviewed PeerJ reviewing PDF \| (2019:10:42050:4:0:NEW 20 Mar 2020) Manuscript to be reviewed Figure 4 Inﬂuence of diﬀerent compounds on bioﬁlm formation in A. baumannii ATCC 17978; (A) amikacin, (B) imipenem, (C) colistin, (D) tigecycline, (E) LL-37, and (F) tannic acid. The addition of amikacin (> 0.5 μg/mL), colistin (> 0.5 μg/mL), LL-37 (> 4 μg/mL), or tannic acid (> 50 μg/mL) decreased the bioﬁlm formation ability of A. baumannii ATCC 17978, whereas the addition of tigecycline (> 62.5 ng/mL) increased the bioﬁlm formation ability of A. baumannii ATCC 17978. The Student’s t-test was used to determine the statistical signiﬁcance of the experimental data. , p < 0.05 and , p < 0.01 and , p < 0.001 between A. baumannii ATCC 17978 with and without chemical compounds. PeerJ reviewing PDF \| (2019:10:42050:4:0:NEW 20 Mar 2020) Manuscript to be reviewed PeerJ reviewing PDF \| (2019:10:42050:4:0:NEW 20 Mar 2020) Manuscript to be reviewed Figure 5 Inﬂuence of diﬀerent compounds on bioﬁlm formation in A. baumannii VGH2; (A) amikacin, (B) imipenem, (C) colistin, (D) tigecycline, (E) LL-37, and (F) tannic acid. The addition of amikacin (> 256 μg/mL), imipenem (> 4 μg/mL), colistin (> 0.125 μg/mL), LL-37 (> 2 μg/mL), or tannic acid (> 50 μg/mL) decreased the bioﬁlm formation ability of VGH2, whereas the addition of tigecycline (> 0.5 μg/mL) increased the bioﬁlm formation ability of VGH2. The Student’s t-test was used to determine the statistical signiﬁcance of the experimental data. , p < 0.05 and **, p < 0.01 between A. baumannii VGH2 with and without chemical compounds. PeerJ reviewing PDF \| (2019:10:42050:4:0:NEW 20 Mar 2020) Manuscript to be reviewed PeerJ reviewing PDF \| (2019:10:42050:4:0:NEW 20 Mar 2020) PeerJ reviewing PDF \| (2019:10:42050:4:0:NEW 20 Mar 2020) Figure 6 SEM and CLSM images of bioﬁlm formation by A. baumannii ATCC 17978 after colistin and tigecycline challenge, (A-F) SEM and (G-O) CLSM. SEM and CLSM images of bioﬁlm formation by A. baumannii ATCC 17978 after colistin and tigecycline challenge, (A-F) SEM and (G-O) CLSM. The SEM image shows the cells become scattered and the CLSM images exhibit decreased thickness of bioﬁlm. In contrast, 125 ng/mL tigecycline exposure has the opposite impact on bioﬁlm formation of A. baumannii ATCC 17978. PeerJ reviewing PDF \| (2019:10:42050:4:0:NEW 20 Mar 2020) Manuscript to be reviewed PeerJ reviewing PDF \| (2019:10:42050:4:0:NEW 20 Mar 2020) Manuscript to be reviewed Table 1(on next page) Table 1(on next page) Bacterial strains used in this study. PeerJ reviewing PDF \| (2019:10:42050:4:0:NEW 20 Mar 2020) Manuscript to be reviewed Manuscript to be reviewed 1 2 Table 1. Bacterial strains used in this study Strain Relevant feature(s) Source or reference 19606 A. baumannii Wild-type strain ATCC 17978 A. baumannii Wild-type strain ATCC ∆baeR Derived from ATCC 17978. baeR mutant obtained by kanr gene replacement (Lin et al. 2014) ∆emrA Derived from ATCC 17978. A1S_1800 mutant obtained by kanr gene replacement (Lin, Lin & Lan 2017a) ∆emrB Derived from ATCC 17978. A1S_1772 mutant obtained not by kanr gene replacement (Lin, Lin & Lan 2017a) ∆ompA Derived from ATCC 17978. ompA mutant obtained by kanr gene replacement (Lin et al. 2015) ABamk Induced amikacin resistant ATCC 17978 (Lin et al. 2017b) ABipm Induced imipenem resistant ATCC 17978 (Lin et al. 2017b) ABcol Induced colistin resistant ATCC 17978 (Lin et al. 2017b) ABtc Induced tigecycline resistant ATCC 17978 (Lin et al. 2014) ABtcm Derived from ABtc. baeR mutant obtained by kanr gene replacement (Lin et al. 2014) ABhl1 Tigecycline resistant clinical isolate from Hualien Tzu Chi Hospital (Lin et al. 2014) ABhl1tc ABhl1 with induced high tigecycline resistance (Lin, Lin & Lan 2015) CT11-14 MDRAB isolate from the Chut-Tung branch of National Taiwan University Hospital (Lin et al. 2017b) CM1-4 MDRAB isolate from Catholic Mercy Hospital (Lin et al. 2017b) HC1-5 MDRAB isolate from the Hsin-Chu branch of National Taiwan University Hospital (Lin et al. 2017b) VGH1-7 MDRAB isolate from the Hsin-Chu branch of Taipei Veterans General Hospital (Lin et al. 2017b) 3 4 PeerJ reviewing PDF \| (2019:10:42050:4:0:NEW 20 Mar 2020) Manuscript to be reviewed Table 2(on next page) Table 2(on next page) Table 2(on next page) PeerJ reviewing PDF \| (2019:10:42050:4:0:NEW 20 Mar 2020) Manuscript to be reviewed Manuscript to be reviewed 1 2 Table 2. MIC and MBEC of A. baumannii ATCC 17978 and VGH2. ATCC 17978 VGH2 　 MIC MBEC MIC MBEC Amikacin (μg/mL) 1 32 1024 >8192 Imipenem (μg/mL) 0.25 1 16 512 Colistin (μg/mL) 1 >8192 0.5 >8192 Tigecycline (μg/mL) 0.5 2 4 32 LL37 (μg/mL) 16 ND 8 ND Tannic acid (μg/mL) >300 ND >300 ND 3 ND: Not detected 2 Table 2. MIC and MBEC of A. baumannii ATCC 17978 and VGH2. PeerJ reviewing PDF \| (2019:10:42050:4:0:NEW 20 Mar 2020)
https://openalex.org/W2026266302	https://zenodo.org/records/1966653/files/article.pdf	German	null	Über die mittlere Parallaxe der Sterne 10. Größe	Astronomische Nachrichten	1,910	public-domain	5,916	ASTRON OM IS CHE NACHRICHTEN. d 185. Nr. 4430. Band 185. 14. I) Publications of the Washburn Observatory. Vol. XI1 (1908). Von H. E. Lou. bisher zmal gemessen. Der \v. F. einer Differenz wiirde so- niit im Mittel &-or18 betragen, was einem w. F. der Eigen- bewegungen voii i. o!'o I entspricht. In den Jahren 1907 -09 habe ich die Engelhardtschen Begleiter der Bradley- Sterne, in der .4bsicht, einen Beitrag zur Ermittelung ihrer Eigenbewegungen zu liefern, am I o- Zoller der Urania-Sternwarte gemessen. Obgleich das Zeit- interval1 im Mittel nur 19.0 Jahre betragt, habe ich doch geglaubt, die Messungen zu einer vorlaufigen Priifung der Cottistockschen Resultate I) benutzeii zu sollen. Eigenbewegungen bei der Ausgleichung zu bestimmen. Die Engelhardtschen Bradley -Sterne diirften in der Tat bei kiinftigen Untersuchungen iiber die Eigenbewegungen der lichtschwachen Sterne von der groilten Bedeutung werden. Die Eigenbeweguugen der Hauptsterne sind mit grofler Ge- nauigkeit bestiniint; nach Prof. Boss betragen die w. F. der Eigenbewegungen im Mittel nur 012 im lahrhundert. Ferner wurden in das Engelhardtsche Verzeichnis nur Sterne init Eigenbewegungen uber o!' I jiihrlich aufgenonimen ; ein end- giiltiges Urteil iiber den Charakter der relativen Bewegun- gen diirfte daher schoii nach wenigen Jahrzehnten moglich sein. Die Engelhardtschen Sterne sind weiter sanitlich von ei n e ni Beobachter gemessen worden, wodurch die Eliinination der konstanten'Messungsfehler erheblich erleichtert wird. Es diirfte schlieillich noch ein grofler Vorteil dieser Sterne sein, dafl bei der Auswahl auf die Groile der relatireii Bewegungen keine Riicksicht genominen wurde, weshalb man zu der An- nahme berechtigt ist, dail die Engelhardtschen Sterne wirklich ein typisches Bild der Verhaltnisse der Sterne 10'" liefern konnen. Die folgende Tafel enthalt die Mittel der Messungen der optischen Paare korrigiert fur Refraktion und bezogen auf das Aqiiinoktiuni 1900.0. Br. 3 2 1 2 Rr. -32 I z a -4ndromedae Br. 90 47 Cassiopeiae o Ceti 9 Persei Br. 536 37 A1 Tauri I 89 2 . I) Publications of the Washburn Observatory. Vol. XI1 (1908). ') Observations astronomiques faites par B. d'Engelhardt. Dresde. 1886 -95. Die Messungen der Bradley-Sterne finden sich teils irn zweiten Bande S. 34-80 teils irn dritten Eande S. 171-189. ) y ( ) ') Observations astronomiques faites par B. d'Engelhardt. Dresde. 1886 -95. Die Messun irn zweiten Bande S. 34-80 teils irn dritten Eande S. 171-189. Von H. E. Lou. 3 4 1908.43 iS91.78 19oS.89 1887.08 I 908.07 1908.65 1887.26 1888.06 1908.42 1888.14 1910.11 I 89 1.58 1887.08 1887.08 189 1.55 1908.94 1908.43 1908.43 1908.43 248' 5 I' 2 5 0 34 I S 2 58 IS5 I 1 273 3 0 2 7 5 40 1 7 1 36 ' 7 ' I4 196 40 198 16 81 2 0 79 ' 5 2 2 0 18 2 2 5 39 '53 5 5 753 42 188 19 189 10 2 41 3 56 353!'96 I 58.38 '55.96 I 53.28 7 1.74 74.83 I 24.00 I 26.89 94.68 95.39 116.27 I 16.94 70.06 74,54 95.71 137.41 168.51 95.08 137.2 I 169.59 Die Messungen dieser lichtschwachen Sterne maren am Urania-Refraktor, namentlich in den hellen Soniniernachten, recht schwierig. Bei ruhigen Bildern ist die Luft hier meistens dunstig, und die Begleiter I 0. bis I 2 . Grofle sind dann kaum sichtbar. Bei sehr reiner Luft sind die Bilder dagegen fast immer wallend, und die Hauptsterne erscheinen dann haufig als explodierende Lichtmassen. Ob die Luftverhaltnisse in Dresden besser gewesen sind, erscheint nach v. Engelhardts Bemerkungen im ersten Teile seiner Beobachtungen ) sehr fraglich. n i e w. F. meiner Messungen, die mit roter Feld- beleuchtung ausgefiihrt werden muflten, zeigen ein rascheres Anwachsen init abnehniender Helligkeit als die Engelhardt- schen, wie aus dem folgenden Tafelchen ersichtlich ist : des Eegleiters s ' dp d s s.dp ds Helligkeit V. Engelhardt Lau 8115- 9m5 L-ol'15 &-0l'13 +o!'r4 =t0!'14 10.5-11.5 r t 0 . 2 5 f0.15 f 0 . 3 4 zt0.28 9.5-10.5 O . Z I t o . 1 1 f o . 2 0 o.zo I) Publications of the Washburn Observatory. Vol. XI1 (1908). 15 15 44 30 2 1 6 2 1 j 2 1 j 45 'I'auri 7 7c1 Orionis 1 3 Orionis CL Aurigac I. Aurigae I I I 'I'auri d Aurigae 7 I Orionis 7 I Orionis 7 4 k Orionis 8 1,yncis 56 U 5 Aurigae 58 qi Aurigae 2 0 Monoceroris lk. 1048 2 2 1.yncis Q Gerninorum 2 5 Monoccrotis 14 Canis min. r q Canis min. o Tirsac maj. o Lrsae niaj. 2 2 y1 Cancri 39 Cancri 39 Cancri y Cancri 10 Crsae maj. 67 Cancri 75 Cancri Br. Von H. E. Lou. 1306 1890.4 I I 908.09 I 891.09 1908.98 I 888.02 I 909.07 1888.26 1888.26 1888.08 1908.92 1908.93 '908.93 I S y 1.2 j '9C8.93 1886.21 , 1909.16 I 886.2 3 I 909.16 1890.24 1909.06 1888.29 I909.04 1888.32 1891.27 I 909.16 I 890.2 j 1909.07 r890.22 1909.16 i y o y . 16 I 909. I 6 I 909. I 6 I90').05 1891.47 1909.07 1891.28 ' 909.30 1909.30 I 886.2 I 1909.26 1909.3 I 1909.25 1890.7 5 1909.26 1888.27 1909.26 I 8 8 8.3 2 1909.26 1887.38 1909.26 I 890.38 1909.3' 1907.1 2 1890.27 1890.24 1890.7 3 189 1.47 IS91.2() 1890.7 5 1890.7 j 43O I 5' 42 16 248 6 248 45 261 15 264 4 2 1 4 j 43 144 29 10 57 5 3' 2 7 1 24 2 7 0 16 2 7 1 6 2 0 2 28 2 0 2 '7 265 o 254 2 2 7 0 58 86 44 87 53 81 j 2 79 5' 2 2 2 5 24 2 7 106 40 1 0 2 46 339 39 339 2 7 '07 44 ' 0 7 55 303 54 303 37 292 34 291 1 7 349 16 349 44 73 5.3 7 5 38 r j r 5 2 1 j 0 j 6 I 5 2 2 2 '5' 1 7 208 6 2 0 7 3' 2 0 56 2 0 30 309 1 2 309 4 110 33 I 1 0 41 6 j j I 65 58 2 0 2 32 324 '9 324 59 41 I 39 18 4 13 4 48 2 0 0 I I 28Y36 1 2 j.67 172.19 171.6j 12j.93 1 j4.98 146.40 127.06 '38.34 7 7 . 1 1 82.08 113.84 1 1 5.49 3 I .(I8 27.78 80.74 87.30 124.14 205.90 '03.99 1 5 2 . 2 7 I .5 7 .O 5 46.63 43.83 '89.93 40.93 40.93 186.62 1 2 7 . j 3 127.99 166.60 168.94 2 12.07 2 13.87 122.5 j 121.66 84.42 86.76 I 14.65 I I 7.82 142.73 143.88 1 7 6.5 9 174.69 I 26.85 I 29.08 134.74 133.88 134.98 134.45 103.26 10 j.23 136.47 129.47 101.47 1 0 2 . 1 2 104.30 147.32 144.39 I 11.57 44 30 I 40 T.yncis 2 3 h I!rsae inaj. 41 Lyncis 41 Lyncis o Leonis y Leonis 62 Lrsac maj. Von H. E. Lou. 6 Lrsac maj. x Virginis x Virginis Virginis 33 Virginis l3r. 1820 Br. 1820 I 8 Bootis 18 Librac 45 Bootis 4 j Bootis o Coronae bor. + Serpcntis I j Hcrculis (r Coronae bor. 4 1 Hcrculis Rr. 2134 30 Ophiuchi Br. 2166 83 Merculis ct 1,yrac a Lyrae 1 1 I Hcrculis I S86.34 1889.06 1909.26 1909.3 I 1888.30 1907.32 1888.30 '907.32 1888.30 1909.26 1888.30 1888.33 I 908.7 6 1890.50 1 909.3 3 I 890.33 1908.74 1908.74 I 8 90.3 3 "08.74 1890.33 190s.74 1887~31 1908.25 '909.33 I 9 09.3 0 1890.96 '909.3 1 1 890.90 1888.53 '909.34 I 889.04 '909.33 I 89 I .ox I 890.96 1891.04 1909.30 '909.47 1887.50 '909.32 189 1.01 1891.14 1891.14 1909.36 1909.29 I909.36 1887.59 1909.38 1891.00 1909.38 '909.38 1891.1 I 1890.66 1909.60 1886.7 5 1905.46 1886.76 1903.55 1886.76 1909.60 36" 7' 37 2 2 2 7 36 2 2 7 49 161 41 161 30 80 39 79 37 41 39 42 26 288 9 288 28 298 59 290 5 2 73 15 73 8 137 44 137 39 1 1 0 39 I I O 40 330 1 2 330 35 187 30 190 42 292 30 293 13 2 ' 5 44 214 50 216 49 2 1 7 3 40 2 0 40 1 2 52 55 49 2 1 43 53 336 42 337 45 279 50 280 37 259 15 258 6 87 6 86 9 2 1 5 8 246 23 219 35 219 1 1 69 4 0 69 3' 267 33 259 2 8 156 53 163 29 38 56 247 8 245 5 5 55 16 266 44 260 18 39 I 1 2 1 6 208105 2 10.80 91.17 93.7 X I . 2 1 79.59 83.58 83.52 82.00 83.86 100.44 98.39 97.59 191.76 I 89.62 '72.95 I 7 2 . g h 2 2 1.06 2 2 2 . 5 3 321.03 321.39 I 7 I..55 130.73 127.02 '59.07 156.04 154.68 I j 6.44 160.32 163.38 108.19 107.37 251.67 246.32 147.78 I 46.8 I 1 7 2 . 1 7 172.28 84.30 178.93 65.97 65.56 j7.89 63.95 165.82 I 58.9 I I 50.6 I 92.17 93.24 I j2.30 159.72 49.5q 53.50 130.66 96.22 99.20 150.42 151.78 '57.96 124.52 45 'I'auri 7 7c1 Orionis 1 3 Orionis CL Aurigac I. Von H. E. Lou. Aurigae I I I 'I'auri d Aurigae 7 I Orionis 7 I Orionis 7 4 k Orionis 8 1,yncis 56 U 5 Aurigae 58 qi Aurigae 2 0 Monoceroris lk. 1048 2 2 1.yncis Q Gerninorum 2 5 Monoccrotis 14 Canis min. r q Canis min. o Tirsac maj. o Lrsae niaj. 2 2 y1 Cancri 39 Cancri 39 Cancri y Cancri 10 Crsae maj. 67 Cancri 75 Cancri Br. 1306 1890.4 I I 908.09 I 891.09 1908.98 I 888.02 I 909.07 1888.26 1888.26 1888.08 1908.92 1908.93 '908.93 I S y 1.2 j '9C8.93 1886.21 , 1909.16 I 886.2 3 I 909.16 1890.24 1909.06 1888.29 I909.04 1888.32 1891.27 I 909.16 I 890.2 j 1909.07 r890.22 1909.16 i y o y . 16 I 909. I 6 I 909. I 6 I90').05 1891.47 1909.07 1891.28 ' 909.30 1909.30 I 886.2 I 1909.26 1909.3 I 1909.25 1890.7 5 1909.26 1888.27 1909.26 I 8 8 8.3 2 1909.26 1887.38 1909.26 I 890.38 1909 3' 1907.1 2 1890.27 1890.24 1890.7 3 189 1.47 IS91.2() 1890.7 5 1890.7 j 43O I 5' 42 16 248 6 248 45 261 15 264 4 2 1 4 j 43 144 29 10 57 5 3' 2 7 1 24 2 7 0 16 2 7 1 6 2 0 2 28 2 0 2 '7 265 o 254 2 2 7 0 58 86 44 87 53 81 j 2 79 5' 2 2 2 5 24 2 7 106 40 1 0 2 46 339 39 339 2 7 '07 44 ' 0 7 55 303 54 303 37 292 34 291 1 7 349 16 349 44 73 5.3 7 5 38 r j r 5 2 1 j 0 j 6 I 5 2 2 2 '5' 1 7 208 6 2 0 7 3' 2 0 56 2 0 30 309 1 2 309 4 110 33 I 1 0 41 6 j j I 65 58 2 0 2 32 324 '9 324 59 41 I 39 18 4 13 4 48 2 0 0 I I 28Y36 1 2 j.67 172.19 171.6j 12j.93 1 j4.98 146.40 127.06 '38.34 7 7 . 1 1 82.08 113.84 1 1 5.49 3 I .(I8 27.78 80.74 87.30 124.14 205.90 '03.99 1 5 2 . 2 7 I .5 7 .O 5 46.63 43.83 '89.93 40.93 40.93 186.62 1 2 7 . Von H. E. Lou. j 3 127.99 166.60 168.94 2 12.07 2 13.87 122.5 j 121.66 84.42 86.76 I 14.65 I I 7.82 142.73 143.88 1 7 6.5 9 174.69 I 26.85 I 29.08 134.74 133.88 134.98 134.45 103.26 10 j.23 136.47 129.47 101.47 1 0 2 . 1 2 104.30 147.32 144 39 I 11.57 2 1 7 I I I Herculis I I I Herculis I I Aquilae E Aquilae E Aquilae 17 Lyrae 17 Lyrae I 7 Lprae 17 Lyrae 31 Aquilae 31 Aquilae p Aquilae p Aquilae 1 7 Cygni ct Aquilae I 9 Cygni 1 5 Sagittae b3 Cygni 15 Delphini 18 Delphini d Equulei 1890.13 1909.60 1890.13 1909.60 I 886.7 I 1908.26 1891.15 1909.6 3 1887.58 1909.63 1892.50 1909.68 1891.29 1909.66 1909.68 1909.66 189 1.33 1891.29 1887.5') 1909.64 1887.59 1909.65 1887.65 1909.66 1890.68 1887.81 1908.6 I 1886.7 I 1905.60 '908.59 I 8 9 1 .3 3 1908.59 1908.90 1886.78 1890.16 1908.66 1890.65 1 890.3 7 1907.68 1886.87 '908.59 1908.59 257'33' 256 2.1 264 33 263 48 187 1 5 261 11 269 I 186 37 163 37 163 6 7 0 7 64 45 296 9 295 33 116 9 116 38 357 34 332 1 2 32' 42 353 '3 343 34 344 19 7 0 3' 142 I 140 3 305 23 1 0 2 30 104 28 2 7 2 2 1 274 53 23 1 7 23 8 81 14 8 2 32 162 28 162 3 2 1 2 5 '7 15 356 59 349 38 7 1 36 309 56 I 19l16 121.32 131.67 733.52 16.66 17.08 132.59 148.57 131.14 149.23 133.49 124.84 I 26.58 152.43 I 50.09 168.16 117.09 113.13 I 18.00 105.57 64.20 182.61 119.27 148.8 I 141.43 161.85 54.1 I 53.72 193.66 2 2 5 . 1 1 109.00 108.11 '97.36 197.24 40.25 45.87 108.97 167.19 59.60 157.56 201.43 2 2 2 . 2 2 4430 Br. 2926 E Cephei 35 Pegasi 35 Pegasi Br. 3077 Br. 3094 1 2 Andromed. Br. Von H. E. Lou. 3107 il Piscium 74 Pegasi 85 Pegasi 1891.41 1908.7 2 189 1.41 1908.7 2 1891.59 1908.7 2 1892.13 1908.72 1887.91 1908.7 2 189 1.42 1890.9s 1908.73 189 1.36 I 908~7 2 I 89 1.36 I 908.7 2 1890.96 1908.87 1886.99 1908.72 1908.90 36" 8' 34 26 '9 54 '7 '9 208 29 2 1 0 2 8 2 4 0 '3 2 4 1 36 183 5' 2 0 7 47 303 ' 5 301 7 '55 40 322 33 319 2 1 344 45 344 1 7 35 47 34 26 366 7 338 18 256 5 2 1 8 79%9 75.02 132.65 129.15 1 0 1 . 5 1 98.31 181.54 88.36 106.57 8 1 . 2 5 84.10 I 19.02 I 20.54 7 5.09 7 7 . 7 7 181.79 159.87 161.77 99.9" 97.74 2 1 . 1 5 45.47 Ein Vergleich der GroOenschatzungen mit den Harvard Messungen l) von 3 5 Sternen gibt folgende Beziehungen : Lau Harvard Engelhardt Harvard 9.41 = 9.53 i 9.32 = 9.47 8m77 = 9mo2 8"9 = 8?49 10.04 = 9.80 9.89 = 9.9s 10.59 = 1 0 . 2 5 I 10.60 = 10.37 Die Ergebnisse der Beobachtungen der optischen Paar sind in der folgenden Tabelle zusanimengestellt. Es enthal die erste Spalte die Bezeichnung des Hauptsternes; die zweit die Harvard-GroOe des Begleiters nach v. Engelhardts und meinen eigenen Schatzungen; die dritte und vierte die jahr liche relative Bewegung in n.cos6 und 6; die fiinfte und sechste die absoluten Eigenbewegungen des Hauptsternes die beiden letzten Spalten die entsprechenden Groflen fii den Begleiter. Die Angaben iiber die Eigenbewegung de Hauptsterne verdanke ich einer brieflichen Mitteilung de Herrn Prof. Boss. 11. E i g e n b e w e g u n g e n d e r E n g e 1 h a r d t s c h en Brad 1 e y - Stern e. Name GroDe Jahrl. rel. Bewegung Br. 3212 9m7 -0!'359 +Or178 >> 10.6 -0.337 +0.181 a Andromedae 9.9 -0.136 +0.143 Br. 90 9.2 + 0 . 0 7 2 -0.160 9 Persei 10.7 -0.359 +0.060 47 Cassiopeiae 9.7 - 0 . 1 2 7 t o . 0 0 6 o Ceti 9.6 -0.004 +0.212 Br. 536 9.4 +0.036 -0.024 39 A? Tauri 8.2 -0.186 -1-0.060 45 Tauri 10.5 -0.184 -0,026 -0.012 + O . ') Annals of Harvard College Observatory. Vol. 64 NO. VI. Von H. E. Lou. I I Z ') Annals of Harvard College Observatory Vol 64 NO VI 37 A1 Tauri 9.4 -0.093 +O.O24 m1 Orionis 9.1 E.B. d. Hauptsternes +or380 -07176 +0.380 -0.176 +o.140 -0.161 +0.339 -0.090- -0.001 -0.237 -0.023 -0.013 +0.093 -0.064 +0.174 -0.138 +0.142 +0.013 - 0 . 0 7 2 +0.174 t 0 . 1 3 3 -0.054 + 0.05 5 -0. I34 E.B. d. Begleiters +or021 t o l o o 2 +0.043 +0.005 +0.004 -0.018 0.000 fo.014 - 0 . o 2 0 -0.030 + 0.006 - 0.048 - 0.005 - 0.02 5 +o.o13 -0.037 0.000 -0.040 -0.042 -0.013 - 0 . 0 1 2 -0.078 +0.043 -0.022 2 1 8 2 1 7 Ein Vergleich der GroOenschatzungen mit den Harvard- Messungen l) von 3 5 Sternen gibt folgende Beziehungen : Lau Harvard Engelhardt Harvard 9.41 = 9.53 i 9.32 = 9.47 8m77 = 9mo2 8"9 = 8?49 10.04 = 9.80 9.89 = 9.9s 10.59 = 1 0 . 2 5 I 10.60 = 10.37 Die Ergebnisse der Beobachtungen der optischen Paare sind in der folgenden Tabelle zusanimengestellt. Es enthalt die erste Spalte die Bezeichnung des Hauptsternes; die zweite die Harvard-GroOe des Begleiters nach v. Engelhardts und meinen eigenen Schatzungen; die dritte und vierte die jahr- liche relative Bewegung in n.cos6 und 6; die fiinfte und sechste die absoluten Eigenbewegungen des Hauptsternes ; die beiden letzten Spalten die entsprechenden Groflen fiir den Begleiter. Die Angaben iiber die Eigenbewegung der Hauptsterne verdanke ich einer brieflichen Mitteilung des Herrn Prof. Boss. 2 2 0 4430 Name 1 3 Orionis a Aurigae 3, Aurigae 1 1 1 Tauri 6 Aurigae 7 I Orionis 74 k Orionis 8 Lyncis 56 q5 Aurigae 58 q7 Aurigae 2 o lLlonocerotis Br. 1048 2 2 I.yncis g Geminoruin 2 5 Monocerotis 14 Canis min. o Ursae rnaj. y1 Cancri 39 Caiicri y Cancri 10 Ursae inaj. 67 Caiicri 7 5 Cancri Br. 1306 40 Lyncis lz Ursae maj. 41 Lyncis o Leonis y Leonis .62 Ursae maj. Q Ursae inaj. x Virginis )> )) 2 )) 8 n >) 33 Virginis Br. 1820 18 Bootis 18 Librae 45 Bootis .o Coronae bor. 9 Serpentis 1 5 Herculis 6 Coronae bor. 41 Herculis Br. 2134 30 Ophiuchi Br. 2166 8 3 Herculis a Lyrae I 11 Herculis >) 2 3 GroOe Jiihrl. rel. Von H. E. Lou. 1 I O ~ I +0!'041 9.8 -0.109 9.7 -0.524 9.4 -0.239 9.8 --0.092 10.6 +0.074 1 1 . 2 - 0 . 1 5 1 9.4 - 0.09 I 9.0 +o.186 8.4 +0.019 10.1 +0.040 9.6 +0.029 10.5 +0.028 9.8 -0.128 10.5 +0.062 9.3 t o . 1 6 7 9.4 +o.r81 10.2 +0.103 10.4 +0.136 10.4 - 0.00 5 9.2 0.000 9.3 -0.033 9.7 +0.106 9.9 +0.380 9.1 +0.028 9.5 f O . 1 0 1 9.8 +0.066 9.5 +0.189 10.8 - a 1 0 5 8.5 -0.010 1 0 . 1 -0.017 10.7 +O.IOO 9.3 +0.104 9.; +0.341 10.0 -0.1 I 5 9.2 + O . O I O 9.0 +0.038 10.7 + 0 . 1 9 2 10.9 t o . 1 7 9 9.8 -0.074 10.4 +0.092 10.5 -0.182 10.3 f 0 . 0 5 5 9.1 - 0.405 10.8 - 0 . 1 5 5 9.7 - 0.995 9.7 +0.157 10.5 t o . 0 1 7 9.8 +0.036 10.0 +0.274 9.6 f o . 2 2 2 9.1 +0.038 9.9 +0.050 9.8 -0.035 9.5 +0.070 9.0 -0.174 10.1 -0.229 10.7 - 0.084 Bewegung +or365 -1-0.430 +0.627 -0.024 +0.096 +0.168 -0.744 - 0 . 2 2 3 +0.296 -0.171 + 0 . 1 5 1 - 0 . 1 7 7 -0.016 +0.032 -0.198 - 0.038 -0.108 - 0. I 04 + 0 . 1 2 7 + 0.047 +0.105 - 0.02 2 0.000 +0.024 +o.z I 0 +0.058 + 0.349 -0.161 fo.010 --- 0.07 2 +0.086 +0.077 +0.030 - 0.006 -0.013 +0.008 +0.006 +0.074 +0.423 +o.org -0.016 -0.056 t o . 1 4 0 +o.148 +0.670 +0.008 i-o.127 - 0.066 +0.068 +0.278 + 0.03 3 +O.I I 7 +0.137 - 0.305 -0.235 -0.135 - 0 . 1 2 1 - 0.044 E.B. d. Ha +01'006 t o . 0 8 6 +0.530 +0.245 +0.086 - 0,098 - 0.098 +0.088 +0.008 -0.018 -0.194 0.000 - 0.00 I +O.I 18 -0.070 -0.159 -0.159 +0.149 - 0 . 1 2 1 - 0 . 1 2 1 -0.030 -0.038 0.038 - 0 . 1 0 2 -- 0.43 I ---0.058 -0.116 - 0.038 -0.217 +o.114 - 0.003 - 0.003 -0.145 - 0 . 1 1 2 -0.346 + O . I I 0 - 0.076 - 0.07 6 -0.076 f - 0 . 2 7 2 -0.164 -0.164 +0.108 -0.106 +o.186 to.186 - 0 . Von H. E. Lou. 1 2 5 - 0.085 - 0.050 -- 0 . 2 7 I -0.218 - 0.03 I - 0.054 - 0.0 I 5 - 0.063 +0.204 +0.204 +0.063 uptsternes - 01'38 I -0.656 -0.126 -0.429 -0.010 - 0 . 2 0 1 -0.201 +0.193 - 0.28 I +o.158 -0.135 +0.217 -0.016 - 0.08 5 +0.183 +0.018 +0.095 +0.095 --o.114 -0.1 14 -0.131 - 0 . 0 2 2 - 0 . 0 2 2 -0.050 -0.261 - 0.088 -0.387 +0.135 + O . O I O + 0 . 0 2 5 -0.136 -0.136 - 0.039 -- 0.044 +0.01 I +0.003 - 0.037 - 0.037 - 0.037 -0.456 - 0.030 -- 0.030 -- 0.038 - 0.076 -0.183 -0.183 - 0.058 -0.150 +0.030 - 0.090 --0.277 -0.019 - 0.088 -0.115 +0.279 +0.279 +0.106 -- 0. I34 E.B. d. Begleiters +0!0$7 -01'016 --0.023 +O.OOI + 0.006 -0.029 +O.OOG -0.034 - 0.006 - 0.030 -0.024 -0.033 -0.249 -0.945 -0.003 -0.030 - 0.008 +o.o I 5 + 0 . 0 2 7 -0.013 + 0 . 0 2 2 +0.016 +0.029 +0.040 + 0 . 0 2 7 --0.032 -0.010 -0.053 -0.033 -0.015 -0.008 - - 0 . 0 2 0 t o . 0 0 8 - 0.0 I 3 +o.oz 2 - 0.009 -0.018 +o.o13 +o.o I 5 - 0.067 - 0 . 0 3 5 -0.026 - 0.038 - 0.044 - 0 . 0 7 1 - 0 . 0 2 2 +0.004 -0.026 -c.og I -0.05 I - 0.030 - 0.030 - 0.0 I 5 -0.038 +0.028 -0.026 -0.028 +o.020 fo.009 -0.047 -0.013 - 0 . 0 5 0 -0.020 -0.059 - 0.045 - 0.009 -0.008 - 0 . 0 5 0 -0.005 -0.110 - 0.00 5 --0.066 -0.029 - 0.02 I - 0.03 I -0.133 -0.033 +0.028 -0.015 +o.o I 5 - 0.046 -0.014 +0.064 +o.o31 -0.035 -0.809 +0.487* +0.032 --o.050 -0.068 -0.023 - 0.0 I 4 - 0.036 + a 0 0 3 - 0 . 0 2 2 +0.004 +O.OOI + O . O O ~ -0.063 - 0.004 -- 0.05 5 -.0.050 --o.oI~ to.0o.i + 0 . Von H. E. Lou. 0 2 2 +0.030 -0.026 -0.021 -0.029 -- 0.0 I 0 - 0.038 +0.037 +0.034 -0.094 -0.025 +0.044 1 4430 S a n e I I I Herculis i I Aquilae t A4quilae 17 Lyrae >> )) >> )) 2 31 Aquilae p Aquilae 1 7 Cygni 01 Aquilae 19 Cygni 1 5 Sagittae b" Cygni 1 5 Delphini 18 Delphini d Equulei Hr. 2926 E Cephei 35 Pegasi Br. 301 7 >> >) )) Br. 3094 GroOe 9'!'9 10.4 9.7 10.3 1 0 . 7 10.8 9.2 1 0 . 2 9.8 10.1 9.3 1 1 . 5 9.5 9.2 9.5 10.6 9. I 9.5 10.4 10.4 9.5 10.1 10.1 1 0 . 2 1 0 . 1 9.6 10.0 1 2 Andromed. 9.4. x Piscium 10.3 74 Pegasi 10.4 Br. 3107 10.5 85 Pegasi 9.5 Jfhrl. rel. Bewegiing --0!'082 -0!'152 -0.088 -0.101 -0.028 +0.104 +o.roo t o . 0 6 7 +0.067 -0.01 I +I.057 + 1 , 1 g 1 -0.123 -0.036 --0.138 +0.009 -0.1 1 1 - 0 . 0 2 5 - 0 . 7 2 0 -0.659 -0.732 -0.724 -0.301 +0.144 -0.238 + O . I ~ O --0.037 +0.426 -0.590 -0.393 -0.046 -0.097 t o . 3 8 1 +0.377 -0.091 -0.131 + 0 . 0 7 7 t o . 0 3 1 -- 0.05 I t o . 2 9 2 -0.260 -0.139 -0.388 -0.083 -- 0.084 f 0 . 2 6 2 --0.110 + 0 . 2 2 5 - 2 . 1 0 2 -0.294 -0.234 -0.063 -0.094 + 0 . 0 2 7 -0.286 -0.037 -0.176 - 0 . 0 2 5 - 0 . 8 7 0 +0.968 -0.030 -0.143 -0.102 to.085 E.B. d. Hauptsternes +or063 +0!'106 +0.063 +0.106 + O . O O I -0.120 -0.064 - 0 . 0 7 7 -0.064 - 0 . 0 7 7 + 0 . 1 2 0 +o.o13 f 0 . 1 2 0 t o . 0 1 3 + O . I Z O +0.013 fo.120 +o.o13 +0.725 +0.630 + 0 . 7 2 5 +0.630 +0.213 -0.152 t o . 2 1 3 - 0 . 1 5 2 +o.oII -0.452 +0.535 +0.380 + O . O I I +0.097 - 0.404 - 0.4 I 3 t o . ') Vergl. Hertzsprung, Zeitschrift ftir wissenschaftliche Photographie. Band 111, S. 43 I ff. Von H. E. Lou. 0 6 5 +0.068 f0.062 +0.096 -0.065 -0.053 + 0.04 I - 0.303 '0.239 +0.127 +0.076 -0.316 +0.076 -0.316 + 2 . 0 8 2 +0.296 +0.204 +0.03 I t o . 1 1 5 -0.078 +0.318 -0.019 +0.085 -0.090 + O . I 2 0 -0.002 +0.835 -0.986 + O . l j O f0.044 E. B. d. Begleiters -0!'019 --Or046 -0.025 +o.oog - 0 . 0 2 7 -0.016 +0.036 -0.010 t o . 0 0 3 - 0.066 +-1.177 +1.184* -0.003 -0.023 -0.018 + 0 . 0 2 2 + 0.009 - 0.0 I 2 +o.005 -0.029 - 0.007 - 0.094 - 0.088 -- 0.008 - 0 . 0 2 5 -0.032 - 0 . 0 5 5 -0.013 -0.035 0.000 - 0.023 - 0.036 -0.026 -0.063 + 0 . 0 3 2 -0.047 + O . O I Z --0.022 --0.026 --0.026 -0.010 -0.011 - 0 . 0 2 1 - 0 . 0 1 2 +0.062 -0.039 -0.008 -0.054 - 0.034 - 0.09 I -0.020 +0.002 -o.ogo -0.032 + O . O Z I -0.051 t o . 0 3 2 -0.056 - 0.01 7 -0.005 -0.056 - 0 . 0 ~ 7 -0.035 -0.018 2 2 2 2 2 1 4430 der hellen HauDtsterne herriihren. Ich fiiide namlich. daO die scheinbaren Eigenbewegungen von dem Positionswinkel p des Begleiters unabhangig sind : P Prccosd W fl 2709 --o!'oo~ -0Or014 9 57.9 -0.011 -0.017 1 0 88.3 - 0 . 0 0 2 -0.019 8 130.3 -0.022 -0.013 9 174.0 0.000 -0.024 I I P P I C C O S J ' Pd n 2 1 0 0 2 -o!'o17 -0Yo41 1 1 256.3 -0.008 - 0 . 0 2 5 1 2 283.9 -0.011 -0.036 1 0 315.8 - 0 . 0 2 0 -0.028 8 351.9 -0.006 -0.027 I 1 Bei der Ausgleichung habe ich daher in den Bedingungs- gleichungen auDer der scheinbaren Sonnenbewegung p und Bei der Ausgleichung habe ich daher in den Bedingungs- gleichungen auDer der scheinbaren Sonnenbewegung p und dem Korrektionsfaktor I + ~ der Newconibschen Pra- zessionskonstante noch die konstanten Fehler z, uiid zd der abgeleiteten Eigenbewegungen aufgenommen. T10 = 0!'0010. T10 = 0!'0010. Dieser Wert ist erheblich kleiner als der von Comstock gefundene. Zugunsten rneiner Bestirnmung mochte ich hervor- heben, dafl die systernatischen Fehler soweit als rnoglich be- riicksichtigt sind, was bei Untersuchungen dieser Art eine unabweisbare Forderung sein mu&. Aus den rohen Eigen- bewegungen der Engelhardt- Sterne wiirde man wegen der systematischen Fehler der pj als Sakularbewegung der Sonne 3" statt 0!'4 erhalten; man hatte somit aus meinen Daten selbst bei einern Zeitintervall von einern halben Jahrhundert bei Vernachlassigung der systernatischen Fehler ein rnit Com- stocks iibereinstimmendes Resultat herausbringen konnen. Aus den Gleichungen in pg -46.09 +30.19 +z.49 = +0.0546 +99.002d - 4 6 . 0 9 p + 0 . 0 7 ~ = -0of0440 + 0.07 + 2.49 t 6 . 6 1 = + O . O I 79 Die Auflosung gibt: = -01'00347 3g = +or0015 ,u = +o.o0412 2, = +0.0028 Bei der Ausgleichung habe ich allen Gleichungen das Gewicht I beigelegt. Dieses Verfahren wird durch die Ver- teilung der Restfehler gerechtfertigt. Als w. F. einer Gleichung findet man narnlich fur verschiedene Helligkeiten der Begleiter : Die vorliegende Untersuchung ergibt in Ubereinstim- rnung rnit Comstock, dafl die Eigenbewegungen der licht- schwachen Sterne grofler sind, als man nach ihrer Helligkeit bezw. Parallaxe erwarten sollte. Die Tatsache, daO die syste- matischen Fehler der Messungen scheinbare Bewegungen von uber 2" im Jahrhundert bewirken konnen, mun jedoch gegen ein allzugrofies Vertrauen in die Ergebnisse der Mikrorneter- messungen warnen. Helligkeit a cos B 6 8m5- 9m5 +0!'024 f0?019 9.5 -10.5 fo.016 k0.020 10.5 - 1 1 . 5 fo.020 t o . 0 1 9 Eine Beziehung zwischen den Restfehlern und der galaktischen Breite ,8 tritt nicht hervor. Die w. F. ftir ver- schiedene Breiten sind : g Nach den bisherigen Untersuchungen scheinen die Sterne des Typus I besonders haufig in der Milchstrafle zu sein. Es durfte daher vielleicht mehr als ein Zufall sein, daf3 die gefundene mittlere Parallaxe der Sterne 10. Crone init der Annahme stimmt, dafl die lichtschwachen Sterne dem Typus I angehocen. Aus Kapteyns Angaben 3, findet man rnit Camp- bells Sonnengeschwindigkeit die mittlere Parallaxe der Sterne des Typus I fur die GroDe 5"s gleich 01'0086; die Parallaxe der Sterne 10. Grofle ware sornit nlo = O!'OOII. Aus den gemessenen Parallaxen der A- Sterne findet Hertzsprung (1. c.) no = 0!'153 oder mlo = 01'001 5 . Von H. E. Lou. ( Bei der Ausgleichung habe ich daher in den Bedingungs- gleichungen auDer der scheinbaren Sonnenbewegung p und dem Korrektionsfaktor I + ~ der Newconibschen Pra- ( Bei der Ausgleichung habe ich daher in den Bedingungs- gleichungen auDer der scheinbaren Sonnenbewegung p und dem Korrektionsfaktor I + ~ der Newconibschen Pra- zessionskonstante noch die konstanten Fehler z, uiid zd der abgeleiteten Eigenbewegungen aufgenommen. ( 8"8 -0!'029 13 9.1 -0.024 I 2 9.4 -0.032 13 9.6 - 0 . 0 ~ 5 1 2 9.9 -0.014 1 2 1 0 . 2 -0.020 1 2 10.4 -0.026 13 - 508 --!'or3 13 + 5.0 -0.025 I 3 f14.1 -0.035 2 2 + 2 5 . 1 - 0 . 0 2 5 15 +34.6 - 0 . 0 2 2 14 +45.2 -0.030 10 + 5 5 . 2 -0.019 6 zessionskonstante noch die konstanten Fehler z, uiid zd der abgeleiteten Eigenbewegungen aufgenommen. ( g g g g g und - ergeben sich aus d(rrcos6) dd Die Koeffizienten d 'I dT g g g g g und - ergeben sich aus den Bossschen Prazessionen in 01 und d. Bei der Berechnung der Koeffizienten von p habe ich auf die verschiedenen Hellig- keiten der Sterne keine Riicksicht genommen. Bei einer vollig strengen Ausgleichung niunte man namlich eher auf die (un- bekannten) Spektraltypen der Begleiter Rucksicht nehmen als auf die nur bis f im5 reichenden Helligkeitsunterschiede. l) d(rrcos6) dd Die Koeffizienten d 'I dT g g g g g und - ergeben sich aus d(rrcos6) dd Die Koeffizienten d 'I dT den Bossschen Prazessionen in 01 und d. Bei der Berechnung der Koeffizienten von p habe ich auf die verschiedenen Hellig- keiten der Sterne keine Riicksicht genommen. Bei einer vollig strengen Ausgleichung niunte man namlich eher auf die (un- bekannten) Spektraltypen der Begleiter Rucksicht nehmen als auf die nur bis f im5 reichenden Helligkeitsunterschiede. l) 4430 4430 2 2 4 2 2 3 stellen sie vielleicht die letzte Stufe ())schwarze Sonnencc) der spektralen Entwickelung dar. Ila die Koeffizienten der Bedingungsgleichungen sich mittels der Kapteynschen Werte ') von x und A leicht herstellen lassen, unterlasse ich es, die I 98 Gleichungen hier anzufiihren. Urn bequeme Arbeitszahlen zu erhalten, habe ich vor der Ausgleichung vorlaufige Korrektionen von + o!'o I o bezw. +o!'oz 5 an die Eigenbewegungen angebracht. Die Korrektion der Newcombschen Prazessionskonstante 9 = -01'0035 stimmt befriedigend rnit der von Cornstock gefundenen iiberein. Von H. E. Lou. Die Prazessionskonstante fur I 850.0 wBie hiernach 50!'2418. Das Hauptinteresse knupft sich jedoch an den Wert von p. Mit Campbells Geschwindigkeit der Sonne ( 4 . 2 Erd- bahnhalbmesser) finden wir aus p = +o!'oo41z als mittlere Parallaxe der Sterne 10. Grone ') : Die partiellen Normalgleichungen sind : Aus den Gleichungen in p,,,,d Aus den Gleichungen in p,,,,d Aus den Gleichungen in p,,,,d +9q.ooa, - o . ~ S p + 7 9 . 8 7 q = -01'007o - 0.28 +z8.56 - 2.37 = t o . 1 1 8 3 +79.87 - 2.37 +68.50 = -0.0620 T10 = 0!'0010. T10 = 0!'0010. Wir sind daher vorlaufig nicht genotigt, eine besondere Konstitution der lichtschwachen Sterie oder eine rnerkbare Absorption im Weltenraume an- zunehmen. Gal. Breite Pin a cos 8 8 6oo-3o0S -41?6 koo"o17 f0!'019 30 - o S - 1 1 . 0 &o.ozz &o.o17 o - 3 o N + 1 3 . 7 fo.015 I o . 0 1 9 30 - 6 0 . N '44.6 f o . 0 2 0 ~ ! ~ o . o 1 9 Bei der Ausgleichung habe ich drei Sterne rnit ano- malen Eigenbewegungen ausgeschlossen. Eine partielle Aus- gleichung gab namlich fur diese Sterne p = + 1?3 2, wahrend die 99 ubrigen p = +o!'oo4 gaben. Diese drei Sterne bilden somit etwas fur sich. Da ihre absolute Helligkeit kaum ein Hunderttausendstel der Helligkeit der ubrigen Sterne betragt, H. E. Lou. Urania-Sternwarte, April 19 10. Urania-Sternwarte, April 19 10. Febr.4 5h55m405 -om19?70 -0' 1816 10,s V 21~44~10156 9.547 +4°52'~6!'4 0.831 - rS85 -io"7 4 5 58 56 7.65 +4 56.9 20,xo V 2 1 44 10.77 9.548 + 4 5 2 19.0 0.832 - 1 - 8 5 - 1 0 . 7 9 1 6 16 2 I yi 47.47 1 +4 41.4 1 I2,6 1 V I 2 1 5 2 35.7il9.558 1 +I 1 2 8.010.832 I -1.82 - 10.9 Urania-Sternwarte, April 19 10. I z 3 H. E. Lou. ') Publications of the Astronomical Laboratory at Groningen no. g (1902). ) Die mittlere Helligkeit der 99 Sterne ist 9.84 Harvard oder 10.01 Potsdarn. ') Astronomische Nachrichten Band 146, S. 110. Beobachtungen des Kometen 1910 a. 1-91, 1 M. Ortszeit I Acz 1 A6 I Vgl. I Bb. I cz app. 1logp.d 1 6 app. llogp.Al Red. ad 1. app. 1 Febr.4 5h55m405 -om19?70 -0' 1816 10,s V 21~44~10156 9.547 +4°52'~6!'4 0.831 - rS85 -io"7 4 5 58 56 7.65 +4 56.9 20,xo V 2 1 44 10.77 9.548 + 4 5 2 19.0 0.832 - 1 - 8 5 - 1 0 . 7 9 1 6 16 2 I yi 47.47 1 +4 41.4 1 I2,6 1 V I 2 1 5 2 35.7il9.558 1 +I 1 2 8.010.832 I -1.82 - 10.9 I z 3 Beobachtungen des Kometen 1910 a. 1-91, 1 M. Ortszeit I Acz 1 A6 I Vgl. I Bb. I cz app. 1logp.d 1 6 app. llogp.Al Red. ad 1. app. 1 *
https://openalex.org/W4249829857	https://f1000research.com/articles/9-1178/v2/pdf	English	null	The effect of non-pharmacologic strategies on prevention or management of intensive care unit delirium: a systematic review	F1000Research	2,021	cc-by	21,308	F1000Research 2021, 9:1178 Last updated: 06 SEP 2022 SYSTEMATIC REVIEW The effect of non-pharmacologic strategies on prevention or management of intensive care unit delirium: a systematic review [version 2; peer review: 1 approved, 1 approved with reservations] Julie S Cupka1,2, Haleh Hashemighouchani 1,2, Jessica Lipori1,2, Matthew M. Ruppert1,2, Ria Bhaskar1,2, Tezcan Ozrazgat-Baslanti1,2, Parisa Rashidi 2,3, Azra Bihorac 1,2 * Equal contributors Open Peer Review Approval Status 1 2 version 3 (revision) 24 Jun 2022 version 2 (revision) 24 Sep 2021 view version 1 28 Sep 2020 view Elizabeth Mahanna-Gabrielli , University of Miami Miller School of Medicine, Miami, USA 1. Claudia D. Spies, Charité - Universitätsmedizin Berlin, Berlin, Germany Fatima Yürek, Charité - Universitätsmedizin Berlin, Berlin, Germany 2. Open Peer Review Approval Status 1 2 version 3 (revision) 24 Jun 2022 version 2 (revision) 24 Sep 2021 view version 1 28 Sep 2020 view First published: 28 Sep 2020, 9:1178 https://doi.org/10.12688/f1000research.25769.1 Second version: 24 Sep 2021, 9:1178 https://doi.org/10.12688/f1000research.25769.2 Latest published: 24 Jun 2022, 9:1178 https://doi.org/10.12688/f1000research.25769.3 v2 First published: 28 Sep 2020, 9:1178 https://doi.org/10.12688/f1000research.25769.1 Second version: 24 Sep 2021, 9:1178 https://doi.org/10.12688/f1000research.25769.2 Latest published: 24 Jun 2022, 9:1178 https://doi.org/10.12688/f1000research.25769.3 v2 Abstract Background: Post-operative delirium is a common complication among adult patients in the intensive care unit. Current literature does not support the use of pharmacologic measures to manage this condition, and several studies explore the potential for the use of non- pharmacologic methods such as early mobility plans or environmental modifications. The aim of this systematic review is to examine and report on recently available literature evaluating the relationship between non-pharmacologic management strategies and the reduction of delirium in the intensive care unit. Methods: Six major research databases were systematically searched for articles analyzing the efficacy of non-pharmacologic delirium interventions in the past five years. Search results were restricted to adult human patients aged 18 years or older in the intensive care unit setting, excluding terminally ill subjects and withdrawal-related delirium. Following title, abstract, and full text review, 27 articles fulfilled the inclusion criteria and are included in this report. Results: The 27 reviewed articles consist of 12 interventions with a single-component investigational approach, and 15 with multi- component bundled protocols. Delirium incidence was the most commonly assessed outcome followed by duration. Family visitation was the most effective individual intervention while mobility Background: Post-operative delirium is a common complication among adult patients in the intensive care unit. Current literature does not support the use of pharmacologic measures to manage this condition, and several studies explore the potential for the use of non- pharmacologic methods such as early mobility plans or environmental modifications. The aim of this systematic review is to examine and report on recently available literature evaluating the relationship between non-pharmacologic management strategies and the reduction of delirium in the intensive care unit. Methods: Six major research databases were systematically searched for articles analyzing the efficacy of non-pharmacologic delirium interventions in the past five years. Search results were restricted to adult human patients aged 18 years or older in the intensive care unit setting, excluding terminally ill subjects and withdrawal-related delirium. Following title, abstract, and full text review, 27 articles fulfilled the inclusion criteria and are included in this report. Results: The 27 reviewed articles consist of 12 interventions with a single-component investigational approach, and 15 with multi- component bundled protocols. Delirium incidence was the most commonly assessed outcome followed by duration. Family visitation was the most effective individual intervention while mobility Elizabeth Mahanna-Gabrielli , University 1. Elizabeth Mahanna-Gabrielli , University 1. SYSTEMATIC REVIEW The effect of non-pharmacologic strategies on prevention or management of intensive care unit delirium: a systematic review [version 2; peer review: 1 approved, 1 approved with reservations] Julie S Cupka1,2, Haleh Hashemighouchani 1,2, Jessica Lipori1,2, Matthew M. Ruppert1,2, Ria Bhaskar1,2, Tezcan Ozrazgat-Baslanti1,2, Parisa Rashidi 2,3, Azra Bihorac 1,2 1Department of Medicine, University of Florida, Gainesville, FL, 32608, USA 2Precision and Intelligent Systems in Medicine (PrismaP), University of Florida, Gainesville, FL, 32608, USA 3Department of Biomedical Engineering, University of Florida, Gainesville, FL, 32608, USA * Equal contributors First published: 28 Sep 2020, 9:1178 https://doi.org/10.12688/f1000research.25769.1 Second version: 24 Sep 2021, 9:1178 https://doi.org/10.12688/f1000research.25769.2 Latest published: 24 Jun 2022, 9:1178 https://doi.org/10.12688/f1000research.25769.3 v2 Abstract Background: Post-operative delirium is a common complication among adult patients in the intensive care unit. Current literature does not support the use of pharmacologic measures to manage this condition, and several studies explore the potential for the use of non- pharmacologic methods such as early mobility plans or environmental modifications. The aim of this systematic review is to examine and report on recently available literature evaluating the relationship between non-pharmacologic management strategies and the reduction of delirium in the intensive care unit. Methods: Six major research databases were systematically searched for articles analyzing the efficacy of non-pharmacologic delirium interventions in the past five years. Search results were restricted to adult human patients aged 18 years or older in the intensive care unit setting, excluding terminally ill subjects and withdrawal-related delirium. Following title, abstract, and full text review, 27 articles fulfilled the inclusion criteria and are included in this report. Results: The 27 reviewed articles consist of 12 interventions with a single-component investigational approach, and 15 with multi- component bundled protocols. Delirium incidence was the most Open Peer Review Approval Status 1 2 version 3 (revision) 24 Jun 2022 version 2 (revision) 24 Sep 2021 view version 1 28 Sep 2020 view Elizabeth Mahanna-Gabrielli , University of Miami Miller School of Medicine, Miami, USA 1. Claudia D. Spies, Charité - Universitätsmedizin Berlin, Berlin, Germany Fatima Yürek, Charité - Universitätsmedizin Berlin, Berlin, Germany 2. F1000Research 2021, 9:1178 Last updated: 06 SEP 2022 SYSTEMATIC REVIEW The effect of non-pharmacologic strategies on prevention or management of intensive care unit delirium: a systematic review [version 2; peer review: 1 approved, 1 approved with reservations] Julie S Cupka1,2, Haleh Hashemighouchani 1,2, Jessica Lipori1,2, Matthew M. Ruppert1,2, Ria Bhaskar1,2, Tezcan Ozrazgat-Baslanti1,2, Parisa Rashidi 2,3, Azra Bihorac 1,2 1Department of Medicine, University of Florida, Gainesville, FL, 32608, USA 2Precision and Intelligent Systems in Medicine (PrismaP), University of Florida, Gainesville, FL, 32608, USA 3Department of Biomedical Engineering, University of Florida, Gainesville, FL, 32608, USA * Equal contributors First published: 28 Sep 2020, 9:1178 https://doi.org/10.12688/f1000research.25769.1 Second version: 24 Sep 2021, 9:1178 https://doi.org/10.12688/f1000research.25769.2 Latest published: 24 Jun 2022, 9:1178 https://doi.org/10.12688/f1000research.25769.3 v2 Abstract Background: Post-operative delirium is a common complication among adult patients in the intensive care unit. Current literature does not support the use of pharmacologic measures to manage this condition, and several studies explore the potential for the use of non- pharmacologic methods such as early mobility plans or environmental modifications. The aim of this systematic review is to examine and report on recently available literature evaluating the relationship between non-pharmacologic management strategies and the reduction of delirium in the intensive care unit. Methods: Six major research databases were systematically searched for articles analyzing the efficacy of non-pharmacologic delirium interventions in the past five years. Search results were restricted to adult human patients aged 18 years or older in the intensive care unit setting, excluding terminally ill subjects and withdrawal-related delirium. Following title, abstract, and full text review, 27 articles fulfilled the inclusion criteria and are included in this report. Results: The 27 reviewed articles consist of 12 interventions with a single-component investigational approach, and 15 with multi- component bundled protocols. Delirium incidence was the most Open Peer Review Approval Status 1 2 version 3 (revision) 24 Jun 2022 version 2 (revision) 24 Sep 2021 view version 1 28 Sep 2020 view Elizabeth Mahanna-Gabrielli , University of Miami Miller School of Medicine, Miami, USA 1. Claudia D. Spies, Charité - Universitätsmedizin Berlin, Berlin, Germany Fatima Yürek, Charité - Universitätsmedizin Berlin, Berlin, Germany 2. F1000Research 2021, 9:1178 Last updated: 06 SEP 2022 Keywords Keywords critical care, delirium, intensive care, non-pharmacologic, systematic review Abstract University of Miami Miller School of Medicine, Miami, Page 1 of 31 F1000Research 2021, 9:1178 Last updated: 06 SEP 2022 Alawi Luetz, Charité - Universitätsmedizin Berlin, Berlin, Germany Any reports and responses or comments on the article can be found at the end of the article. interventions were the least effective. Two of the three family studies significantly reduced delirium incidence, while one in five mobility studies did the same. Multi-component bundle approaches were the most effective of all; of the reviewed studies, eight of 11 bundles significantly improved delirium incidence and seven of eight bundles decreased the duration of delirium. Any reports and responses or comments on the article can be found at the end of the article. Conclusions: Multi-component, bundled interventions were more effective at managing intensive care unit delirium than those utilizing an approach with a single interventional element. Although better management of this condition suggests a decrease in resource burden and improvement in patient outcomes, comparative research should be performed to identify the importance of specific bundle elements. Corresponding author: Azra Bihorac (abihorac@ufl.edu) Author roles: Cupka JS: Data Curation, Formal Analysis, Investigation, Visualization, Writing – Original Draft Preparation, Writing – Review & Editing; Hashemighouchani H: Data Curation, Formal Analysis, Investigation, Methodology, Validation, Visualization, Writing – Original Draft Preparation, Writing – Review & Editing; Lipori J: Formal Analysis, Investigation, Visualization, Writing – Original Draft Preparation, Writing – Review & Editing; Ruppert MM: Visualization, Writing – Review & Editing; Bhaskar R: Investigation, Writing – Review & Editing; Ozrazgat-Baslanti T: Conceptualization, Data Curation, Formal Analysis, Funding Acquisition, Methodology, Resources, Supervision, Validation, Visualization, Writing – Review & Editing; Rashidi P: Conceptualization, Funding Acquisition, Methodology, Project Administration, Resources, Supervision, Validation, Writing – Review & Editing; Bihorac A: Conceptualization, Funding Acquisition, Methodology, Project Administration, Resources, Supervision, Validation, Writing – Review & Editing Competing interests: No competing interests were disclosed. Grant information: A.B. was supported by R01 GM110240 from the National Institute of General Medical Sciences (NIH/NIGMS), R01 EB029699 and R21 EB027344 from the National Institute of Biomedical Imaging and Bioengineering (NIH/NIBIB), R01 NS120924 from the National Institute of Neurological Disorders and Stroke (NIH/NINDS), and by R01 DK121730 from the National Institute of Diabetes and Digestive and Kidney Diseases (NIH/NIDDK). T.O.B. was supported by K01 DK120784, R01 DK123078, and R01 DK121730 from the National Institute of Diabetes and Digestive and Kidney Diseases (NIH/NIDDK), R01 GM110240 from the National Institute of General Medical Sciences (NIH/NIGMS), R01 EB029699 from the National Institute of Biomedical Imaging and Bioengineering (NIH/NIBIB), and R01 NS120924 from the National Institute of Neurological Disorders and Stroke (NIH/NINDS). P.R. was supported by National Science Foundation CAREER award 1750192, 1R01EB029699 and 1R21EB027344 from the National Institute of Biomedical Imaging and Bioengineering (NIH/NIBIB), R01GM-110240 from the National Institute of General Medical Science (NIH/NIGMS), 1R01NS120924 from the National Institute of Neurological Disorders and Stroke (NIH/NINDS), and by R01 DK121730 from the National Institute of Diabetes and Digestive and Kidney Diseases (NIH/NIDDK). g y e funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript. C i h © 2021 C k JS l Thi i i l di ib d d h f h C i C A ib i Li pka JS et al. Introduction Delirium is a multifactorial, acute, confusional state character- ized by the disturbance of consciousness and cognition; it is par- ticularly common in the intensive care unit (ICU) with incidence ranging from 19 to 87% with especially high rates in mechani- cally ventilated patients1–3. ICU delirium is associated with adverse outcomes including increased mortality, prolonged mechani- cal ventilation and hospitalization, increased risk of cognitive dysfunction after discharge, and increased cost of care4–7. In addition to our search terms, search results were restricted to articles published in English within five years of the date of the search (Jan 1, 2014 to May 15, 2019). After retrieving articles within a ten year window as outlined in Extended Data20, the authors decided to further reduce the range to within five years. This date range was chosen in order to pro- vide a review of the most recent developments in this field of research. After search results were compiled and duplicates were removed, a total of 5234 articles were selected for title and abstract review. Four authors (JSC, HH, MMR, RB) screened the titles and abstracts and retrieved articles for eligibility, result- ing in 113 articles selected for full text review. Three authors (JSC, HH, JL) then independently reviewed the full text of eli- gible articles, completed data extraction worksheets adapted from the Cochrane Review Group’s Data Extraction Form (https:// dplp.cochrane.org/data-extraction-forms), and assessed the arti- cles for risk of bias using the Cochrane risk of bias tool21. Ele- ments of the data extraction worksheet included study design and setting, participant characteristics, details of the interven- tion and control groups, diagnostic tools, and patient outcomes (Extended data, Supplementary Table 1)20. Any disagreements were resolved by thoroughly discussing any points of concern. During the full text review 86 articles were removed because they failed to meet our inclusion criteria resulting in a total of 27 included articles (Figure 1). While the pathophysiology of delirium is not well understood, there are multiple factors associated with an increased risk for developing delirium including age, neurologic or psycho- logical disorders, polypharmacy, medications, and sensory impairment8–11. Modifiable environmental risk factors includ- ing immobilization, use of restraints, isolation, and levels of environmental light and sound are also considered risk factors for the development of delirium in the ICU8,12. Amendments from Version 1 REVISED Search strategy and data extraction This version of the article updated the following sections: the timeframe surrounding the search strategy was clarified and specific dates were provided within the Methods section; the authors who screened articles and extracted data were specified within the Methods section; Table 2 was updated to provide more clarity to the reader. The Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines were followed in this review and included as Reporting guidelines19,20. The electronic data- bases of PubMed, Embase, Cochrane Central, Web of Science, The Cumulative Index to Nursing and Allied Health Literature (CINAHL), and ClinicalTrials.gov were systematically searched on May 15, 2019 for articles concerning non-pharmacologic treatments for delirium in the ICU. Search terms were tailored to each database in order to best utilize the individual subject headings, keywords, and medical subject headings (MeSH) terms included in the individual databases. A full list of search terms is shown in the Extended data20. Any further responses from the reviewers can be found at the end of the article Any further responses from the reviewers can be found at the end of the article Corresponding author: Azra Bihorac (abihorac@ufl.edu) This is an open access article distributed under the terms of the Creative Commons Attribution Licen tricted use, distribution, and reproduction in any medium, provided the original work is properly cited. pyright: © 2021 Cupka JS et al. This is an open access article distributed under the terms of the Creative Common hich permits unrestricted use, distribution, and reproduction in any medium, provided the original work is prope How to cite this article: Cupka JS, Hashemighouchani H, Lipori J et al. The effect of non-pharmacologic strategies on prevention or management of intensive care unit delirium: a systematic review [version 2; peer review: 1 approved, 1 approved with reservations] F1000Research 2021, 9:1178 https://doi.org/10.12688/f1000research.25769.2 First published: 28 Sep 2020, 9:1178 https://doi.org/10.12688/f1000research.25769.1 Page 2 of 31 F1000Research 2021, 9:1178 Last updated: 06 SEP 2022 Introduction The morbidity associated with delirium as well as the multitude of delirium risk factors present in the ICU make delirium pre- vention and management strategies essential. These strate- gies have included pharmacological, non-pharmacological, and multicomponent interventions with the aim of decreasing the incidence and duration of delirium. Research into pharmacologi- cal interventions has focused on haloperidol and dexmedetomi- dine, though there has also been limited research into the effects of ramelteon, melatonin, and ziprasidone13–16. Despite continued research, current literature does not support the use of anti-psychotic agents, benzodiazepines, or melatonin in the management of delirium13,17. Inclusion and exclusion criteria Our review addresses non-pharmacological management strategies for delirium in the ICU. Included articles were those investigating non-pharmacologic interventions and their impact on delirium incidence, prevalence, duration, or severity in an adult (≥ 18 years) intensive care unit setting. Articles were excluded if they focused on non-human subjects, pediatrics, terminally ill subjects, withdrawal related delirium, case reports, or where no full-text article was available (abstract only). There were no restrictions on study design. Studies solely investigating delirium-free-coma-free days were excluded since it is not pos- sible to review as a delirium-specific result. One multi-center study was excluded as both the frequency and method of assess- ment for delirium were not specified for all study centers, mak- ing it difficult to reliably compare the results with other trials22. Another study was excluded because neither the screening proc- ess nor the cohort were described other than total number of Given the lack of evidence supporting pharmacological measures, further research into the efficacy of non-pharmacologic interventions such as early mobilization, environmental modifi- cations, or management bundles is crucial. The implementation of effective delirium management shows promise in decreas- ing morbidity, mortality, length of stay, and resource burden in the ICU setting. In terms of the PICOS framework (Popula- tion, Interventions, Comparisons, Outcomes, Study Design)18, our systematic review aims to address the effects of any non- pharmacologic prevention or management strategy on the inci- dence, prevalence, duration, or severity of delirium in critically ill adult patients compared to control patients, with no restrictions on study design. Page 3 of 31 F1000Research 2021, 9:1178 Last updated: 06 SEP 2022 Figure 1. PRISMA Record Screening Flow Chart. Figure 1. PRISMA Record Screening Flow Chart. trial are located in Table 1. The 27 included studies provide results on many distinct outcomes; however, only the delirium- related outcomes of incidence, prevalence, duration, and severity were reviewed (Table 2–Table 5). Outcomes combining delir- ium and coma into the same statistic were excluded, as no delirium-specific results could be assessed outright. An overall summary of delirium outcomes can be found in Table 2. patients enrolled, and there were no exclusion criteria noted to infer any characteristics of the selected population23. Risk of bias assessment In addition to data extraction using the Cochrane Review Group’s Data Extraction Form, a risk of bias assessment was performed by all authors on all included randomized control- led trials (RCTs) and randomized pilot studies. A risk of bias worksheet was developed by modifying Cochrane’s Risk of Bias Tool and articles were ranked as having a low, high, or unclear risk of bias21. Disagreements were settled by discussion between the authors. A total of eleven included studies underwent this assessment. Details of the risk of bias assessment categories can be found in the Extended data, Supplementary Table 220. Of these 27 articles, 24 assessed incidence and/or prevalence within their cohorts24–43,45–48,50, 16 assessed for duration24–28,34,35,38,42,44–46,48–50, and four for severity24,36,42,50. Additionally, 12 focused on the effect of single interventions24,29,31,32,34,38,41,42,44,45,47,48 while 15 considered bundled, multicomponent interventions25–28,30,33,35–37,39,40,43,46,49,50. Individual interventions included mobility protocols, dis- tinct family visiting policies, dynamic lighting, music therapy, automated reorientation messages, mindfulness exercises, and the structured use of mirrors in recovery. These individ- ual interventions also comprised multiple components of the Results After searching the literature, 27 articles are included in our systematic review24–50 (Figure 1). Study details of each reviewed Page 4 of 31 F1000Research 2021, 9:1178 Last updated: 06 SEP 2022 Study Design Population, n Patient characteristics Notable exclusion criteria Intervention characteristics Control characteristics Method(s) of assessment & frequency Pharmacologic aspect? Álvarez et al. 2017 RCT, pilot n=140 Non-intubated patients ≥60 years old, hospitalized within 24 hours in the ICU for postsurgical observation or for acute or chronic decompensated illness Severe communication disorders; delirium before ICU admission; cognitive decline (defined with score >3.3 in the IQCODE and >6 points in the Spanish version of FAQ) Early and intensive OT Details: two 40 minute sessions per day (one in the morning and one in the evening) for 5 consecutive days Components: polysensory stimulation; body positioning; cognitive stimulation; BADLs; stimulation of upper extremities; family engagement Standard non- pharmacological prevention Components: reorientation twice daily; early mobilization 3 times daily; sight and hearing aids; environmental management (minimize physical restraints; clock and calendar); sleep protocols; caution with use of medications with the potential to cause delirium CAM: duration & incidence of delirium (twice daily for 5 consecutive days) DRS: severity of delirium Yes: avoidance of nighttime medications, anticholinergic drugs, and benzodiazepines Arbabi et al. 2018 quasi- experimental n=148 Patients >18 years admitted to general ICU History of cognitive disorders such as dementia; alcohol abuse; impaired consciousness; history of psychiatric drugs; must not interrupt the admission drug use Multicomponent bundle Components: staff education; educational posters; environmental changes: clock, calendar, family visits, appropriate lighting, eye & ear aids, light alarms instead of audio alarms, staff-patient interaction, hydration, early mobility Usual care, details not specified. CAM-ICU (twice daily, once in morning and once in evening) Yes: staff educated on pharmacologic delirium treatments Balas et al. 2014 pre-post, prospective cohort study n=296 Patients ≥19 years old admitted to medical or surgical critical care service No LAR present to provide consent within 48 hours of ICU admission; those who failed the safety screening or did not receive clearance from their physician ABCDE bundle Components: (A) sedation awakening trial; (B) spontaneous breathing trial; (C) interdisciplinary coordination; (D) delirium monitoring & management; (E) early mobility Usual care, details not specified. CAM-ICU (every 8 hours) None. Results F1000Research 2021, 9:1178 Last updated: 06 SEP F1000Research 2021, 9:1178 Last updated: 06 SEP 2022 Study Design Population, n Patient characteristics Notable exclusion criteria Intervention characteristics Control characteristics Method(s) of assessment & frequency Pharmacologic aspect? Bounds et al. 2016 retrospective study n=159 Patients ≥18 years old and have stayed in the ICU >24 hours Intracranial pressure increased >50% from first ICU measure for hospitalization; quadriplegia; GCS score < 8 without use of sedatives; comfort measures only as documented in the medical record as a medical order for life- sustaining treatment and/or palliative care; cardiopulmonary arrest resulting in death. ABCDE bundle Components: (A) sedation awakening trial; (B) spontaneous breathing trial; (C) choice of medication & interdisciplinary coordination; (D) delirium prevention & management; (E) early mobility Usual care, details not specified. ICDSC (twice daily at 05:00 and 17:00) Yes: medication management Bryczkowski et al. 2014 pre-post, prospective intervention cohort study n=123 All patients >50 years old consecutively admitted to the SICU for ≥24 h Diagnosed with moderate to severe TBI, defined as a head AIS score of ≥3; transfer from jail or in police custody; history of dementia; patients whose delirium statuses were recorded as “unobtainable” or undocumented during their time in the SICU Multicomponent bundle Components: protocol to lighten sedation; limit medications associated with delirium; sleep protocol: limit unnecessary PM care, lower lights and volume, massage, music, quiet time therapy; staff- patient-family education program Usual care, details not specified. CAM-ICU (at least every 12 hours, additional measurements after changes to routine, i.e. after a visit to radiology) Yes: medication management Campbell 2014 evidence-based project n=58 Patients ≥18 years old, admitted to or in the ICU and placed on mechanical ventilation for at least 48 hours Primary diagnosis of stroke with coma, or myocardial infarction with coma; pregnant; history of developmental disability or dementia; receiving therapeutic hypothermia; those dependent in activities of daily living prior to admission; actively dying Early mobility protocol Details: introduction of progressive mobility exercises following successful interruption of sedation and awakening Usual care, details not specified. CAM-ICU (once daily at 08:00) None. P F1000Research 2021, 9:1178 Last updated: 06 SE Page 6 of 31 F1000Research 2021, 9:1178 Last updated: 06 SEP 2022 Study Design Population, n Patient characteristics Notable exclusion criteria Intervention characteristics Control characteristics Method(s) of assessment & frequency Pharmacologic aspect? Results Chai 2017 pre-post, quasi- experimental QI project n=301 Any patient >18 years old admitted to a comprehensive mixed-ICU (medical, surgical, cardiac, neurology) No exclusion criteria. ABCDEF bundle Components: (A) assessment/ management of pain; (B) spontaneous awakening & breathing trials; (C) choice of sedation; (D) delirium monitoring & management; (E) early mobility; (F) family engagement Details not specified. CAM-ICU (at least once per shift) Yes: pain management and choice of sedation Damshens et al. 2018 RCT n=80 Any patient >15 years old admitted to ICU trauma service History of cognitive impairment, depression, taking psychotropic drugs, drug abuse, and alcohol abuse; patients with visual or hearing loss; ICU admission for <48 hours Music therapy Details: two 45 minute sessions per day (once in morning and once at night); music tracks were mainly instrumental and selected by a music expert Details not specified. CAM-ICU (at least once per shift) None. Eghbali-Babadi et al. 2017 RCT n=68 Non-intubated ICU patients 18–70 years old after elective open heart surgery, with immediate family available for intervention History of addiction to drugs, alcohol, and cigarette smoking; presence of delirium, consciousness level disorder, or mental disease before surgery; history of blindness or deafness; family history for surgery; death; post-surgery acute complications Family visitation Details: additional 30–40 minute morning visitation session by an approved, study-educated immediate family member Components: visual and hearing aids; hand holding; reorientation; presence of meaningful personal items; encouragement to express feelings and regain independence Visitation by hospital regulations, and visits were recorded. CAM-ICU (twice daily, once in evening and once in morning after family visit) None. Pa Study Design Population, n Patient characteristics Notable exclusion criteria Intervention characteristics Control characteristics Method(s) of assessment & frequency Pharmacologic aspect? Fallahpoor et al. (2016) action research study n=100 Patients >18 years old admitted to ICU after elective CABG surgery Blindness or deafness; history of mental illness, CVA, or kidney failure; return to operating room; bleeding or tamponade; requiring balloon pump or mitral valve repair; death 3-part management program: Before surgery: effective patient communication; identify and control delirium risk factors; staff and patient education; During surgery: identify delirium risk factors; stabilize hemodynamics; time management; After surgery: identify delirium risk factors; optimize environment; reorientation; effective communication; psychological & medical support; physical stability & activity; safety; sleep and sensory experiences Details not specified. CAM-ICU (every 8 hours) Not specified. Giraud et al. Results 2016 RCT, pilot time- cluster n=223 ICU patients ≥70 years old after elective or urgent cardiac surgery Severe visual impairment preventing self- recognition in a mirror; physical or communication barriers impeding intervention; severe mental disability; history of psychiatric illness requiring prior hospitalization Structured mirror usage Details: use of and education on using mirrors to support mental status, attention, physical mobilization, and multisensory feedback and integration; use on awakening, change in mental status, and during nursing procedures and activity therapies Mirror details: (a) 23x41 cm personal mirror for viewing of the face; (b) 160x50 cm posture mirror to provide visual, proprioceptive feedback Current standard post-surgical ICU care, including no prescriptions around the use of mirrors, and allowing control patients who brought a mirror from home to use it per their normal habits. CAM-ICU (twice daily) None. F1000Research 2021, 9:1178 Last updated: 06 SE F1000Research 2021, 9:1178 Last updated: 06 SEP 2022 Page 8 of 31 F1000Research 2021, 9:1178 Last updated: 06 SEP 2022 Study Design Population, n Patient characteristics Notable exclusion criteria Intervention characteristics Control characteristics Method(s) of assessment & frequency Pharmacologic aspect? Guo et al. 2016 RCT n=160 Oral cancer patients aged 65–80 years scheduled for tumor resection surgery History of CNS disorders or mental illness; MMSE score <24 or dementia of various etiologies; history of endocrine and metabolic disorders; recent use of glucocorticoids; alcohol or drug dependence; secondary surgery or severe infectious complications; illiteracy or language barriers; severe hearing or visual impairment; projected SICU stay ≤ 48 hours Multicomponent non-pharmacologic intervention: Components: cognitive prehabilitation; post- operative reorientation; eye & ear aids; additional communication avenues for verbally impaired patients; decreased environmental noise; improved sleep protocols; eye masks & earplugs; limited restraint & catheter use; music therapy; immediate nasal feeding if able Usual care consisted of standard hospital services provided by physicians, nurses, and support staff (e.g., physical therapists, pharmacists, and dietitians). CAM-ICU (twice daily, once in morning and once in evening) None. Hamzehpour et al. 2017 RCT n=100 ICU patients >18 years old GCS ≤ 7; history of mental illness; death; ICU stay <24 hours; sedative injection Roy adaptation model Components: delirium education; fluid and electrolyte balance; nutrition protocol; sleep protocol; activity and mobility protocol; monitoring of excretion, oxygen, circulatory, and endocrine condition Routine care: performing physician orders; daily assessment of consciousness; systematic patient evaluation NEECHAM confusion scale (twice daily, once in morning and once in evening) Not specified. Results Karadas & Ozdemir 2016 RCT n=94 Patients ≥65 years old with an ICU stay of at least 24 hours Presence of delirium before the procedure; amputated extremity; invasive mechanical ventilation; procedures limiting mobility; RASS score of -4 and -5; advanced osteoporosis; terminal illness; known cognitive disorders (dementia and psychosis); increased intracranial pressure; active gastrointestinal system bleeding; arrhythmia; active myocardial ischemia ROM exercises Details: 10 repetitions per exercise for approximately 30 min once daily; if patient did not tolerate the intervention, it ceased & continued the next day Components: passive, assisted-active, or active ROM exercises performed on the four extremities in the supine position Usual care, details not specified. CAM-ICU (once daily) None. P 9 f 3 F1000Research 2021, 9:1178 Last updated: 06 SEP 202 Page 9 of 31 Page 9 of 31 Page 9 of 31 F1000Research 2021, 9:1178 Last updated: 06 SEP 2022 Study Design Population, n Patient characteristics Notable exclusion criteria Intervention characteristics Control characteristics Method(s) of assessment & frequency Pharmacologic aspect? Khan et al. (2014) pre-post, implementation study n=702 Mechanically ventilated ICU patients ≥ 18 years old Hearing impaired; legally blind; admitted with alcohol intoxication; prisoners; diagnosed as having an Axis 1 Psychiatric disorder ‘Wake Up and Breathe’ protocol Components: daily sedation interruption & spontaneous awakening trial, followed by spontaneous breathing trial depending on the patient’s response to the former Continuous analgesic and sedative infusions based on physicians’ discretion. CAM-ICU (twice daily) Yes: sedation management Kram et al. (2015) pre-post, implementation study n=83 All adult patients admitted to the ICU Safety screening failures; patients without clearance from their attending surgeon; physicians could write an order out of any bundle component they believed was not in the patient’s best interest ABCDE bundle Components: (A & B) spontaneous awakening and breathing trials; (C) interdisciplinary coordination; (D) delirium screening & management; (E) early mobility Details not specified. ICDSC (twice daily at 05:00 and 17:00) Yes: sedation management Lisann- Goldman et al. Results 2017 mixed-methods pilot study n=25 Patients ≥40 years old, with planned ICU admission after elective cardiac surgery with cardiopulmonary bypass, and a MMSE score of ≥24 Intracranial or emergency surgery; severe visual or auditory disorder or handicaps; illiteracy or minimal reading or writing ability; presence of a major, uncontrolled psychiatric condition; severe cognitive impairment Mindfulness exercises Pre-operative: discussion exercises with the study PI; listen to audio discussing Langerian mindfulness (8 minutes 48 seconds) Post-operative: discussion exercises with the study PI, twice daily in the morning and afternoon Preoperative: Engaged in discussion related to cardiac surgery and then listened to audio describing the step-by- step process of undergoing cardiac surgery (8 minutes 34 seconds) Post-operative: neutral conversation on the same twice daily schedule CAM-ICU (twice daily) None. Page 10 Page 10 of 31 caregiver available for scheduled bedside visits (RNAO, 2016). ICDSC (once daily) Moon & Lee 2015 RCT, single-blind n=123 Patients ≥18 years old on day of ICU admission with ICU stay of ≥48 hours Persistent RASS score of -4 or -5; severe visual and auditory problems; serious psychiatric or neurologic diagnosis; score of ≤23 on the MMSE-Korean; admission to the isolation ward due to infection; death or discharge on day of admission; inability to conduct CAM-ICU when in a very violent status & RASS +3 or +4 Multicomponent bundle Components: cognitive assessment & reorientation; sensory aids; indirect night lighting; consistent care staff; items from home; minimize bed relocation; address & manage delirium risk factors early; nutritional, fluid, electrolyte balance; early ambulation; increased awareness of drugs associated with delirious side effects; early infection detection; limit catheter use; monitor for hypoxemia; pain control Typical nursing care included regular checking of consciousness and orientation without attempting to (1) reorient the patient; (2) communicate using nonverbal communication skills; (3) provide visual or hearing aids; (4) assign the same nurse in charge throughout stay; (5) minimize bed movement; or (6) carefully use particular medications (e.g., anticholinergic agents and opiates) CAM-ICU (frequency not specified) Yes: careful use of drugs associated with delirious side effects; pain control F1000Research 2021, 9:1178 Last updated: 0 F1000Research 2021, 9:1178 Last updated: 06 SEP 2022 y Design Population, n Patient characteristics Notable exclusion criteria Intervention characteristics Control characteristics Method(s) of assessment & frequency Pharmacologic aspect? ot et al. Results randomized pilot study n=30 All patients admitted to the surgery ICU or surgery unit with: (1) post- surgical delirium confirmed by a DSM-V medical diagnosis; (2) undergoing either CABG or heart valve surgery; and (3) has a family caregiver available for scheduled bedside visits Planned transfer to another hospital less than 3 days after delirium onset; preoperative diagnosis of cognitive impairment; irreversible postoperative cognitive damage Family caregiver Details: an approved, study-educated family member applied bedside strategies to aid the patient in reorientation, observe and communicate signs of delirium with nursing staff, and present family memories Both groups received usual care, including pharmacological and non- pharmacological interventions suggested in best practice guidelines described in the Registered Nurses’ Association of Ontario guidelines (RNAO, 2016). CAM-ICU (once daily) DI (once positive delirium result is recorded) DSM-V (confirm delirium diagnosis once positive ICDSC result is recorded) ICDSC (once daily) Yes: use of pharmacological interventions in usual care; usual care is applied to both study groups. & Lee RCT, single-blind n=123 Patients ≥18 years old on day of ICU admission with ICU stay of ≥48 hours Persistent RASS score of -4 or -5; severe visual and auditory problems; serious psychiatric or neurologic diagnosis; score of ≤23 on the MMSE-Korean; admission to the isolation ward due to infection; death or discharge on day of admission; inability to conduct CAM-ICU when in a very violent status & RASS +3 or +4 Multicomponent bundle Components: cognitive assessment & reorientation; sensory aids; indirect night lighting; consistent care staff; items from home; minimize bed relocation; address & manage delirium risk factors early; nutritional, fluid, electrolyte balance; early ambulation; increased awareness of drugs associated with delirious side effects; early infection detection; limit catheter use; monitor for hypoxemia; pain control Typical nursing care included regular checking of consciousness and orientation without attempting to (1) reorient the patient; (2) communicate using nonverbal communication skills; (3) provide visual or hearing aids; (4) assign the same nurse in charge throughout stay; (5) minimize bed movement; or (6) carefully use particular medications (e.g., anticholinergic agents and opiates) CAM-ICU (frequency not specified) Yes: careful use of drugs associated with delirious side effects; pain control F1000Research 2021, 9:1178 Last updated: 06 Study Design Population, n Patient characteristics Notable exclusion criteria Intervention characteristics Control characteristics Method(s) of assessment & frequency Pharmacologic aspect? Mailhot et al. Results 2017 randomized pilot study n=30 All patients admitted to the surgery ICU or surgery unit with: (1) post- surgical delirium confirmed by a DSM-V medical diagnosis; (2) undergoing either CABG or heart valve surgery; and (3) has a family caregiver available for scheduled Planned transfer to another hospital less than 3 days after delirium onset; preoperative diagnosis of cognitive impairment; irreversible postoperative cognitive damage Family caregiver Details: an approved, study-educated family member applied bedside strategies to aid the patient in reorientation, observe and communicate signs of delirium with nursing staff, and present family memories Both groups received usual care, including pharmacological and non- pharmacological interventions suggested in best practice guidelines described in the Registered Nurses’ Association of Ontario guidelines (RNAO, 2016). CAM-ICU (once daily) DI (once positive delirium result is recorded) DSM-V (confirm delirium diagnosis once positive ICDSC result is recorded) ICDSC (once daily) Yes: use of pharmacological interventions in usual care; usual care is applied to both study groups. Page 11 of 31 Study Design Population, n Patient characteristics Notable exclusion criteria Intervention characteristics Control characteristics Method(s) of assessment & frequency Pharmacologic aspect? Munro et al. 2017 RCT, prospective n=30 Patients >18 years old within 24 hours of ICU admission Anticipation of imminent patient death; medical contraindication to the intervention (i.e. psychiatric auditory hallucinations, or profound deafness); inability to speak either English or Spanish Automated reorientation messages Details: reorientation script recorded by either a family member or a bilingual female unknown to the subject; played 8x from 09:00-16:00 stating the subject’s preferred name & that the message is recorded; the following details were randomly ordered to prevent repetition: information about ICU environment, expected visual and auditory stimuli, availability of staff & family Usual care, details not specified. CAM-ICU (twice daily) None. Parry et al. Results 2014 case-matched control study n=16 Patients ≥18 years old admitted to the ICU with a diagnosis of sepsis, mechanically- ventilated for >48 hours, and in ICU for ≥4 days Physical limitations (presence of external fixator, pacemaker, defibrillator, open wound or skin abrasions); obesity (BMI>40, too heavy for cycling machine); anticipation of imminent patient death Functional electrical stimulation (FES) cycling Details: supine, motorized cycle ergometer attached to a current-controlled stimulator(20-60 min once daily 5x a week); stimulation occurred at specific times, based on normal activation patterns regulated by the bicycle software & causing visible muscle contraction (quadriceps, hamstrings, gluteals, calves) Usual care following institutional protocols for resuscitation & sepsis management, routine physiotherapy once subject was considered awake using De Jonghe 5-point criteria CAM-ICU (once daily) None. Pun et al. 2019 prospective, multicenter cohort study via national QI collaborative n=10840 Any adult patient admitted to medical, surgical, cardiac, or neurologic ICU for at least 2 consecutive 24- hour days Death or ICU discharge within 24 hours of ICU admission, active life support withdrawal and/or comfort-care measures within 24 hours of ICU admission ABCDEF bundle Components: (A) assess/prevent/manage pain; (B) spontaneous awakening & breathing trials; (C) choice of analgesia/sedation; (D) delirium monitoring & management; (E) early mobility; (F) family engagement NA: analytic comparison is between complete bundle performance and partial bundle performance CAM-ICU or ICDSC (≥2 times daily) Yes: pain management and choice of analgesia/ sedation F1000Research 2021, 9:1178 Last updated: 06 SEP 2022 F1000Research 2021, 9:1178 Last updated: 06 SEP 2022 Page 12 of 31 Page 12 of 31 F1000Research 2021, 9:1178 Last updated: 06 SEP 2022 Study Design Population, n Patient characteristics Notable exclusion criteria Intervention characteristics Control characteristics Method(s) of assessment & frequency Pharmacologic aspect? Rivosecchi et al. 2016 pre-post, prospective observational QI project n=483 Any patient admitted to the medical ICU Any amount of time admitted to an ICU (internal or external) before MICU admission; documented history of cognitive impairment; already admitted to MICU when study period began; MICU stay ≤24 hours; presented to the MICU delirious; no recorded ICDSC measurements M.O.R.E. bundle + staff education Components: (M) music; (O) opening/closing of blinds; (R) reorientation/ cognitive stimulation; (E) eye/ear care Daily bedside multidisciplinary rounds, sedation algorithms, mobilization protocols, and every 4-hour delirium screening using the ICDSC. ICDSC (every 4 hours) Yes: sedation algorithms Rosa et al. Results ICDSC (every 4 hours) Yes: sedation algorithms R t l t t d 286 ti t ≥18 P ti t ith h i E t d d i it ti h R l i it ti CAM ICU (t i N Page 13 of 31 Page 13 of 31 F1000Research 2021, 9:1178 Last updated: 06 SEP 2022 Study Design Population, n Patient characteristics Notable exclusion criteria Intervention characteristics Control characteristics Method(s) of assessment & frequency Pharmacologic aspect? Sullinger et al. 2017 pre-post study, retrospective, observational n=89 critically ill patients ≥18 years old with acute delirium in an open surgical- trauma ICU Missing or incomplete CAM- ICU assessments; pregnancy; history of dementia, schizophrenia, bipolar disease or a prolonged QTc at baseline (>500 ms regardless of gender). Management bundle + nursing education Components: non-pharm interventions (replacing vision/hearing aids, healing arts consult [massage therapy, craniosacral therapy, therapeutic touch, music therapy], mobility plan [repositioning, sitting, ambulation], light and blinds adjustment, revised quiet hours [13:00- 15:00 and 22:00- 04:00], earplugs, family presence); initiate anti- psychotic medications if non-pharmacologic therapies fail Details not specified. CAM-ICU (twice daily, once per 12- hour shift) Yes: initiate anti-psychotic medications if non- pharmacologic therapies fail Zhang et al. 2017 pre-post study, prospective n=278 any patient ≥18 years old admitted to the cardiothoracic ICU after CABG ICU stay of ≤24 hours; history of mental disease or delirium at admission; patient could not awaken from surgery within the first 24 hours after surgery Management bundle Components: delirium risk factor screening & modification (pain control, early catheter removal, reorientation, increased family visits, minimizing care-related interruptions, comfortable nursing, monitoring for sleeping difficulties) Details not specified. CAM-ICU: delirium (not specified) DSR-R-98: severity (when a positive CAM-ICU is recorded) Yes: medication management Abbreviations: BADLs: basic activities of daily living; CABG: coronary artery bypass graft; CAM: Confusion Assessment Method; CNS: central nervous system; CVA: cerebrovascular accident; DI: Delirium Index; DOS scale: Delirium Observation Screening scale; DRS/DRS-R-98: Delirium Rating Scale-Revised-98; DSM-V: Diagnostic and Statistical Manual of Mental Disorders, 5th Edition; FAQ: Functional Activity Questionnaire; GCS: Glasgow Coma Score; ICDSC: Intensive Care Delirium Screening Checklist; ICU: intensive care unit; IQCODE: Informant Questionnaire on Cognitive Decline in the Elderly; LOS: length of stay; MMSE: Mini-Mental State Examination; OT: occupational therapy; QI: quality improvement; RASS: Richmond Agitation-Sedation Scale; RCT: randomized controlled trial; SICU: surgical ICU. Results 2017 pre-post study, prospective n=286 any patient ≥18 years old admitted to the ICU Patients with aphasia, pre-existing delirium, exclusively palliative treatment at ICU admission, expected ICU stay <24 hours, those who remained unarousable ≥48 hours (RASS –4 or –5), and those readmitted to the ICU after enrollment in the study. Extended visitation hours Details: two or fewer visitors at a time, for 12 hours/day (09:00- 21:00); families allowed to participate in multidisciplinary bedside rounds; no time limit for terminal patients, conflicts, or delirium Regular visitation schedule Details: two or fewer family visitors at a time for 4.5 hours per day, broken up into three visitation blocks (09:00-11:00, 16:00-17:30, 21:00-22:00); no time limit for terminal patients, conflicts, or delirium CAM-ICU (twice daily, once every 12 hours) None. Simons et al. 2016 RCT n=734 any patient ≥18 years old admitted to the ICU An expected ICU stay of <24 hours; anticipated life expectancy <48 hours; severe hearing or visual impairment; severe mental impairment Dynamic Lighting Application (DLA) Details: gradual exposure to blueish-white light from 07:00-11:30 and 13:30- 16:00 (peak 1700 lux, 4300 K); from 11:30-13:30 the light was dimmed to 300 lux, 3000 K. The control group was consistently exposed to 300 lux, 3000 K lighting. CAM-ICU (at least 3 times daily) non-ICU patients: DOS scale, consultation with geriatrician, treatment with pharmacological agents for suspected delirium (frequency not specified) None. F1000Research 2021, 9:1178 Last updated: 06 SEP 202 Study Design Population, n Patient characteristics Notable exclusion criteria Intervention characteristics Control characteristics Method(s) of assessment & frequency Pharmacologic aspect? Rivosecchi et al. 2016 pre-post, prospective observational QI project n=483 Any patient admitted to the medical ICU Any amount of time admitted to an ICU (internal or external) before MICU admission; documented history of cognitive impairment; already admitted to MICU when study period began; MICU stay ≤24 hours; presented to the MICU delirious; no recorded ICDSC measurements M.O.R.E. bundle + staff education Components: (M) music; (O) opening/closing of blinds; (R) reorientation/ cognitive stimulation; (E) eye/ear care Daily bedside multidisciplinary rounds, sedation algorithms, mobilization protocols, and every 4-hour delirium screening using the ICDSC. Results exclusion criteria of ‘family history for surgery’: cannot confirm whether for elective open heart surgery, or surgery in general. p F1000Research 2021, 9:1178 Last updated: 06 SEP 2022 Table 2. Summary of delirium outcomes. Table 2. Summary of delirium outcomes. Individual interventions Incidence Prevalence Duration Severity automated reorientation44 -- -- yes* -- dynamic lighting48 n.s. -- n.s. -- family, caregiver42 NA -- NA n.s. family, RVM vs EVM47 yes -- -- -- family, additional structured visit32 yes -- -- -- mindfulness exercises41 NA -- -- -- mirrors34 n.s. -- n.s. -- mobility, early and intensive OT24 yes -- yes n.s. mobility, early29 n.s. -- n.s. -- mobility, ROM exercises38 n.s. -- n.s. -- mobility, FES45 n.s. -- yes -- music therapy31 n.s. -- -- -- Bundled interventions Incidence Prevalence Duration Severity ABCDE26 -- yes yes* -- ABCDE27 -- yes yes -- ABCDE40 -- NA -- -- ABCDEF30 yes -- -- -- ABCDEF37 yes -- -- -- M.O.R.E.46 yes -- yes -- multicomponent25 yes -- yes -- multicomponent28 n.s. -- yes -- multicomponent43 n.s. -- -- -- multicomponent49 -- -- yes -- multicomponent non-pharmacologic35 yes* -- yes -- post-CABG delirium management33 yes -- -- -- risk factor screening & target modification50 yes -- n.s. n.s. Roy adaptation model36 yes* -- -- yes* ‘Wake Up and Breathe’ protocol39 n.s. n.s. -- -- Legend: n.s., not significant, p>0.05; yes, significant, p<0.05; NA, not analyzed for significance; --, not measured in this study; yes, some measured time-points are significant. bundled interventions. A summary of study details can be found in Table 1. bundled interventions. A summary of study details can be found in Table 1. bundled interventions. A summary of study details can be found in Table 1. and Champagne (NEECHAM) scales. Incidence and preva- lence were similarly defined in all studies except for one, and are recorded separately in Table 3; only one study looked at both incidence and prevalence39. Severity was assessed by using the Delirium Index (DI), the Delirium Rating Scale (DRS), the Revised Delirium Rating Scale (DRS-R-98), and NEECHAM scale (Table 5). Measurements for incidence, prevalence, and duration were based upon multiple methods of delirium screening, including the Confusion Assessment Method (CAM), CAM-ICU, Inten- sive Care Delirium Screening Checklist (ICDSC), and Neelon Page 15 of 31 Page 15 of 31 F1000Research 2021, 9:1178 Last updated: 06 SEP 2022 Table 3. Results on incidence and prevalence of delirium. Results Incidence Intervention Number of patients enrolled Group Results Statistical Results ABCDEF bundle30b n=351 (CG n=151; IG n=150) Day 08:00–11:00 CG 77 (51%) IG 16 (10.7%) χ2(1) = 57.32 p <.001 Night 20:00–23:00 CG 89 (58.9%) IG 27 (18%) χ2(1) = 53.25 p <.001* ABCDEF bundle, complete vs partial performance37i n=10840 Not reported AOR 0.60 (0.49–0.72) p < 0.0001 Dynamic lighting48c n=734 (CG n=373; IG n=361) CG 123 (33%) IG 137 (38%) OR 1.24 (0.92, 1.68) † p=0.16 Family, additional structured visit32c n=68 (CG n=34; IG n=34) Day 2 10:00 CG 8 (23.53%) IG 4 (11.76%) χ2=3.98 p=0.04* Day 2 17:00 CG 11 (32.35%) IG 4 (11.76%) χ2=8.38 p<0.05* Day 3 10:00 CG 7 (20.58%) IG 3 (8.83%) χ2=4.12 p=0.03* Family, caregiver42a n=30 (CG n=14; IG n=16) CG 71.4% IG 43.8% NA Family, restricted visitation vs extended visitation47c n=286 (CG n=141; IG n=145) CG 29 (20.5%) IG 14 (9.6%) RR 0.50 (0.26, 0.95) † p=0.03* Mindfulness exercises41d n=25 (CG n=13; IG n=12) CG 0 (0%) IG 0 (0%) NA Mirrors34c n=223 (CG n=108; IG n=115) CG 17 (16%) IG 20 (17%) OR 1.15 (0.54,2 .43)† p=0.705 Mobility, early & intensive OT24c n=140 (CG n=70; IG n=70) CG 14 (20%) IG 2 (3%) p=0.001* Mobility, early mobility29c n=58 (CG n=27; IG n=31) CG 24 (89%) IG 29 (93.5%) χ2(1, N=38) = 0.398 p=0.53 Mobility, FES cycling therapy45c n=16 (CG n=8; IG n=8) CG 7 (87%) IG 2 (25%) p>0.05 Mobility, ROM exercises38c n=94 (CG n=47; IG n=47) CG 10 (21.3%) IG 4 (8.5%) χ2=3.02 p>0.05 M.O.R.E. bundle + nursing education46e n=483 (CG n=230; IG n=253) CG 36 (15.7%) IG 24 (9.4%) p=0.04* Multicomponent bundle (staff education & environmental changes)25c n=148 (CG n=69; IG n=79) CG 50 (72.46%) IG 30 (37.97%) p=0.01* Multicomponent bundle + education28c n=123 (CG n=57; IG n=66) CG 27 (47%) IG 38 (58%) p=0.26 Multicomponent bundle43c n=121 (CG n=63; IG n=60) CG 21 (33.3%) IG 12 (20.0%) p=0.10 Multicomponent, non- pharmacologic bundle35c n=160 (CG n=79; IG n=81) Day 1 CG 13 (16.25%) IG 4 (7.50%) p=0.035* Day 2 CG 9 (11.25%) IG 5 (6.25%) p=0.374 Day 3 CG 4 (5.00%) IG 1 (1.25%) p=0.364 Total CG 25 (31.25%) IG 10 (15.00%) p=0.006* Table 3. Results on incidence and prevalence of delirium. Results NA F1000Research 2021, 9:1178 Last updated: 06 SEP 2022 Incidence Intervention Number of patients enrolled Group Results Statistical Results Music therapy31c n= 80 (CG n=40; IG n= 40) CG 16 (40%) IG 15 (37.5%) χ2=0.053 p>0.818 Post-CABG delirium management bundle33b,c n=100 (CG n=50; IG n=50) Overall incidence CG 34 (68%) IG 19 (38%) p=0.001* -30% change Patients with ‘x’ number of delirious events 0 CG 16 (32%) IG 31 (62%) 1–3 CG 24 (48%) IG 12 (24%) 4–7 CG 8 (16%) IG 5 (10%) 8+ CG 2 (4%) IG 2 (4%) Overall NA p=0.008 Risk factor screening & target modification50c n=278 (CG n=137; IG n=141) CG 41 (29.93% IG 19 (13.48%) χ2=11.112 p=0.001* Roy adaptation model36f n=100 (CG n=50; IG n=50) Days 1–6 (AM & PM) See full reference text. p>0.05 Day 7 AM CG 31 (68.9%)‡ IG 15 (36.6%)‡ p<0.008* Day 7 PM CG 30 (61.9%)‡ IG 20 (42.9%)‡ p<0.05* ‘Wake Up and Breathe’ protocol39g n=702 (CG n=262; IG n=440) CG 14 (23.0%) IG 33 (19.6%) AOR 0.718 (0.326,1.578)† p=0.40 Prevalence Intervention Number of patients enrolled Group Results Statistical Results ABCDE bundle26c n=296 (CG n=146; IG n=150) CG 91 (62.3%) IG 73 (48.7%) p=0.03* ABCDE bundle27e n=159 (CG n=80; IG n=79) CG 38% IG 23% p=0.01* ABCDE bundle40e n=83 (CG n=47; IG n=36) CG NA IG 7 (19%) NA ‘Wake Up and Breathe’ protocol39h n=702 (CG n=262; IG n=440) CG 94 (66.7%) IG 167 (55.3%) AOR 0.650 (0.413, 1.022)† p=0.06 Abbreviations: AOR: adjusted odds ratio; CABG: coronary artery bypass graft; CG: control group; FES: functional electrical stimulation; IG: intervention group; FES: functional electrical stimulation; NA: not analyzed; OR: odds ratio; OT: occupational therapy; ROM: range of motion; RR: relative risk. Results Legend: a = Percent of patients with at least one positive CAM-ICU screening; b = Number of recorded delirium events; c = Number of patients with at least one positive CAM-ICU screening; d = Number of patients with at least one positive CAM-ICU screening, patients with RASS -4 or -5 counted as ‘not delirious’; e = Number or percent of patients with at least one positive ICDSC screening; f = Number of patients with at least one NEECHAM score of < 25; g = Number of patients with a positive CAM-ICU after an initial negative result; h = Number of patients with any CAM-ICU positive result; i = not reported; * = significant difference, p<0.05; † = 95% confidence interval; ‡ = contradictory numbers reported, see referenced text. bbreviations: AOR: adjusted odds ratio; CABG: coronary artery bypass graft; CG: control group; FES: functional electrical stimulation nctional electrical stimulation; NA: not analyzed; OR: odds ratio; OT: occupational therapy; ROM: range of motion; RR: relative risk. Legend: a = Percent of patients with at least one positive CAM-ICU screening; b = Number of recorded delirium events; c = Number of patients with at least one positive CAM-ICU screening; d = Number of patients with at least one positive CAM-ICU screening, patients with RASS -4 or -5 counted as ‘not delirious’; e = Number or percent of patients with at least one positive ICDSC screening; f = Number of patients with at least one NEECHAM score of < 25; g = Number of patients with a positive CAM-ICU after an initial negative result; h = Number of patients with any CAM-ICU positive result; i = not reported; * = significant difference, p<0.05; † = 95% confidence interval; ‡ = contradictory numbers reported, see referenced text. Page 17 of 31 Page 17 of 31 Page 17 of 31 F1000Research 2021, 9:1178 Last updated: 06 SEP 2022 Table 4. Results on duration of delirium. Results Duration Intervention Number of patients enrolled Duration measurement Group Results Statistical Results ABCDE bundle26 n=296 (CG n=146; IG n=150) duration in days median (IQR) CG 3 (1, 6) IG 2 (1, 4) p=0.52 % ICU days spent delirious median (IQR) CG 50% (30, 64.3) IG 33.3% (18.8, 50) p=0.003* ABCDE bundle27 number of days a patient had positive ICDSC score mean ± SD (range) CG 3.8 ± 2.9 (1.0, 14.0) IG 1.72 ± 0.8 (1.0, 4.0) p<0.001* Automated reorientation44 n=30 (CG n=10; UG n=10; FG n=10) delirium-free days mean (SD) CG 1.6 (1.13) UG 1.6 (1.07) FG 1.9 (0.99) p=0.0437* days of delirium mean (SD) CG 0.9 (1.28) UG 0.6 (0.84) FG 0.3 (0.48) p>0.05 Dynamic lighting48 n=734 (CG n=373; IG n=361) duration in hours median (IQR) CG 2 (1, 5) IG 2 (2, 5) p=0.87 Family, caregiver42 n=30 (CG n=14; IG n=16) duration in days mean (SD) CG 4.14 (4.04) IG 1.94 (1.34) -- Mirrors34 n=223 (CG n=108; IG n=115) duration in days median (IQR [range]) CG 2 (1, 8 [1, 3]) IG 1 (1, 3 [1, 25]) RR 0.66 (0.26, 1.70) † p=0.393 proportion of ICU stay mean (SD) CG 0.65 (0.29) IG 0.54 (0.30) Co-eff -0.10 (-0.67, 0.47) † p=0.729 Mobility, early & intensive OT24 n=140 (CG n=70; IG n=70) ratio of delirium duration to exposure time CG IRR 6.66 (5.23, 8.3) † IG IRR 0.15 (0.12, 0.19) † CG p=0.000* IG p=0.000* Mobility, early29 n=58 (CG n=27; IG n=31) duration in days mean ± SD (range) CG 2.70 ± 2.18 (0, 9) IG 3.58 ± 2.68 (0, 9) p=0.18 Mobility, FES45 n=16 (CG n=8; IG n=8) duration in days median (IQR) CG 6.0 (3.3, 13.3) IG 0.0 (0.0, 3.0) p=0.042* Mobility, ROM exercises38 n=94 (CG n=47; IG n=47) duration in hours median (range) CG 38 (9, 120) IG 15 (3, 144) Z= -0.997 p>0.05 M.O.R.E. Results p>0.05 Day 4–7, AM & PM (8 time points) See full reference text. p≤0.028* (range p=0.000 to p=0 028) Duration Intervention Number of patients enrolled Duration measurement Group Results Statistical Results Risk factor screening & target modification50 n=278 (CG n=137; IG n=141) duration in days, from first positive CAM-ICU to recovery (2 consecutive days with negative CAM-ICU) all durations (1–5 days), both groups p=0.876 Abbreviations: CG: control group; FES: functional electrical stimulation; FG: family voice group; IG: intervention group; IQR: interquartile range; IRR: incidence risk ratio; OT: occupational therapy; ROM: range of motion; RR: relative risk; SD: standard deviation; UG: unknown voice group. Legend: * = significant p<0 05; † = 95% confidence interval Table 5. Results on severity of delirium. Table 5. Results on severity of delirium. Severity Intervention Number of patients enrolled Measurement Group Results Statistical Results Family, caregiver42 n=30 (CG n=14; IG n=16) DI mean (SD) Overall (Days 1–3) NA p=0.27 Day 1 CG 2.07 (4.05) IG 10.56 (3.5) NA Day 2 CG 8 (6.34) IG 5.38 (5.45) NA Day 3 CG 5.5 (7) IG 3.43 (4.96) NA Mobility, early & intensive OT24 n=140 (CG n=70; IG n=70) DRS mean (range) CG 10 (8, 13) † IG 9 (6, 12)† p=0.7 Risk factor screening & target modification50 n=278 (CG n=137; IG n=141) DRS-R-98 Number (%) Overall (Results from all 3 severity groups) NA Z= -0.792 p=0.428 Mild CG 10 (7.30) IG 7 (4.96) NA Moderate CG 21 (15.33) IG 8 (5.67) NA Severe CG 10 (7.30) IG 4 (2.84) NA Roy adaptation model36 n=100 (CG n=50; IG n=50) NEECHAM Confusion Scale Day 1–3, AM & PM (6 time points) See full reference text. p>0.05 Day 4–7, AM & PM (8 time points) See full reference text. p≤0.028* (range p=0.000 to p=0.028) Abbreviations: DI: Delirium Index; DRS: Delirium Rating Scale; DRS-R-98: Delirium Rating Scale-Revised-98; NA: not analyzed; OT: occupational therapy. Legend: * = significant difference, p<0.05; † = confidence interval not specified. Results bundle46 n=483 (CG n=230; IG n=253) duration in hours median (IQR) CG 20 (9.5, 37) 16.1% of ICU LOS IG 16 (8, 24) 9.6% of ICU LOS p<0.001* Multicomponent bundle (staff education & environmental changes)25 n=148 (CG n=69; IG n=79) % of days with delirium mean ± SD CG 35.84 ± 39.31 IG 26.18 ± 35.38 p=0.001* Multicomponent bundle + education28 n=123 (CG n=57; IG n=66) delirium-free days out of 30 mean (range) CG 24 (22, 26) † IG 27 (25, 28) † p=0.002* Multicomponent non- pharmacologic bundle35 n=160 (CG n=79; IG n=81) duration in hours mean ± SD CG 60.2 ± 15.8 IG 28.1 ± 8.6 p<0.001* Multicomponent bundle + nursing education49 n=89 (CG n=38; IG n=51) number of delirious days mean ± SD CG 8.2 ± 5.7 IG 4.5 ± 4.4 p<0.001* F1000Research 2021, 9:1178 Last updated: 06 SEP 2022 Duration Intervention Number of patients enrolled Duration measurement Group Results Statistical Results Risk factor screening & target modification50 n=278 (CG n=137; IG n=141) duration in days, from first positive CAM-ICU to recovery (2 consecutive days with negative CAM-ICU) all durations (1–5 days), both groups p=0.876 Abbreviations: CG: control group; FES: functional electrical stimulation; FG: family voice group; IG: intervention group; IQR: interquartile range; IRR: incidence risk ratio; OT: occupational therapy; ROM: range of motion; RR: relative risk; SD: standard deviation; UG: unknown voice group. Legend: * = significant, p<0.05; † = 95% confidence interval. Table 5. Results on severity of delirium. Severity Intervention Number of patients enrolled Measurement Group Results Statistical Results Family, caregiver42 n=30 (CG n=14; IG n=16) DI mean (SD) Overall (Days 1–3) NA p=0.27 Day 1 CG 2.07 (4.05) IG 10.56 (3.5) NA Day 2 CG 8 (6.34) IG 5.38 (5.45) NA Day 3 CG 5.5 (7) IG 3.43 (4.96) NA Mobility, early & intensive OT24 n=140 (CG n=70; IG n=70) DRS mean (range) CG 10 (8, 13) † IG 9 (6, 12)† p=0.7 Risk factor screening & target modification50 n=278 (CG n=137; IG n=141) DRS-R-98 Number (%) Overall (Results from all 3 severity groups) NA Z= -0.792 p=0.428 Mild CG 10 (7.30) IG 7 (4.96) NA Moderate CG 21 (15.33) IG 8 (5.67) NA Severe CG 10 (7.30) IG 4 (2.84) NA Roy adaptation model36 n=100 (CG n=50; IG n=50) NEECHAM Confusion Scale Day 1–3, AM & PM (6 time points) See full reference text. Risk of bias assessment Of the 27 included studies, 11 were RCTs or randomized pilot studies24,31,32,34–36,38,42–44,48, eight were pre-post prospec- tive studies26,28,39,40,46,47,49,50, and two were quasi-experimental25,30. The remaining six were a case-matched control study, an evidence based protocol, a mixed-methods pilot study, a prospective multicenter cohort study, a retrospective cohort study, and an action research study27,29,33,37,41,45. The ten RCTs and the randomized pilot study underwent a risk of bias assessment performed by all authors. Risks of bias fell into five major groups (selection bias, performance bias, detection bias, attrition bias, and reporting bias), and based on a study’s scores in each of these groups it was labeled as having an overall high, low, or unclear risk of bias. Four were considered Page 19 of 31 F1000Research 2021, 9:1178 Last updated: 06 SEP 2022 F1000Research 2021, 9:1178 Last updated: 06 SEP 2022 low risk of bias24,34,42,48, two had a high risk of bias32,44, and five had an unclear risk of bias31,35,36,38,43. The most common source of bias was performance bias due to the impossibility of blind- ing participants or personnel to certain treatments. Common sources of unclear and high risk of bias included the methods of randomization and allocation concealment, as well as how missing data was handled. low risk of bias24,34,42,48, two had a high risk of bias32,44, and five had an unclear risk of bias31,35,36,38,43. The most common source of bias was performance bias due to the impossibility of blind- ing participants or personnel to certain treatments. Common sources of unclear and high risk of bias included the methods of randomization and allocation concealment, as well as how missing data was handled. Family involvement. Of the 12 studies in our review which focused on individual interventions, three studied the effect of family involvement on delirium in adult ICU patients. One was a randomized pilot study42, one an RCT32, and one was a pre-post study47. All three studies utilized CAM-ICU in their assessment of delirium. Mailhot et al. constructed a randomized pilot study to explore the effect of a family caregiver (FC) assisting with delirium manage- ment after being ‘mentored’ (MENTOR_D) by nurses in the ‘MEN- TOR_D’ intervention42. They assessed the efficacy of this interven- tion on all delirious, adult coronary artery bypass graft (CABG) patients admitted to the surgical ICU by measuring the outcomes of duration, occurrence, and severity of delirium over three days42. Risk of bias assessment This intervention enrolled 14 patient-nurse (control) care dyads and 16 patient-FC care dyads, which had the FC apply bedside strategies to aid the patient in reorientation. In addition to reorientation, the FC was asked to observe and communicate signs of delirium with nursing staff, present family memories, and speak clearly and simply. Delirium duration and occurrence on post-operative Day 2 improved clinically between groups (dura- tion, mean days from 4.14 to 1.94; occurrence, from 71.40% to 43.80%); however, this result was not assessed for statistical sig- nificance and the severity result was not found to be significant (Table 3–Table 5). Individual interventions Early mobility. The effect of early mobility protocols on delir- ium was the most commonly studied individual intervention. Four of the studies included in our review individually assessed the efficacy of early mobility24,29,38,45 in treating and prevent- ing delirium; of these, two were RCTs24,38, one was an evidence- based project29, and one was a case-matched control study45. They assessed delirium through CAM24 and CAM-ICU29,38,45. The pilot RCT performed by Álvarez et al. investigated the effect of early mobilization through early and intensive occupational therapy (OT), including polysensory stimulation, body position- ing, cognitive stimulation exercises, basic activities of daily liv- ing, upper extremity motor exercises, and family involvement, on non-intubated, elderly patients (≥ 60) in addition to the study center’s standard, non-pharmacological delirium prevention care24. Delirium associated outcomes included incidence, dura- tion, and severity; they found significant differences in inci- dence and duration of delirium, with both p-values ≤ 0.001, but no significant difference in severity (Table 3–Table 5). The RCT performed by Eghbali-Babadi et al. investigated a modified family visitation policy, implementing an additional 30–40 minute special visit by an approved family member, and its effect on delirium incidence in non-intubated adults aged 18–70 after elective open heart surgery32. They found a statis- tically significant reduction in delirium incidence in the inter- vention group with a p-values of 0.04, <0.05, and 0.03 at three different time points (Table 3). Another RCT by Karadas and Ozdemir assessed the effect of range of motion (ROM) exercises on delirium in elderly ICU patients (≥ 65 years)38. Interventional care included ROM exer- cises for 30 minutes daily after establishing the patient’s ability to complete 10 repetitions on each of the four extremities while lying in bed. They reported no statistically significant differ- ences between cohorts for delirium associated outcomes (Table 3 & Table 4). Rosa et al. also measured the effect of a modified family visita- tion policy on delirium incidence, although their population was less restrictive and included any adult ICU patient47. Their pre-post study included the extension of visitation hours from 4.5 hours per day over three visitation blocks to 12 hours per day between 09:00–21:00. This resulted in a statistically sig- nificant difference in delirium incidence, improving from 20.5% to 9.6% (Table 3). & Table 4). Three measured delirium out- comes using CAM-ICU26,30,40, one utilized ICDSC27, and one multicenter study used either CAM-ICU or ICDSC37. Self-involvement approaches (mirror usage, mindfulness exercises). The remaining two studies on the effect of indi- vidual interventions assessed the impact of self-involvement approaches, including mirror usage34 and mindfulness exercises41, on ICU delirium measured by CAM-ICU. One study was a pilot RCT34, while the other was a mixed-methods pilot study41. In a pilot time-cluster RCT, Giraud et al. tested the effect of introducing structured mirror usage into post-operative recov- ery in elderly ICU patients (≥70 years) after cardiac surgery34. Mirror usage was standardized by developing a protocol for nurses and physiotherapists, aiming to use both small, personal mir- rors as well as larger posture mirrors in order to help the patient with reorientation and self-awareness, enhance multisensory feedback on minor procedures, and augment passive and active physical therapies. The control cohort received usual care, includ- ing allowing control patients who brought a mirror from home to use it per their normal habits. After comparing the usual care group with the mirrors group, no significant improvement was found in delirium incidence, ICU days with delirium, or the pro- portion of the total ICU length of stay that the patient spent delirious (Table 3 & Table 4). Balas et al. assessed the impact of an ABCDE bundle on adult ICU patients, evaluating the prevalence and duration of delir- ium in both total days and percent of ICU days spent deliri- ous, with a pre-post study26. The prevalence and percent of ICU days spent delirious were improved in the post period with p-values of 0.03 and 0.003 respectively (Table 3 & Table 4). However, the overall duration of delirium was not significantly different (Table 4). The retrospective assessment of an ABCDE bundle by Bounds et al. evaluated its effect on delirium prevalence and duration in an adult ICU population27. Both the prevalence and dura- tion were significantly decreased in the ABCDE bundle group (p= 0.01 and 0.001 respectively; (Table 3 & Table 4). The mixed-methods study by Lisann-Goldman et al. had subjects who were 40 years of age or older participate in Langerian mindfulness discussion exercises both prior to and after elective cardiac surgery with cardiopulmonary bypass41. In addition to discussion exercises, patients listened to an audio file before surgery. & Table 4). In contrast, the ‘informational control’ group went through normal pre-operative discussions followed by an audio file describing the process of cardiac surgery. They found that no subject developed delirium in either the interventional or the ‘informational control’ group so the effectiveness of the treatment could not be assessed. In another RCT, Damshens et al. introduced therapeutic music selected by a music expert, twice a day for 45 minutes to assess the effect on delirium incidence in adults admitted to ICU trauma service31. Patients in the control group received con- ventional care for the duration of their admission. There was no resultant change to delirium incidence between the two groups (Table 3). p ‘Wake Up and Breathe’ protocol. Khan et al. designed a ‘Wake Up and Breathe’ protocol in a pre-post interventional study to assess for any change in delirium and sedation in mechanically ventilated, adult ICU patients39. They modified elements of the Awakening and Breathing Controlled trial (ABC) to implement a spontaneous awakening trial and daily sedation vacation fol- lowed by a spontaneous breathing trial, depending on the patient’s response51. Delirium was assessed by CAM-ICU, and both the incidence and prevalence of delirium were analyzed, with the study finding no significant change in either measured outcome (Table 3). Munro et al. developed a novel patient reorientation strategy in an RCT, which utilized bilingual (Spanish or English) messages pre-recorded by either family members or females unknown to the adult ICU subjects44. The recordings included an intro- duction with the patient’s name and location, with several additional randomly ordered statements in order to reorient the patient to their unfamiliar surroundings and reason for hospi- talization. All three arms (two intervention groups and one con- trol group) were compared and it was found that the family voice group had a significant improvement in delirium free days (p= 0.0437) but not mean days of delirium (Table 4). ABCDE(F) bundles. Five of the 15 studies which examined delirium bundles studied the effectiveness of ABCDE(F) bun- dle protocols on reducing delirium. ABCDE(F) bundles have multiple components including: spontaneous awakening (A) and breathing (B) trials, interdisciplinary coordination of seda- tives and medications (C), delirium screening and management (D), early mobilization (E), and family engagement and involve- ment (F)52. Of these five studies, two were pre-post studies26,40, one was a prospective multicenter cohort study37, one was a quasi-experimental quality improvement project30, and one was a retrospective cohort study27. & Table 4). Campbell addressed early mobilization in mechanically ven- tilated ICU patients with an evidence-based project29. They measured the effect of a tiered protocol of ROM exercises, bed mobility exercises, seated balance activities, transfer activities (such as bed to chair), standing exercises, and ambula- tion on delirium incidence and duration but found neither to be significant (Table 3 & Table 4). Environmental approaches (lighting, music therapy, automated reorientation). Three RCTs assessed the impact of environ- mental factors on delirium in the ICU, assessed by CAM-ICU, through manipulation of light48, music therapy31, or automated reorientation44. Environmental approaches (lighting, music therapy, automated reorientation). Three RCTs assessed the impact of environ- mental factors on delirium in the ICU, assessed by CAM-ICU, through manipulation of light48, music therapy31, or automated reorientation44. The effectiveness of functional electrical stimulation (FES) to pro- mote mobility and recovery in mechanically ventilated patients with sepsis was evaluated by Parry et al. in a case-matched con- trol study45. The intervention included use of a motorized cycle ergometer to directly stimulate four major lower limb mus- cles (quadriceps, hamstrings, gluteals, and calves) five times weekly for 20–60 minutes a session dependent on the individ- ual patient’s tolerance. While delirium incidence was not sig- nificantly affected (Table 3), the median days of delirium dif- fered between arms (6.0 in control and 0.0 in intervention) (Table 4). In Simons et al.’s dynamic lighting application RCT, adult ICU patients were exposed to variations in high intensity, blueish-white lighting while delirium incidence and dura- tion were measured48. The intervention group was exposed to a peak of 1700 lux (brightness)/4300 K (color temperature) from 09:00–11:30 and 13:30–16:00, and a daytime minimum of 300 lux/3000 K from 11:30–13:30; the control group was exposed Page 20 of 31 Page 20 of 31 F1000Research 2021, 9:1178 Last updated: 06 SEP 2022 F1000Research 2021, 9:1178 Last updated: 06 SEP 2022 solely to 300 lux/3000 K (Table 1). Neither the cumulative inci- dence of ICU-acquired delirium nor the duration were sig- nificantly affected, and the trial was ended early after the intervention was deemed futile (Table 3 & Table 4). solely to 300 lux/3000 K (Table 1). Neither the cumulative inci- dence of ICU-acquired delirium nor the duration were sig- nificantly affected, and the trial was ended early after the intervention was deemed futile (Table 3 & Table 4). unknown. The discussion exercises continued post-operatively twice daily. & Table 4). This audio file walked them through tech- niques on how to re-assess one’s situation and improve their outlook by taking emotional control of the situation, encourag- ing the patients to focus on the process of change and allow- ing oneself to accept new ideas and remain confident about the Kram et al. also looked at a similar patient cohort, all adult patients 18 or older admitted to the ICU, in a pre-post ABCDE bundle study with a smaller subject population (Kram, n=83; Balas, n=296; Bounds, n=159)40. They assessed the effective- ness of the ABCDE bundle on delirium by measuring delirium prevalence and comparing it to a control based on literature val- ues. The measured delirium prevalence of 19% (Table 3) fell outside their cited literature values of 20–80%. Page 21 of 31 F1000Research 2021, 9:1178 Last updated: 06 SEP 2022 F1000Research 2021, 9:1178 Last updated: 06 SEP 2022 Chai initiated an ABCDEF bundle in a mixed ICU setting and analyzed delirium incidence in the adult patients in a pre-post, quasi-experimental quality improvement project30. Delirium inci- dence was compared between morning and night occurrences (morning 08:00–11:00; night 20:00–23:00); both showed sig- nificant improvement in the intervention group with a p-value <0.001 for both morning and evening measurements (Table 3). methods to identify delirium risk factors, optimize time spent in surgery, and introduce staff education and post-operative environ- mental changes33. Delirium related outcomes included the inci- dence ratio and total number of recorded delirium events, with sig- nificant differences found in both (p=0.001 and 0.008 respectively; Table 3). In the RCT conducted by Guo et al., the effect of a bundle consisting of cognitive prehabilitation, post-operative cognitive stimulation activities, environmental changes, music therapy, and non-pharmacologic care changes on delirium incidence and duration after oral tumor resection in patients aged 65–80 years was studied35. The incidence of delirium improved significantly overall, but was only significantly different on post-operative day one compared to days two and three (p=0.035, p=0.374, p=0.364 respectively; Table 3); the duration of delirium also differed significantly (p< 0.001; Table 4). A prospective cohort study performed by Pun et al. through a national quality improvement initiative compared complete ABCDEF bundle performance with proportional ABCDEF bun- dle performance in adult ICU patients with an ICU stay of at least 48 consecutive hours37. & Table 4). Complete bundle performance was defined as a patient-day where 100% of the eligible bundle ele- ments were performed, whereas proportional performance was anything less37. Their study was comprehensive, including 10,840 patients for delirium outcome analysis across 68 ICUs in the United States and Puerto Rico37. When comparing the incidence of delirium between patients with complete and proportional ABCDEF bundle performance, they found that patients with complete performance were significantly less likely to develop delirium (Table 3)37. In an additional analysis, Pun et al. found a dose-dependent reduction of delirium incidence when the more eligible ABCDEF bundle elements were performed (p < 0.0001)37. It is worth noting that this study had a high rate of ‘missingness’ for delirium data and the analysis team chose not to perform multiple imputations37. Hamzehpour et al. designed an RCT and implemented the Roy adaptation nursing model for all adult ICU patients, which focuses on balance of nutrition, electrolytes, and fluids while promoting activity, sleep hygiene, and monitoring of circula- tion and endocrine function36. Their primary delirium-specific outcomes were incidence and severity, and they analyzed both outcomes for two time points (morning & night) for seven days. Their research only showed significant improvements to inci- dence on day seven, both morning and night (p<0.008 and p<0.05; Table 3), but delirium severity, assessed with NEECHAM, improved through the morning of day four to the night of day seven at all measured time points (every time point, p≤0.028; Table 5). Other bundled protocols. The remaining nine bundle studies developed new, unique bundles. They included four pre-post studies28,46,49,50, three RCTs35,36,43, one quasi-experimental study25, and one action research study33. Seven assessed delir- ium incidence and duration using CAM-ICU25,28,33,35,43,49,50, one used NEECHAM36, and one used ICDSC46. Other bundled protocols. The remaining nine bundle studies developed new, unique bundles. They included four pre-post studies28,46,49,50, three RCTs35,36,43, one quasi-experimental study25, and one action research study33. Seven assessed delir- ium incidence and duration using CAM-ICU25,28,33,35,43,49,50, one used NEECHAM36, and one used ICDSC46. Moon and Lee implemented a bundle which included early cognitive assessments and reorientation, sensory aids, envi- ronmental changes, consistent care staff and location, familiar items from home, nursing care changes, and early mobility as part of an RCT aimed at assessing delirium incidence in adult ICU patients with at least a 48 hour stay43. F1000Research 2021, 9:1178 Last updated: 06 SEP 2022 F1000Research 2021, 9:1178 Last updated: 06 SEP 2022 number of days spent delirious, resulting in a significant decrease from 8.2 to 4.5 median days (Table 4). number of days spent delirious, resulting in a significant decrease from 8.2 to 4.5 median days (Table 4). The pilot RCT performed by Alvarez et al. utilized a unique method for assessing the performance of their intervention. In addition to assessing delirium incidence, they measured the ratio of delirium duration to the amount of time exposed to the treatment (IRR)24. They found that IRR decreased as the time exposed to treatment increased to a significant degree (p= 0.000). This ratio could be explained in three ways. Either the dura- tion of delirium stayed the same as the time exposed to treat- ment increased, the duration of delirium increased slower than the time exposed increased, or the duration of delirium decreased while the time of exposure increased. However, the last explanation is impossible, due to the duration of delirium being a sum overtime which could not decrease, such that the result must be explained by either a small increase or no increase in the duration of delirium. If the decrease in the IRR is explained by smaller and smaller increases to delirium duration it is likely that the IRR results from either the trend of patients slowly becoming healthier over time, or the conjunction of that with the intervention. However, if the IRR is explained by the delir- ium duration ceasing to increase, then once it stops the treat- ment may still be effective, but it is not becoming more effective over time and plateaus in effectiveness. Any patient 18 years or older admitted to a cardiothoracic ICU after CABG surgery was analyzed for incidence, duration, and severity of delirium by Zhang et al. in a prospective pre-post study50. Their delirium bundle targeted risk factor screening and modifications, including increased family visits, reorienta- tion, and changes to nursing care. The only significant improve- ment was to incidence of delirium which dropped from 29.93% to 13.48% (Table 3), while the intervention had no impact on duration or severity (Table 4 & Table 5). & Table 4). Their study did not show a significant difference between the intervention and the control group who received usual care (Table 3). A quasi-experimental study designed by Arbabi et al. developed a multi-component delirium management bundle comprised of staff education and environmental and non-pharmacologic care changes25. They measured the effectiveness of their bun- dle by assessing delirium incidence and duration in all adult patients admitted to the general ICU, finding a significant dif- ference in both outcomes (p = 0.01 and 0.001 respectively; Table 3 & Table 4). In a pre-post, observational quality improvement project, Rivosecchi et al. combined staff education with a non-pharma- cologic bundle to look at incidence and duration of delirium in any adult patient aged 18 or older admitted to the medi- cal ICU46. Their M.O.R.E. bundle included (M)usic, (O)pening blinds, (R)eorientation and cognitive stimulation, and (E)ye and ear care. Both delirium incidence and duration were signifi- cantly impacted, with incidence decreasing from 15.7% to 9.4% and a reduction in duration from 16.1% of the ICU stay to 9.6% (p = 0.04 and <0.001 respectively; Table 3, Table 4). Bryczkowski et al. assessed the effectiveness of their bundle, which included a staff-patient-family education program, medi- cation management strategies, and non-pharmacological sleep enhancement protocols, on delirium incidence and delirium free days in patients over the age of 50 years28. The research team found no significant improvement in delirium incidence (Table 3), although the average total number of delirium-free days out of 30 changed significantly from 24 to 27 between groups (p=0.002; Table 4). Sullinger et al. enrolled adult surgical-trauma ICU patients with acute delirium in a pre-post retrospective study, tailor- ing their bundle to incorporate staff education with sensory aids, healing arts techniques, mobility, environmental changes, and family presence49. Their bundle also included the initia- tion of anti-psychotic medications if non-pharmacologic tactics failed. The only specifically delirium related outcome was the Another bundle study developed by Fallahpoor et al. focused specifically on adults admitted to the ICU after elective CABG in an action research study. Their post-CABG delirium man- agement bundle was assessed in an action research study and had three elements focusing on pre-, intra-, and post-operative Page 22 of 31 F1000Research 2021, 9:1178 Last updated: 06 SEP 2022 Implications of results and application to practice Implications of results and application to practice The reviewed studies focused on individual interventions that had a wide range of limitations and were, on the whole, less effective than bundled protocols in the treatment and man- agement of delirium. Many of these studies had limited reli- ability due to small or extremely small sample sizes29,41,44,45. Additionally, even when results were significant, they often had limited applications to practice due to the prevalence of restricted populations. Three of the individual intervention studies lim- ited their study cohort to elderly adults15,24,38, a population which is at an increased risk of delirium. It is unclear whether these results would apply to younger patients. Studied populations were also commonly narrowed to either exclusively intubated patients45, non-intubated patients24,32 or patients with a particular illness34,42,45. Another possible limitation was in the question- able reliability of delirium assessment criteria. This is mentioned by Campbell who stated that 35% of CAM-ICU were incorrectly labeled as ‘unable to assess’29. The question of reliability was also raised in Lisann-Goldman et al.. This study could not assess the effectiveness of their intervention as no patients developed delirium41. However, this could be explained by the fact that fully sedated patients were considered ‘not delirious’ since CAM-ICU could not be performed. The authors also noted that since it often takes weeks or months to fully integrate new behavioral thought techniques, a study focused on chang- ing thought patterns in days would not entirely reflect the full benefit if any were present41. In addition to incidence and prevalence, another common outcome was a change in the duration of delirium. Sixteen of the reviewed studies evaluated the duration of delirium, eight focus- ing on individual interventions and eight introducing bundled protocols. Of these 16 studies, eight found significant changes overall24,25,27,28,35,45,46,49, two had significant improvements at select time points26,44, and five did not have significant results29,34,38,48,50; the one remaining study did not assess for statistical significance42. Six of the eight successful trials were bundles and both of the two effective individual therapies were mobility-focused (early and intensive OT, and FES)24,45. Only four studies looked at delirium severity24,36,42,50 with only one finding any signifi- cant results, and only finding them at select time-points (Roy Adaptation Model)36. In addition to incidence and prevalence, another common outcome was a change in the duration of delirium. Summary of findings yi g Our review included 27 trials that evaluated the effect of various non-pharmacological treatment and management pro- tocols on delirium in an ICU setting. Assessment of the efficacy of these protocols in the last five years was most commonly done by considering incidence and/or prevalence. A total of 25 studies assessed for the effects of these protocols on incidence and/or prevalence, with 11 studying individual approaches and 14 studying bundles. Of these 25 trials, 11 reported significant improvements overall24–27,30,32,33,37,46,47,50, nine found no significant improvement28,29,31,34,38,39,43,45,48, and two only found significant change at certain time points35,36; the remaining three did not ana- lyze for statistical significance of their results40–42. The 11 effec- tive interventions for incidence and/or prevalence were primarily bundled protocols (eight trials)25–27,30,33,37,46,50, followed by fam- ily approaches (two trials)32,47, and early and intensive OT (one trial)24. The two studies with time point dependent changes were both bundles, one non-pharmacologic35 and the other intro- duced the Roy adaptation nursing model36. The multicomponent non-pharmacologic bundle found improvements in incidence both overall and on day one, while the Roy adaptation trial only saw a change in incidence on day seven in both the morn- ing and the evening. Three studies did not analyze for statistical significance40–42; however, the family caregiver intervention saw an overall reduction in the percent of subjects who devel- oped delirium from 71.40% to 43.80%42. The study on mindful- ness exercises had no subjects in either investigational group develop delirium41, and an ABCDE bundle reported a post-bundle incidence of 19% but stated there was no pre-bundle data with which to compare40. No other studies that looked at delirium incidence were effective. Implications of results and application to practice Sixteen of the reviewed studies evaluated the duration of delirium, eight focus- ing on individual interventions and eight introducing bundled protocols. Of these 16 studies, eight found significant changes overall24,25,27,28,35,45,46,49, two had significant improvements at select 6ii 5 Eight of the individual intervention studies were RCTs. This type of study introduces the possibility of additional limita- tions due to the nature of its design. Two of these RCTs had a high risk of bias32,44 due to a failure to blind patients and person- nel, as well as blinding of the outcome assessment and improper allocation concealment. The question of blinding raises another possible limitation of many of these studies, namely the pos- sibility of the Hawthorne effect in patients who knew that they Page 23 of 31 F1000Research 2021, 9:1178 Last updated: 06 SEP 2022 were being observed and receiving an intervention for the treatment/prevention of delirium. Strengths and limitations The strengths of our systematic review include thorough search terms and the methodology to assess a vast majority of recent literature in this field. Strengths and limitations The 15 studies which investigated bundled protocols had, overall, larger sample sizes, fewer cohorts with limited popula- tions, and indicated better reliability in their delirium assessment than the studies which focused on individual interventions. The smallest sample size was 83 patients40; however, this study did not split the sample into multiple cohorts and all patients received the intervention. One study had a sample size of 8949, and all other studies had a sample size of at least 100 patients. A total of five studies restricted their studied population beyond adult ICU patients28,33,35,39,50. Two of these limited their population by age28,35; however, Bryczkowski et al., despite limiting their population by age, included any patients greater than 50 years old, younger than the age when delirium risk is noted to increase53,54. One study limited its population to mechanically ventilated patients39, two considered only patients undergoing CABG33,50, and one studied patients after oral tumor resection35. These population restrictions could limit the generalizability and applicability of the interventions; however, this risk is reduced since bundles were often investigated in multiple studies with similar results. One limitation of this systematic review is that we only focused on trials within the past five years which excluded some well-cited early studies on delirium. Implications of results and application to practice We also did not evaluate other listed outcomes which could provide additional insight into any change in delirium status. Since the condition can be transient and delir- ium screenings are not performed as frequently throughout the ICU day as other measurements, outcomes such as restraint use or amount of prescribed sedatives or anti-psychotic medications would be beneficial to assess in this setting. While the decision to omit exclusion criteria on study design allowed for assess- ment of a broader range of trials, it was difficult to compare out- comes when multiple differing designs and measurement tools were used. Although the CAM-ICU was widely used, some stud- ies used alternative tools and there was no standardized way of defining or measuring delirium duration or severity. A different measurement tool was used to evaluate severity in each of the four studies reviewing this outcome, and duration was defined in a multitude of fashions. Combining this realization with the fact that some studies focused on highly specific subpopulations sug- gests that some trials may need to be replicated in a standard- ized fashion to account for any differences in methodology or subjective assessments. Only three of the bundle studies were RCTs35,36,43. Each of these RCTs had an unclear risk of bias with the most common risk being the inability to blind participants and personnel. The impossibility of blinding in delirium intervention studies makes RCTs a questionable approach. Eight of the bundle studies, rec- ognizing blinding as an impossibility, chose to conduct pre-post prospective studies rather than RCTs26,28,30,39,40,46,49,50. These stud- ies carried a lower risk of introducing bias to their studies and avoided crossover between arms. The pre-post study performed by Kram et al. had the major limitation of not including a pre cohort and only comparing the results of their intervention with literature values40. Additionally, while they found their meas- ured delirium prevalence to fall outside their included litera- ture values (19%), this prevalence falls within the values pro- vided by the current literature (19- 87%)2. Chai’s pre-post study had a delirium assessment with questionable reliability. While all other studies assessed delirium whenever the Richmond Agitation-Sedation Scale (RASS) was ≥ -3, they reported that patients were unable to be assessed whenever RASS was < - 2, resulting in a greater proportion of patients not assessed for delirium30. Only three of the bundle studies were RCTs35,36,43. Conclusions M ICU d li Many ICU delirium treatment and management protocols were developed and tested within the last five years in a variety of study designs. Few trials on individual interventions had posi- tive effects on delirium incidence and duration, but multicom- ponent bundles were found to be more effective overall while incorporating the effective individual intervention of family engagement. Based on the results of bundle studies, the imple- mentation of multi-component protocols in ICUs can reduce ICU delirium, thereby reducing cost of care, improving overall outcomes, and limiting time spent mechanically ven- tilated, medicated, or admitted. Despite these results, further research is needed on individual interventions in order to improve specific elements of multicomponent bundles by adding or removing ineffective therapies. Additional research is also war- ranted to evaluate for any positive effects in more generalized hospital populations. ies carried a lower risk of introducing bias to their studies and avoided crossover between arms. The pre-post study performed by Kram et al. had the major limitation of not including a pre cohort and only comparing the results of their intervention with literature values40. Additionally, while they found their meas- ured delirium prevalence to fall outside their included litera- ture values (19%), this prevalence falls within the values pro- vided by the current literature (19- 87%)2. Chai’s pre-post study had a delirium assessment with questionable reliability. While all other studies assessed delirium whenever the Richmond Agitation-Sedation Scale (RASS) was ≥ -3, they reported that patients were unable to be assessed whenever RASS was < - 2, resulting in a greater proportion of patients not assessed for delirium30. Given the findings of this systematic review, further research is warranted in order to confirm these results and apply them to other patient populations. Multicomponent, bundled approaches were more successful at improving delirium outcomes compared to individual techniques; however, the effective individual tactic of family engagement was included as a component in the effec- tive bundles. Although a majority of the reviewed bundles were effective, it is difficult to compare results as the trials had large differences in study design, enrollment numbers, and delirium assessment measures. Implications of results and application to practice Each of these RCTs had an unclear risk of bias with the most common risk being the inability to blind participants and personnel. The impossibility of blinding in delirium intervention studies makes RCTs a questionable approach. Eight of the bundle studies, rec- ognizing blinding as an impossibility, chose to conduct pre-post prospective studies rather than RCTs26,28,30,39,40,46,49,50. These stud- Registration DOI: 10.17605/OSF.IO/C3RHF • Supplementary Table 1 – Data extraction form Data are available under the terms of the ‘Creative Commons Zero “No rights reserved” data waiver’ (CC0 1.0 Public domain dedication). • Supplementary Table 2 – Risk of bias assessment Data are available under the terms of the ‘Creative Commons Zero “No rights reserved” data waiver’ (CC0 1.0 Public domain dedication). Acknowledgements A previous version of this article is available from medRxiv: https://doi.org/10.1101/2020.05.20.20100552. Acknowledgements A previous version of this article is available from medRxiv: https://doi.org/10.1101/2020.05.20.20100552. Data availability Underlying data All data underlying the results are available as part of the article and no additional source data are required. Extended data OSF: The Effect of Non-Pharmacologic Strategies on Prevention or Management of Intensive Care Unit Delirium: A Systematic Review. https://doi.org/10.17605/OSF.IO/Y7NQ620. Extended data OSF: The Effect of Non-Pharmacologic Strategies on Prevention or Management of Intensive Care Unit Delirium: A Systematic Review. https://doi.org/10.17605/OSF.IO/Y7NQ620. Page 24 of 31 Page 24 of 31 F1000Research 2021, 9:1178 Last updated: 06 SEP 2022 F1000Research 2021, 9:1178 Last updated: 06 SEP 2022 Reporting guidelines OSF: PRISMA checklist for ‘The Effect of Non-Pharmacologic Strategies on Prevention or Management of Intensive Care Unit Delirium: A Systematic Review’. https://doi.org/10.17605/OSF. IO/Y7NQ620 Delirium: A Systematic Review’. https://doi.org/10.17605/OSF. IO/Y7NQ620 This project contains the following extended data: This project contains the following extended data: • Supplementary Appendix 2 – Database search terms R77. PubMed Abstract \| Publisher Full Text \| Free Full Text 25. Arbabi M, Zebardast J, Noorbala AA, et al.: Efficacy of Liaison Education and Environmental Changes on Delirium Incidence in ICU. Arch Neurosci. 2018; 5(2): e56019. Publisher Full Text 13. Thom RP, Bui MP, Rosner B, et al.: A Comparison of Early, Late, and No Treatment of Intensive Care Unit Delirium With Antipsychotics: A Retrospective Cohort Study. Prim Care Companion CNS Disord. 2018; 20(6): 18m02320. P bM d Ab t t \| P bli h F ll T t 26. Balas MC, Vasilevskis EE, Olsen KM, et al.: Effectiveness and safety of the awakening and breathing coordination, delirium monitoring/ management, and early exercise/mobility bundle. Crit Care Med. 2014; 42(5): 3 9. PubMed Abstract \| Publisher Full Text 2. Kalabalik J, Brunetti L, El-Srougy R: Intensive care unit delirium: a review of the literature. J Pharm Pract. 2014; 27(2): 195–207. PubMed Abstract \| Publisher Full Text 2. Kalabalik J, Brunetti L, El-Srougy R: Intensive care unit delirium: a review of the literature. J Pharm Pract. 2014; 27(2): 195–207. PubMed Abstract \| Publisher Full Text 16. Chen S, Shi L, Liang F, et al.: Exogenous Melatonin for Delirium Prevention: a Meta-analysis of Randomized Controlled Trials. Mol Neurobiol. 2016; 53(6): 4046–53. PubMed Abstract \| Publisher Full Text 3. Page VJ, Ely EW, Gates S, et al.: Effect of intravenous haloperidol on the duration of delirium and coma in critically ill patients (Hope-ICU): a randomised, double-blind, placebo-controlled trial. Lancet Respir Med. 2013; 1(7): 515–23. 3. Page VJ, Ely EW, Gates S, et al.: Effect of intravenous haloperidol on the duration of delirium and coma in critically ill patients (Hope-ICU): a randomised, double-blind, placebo-controlled trial. Lancet Respir Med. 2013; 1(7): 515–23. 17. Barbateskovic M, Krauss SR, Collet MO, et al.: Pharmacological interventions for prevention and management of delirium in intensive care patients: a systematic overview of reviews and meta-analyses. BMJ open. 2019; 9(2): e024562. PubMed Abstract \| Publisher Full Text \| Free Full Text References PubMed Abstract \| Publisher Full Text \| Free Full Text 10. Fong TG, Tulebaev SR, Inouye SK: Delirium in elderly adults: diagnosis, prevention and treatment. Nat Rev Neurol. 2009; 5(4): 210–20. PubMed Abstract \| Publisher Full Text \| Free Full Text 11. Davoudi A, Ebadi A, Rashidi P, et al.: Delirium Prediction using Machine Learning Models on Preoperative Electronic Health Records Data. Proc IEEE Int Symp Bioinformatics Bioeng. 2017; 2017: 568–73. PubMed Abstract \| Publisher Full Text \| Free Full Text 12. Van Rompaey B, Elseviers MM, Schuurmans MJ, et al.: Risk factors for delirium in intensive care patients: a prospective cohort study. Crit Care. 2009; 13(3): R77. PubMed Abstract \| Publisher Full Text \| Free Full Text 13. Thom RP, Bui MP, Rosner B, et al.: A Comparison of Early, Late, and No Treatment of Intensive Care Unit Delirium With Antipsychotics: A Retrospective Cohort Study. Prim Care Companion CNS Disord. 2018; 20(6): 18m02320. PubMed Abstract \| Publisher Full Text 14. Knauert MP, Pisani MA: Dexmedetomidine for hyperactive delirium: worth References Álvarez EA, Garrido MA, Tobar EA, et al.: Occupational therapy for delirium management in elderly patients without mechanical ventilation in an intensive care unit: A pilot randomized clinical trial. J Crit Care. 2017; 37: 85–90. PubMed Abstract \| Publisher Full Text 25. Arbabi M, Zebardast J, Noorbala AA, et al.: Efficacy of Liaison Education and Environmental Changes on Delirium Incidence in ICU. Arch Neurosci. 2018; 5(2): e56019. Publisher Full Text 26. Balas MC, Vasilevskis EE, Olsen KM, et al.: Effectiveness and safety of the awakening and breathing coordination, delirium monitoring/ management, and early exercise/mobility bundle. Crit Care Med. 2014; 42(5): 1. Lin WL, Chen YF, Wang J: Factors Associated With the Development of Delirium in Elderly Patients in Intensive Care Units. J Nurs Res. 2015; 23(4): 322–9. PubMed Abstract \| Publisher Full Text 2. Kalabalik J, Brunetti L, El-Srougy R: Intensive care unit delirium: a review of the literature. J Pharm Pract. 2014; 27(2): 195–207. PubMed Abstract \| Publisher Full Text 3. Page VJ, Ely EW, Gates S, et al.: Effect of intravenous haloperidol on the duration of delirium and coma in critically ill patients (Hope-ICU): a randomised, double-blind, placebo-controlled trial. Lancet Respir Med. 2013; 1(7): 515–23. PubMed Abstract \| Publisher Full Text \| Free Full Text 4. Parker AM, Sricharoenchai T, Raparla S, et al.: Posttraumatic stress disorder in critical illness survivors: a metaanalysis. Crit Care Med. 2015; 43(5): 1121–9. PubMed Abstract \| Publisher Full Text 5. Lat I, McMillian W, Taylor S, et al.: The impact of delirium on clinical outcomes in mechanically ventilated surgical and trauma patients. Crit Care Med. 2009; 37(6): 1898–905. PubMed Abstract \| Publisher Full Text 6. Maldonado JR: Delirium pathophysiology: An updated hypothesis of the etiology of acute brain failure. Int J Geriatr Psychiatry. 2018; 33(11): 1428–57. PubMed Abstract \| Publisher Full Text 7. Milbrandt EB, Deppen S, Harrison PL, et al.: Costs associated with delirium in mechanically ventilated patients. Crit Care Med. 2004; 32(4): 955–62. PubMed Abstract \| Publisher Full Text 8. Micek ST, Anand NJ, Laible BR, et al.: Delirium as detected by the CAM-ICU predicts restraint use among mechanically ventilated medical patients. Crit Care Med 2005; 33(6): 1260–5. PubMed Abstract \| Publisher Full Text 9. Ahmed S, Leurent B, Sampson EL: Risk factors for incident delirium among older people in acute hospital medical units: a systematic review and meta-analysis. Age Ageing. 2014; 43(3): 326–33. ( ) PubMed Abstract \| Publisher Full Text \| Free Full Text 4. Parker AM, Sricharoenchai T, Raparla S, et al.: Posttraumatic stress disorder in critical illness survivors: a metaanalysis. Crit Care Med. 2015; 43(5): 1121–9. PubMed Abstract \| Publisher Full Text 4. Parker AM, Sricharoenchai T, Raparla S, et al.: Posttraumatic stress disorder in critical illness survivors: a metaanalysis. Crit Care Med. 2015; 43(5): 1121–9. PubMed Abstract \| Publisher Full Text 18. Huang X, Lin J, Demner-Fushman D: Evaluation of PICO as a knowledge representation for clinical questions. AMIA Annu Symp Proc. 2006; 2006: 359–63. PubMed Abstract \| Free Full Text 5. Lat I, McMillian W, Taylor S, et al.: The impact of delirium on clinical outcomes in mechanically ventilated surgical and trauma patients. Crit Care Med. 2009; 37(6): 1898–905. PubMed Abstract \| Publisher Full Text ( ) PubMed Abstract \| Publisher Full Text 10. Fong TG, Tulebaev SR, Inouye SK: Delirium in elderly adults: diagnosis, prevention and treatment. Nat Rev Neurol. 2009; 5(4): 210–20. PubMed Abstract \| Publisher Full Text \| Free Full Text 10. Fong TG, Tulebaev SR, Inouye SK: Delirium in elderly adults: diagnosis, prevention and treatment. Nat Rev Neurol. 2009; 5(4): 210–20. PubMed Abstract \| Publisher Full Text \| Free Full Text 23. Patel J, Baldwin J, Bunting P, et al.: The effect of a multicomponent multidisciplinary bundle of interventions on sleep and delirium in medical and surgical intensive care patients. Anaesthesia. 2014; 69(6): 540–9. PubMed Abstract \| Publisher Full Text 11. Davoudi A, Ebadi A, Rashidi P, et al.: Delirium Prediction using Machine Learning Models on Preoperative Electronic Health Records Data. Proc IEEE Int Symp Bioinformatics Bioeng. 2017; 2017: 568–73. PubMed Abstract \| Publisher Full Text \| Free Full Text 24. Álvarez EA, Garrido MA, Tobar EA, et al.: Occupational therapy for delirium management in elderly patients without mechanical ventilation in an intensive care unit: A pilot randomized clinical trial. J Crit Care. 2017; 37: 85–90. PubMed Abstract \| Publisher Full Text 12. Van Rompaey B, Elseviers MM, Schuurmans MJ, et al.: Risk factors for delirium in intensive care patients: a prospective cohort study. Crit Care. 2009; 13(3): R77. 12. Van Rompaey B, Elseviers MM, Schuurmans MJ, et al.: Risk factors for delirium in intensive care patients: a prospective cohort study. Crit Care. 2009; 13(3): R77. PubMed Abstract \| Free Full Text 19. Moher D, Liberati A, Tetzlaff J, et al.: Preferred reporting items for systematic reviews and meta-analyses: the PRISMA statement. Ann Intern Med. 2009; 151(4): 264–9, w64. PubMed Abstract \| Publisher Full Text 7. Milbrandt EB, Deppen S, Harrison PL, et al.: Costs associated with delirium in mechanically ventilated patients. Crit Care Med. 2004; 32(4): 955–62. PubMed Abstract \| Publisher Full Text 20. Bihorac A: The Effect of Non-Pharmacologic Strategies on Prevention or Management of Intensive Care Unit Delirium: A Systematic Review. OSF. 2020. http://www.doi.org/10.17605/OSF.IO/C3RHF 21. Higgins JP, Sterne JA, Savovic J, et al.: A revised tool for assessing risk of bias in randomized trials. Cochrane database of systematic reviews. 2016; 10(Suppl 1): 29–31. Reference Source 9. Ahmed S, Leurent B, Sampson EL: Risk factors for incident delirium among older people in acute hospital medical units: a systematic review and meta-analysis. Age Ageing. 2014; 43(3): 326–33. PubMed Abstract \| Publisher Full Text \| Free Full Text 9. Ahmed S, Leurent B, Sampson EL: Risk factors for incident delirium among older people in acute hospital medical units: a systematic review and meta-analysis. Age Ageing. 2014; 43(3): 326–33. PubMed Abstract \| Publisher Full Text \| Free Full Text 22. Schaller SJ, Anstey M, Blobner M, et al.: Early, goal-directed mobilisation in the surgical intensive care unit: a randomised controlled trial. Lancet. 2016; 388(10052): 1377–88. P bM d Ab \| P bli h F ll T References 1. Lin WL, Chen YF, Wang J: Factors Associated With the Development of Delirium in Elderly Patients in Intensive Care Units. J Nurs Res. 2015; 23(4): 322–9. PubMed Abstract \| Publisher Full Text 2. Kalabalik J, Brunetti L, El-Srougy R: Intensive care unit delirium: a review of the literature. J Pharm Pract. 2014; 27(2): 195–207. PubMed Abstract \| Publisher Full Text 3. Page VJ, Ely EW, Gates S, et al.: Effect of intravenous haloperidol on the duration of delirium and coma in critically ill patients (Hope-ICU): a randomised, double-blind, placebo-controlled trial. Lancet Respir Med. 2013; 1(7): 515–23. PubMed Abstract \| Publisher Full Text \| Free Full Text 4. Parker AM, Sricharoenchai T, Raparla S, et al.: Posttraumatic stress disorder in critical illness survivors: a metaanalysis. Crit Care Med. 2015; 43(5): 1121–9. PubMed Abstract \| Publisher Full Text 5. Lat I, McMillian W, Taylor S, et al.: The impact of delirium on clinical outcomes in mechanically ventilated surgical and trauma patients. Crit Care Med. 2009; 37(6): 1898–905. PubMed Abstract \| Publisher Full Text 6. Maldonado JR: Delirium pathophysiology: An updated hypothesis of the etiology of acute brain failure. Int J Geriatr Psychiatry. 2018; 33(11): 1428–57. PubMed Abstract \| Publisher Full Text 7. Milbrandt EB, Deppen S, Harrison PL, et al.: Costs associated with delirium in mechanically ventilated patients. Crit Care Med. 2004; 32(4): 955–62. PubMed Abstract \| Publisher Full Text 8. Micek ST, Anand NJ, Laible BR, et al.: Delirium as detected by the CAM-ICU predicts restraint use among mechanically ventilated medical patients. Crit Care Med 2005; 33(6): 1260–5. PubMed Abstract \| Publisher Full Text 9. Ahmed S, Leurent B, Sampson EL: Risk factors for incident delirium among older people in acute hospital medical units: a systematic review and meta-analysis. Age Ageing. 2014; 43(3): 326–33. PubMed Abstract \| Publisher Full Text \| Free Full Text 10. Fong TG, Tulebaev SR, Inouye SK: Delirium in elderly adults: diagnosis, prevention and treatment. Nat Rev Neurol. 2009; 5(4): 210–20. PubMed Abstract \| Publisher Full Text \| Free Full Text 11. Davoudi A, Ebadi A, Rashidi P, et al.: Delirium Prediction using Machine Learning Models on Preoperative Electronic Health Records Data. Proc IEEE Int Symp Bioinformatics Bioeng. 2017; 2017: 568–73. PubMed Abstract \| Publisher Full Text \| Free Full Text 12. Van Rompaey B, Elseviers MM, Schuurmans MJ, et al.: Risk factors for delirium in intensive care patients: a prospective cohort study. Crit Care. References 2009; 13(3): R77. PubMed Abstract \| Publisher Full Text \| Free Full Text 13. Thom RP, Bui MP, Rosner B, et al.: A Comparison of Early, Late, and No Treatment of Intensive Care Unit Delirium With Antipsychotics: A Retrospective Cohort Study. Prim Care Companion CNS Disord. 2018; 20(6): 18m02320. PubMed Abstract \| Publisher Full Text 14. Knauert MP, Pisani MA: Dexmedetomidine for hyperactive delirium: worth further study. J Thorac Dis. 2016; 8(9): E999–e1002. PubMed Abstract \| Publisher Full Text \| Free Full Text 15. Girard TD, Exline MC, Carson SS, et al.: Haloperidol and Ziprasidone for Treatment of Delirium in Critical Illness. N Engl J Med. 2018; 379(26): 2506–16. PubMed Abstract \| Publisher Full Text \| Free Full Text 16. Chen S, Shi L, Liang F, et al.: Exogenous Melatonin for Delirium Prevention: a Meta-analysis of Randomized Controlled Trials. Mol Neurobiol. 2016; 53(6): 4046–53. PubMed Abstract \| Publisher Full Text 17. Barbateskovic M, Krauss SR, Collet MO, et al.: Pharmacological interventions for prevention and management of delirium in intensive care patients: a systematic overview of reviews and meta-analyses. BMJ open. 2019; 9(2): e024562. PubMed Abstract \| Publisher Full Text \| Free Full Text 18. Huang X, Lin J, Demner-Fushman D: Evaluation of PICO as a knowledge representation for clinical questions. AMIA Annu Symp Proc. 2006; 2006: 359–63. PubMed Abstract \| Free Full Text 19. Moher D, Liberati A, Tetzlaff J, et al.: Preferred reporting items for systematic reviews and meta-analyses: the PRISMA statement. Ann Intern Med. 2009; 151(4): 264–9, w64. PubMed Abstract \| Publisher Full Text 20. Bihorac A: The Effect of Non-Pharmacologic Strategies on Prevention or Management of Intensive Care Unit Delirium: A Systematic Review. OSF. 2020. http://www.doi.org/10.17605/OSF.IO/C3RHF 21. Higgins JP, Sterne JA, Savovic J, et al.: A revised tool for assessing risk of bias in randomized trials. Cochrane database of systematic reviews. 2016; 10(Suppl 1): 29–31. Reference Source 22. Schaller SJ, Anstey M, Blobner M, et al.: Early, goal-directed mobilisation in the surgical intensive care unit: a randomised controlled trial. Lancet. 2016; 388(10052): 1377–88. PubMed Abstract \| Publisher Full Text 23. Patel J, Baldwin J, Bunting P, et al.: The effect of a multicomponent multidisciplinary bundle of interventions on sleep and delirium in medical and surgical intensive care patients. Anaesthesia. 2014; 69(6): 540–9. PubMed Abstract \| Publisher Full Text 24. PubMed Abstract \| Publisher Full Text 14. Knauert MP, Pisani MA: Dexmedetomidine for hyperactive delirium: worth Page 25 of 31 Page 25 of 31 F1000Research 2021, 9:1178 Last updated: 06 SEP 2022 F1000Research 2021, 9:1178 Last updated: 06 SEP 2022 1024–36. PubMed Abstract \| Publisher Full Text \| Free Full Text Community Hospital. Dimens Crit Care Nurs 2015; 34(5): 250–8. PubMed Abstract \| Publisher Full Text 41. Lisann-Goldman LR, Pagnini F, Deiner SG, et al.: Reducing Delirium and Improving Patient Satisfaction With a Perioperative Mindfulness Intervention: A Mixed-Methods Pilot Study. Holist Nurs Pract. 2019; 33(3): 163–76. PubMed Abstract \| Publisher Full Text 27. Bounds M, Kram S, Speroni KG, et al.: Effect of ABCDE Bundle Implementation on Prevalence of Delirium in Intensive Care Unit Patients. Am J Crit Care. 2016; 25(6): 535–44. PubMed Abstract \| Publisher Full Text 42. Mailhot T, Cossette S, Côté J, et al.: A post cardiac surgery intervention to manage delirium involving families: a randomized pilot study. Nurs Crit Care. 2017; 22(4): 221–8. PubMed Abstract \| Publisher Full Text 28. Bryczkowski SB, Lopreiato MC, Yonclas PP, et al.: Delirium prevention program in the surgical intensive care unit improved the outcomes of older adults. J Surg Res. 2014; 190(1): 280–8. PubMed Abstract \| Publisher Full Text ; PubMed Abstract \| Publisher Full Text \| Free Full Text 35. Guo Y, Sun L, Li L, et al.: Impact of multicomponent, nonpharmacologic interventions on perioperative cortisol and melatonin levels and postoperative delirium in elderly oral cancer patients. Arch Gerontol Geriatr. 2016; 62: 112–7. PubMed Abstract \| Publisher Full Text 50. Zhang W, Sun Y, Liu Y, et al.: A nursing protocol targeting risk factors for reducing postoperative delirium in patients following coronary artery bypass grafting: Results of a prospective before-after study. Int J Nurs Sci. 2017; 4(2): 81–7. PubMed Abstract \| Publisher Full Text 32. Eghbali-Babadi M, Shokrollahi N, Mehrabi T: Effect of Family-Patient Communication on the Incidence of Delirium in Hospitalized Patients in Cardiovascular Surgery ICU. Iran J Nurs Midwifery Res. 2017; 22(4): 327–31. PubMed Abstract \| Publisher Full Text \| Free Full Text 46. Rivosecchi RM, Kane-Gill SL, Svec S, et al.: The implementation of a nonpharmacologic protocol to prevent intensive care delirium. J Crit Care. 2016; 31(1): 206–11. PubMed Abstract \| Publisher Full Text PubMed Abstract \| Publisher Full Text 49. Sullinger D, Gilmer A, Jurado L, et al.: Development, Implementation, and Outcomes of a Delirium Protocol in the Surgical Trauma Intensive Care Unit. Ann Pharmacother. 2017; 51(1): 5–12. PubMed Abstract \| Publisher Full Text 2016; 31(1): 206 11. PubMed Abstract \| Publisher Full Text 33. Fallahpoor S, Abedi H, Mansouri M: Development and Evaluation of Care Programs for the Delirium Management in Patients after Coronary Artery Bypass Graft Surgery (CABG). International Journal of Medical Research Health Sciences. 2016; 5(7): 547–53. Reference Source 47. Rosa RG, Tonietto TF, da Silva DB, et al.: Effectiveness and Safety of an Extended ICU Visitation Model for Delirium Prevention: A Before and After Study. Crit Care Med. 2017; 45(10): 1660–7. PubMed Abstract \| Publisher Full Text 48. Simons KS, Laheij RJ, van den Boogaard M, et al.: Dynamic light application therapy to reduce the incidence and duration of delirium in intensive- care patients: a randomised controlled trial. Lancet Respir Med. 2016; 4(3): 194–202. 34. Giraud K, Pontin M, Sharples LD, et al.: Use of a Structured Mirrors Intervention Does Not Reduce Delirium Incidence But May Improve Factual Memory Encoding in Cardiac Surgical ICU Patients Aged Over 70 Years: A Pilot Time-Cluster Randomized Controlled Trial. Front Aging Neurosci. 2016; 8: 228. 52(9): 1423 32. PubMed Abstract \| Publisher Full Text 44. Munro CL, Cairns P, Ji M, et al.: Delirium prevention in critically ill adults through an automated reorientation intervention - A pilot randomized controlled trial. Heart Lung. 2017; 46(4): 234–8. PubMed Abstract \| Publisher Full Text 30. Chai J: The effect of the ABCDEF bundle on incidence of delirium in critically ill patients. Brandman University. 2017. Reference Source 31. Damshens MH, Sanie MS, Javadpour S, et al.: The Role of Music on the Delirium in Traumatic Patients: A Case Study in the ICU of Peymanieh Hospital of Jahrom, Fars Province, Iran. Ambient Science. 2018; 5: 97–101. Reference Source 45. Parry SM, Berney S, Warrillow S, et al.: Functional electrical stimulation with cycling in the critically ill: a pilot case-matched control study. J Crit Care. 2014; 29(4): 695.e1–7. PubMed Abstract \| Publisher Full Text g Full Text Publisher Full Text 37. Pun BT, Balas MC, Barnes-Daly MA, et al.: Caring for Critically Ill Patients with the ABCDEF Bundle: Results of the ICU Liberation Collaborative in Over 15,000 Adults. Crit Care Med. 2019; 47(1): 3–14. PubMed Abstract \| Publisher Full Text \| Free Full Text 38. Karadas C, Ozdemir L: The effect of range of motion exercises on delirium prevention among patients aged 65 and over in intensive care units. Geriatr Nurs 2016; 37(3): 180–5. PubMed Abstract \| Publisher Full Text 39. Khan BA, Fadel WF, Tricker JL, et al.: Effectiveness of implementing a wake up and breathe program on sedation and delirium in the ICU. Crit Care Med. 2014; 42(12): e791–5. PubMed Abstract \| Publisher Full Text \| Free Full Text 40. Kram SL, DiBartolo MC, Hinderer K, et al.: Implementation of the ABCDE Bundle to Improve Patient Outcomes in the Intensive Care Unit in a Rural 37. Pun BT, Balas MC, Barnes-Daly MA, et al.: Caring for Critically Ill Patients with the ABCDEF Bundle: Results of the ICU Liberation Collaborative in Over 15,000 Adults. Crit Care Med. 2019; 47(1): 3–14. PubMed Abstract \| Publisher Full Text \| Free Full Text 52. Morandi A, Brummel NE, Ely EW: Sedation, delirium and mechanical ventilation: the ‘ABCDE’ approach. Curr Opin Crit Care. 2011; 17(1): 43–9. PubMed Abstract \| Publisher Full Text 38. Karadas C, Ozdemir L: The effect of range of motion exercises on delirium prevention among patients aged 65 and over in intensive care units. Geriatr Nurs 2016; 37(3): 180–5. PubMed Abstract \| Publisher Full Text 53. Kim MY, Park UJ, Kim HT, et al.: DELirium Prediction Based on Hospital Information (Delphi) in General Surgery Patients . Medicine (Baltimore). 2016; 95(12): e3072. PubMed Abstract \| Publisher Full Text \| Free Full Text 39. Khan BA, Fadel WF, Tricker JL, et al.: Effectiveness of implementing a wake up and breathe program on sedation and delirium in the ICU. Crit Care Med. 2014; 42(12): e791–5. PubMed Abstract \| Publisher Full Text \| Free Full Text 54. van den Boogaard M, Pickkers P, Slooter AJ, et al.: Development and validation of PRE-DELIRIC (PREdiction of DELIRium in ICu patients) delirium prediction model for intensive care patients: observational multicentre study. BMJ. 2012; 344: e420. PubMed Abstract \| Publisher Full Text \| Free Full Text ( ) PubMed Abstract \| Publisher Full Text 43. Moon KJ, Lee SM: The effects of a tailored intensive care unit delirium prevention protocol: A randomized controlled trial. Int J Nurs Stud. 2015; 52(9): 1423–32. PubMed Abstract \| Publisher Full Text 29. Campbell MR: The Effect of an Early Mobility Protocol in Critically Ill Mechanically Ventilated Patients on Incidence and Duration of Delirium and Length of Stay. 2014. Reference Source ; : . PubMed Abstract \| Publisher Full Text 36. Hamzehpour H, Valiee S, Majedi MA, et al.: The Effect of Care Plan Based on Roy Adaptation Model on the Incidence and Severity of Delirium in Intensive Care Unit Patients: A Randomised Controlled Trial. Journal of Clinical Diagnostic Research. 2018; 12(11): LC21–LC25. Publisher Full Text 2017; 4(2): 81 7. PubMed Abstract \| Publisher Full Text \| Free Full Text 51. Girard TD, Kress JP, Fuchs BD, et al.: Efficacy and safety of a paired sedation and ventilator weaning protocol for mechanically ventilated patients in intensive care (Awakening and Breathing Controlled trial): a randomised controlled trial. Lancet. 2008; 371(9607): 126–34. PubMed Abstract \| Publisher Full Text https://doi.org/10.5256/f1000research.30705.r74838 © 2022 Spies C et al. This is an open access peer review report distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. Claudia D. Spies Department of Anesthesiology and Operative Intensive Care Medicine (CCM, CVK), Charité - Universitätsmedizin Berlin, Berlin, Germany Fatima Yürek Charité - Universitätsmedizin Berlin, Berlin, Germany Alawi Luetz Department of Anesthesiology and Operative Intensive Care Medicine (CCM, CVK), Charité - Universitätsmedizin Berlin, Berlin, Germany Department of Anesthesiology and Operative Intensive Care Medicine (CCM, CVK), Charité - Universitätsmedizin Berlin, Berlin, Germany 2014; 42(12): e791 5. PubMed Abstract \| Publisher Full Text \| Free Full Text 40. Kram SL, DiBartolo MC, Hinderer K, et al.: Implementation of the ABCDE Bundle to Improve Patient Outcomes in the Intensive Care Unit in a Rural Page 26 of 31 Page 26 of 31 F1000Research 2021, 9:1178 Last updated: 06 SEP 2022 Open Peer Review Current Peer Review Status: Major issues: The evaluations do not clearly differentiate between ICU delirium in critically ill patients and postoperative delirium in the ICU. The risk factors for critically ill patients with acute organ failure requiring ICU treatment and patients who have had elective or emergency surgery may differ. It would be better if the populations of critically ill patients were considered separately from the population of elective and emergency surgery patients. Given the predisposing and precipitating risk factors that are additional to surgery, this could impact the effectiveness of nonpharmacologic preventive measures. 1. Are the conclusions drawn adequately supported by the results presented in the review? Partly We confirm that we have read this submission and believe that we have an appropriate level of expertise to confirm that it is of an acceptable scientific standard, however we have significant reservations, as outlined above. Author Response ( ) 03 Jun 2022 Azra Bihorac, University of Florida, Gainesville, USA Response to Major Issue #1: Thank you so much for pointing this out – this distinction is imperative. We added a brief segment to the limitations portion of our discussion section discussing this important delineation. Response to Minor Issue #1: This is a great consideration, thank you for mentioning this limitation. We added related information to the future directions portion in the article’s discussion section. Response to Minor Issue #2: Thank you for also bringing this distinction to our attention. This is a valid concern – certain strategies may be more effective for older versus younger ICU delirium patients and these sub-populations may have different underlying risk factors and etiologies. We added a segment into our discussion section regarding this aspect. Competing Interests: No competing interests were disclosed. Reviewer Report 05 November 2020 https://doi.org/10.5256/f1000research.28439.r73595 Minor issues: The effectiveness and efficiency of non-pharmacological measures is limited to the intensive care unit. It remains open how non-pharmacological measures affect patients in the normal ward or to what extent the results can be transferred to structures in the normal ward. The effect of mobility in the normal ward offers a much greater advantage for the patients (preservation of authority, self-care, and independence) than it could be shown in the ICU. Therefore, this should be emphasized more clearly. 1. The age distribution in the studies considered is very broad. The question arises whether the integration of relatives into the treatment concept of non-pharmacological preventive measures for delirium has shown an independent preventive effect or whether this effect could be shown specifically in certain age groups. 2. Are the rationale for, and objectives of, the Systematic Review clearly stated? Page 27 of 31 Yes Are sufficient details of the methods and analysis provided to allow replication by others? Yes Is the statistical analysis and its interpretation appropriate? Partly Are the conclusions drawn adequately supported by the results presented in the review? Partly Competing Interests: No competing interests were disclosed. We confirm that we have read this submission and believe that we have an appropriate level of expertise to confirm that it is of an acceptable scientific standard, however we have significant reservations, as outlined above. Author Response ( ) 03 Jun 2022 Azra Bihorac, University of Florida, Gainesville, USA Response to Major Issue #1: Thank you so much for pointing this out – this distinction is imperative. We added a brief segment to the limitations portion of our discussion section discussing this important delineation. Response to Minor Issue #1: This is a great consideration, thank you for mentioning this limitation. We added related information to the future directions portion in the article’s discussion section. Response to Minor Issue #2: Thank you for also bringing this distinction to our attention. This is a valid concern – certain strategies may be more effective for older versus younger ICU delirium patients and these sub-populations may have different underlying risk factors and etiologies. We added a segment into our discussion section regarding this aspect. Competing Interests: No competing interests were disclosed. F1000Research 2021, 9:1178 Last updated: 06 SEP 2022 F1000Research 2021, 9:1178 Last updated: 06 SEP 2022 Yes Are sufficient details of the methods and analysis provided to allow replication by others? Minor issues: Yes Is the statistical analysis and its interpretation appropriate? Partly Are the conclusions drawn adequately supported by the results presented in the review? Partly Competing Interests: No competing interests were disclosed. We confirm that we have read this submission and believe that we have an appropriate level of expertise to confirm that it is of an acceptable scientific standard, however we have significant reservations, as outlined above. Author Response ( ) 03 Jun 2022 Azra Bihorac, University of Florida, Gainesville, USA R t M j I #1 Th k h f i ti thi t thi di ti ti i For the outcomes that were measured but not analyzed, it may be clearer to write “not analyzed” or “NA” rather than “o.” 3. For Table 2, I found it confusing using x’s, y’s and o’s. It may be better to title the Table “Summary of delirium outcomes by statistical significance (p<0.05),” and then spell out “yes” instead of y, and either “no” or “not significant” instead of x. For the outcomes that were measured but not analyzed, it may be clearer to write “not analyzed” or “NA” rather than “o.” 3. The other tables were clear and appropriately organized. The discussion is well written and discusses the limitations of the included studies and of this systematic review. I have no suggestions for improvement. F1000Research 2021, 9:1178 Last updated: 06 SEP 2022 Elizabeth Mahanna-Gabrielli Department of Anesthesiology, Perioperative and Pain Medicine, University of Miami Miller School of Medicine, Miami, FL, USA Elizabeth Mahanna-Gabrielli Department of Anesthesiology, Perioperative and Pain Medicine, University of Miami Miller School of Medicine, Miami, FL, USA This manuscript is a systematic review of the effect of non-pharmacologic strategies on prevention or management of delirium in the ICU. The manuscript is comprehensive, well written, and follows appropriate PICOS framework and PRISMA guidelines for a systematic review. The introduction is well written, succinct and appropriately gives background information on the importance of ICU delirium prevention and management. The methods clearly state how the authors followed the PRISMA guidelines. The authors searched Pubmed, Embase, Cochrane Central, Web of Science, CINAHL, and Clinciattrials.gov. MeSH terms appropriately included delirium and other terms that are often used to describe the delirium syndrome such as “acute brain dysfunction” and “psychosis”. A full list of the search terms for each database is included in supplementary appendix 2 to allow for reproducibility. Could the authors clarify if the date range was 5 years from the date of search (Jan 1, 2014 to May 15, 2019) as stated in the manuscript? Or was it 10 years, as is listed in the search terms of the supplementary appendix 2? 1. Did all the authors screen the titles and abstracts? And did all the authors independently review the full text of the 113 eligible articles, or only some of the authors? 2. Did all the authors screen the titles and abstracts? And did all the authors independently review the full text of the 113 eligible articles, or only some of the authors? 2. Bias was appropriately assessed by the Cochrane risk of bias tool. Disagreements between authors were adjudicated as a group through discussion. Bias was appropriately assessed by the Cochrane risk of bias tool. Disagreements between authors were adjudicated as a group through discussion. The results are clearly written and concise. The results are clearly written and concise. For Table 2, I found it confusing using x’s, y’s and o’s. It may be better to title the Table “Summary of delirium outcomes by statistical significance (p<0.05),” and then spell out “yes” instead of y, and either “no” or “not significant” instead of x. https://doi.org/10.5256/f1000research.28439.r73595 © 2020 Mahanna-Gabrielli E. This is an open access peer review report distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. Page 28 of 31 Competing Interests: No competing interests were disclosed. Reviewer Expertise: perioperative brain health, delirium - ICU and postoperative, sleep I confirm that I have read this submission and believe that I have an appropriate level of expertise to confirm that it is of an acceptable scientific standard. eviewer Expertise: perioperative brain health, delirium - ICU and postoperative, sleep Is the statistical analysis and its interpretation appropriate? Yes Are the conclusions drawn adequately supported by the results presented in the review? Page 29 of 31 F1000Research 2021, 9:1178 Last updated: 06 SEP 2022 Yes Competing Interests: No competing interests were disclosed. I confirm that I have read this submission and believe that I have an appropria expertise to confirm that it is of an acceptable scientific standard. Azra Bihorac, University of Florida, Gainesville, USA Could the authors clarify if the date range was 5 years from the date of search (Jan 1, 2014 to May 15, 2019) as stated in the manuscript? Or was it 10 years, as is listed in the search terms of the supplementary appendix 2? Thank you so much for this comment so that we can clarify. The searches for each database listed in Supplementary Appendix 2 were initially performed for 10 years from the date of the search. During the title and abstract screening process, we ultimately decided to narrow the results further to within 5 years of the dated search. This step can be seen in Fig 1 PRISMA Record Screening Flow Chart, where we removed 3039 records prior to 2014. We added a statement to our methods section to further clarify this aspect. 1. Did all the authors screen the titles and abstracts? And did all the authors independently review the full text of the 113 eligible articles, or only some of the authors? Thank you for informing us of this insufficiency. Four authors (JSC, HH, MMR, RB) screened the titles and abstracts, and three authors independently reviewed the full text (JSC, HH, JL). We added a statement to our methods section clarifying this point. 2. Did all the authors screen the titles and abstracts? And did all the authors independently review the full text of the 113 eligible articles, or only some of the authors? Thank you for informing us of this insufficiency. Four authors (JSC, HH, MMR, RB) screened the titles and abstracts, and three authors independently reviewed the full text (JSC, HH, JL). We added a statement to our methods section clarifying this point. 2. For Table 2, I found it confusing using x’s, y’s and o’s. It may be better to title the Table “Summary of delirium outcomes by statistical significance (p<0.05),” and then spell out “yes” instead of y, and either “no” or “not significant” instead of x. For the outcomes that were measured but not analyzed, it may be clearer to write “not analyzed” or “NA” rather than “o.” Thank you for the feedback about this table. This change will definitely make our table more clear to the reader, and Table 2 has been updated to reflect this amendment. 3. For Table 2, I found it confusing using x’s, y’s and o’s. Competing Interests: No competing interests were disclosed. Reviewer Expertise: perioperative brain health, delirium - ICU and postoperative, sleep I confirm that I have read this submission and believe that I have an appropriate level of expertise to confirm that it is of an acceptable scientific standard. I confirm that I have read this submission and believe that I have an appropriate level of expertise to confirm that it is of an acceptable scientific standard. I confirm that I have read this submission and believe that I have an appropria expertise to confirm that it is of an acceptable scientific standard. It may be better to title the Table “Summary of delirium outcomes by statistical significance (p<0.05),” and then spell out “yes” instead of y, and either “no” or “not significant” instead of x. For the outcomes that were measured but not analyzed, it may be clearer to write “not analyzed” or “NA” rather than “o.” Thank you for the feedback about this table. This change will definitely make our table more clear to the reader, and Table 2 has been updated to reflect this amendment. 3. Competing Interests: No competing interests were disclosed. Competing Interests: No competing interests were disclosed. Page 30 of 31 F1000Research 2021, 9:1178 Last updated: 06 SEP 2022 The benefits of publishing with F1000Research: Your article is published within days, with no editorial bias • You can publish traditional articles, null/negative results, case reports, data notes and more • The peer review process is transparent and collaborative • Your article is indexed in PubMed after passing peer review • Dedicated customer support at every stage • For pre-submission enquiries, contact research@f1000.com The benefits of publishing with F1000Research: Your article is published within days, with no editorial bias • You can publish traditional articles, null/negative results, case reports, data notes and more • The peer review process is transparent and collaborative • Your article is indexed in PubMed after passing peer review • Dedicated customer support at every stage • For pre-submission enquiries, contact research@f1000.com The benefits of publishing with F1000Research: The benefits of publishing with F1000Research: Page 31 of 31
https://openalex.org/W3164725762	https://www.e3s-conferences.org/10.1051/e3sconf/202126102062/pdf	English	null	Design and application of purified water preparation device for chemical analysis	E3S web of conferences	2,021	cc-by	1,649	1 Overview A compact, efficient, and well-operated water treatment equipment is one of the infrastructure of the analysis laboratory. This water treatment equipment mainly provides all levels of water demand for analysis laboratories. The equipment uses ordinary tap water as the water source, and reverse osmosis is the pre- demineralization system. It is composed of a total of 3 RO membranes in the first and second stages to separate and remove 97% of total dissolved solids. The ion exchange water treatment equipment removes residual impurities. Among them, the RO effluent can be passed through the cation exchanger and anion exchanger in turn to reach the secondary water standard for analytical laboratories, and after passing through the mixed ion exchanger, it can reach the primary water standard. In the trace analysis, the rest of the very small amount of dissolved solids by ion Design and application of purified water preparation device for chemical analysis 1Shandong Electric Power Technical College, Jinan, 250002 Abstract. Qualified purified water is essential for all chemical analysis. Spectral analysis, trace analysis and general chemical analysis require different purity of water. This paper introduces a kind of purified water preparation device which can be widely used in all chemical analysis work. The purified water preparation device consists of two-stage reverse osmosis device and water treatment device of ion exchange system. The two-stage reverse osmosis water treatment unit can be operated in two modes: program-controlled and manual mode, and the water production rate is about 80%. The ion exchange water treatment unit consists of two cation exchangers, two cation exchangers and one mixed ion exchanger, which can prepare silicon free water and sodium free water and meet the requirements of trace analysis. The ion exchange water treatment unit is equipped with independent purified water pipeline system and regenerated liquid pipeline system to avoid the pollution of purified water. The regeneration operation of ion exchange resin is simple, and the discontinuous water production can be easily realized. This set of purified water preparation device can adapt to various water sources to complete purification treatment, and produce purified water meeting various chemical analysis. exchange resin to remove. Equipped with effluent conductivity and pH on-line instrument monitor to monitor water quality changes at any time to ensure stable and qualified water quality. * L.Yanping li9390@126.com © The Authors, published by EDP Sciences. This is an open access article distributed under the terms of the Creative Commons Attribution License 4.0 (http://creativecommons.org/licenses/by/4.0/). https://doi.org/10.1051/e3sconf/202126102062 https://doi.org/10.1051/e3sconf/202126102062 E3S Web of Conferences 261, 02062 (2021) ICEMEE 2021 * L.Yanping li9390@126.com 4 Application effectiveness The regeneration system is relatively independent, adopting the design of low-position acid-base tank for liquid distribution and high-position acid-base tank for self-flow regeneration, as shown in Figure 3. The regeneration process does not require continuous operation of acid-base pumps. During the regeneration process, the regeneration liquid flows smoothly, the resin does not mess up, the regeneration degree is high, and the effluent quality is good. 2.1 RO system process The inlet water source of this RO system is tap water. It adopts a one-stage and two-stage arrangement, with one ultrafiltration membrane(UF), one coconut shell activated carbon filter, security filter and three RO membranes. The system process is shown in Figure 1. All operation buttons and indicators are integrated in the operation panel. 1—UF; 2—security filter; 3—high pressure pump; 4—RO; 5—product water; 6—concentrated water Fig.1 Reverse osmosis system 1—UF; 2—security filter; 3—high pressure pump; 4—RO; 5—product water; 6—concentrated water Fig.1 Reverse osmosis system Fig.1 Reverse osmosis system L.Yanping li9390@126.com © The Authors, published by EDP Sciences. This is an open access article distributed under the terms of the Creative Commons Attribution License 4.0 (http://creativecommons.org/licenses/by/4.0/). E3S Web of Conferences 261, 02062 (2021) ICEMEE 2021 https://doi.org/10.1051/e3sconf/202126102062 2.3 System settings and functions The outlet water tank is equipped with high water level and low water level protection equipments. The high water level protection is linked with the RO body, and it will automatically stop when the set water level is reached. The low water level protection is linked with the subsequent ion exchange system water inlet pump, and the pump is automatically stopped when the water level is lower than the set water level. Set the concentrated water return valve and pipeline to adjust the concentrated water return flow rate, water production rate, and pressure between sections. The antiscalant dosing system is started in conjunction with the high-pressure pump. 3.1 Process of ion exchange system Security filter: Polypropylene (PP) hot-melt filter element, filtration accuracy 5μm Security filter: Polypropylene (PP) hot-melt filter element, filtration accuracy 5μm The ion exchange water treatment system consists of five ion exchange columns, including two cation exchangers, two anion exchangers and one mixed ion exchanger. It is arranged in the order of one cation exchanger + one anion exchanger + one mixed ion exchanger + one cation exchanger + one anion exchanger, as shown in Figure 2. RO water outlet tank: 1m3, polyethylene material, with high and low water level linkage protection. 3.2 Main parameters of ion exchange system Cation exchange resin: 001 × 7 Strong acid cation exchange resin for water treatment; Anion exchange resin: 201×7 Strong basic cation exchange resin for water treatment; Ion exchange column: φ200mm×1000mm; Ion exchange column: φ200mm×1000mm; High-level water tank: V=0.25m3; Acid and alkali tank: V=0.05m3; Acid and alkali tank: V=0.05m3; Acid (/alkali pump): polypropylene Acid (/alkali pump): polypropylene corrosion-resistant centrifugal pump, Q=3m3/h, H=6m; Acid (/alkali pump): polypropylene corrosion-resistant centrifugal pump, Q=3m3/h, H=6m; g p p, Q , ; Resin mixing ratio: (positive: negative resin=1:2); Resin mixing ratio: (positive: negative resin 1:2) Legend: 1—Cation exchanger; 2—Cation exchanger; 3—Mixed ion exchanger； 4—level 1 outlet; 5—level 2 outlet;6—level 3 outlet Fig. 2 Structure of ion exchange equipment Legend: 1—Cation exchanger; 2—Cation exchanger; 3—Mixed ion exchanger； 4—level 1 outlet; 5—level 2 outlet;6—level 3 outlet Fig. 2 Structure of ion exchange equipment Fig. 2 Structure of ion exchange equipment 2.2 Equipment parameters of reverse osmosis system The water inlet of the ion exchange system can switch the water source between RO outlet water and tap water as needed. RO membrane: Dow BW30-4040; Ultrafiltration membrane: LH3-8Hd High pressure pump: Q=2m3/h, H=98m; Security filter: Polypropylene (PP) hot-melt filter element, filtration accuracy 5μm RO membrane: Dow BW30-4040; Ultrafiltration membrane: LH3-8Hd High pressure pump: Q=2m3/h, H=98m; Ultrafiltration membrane: LH3-8Hd High pressure pump: Q=2m3/h, H=98m; 4.2Water quality of water treatment plant 4.2Water quality of water treatment plant Tab.2 Water quality of water treatment plant Detection location Conductivity(25℃) mS/m pH(25℃) Raw water 62~68 7.49~7.56 RO Produced water 1.2~2.0 6.9~7.3 level 1 Outlet 0.006~0.02 6.7~6.8 level 2 Outlet 0.006~0.01 6.5~6.8 level 3 Outlet 0.0056~0.01 6.8~7.0 exchange water treatment equipment in each operation cycle is related to the dissolved solid content of reverse osmosis produced water. Under the condition that the effluent conductivity is not higher than 1.5ms /m, the system can produce about 500m3 of pure water for analysis laboratory each operation cycle, and the ion exchange resin needs to be regenerated once a year. Therefore, the water treatment equipment has a high prospect of popularization and application. 5 Conclusion The purifying water treatment equipment has been used in many research institutions and university analysis laboratories. Raw water can be used tap water or other water sources, after the reverse osmosis purification treatment to remove most of the dissolved solids, and then through the ion exchange resin to remove the residual dissolved solids, appropriately increase the number of positive and negative beds, can improve the quality of water production, to meet the requirements of more precise analysis and test. This water treatment and purification device has been in use for more than 3 years. At present, the desalination rate of reverse osmosis system is stable at more than 97%. The water yield of ion- 4.1 Water requirements for analytical laboratories According to the national standard GB/T6682-2008, laboratory water is divided into three specifications, as shown in Table 1. Among them, tertiary water is used for general chemical analysis experiments, secondary water 2 2 E3S Web of Conferences 261, 02062 (2021) ICEMEE 2021 https://doi.org/10.1051/e3sconf/202126102062 Legend: 1 - high acid tank; 2 - high alkali tank; 3 - low acid tank;4 - low alkali tank；5-Acid pump;6-alkali pump. Fig.3 Reborn aqua system Legend: 1 - high acid tank; 2 - high alkali tank; 3 - low acid tank;4 - low alkali tank；5-Acid pump;6-alkali pump. Fig.3 Reborn aqua system is used for precision chemical analysis such as inorganic trace analysis, and primary water is used for analytical experiments with strict requirements. The water quality of the water treatment unit is shown in Table 2. Tab.1 Water for analytical laboratory use Specification Test item Level 1 Level 2 Level 3 pH(25℃) — — 5.0~7.5 Conductivity(25℃) mS/m ≤0.01 ≤0.10 ≤0.50 Oxidable (calculated as O) mg/L — ≤0.08 ＜0.40 Absorbance (254nm, 1cm optical path) ≤0.001 ≤0.01 — Evaporation residue （105℃） mg/L — ≤1.0 ≤2.0 Soluble silicon (calculated as SiO2）mg/L ≤0.01 ≤0.02 — 4.2Water quality of water treatment plant References 1. H. Chengqun. Water treatment equipment and operation of power plant, China Electric Power Press, (2015) 2. G. Jiaocheng. Ion exchange membrane and its 2. G. Jiaocheng. Ion exchange membrane and its 3 E3S Web of Conferences 261, 02062 (2021) E3S Web of Conferences 261, 02062 (2021) ICEMEE 2021 https://doi.org/10.1051/e3sconf/202126102062 application in water treatment[J]. Science and Technology Information, 12(31):44(2014) application in water treatment[J]. Science and Technology Information, 12(31):44(2014) 3. M. Shuqiang. Comparison of water treatment with full membrane method and ion exchange desalination system[J]. Journal of Chongqing Electric Power College, 17(02): 85-87(2012) 4. Z. Xiaoping. Wen Quan, Xu Bin. Application of ion exchange real bed in the preparation of pure water[J]. Guangdong Chemical Industry, 44(11): 172- 174(2017) 5. L.Yanping i,An ion exchange water treatment training device for teaching[P].Chinese patent,20180907 5. L.Yanping i,An ion exchange water treatment training device for teaching[P].Chinese patent,20180907 6. Yanping Li,An ion exchange resin treatment device[P].Chinese patent, 20200619 7. Water for analytical laboratory use Specification and test methods，GB/T6682-2008 （IOS 3696:1987， MOD） 4 4 4
https://openalex.org/W3131653920	https://www.researchsquare.com/article/rs-151525/latest.pdf	English	null	Tumor suppressor SMAR1 regulates PKM alternative splicing by HDAC6-mediated deacetylation of PTBP1	Cancer & metabolism	2,021	cc-by	13,289	Page 1/30 Tumor Suppressor SMAR1 Regulates PKM Alternative Splicing by HDAC6 Mediated Deacetylation of PTBP1 Arpankumar Choksi NCCS: National Centre For Cell Science Apoorva Parulekar NCCS: National Centre For Cell Science Richa Pant NCCS: National Centre For Cell Science Vibhuti Kumar Shah NCCS: National Centre For Cell Science Ramakrishna Nimma NCCS: National Centre For Cell Science Priyanka Firmal NCCS: National Centre For Cell Science Smriti Singh IISER Bhopal: Indian Institute of Science Education and Research Bhopal Gopal C. Kundu Kalinga Institute of Industrial Technology Sanjeev Shukla IISER Bhopal: Indian Institute of Science Education and Research Bhopal Samit Chattopadhyay (  samitchatterji@yahoo.com ) Birla Institute of Technology and Science Pilani - K K Birla Goa Campus https://orcid.org/0000-0003- 3581-9625 Research Keywords: SMAR1, PKM1, PKM2, HDAC6, PTBP1, alternative splicing, Warburg effect Posted Date: February 1st, 2021 DOI: https://doi.org/10.21203/rs.3.rs-151525/v1 License:   This work is licensed under a Creative Commons Attribution 4.0 International License. Read Full License Tumor Suppressor SMAR1 Regulates PKM Alternative Splicing by HDAC6 Mediated Deacetylation of PTBP1 Arpankumar Choksi NCCS: National Centre For Cell Science Vibhuti Kumar Shah NCCS: National Centre For Cell Science Vibhuti Kumar Shah NCCS: National Centre For Cell Science Ramakrishna Nimma NCCS: National Centre For Cell Science Ramakrishna Nimma NCCS: National Centre For Cell Science Priyanka Firmal NCCS: National Centre For Cell Science Tumor Suppressor SMAR1 Regulates PKM Alternative Splicing by HDAC6 Mediated Deacetylation of PTBP1 Arpankumar Choksi NCCS: National Centre For Cell Science Apoorva Parulekar NCCS: National Centre For Cell Science Richa Pant NCCS: National Centre For Cell Science Vibhuti Kumar Shah NCCS: National Centre For Cell Science Ramakrishna Nimma NCCS: National Centre For Cell Science Priyanka Firmal NCCS: National Centre For Cell Science Smriti Singh IISER Bhopal: Indian Institute of Science Education and Research Bhopal Gopal C. Kundu Kalinga Institute of Industrial Technology Sanjeev Shukla IISER Bhopal: Indian Institute of Science Education and Research Bhopal Samit Chattopadhyay (  samitchatterji@yahoo.com ) Birla Institute of Technology and Science Pilani - K K Birla Goa Campus https://orcid.org/0000 3581-9625 Research Page 1/30 Tumor Suppressor SMAR1 Regulates PKM Alternative Splicing by HDAC6 Mediated Deacetylation of PTBP1 Arpankumar Choksi NCCS: National Centre For Cell Science Apoorva Parulekar NCCS: National Centre For Cell Science Richa Pant NCCS: National Centre For Cell Science Vibhuti Kumar Shah NCCS: National Centre For Cell Science Ramakrishna Nimma NCCS: National Centre For Cell Science Priyanka Firmal NCCS: National Centre For Cell Science Smriti Singh IISER Bhopal: Indian Institute of Science Education and Research Bhopal Gopal C. Kundu Kalinga Institute of Industrial Technology Sanjeev Shukla IISER Bhopal: Indian Institute of Science Education and Research Bhopal Samit Chattopadhyay (  samitchatterji@yahoo.com ) Birla Institute of Technology and Science Pilani - K K Birla Goa Campus https://orcid.org/0000-0003- 3581-9625 Research Keywords: SMAR1, PKM1, PKM2, HDAC6, PTBP1, alternative splicing, Warburg effect Posted Date: February 1st, 2021 DOI: https://doi.org/10.21203/rs.3.rs-151525/v1 License:   This work is licensed under a Creative Commons Attribution 4.0 International License. Read Full License Version of Record: A version of this preprint was published at Cancer & Metabolism on April 16th, 2021. See the published version at https://doi.org/10.1186/s40170-021-00252-x. Results: The expression profile of SMAR1 and PKM isoforms in breast cancer patients revealed that SMAR1 has an inverse correlation with PKM2 and a positive correlation with PKM1. Further quantitative PKM isoform expression upon modulation in SMAR1 expression also reflects that SMAR1 promotes the expression of PKM1 over tumorigenic isoform PKM2. SMAR1 deacetylates PTBP1 via recruitment of HDAC6 resulting in reduced enrichment of PTBP1 on PKM pre-mRNA. SMAR1 inhibits the Warburg effect, tumorigenic potential of cancer cells and in vivo tumor generation in PKM2 dependent manner. Conclusions: SMAR1 regulates PKM alternative splicing by causing HDAC6 dependent deacetylation of PTBP1, resulting in reduced enrichment of PTBP1 on PKM pre-mRNA. Additionally, SMAR1 suppresses glucose utilization and lactate production via repression of PKM2 expression. This suggests that tumor suppressor SMAR1 inhibits tumor cell metabolism and tumorigenic properties of cancer cells via regulation of PKM alternative splicing. Research Page 1/30 Version of Record: A version of this preprint was published at Cancer & Metabolism on April 16th, 2021. See the published version at https://doi.org/10.1186/s40170-021-00252-x. Version of Record: A version of this preprint was published at Cancer & Metabolism on April 16th, 2021. See the published version at https://doi.org/10.1186/s40170-021-00252-x. Page 2/30 Background: Highly proliferating cancer cells exhibit the Warburg effect by regulation of PKM alternative splicing and promoting the expression of PKM2. Majority of the alternative splicing events are known to occur in the nuclear matrix where various MARBPs actively participate in the alternative splicing events. SMAR1, being a MARBP and an important tumor suppressor, is known to regulate the splicing of various cancer- associated genes. This study focuses on the regulation of PKM alternative splicing and inhibition of the Warburg effect by SMAR1. Highly proliferating cancer cells exhibit the Warburg effect by regulation of PKM alternative splicing and promoting the expression of PKM2. Majority of the alternative splicing events are known to occur in the nuclear matrix where various MARBPs actively participate in the alternative splicing events. SMAR1, Methods: Immunohistochemistry was performed in breast cancer patient samples to establish the correlation between SMAR1 and PKM isoform expression. Further, expression of PKM isoforms upon modulation in SMAR1 expression in breast cancer cell lines was quantified by qRT-PCR and western blot. The acetylation status of PTBP1 was estimated by immunoprecipitation along with its enrichment on PKM pre-mRNA by CLIP in SMAR1 knockdown conditions. The role of SMAR1 in tumor metabolism and tumorigenesis was explored by in vitro enzymatic assays and functional assays upon SMAR1 knockdown. Besides, in vivo tumor formation by injecting adeno-SMAR1 transduced MDA-MB-231 cells in NOD/SCID mice was performed. Results: Background Reprogramming cellular metabolism is one of the key hallmarks of cancer cells. In the presence of oxygen, somatic cells convert glucose to pyruvate by glycolysis and then redirect this pyruvate to the TCA Page 3/30 Page 3/30 cycle in mitochondria for ATP production. In the absence of oxygen, these cells convert pyruvate to lactate which is an energetically less efficient but relatively fast reaction. Whereas highly proliferating cancer cells utilize more glucose and produce more lactate compared to normal cells, independent of the presence or absence of oxygen. This phenomenon is known as the Warburg effect or aerobic glycolysis1,2. Cancer cells reprogram their cellular metabolism by regulating alternative splicing of Pyruvate Kinase Muscle (PKM) isoforms3. Pyruvate kinase is one of the rate-determining enzymes of glycolysis and facilitates the conversion of phosphoenolpyruvate to pyruvate. There are total four isoforms of pyruvate kinase enzyme: PKL (Pyruvate kinase liver), PKR (Pyruvate kinase Red blood cells), PKM1 and PKM2. PKL is expressed in the liver, PKR is expressed in the erythrocytes, PKM1 is expressed predominantly in terminally differentiated tissues and PKM2 is highly expressed in the proliferating cells such as cancer cells and stem cells3,4. PKM gene contains 12 exons, wherein the incorporation of exon 9 and exon 10 are regulated by mutually exclusive alternative splicing resulting in the expression of PKM1 and PKM2 isoforms, respectively5,6. Exon 9 and exon 10 both codes for 56 amino acids attributing to the distinctive regulatory and enzymatic activities of PKM isoforms. PKM1 forms a tetramer and is constitutively active, whereas PKM2 forms a tetramer (enzymatically efficient), as well as a dimer (enzymatically less efficient). Enzymatic activity of PKM2 dimer is influenced by allosteric regulation due to fructose-1,6-bisphosphate levels and interaction with phosphotyrosine-binding proteins3,4,7. Higher expression of PKM2 provides cancer cells with a metabolic advantage over normal cells. Due to the lower enzymatic activity of PKM2 as compared to that of PKM1, the major amount of glucose present in the cell remains as glycolytic intermediates which provide building blocks, such as amino acids, nucleotides and fatty acids to the highly proliferating cancer cells8. The remaining glucose gets converted to pyruvate and this pyruvate instead of entering the TCA cycle gets converted to lactate. To meet the high energy and carbon demands these cancer cells utilize more and more glucose compared to normal cells2. Background Apart from its role in glucose metabolism, PKM2 also regulates other cellular processes by getting translocated to the nucleus from the cytoplasm and affecting a variety of signaling pathways leading to oncogenesis9,10. Cancer cells achieve a metabolic edge over normal cells by regulating PKM alternative splicing and promoting the expression of PKM23,8. hnRNPs (heterogeneous ribonucleoproteins); hnRNP A1, hnRNP A2 and PTBP1 (hnRNP I) are the key regulators of PKM alternative splicing11,12. c-Myc binds to promoters of these three genes and regulates their expression12. In majority of cancers, high expression of c-Myc promotes the higher expression of these three hnRNPs. When present in abundance, these three proteins bind on the intronic region flanking exon 9 and inhibit the incorporation of exon 9 and as an effect, there is the inclusion of exon 10 of PKM gene13. Molecular targeting of regulation of PKM alternative splicing by these three proteins might prove to be an effective strategy in eradicating the cancer cells14. The nuclear matrix is the site for many important nuclear events such as DNA replication, DNA repair, transcription and post-transcriptional modifications15,16. Post-transcriptional modifications include RNA splicing, RNA capping and poly-A tail addition. The nuclear matrix is a scaffold around which chromatin Page 4/30 folds and many nuclear matrix binding proteins are involved in chromatin organization15. Apart from chromatin compaction, nuclear matrix binding proteins which are also known as the matrix-associated region (MAR) binding proteins (MARBPs), actively participate in the regulation of DNA replication, DNA repair, transcription and post-transcriptional modifications15,16. One such MARBP is SMAR1 (Scaffold/Matrix attachment region binding protein 1)17. The human homolog of SMAR1 is also known as BANP (BTG3 Associated Nuclear Protein). SMAR1 has been reported to play an important role as a tumor suppressor protein and in the majority of higher grades of cancer, SMAR1 has been reported to be dysregulated18–24. SMAR1 is located on human chromosome 16q24 and loss of heterozygosity (LOH) has been reported for this chromosomal region in various cancers25–27. SMAR1 has been reported to colocalize with SC35 which is a marker of nuclear splicing speckles and an important regulator of alternative splicing28. Moreover, SMAR1 has an RS domain that facilitates the RNA binding and it interacts with snRNAs which are the core components of splicing machinery28. SMAR1 has been reported to regulate splicing of CD44 variants by deacetylation of Sam68 with help of HDAC6 in breast cancer cell lines. Cell culture MCF7, MDA-MB-231, MDA-MB-468 and T47D cells were obtained from the NCCS cell repository, Pune, India. MCF7, MDA-MB-231 and MDA-MB-468 were cultured in DMEM (Gibco) and T47D was cultured in RPMI (Gibco). All the cell lines were supplemented with 10% Fetal Bovine Serum (FBS) (Gibco) and 100 units/ml Penicillin and Streptomycin (Gibco) and incubated in a humidified 5% CO2 incubator at 37ºC. Background In addition to CD44 variants, SMAR1 also regulates alternative splicing of FAS ligand28. ChIP-sequencing study of SMAR1 in HCT116 cells suggests that there are several global gene targets of SMAR1 which are involved in the regulation of RNA processing and alternative splicing29. This suggests that SMAR1 might be involved in alternative splicing regulation of other cancer-associated genes. This study focuses on the role of SMAR1 in the regulation of PKM alternative splicing via HDAC6 dependent deacetylation of PTBP1 and its implication in inhibition of the Warburg effect and tumorigenesis. Our study thus demonstrates the inhibition of cancer cell metabolism and breast cancer progression by SMAR1 via suppression of oncogenic isoform PKM2. Cloning of dual reporter PKM minigene system To develop a dual chromatic PKM minigene system, eGFP was cloned into mCherry-N1 vector between SalI (NEB) and AgeI (NEB) restriction sites in such a way that both eGFP and mCherry remain in two different frames due to difference in one base pair and eGFP contains stop-codon at the end when in the frame. Further, Exon 8 - Exon 11 of the PKM gene along with introns was amplified from PKM minigene construct31 (a kind gift from Dr. Adrian R. Krainer) using Platinum™ SuperFi™ DNA Polymerase (Invitrogen) and cloned into this vector between XhoI (NEB) and NdeI (NEB) sites by Infusion cloning (NEB). One base pair insertion was further introduced in the cloned plasmid in Exon 10 by Infusion cloning. Due to one base pair insertion, the incorporation of Exon 9 resulted in the expression of mCherry and the incorporation of Exon 10 led to the expression of eGFP. The expressions of mCherry and eGFP were analyzed by confocal microscopy. Plasmids, siRNAs and shRNA constructs 3xFlag26-SMAR1 was used to over-express SMAR1 and 3xFlag26-Vector was used as a control. Silencer™ Select Negative Control No. 1 siRNA (4390843) and si-RNA against human-SMAR1 (BANP) (s29889) were obtained from Thermo Fischer Scientific Silencer® select siRNA range. shSMAR1-eGFP (ULTRA-3344235) was obtained from TransOMIC and used for shRNA-mediated SMAR1 knockdown. shNon translated-1- eGFP (shNT1) (TLNSU1420) was also obtained from TransOMIC and used as a control for all shRNA- Page 5/30 Page 5/30 Page 5/30 mediated knockdown studies. shHDAC6 clone TRCN0000004839 was used for HDAC6 knockdown28. shPKM2 was cloned in pLKO.1-TRC vector and the sequence targeting only PKM2 were obtained from Cortés-Cros et al.30. The sequence of shPKM2 used: mediated knockdown studies. shHDAC6 clone TRCN0000004839 was used for HDAC6 knockdown28. shPKM2 was cloned in pLKO.1-TRC vector and the sequence targeting only PKM2 were obtained from Cortés-Cros et al.30. The sequence of shPKM2 used: 5′CCGGCTACCACTTGCAATTATTTGACTCGAGTCAAATAATTGCAAGTGGTAGTTTTTG3′. 5′CCGGCTACCACTTGCAATTATTTGACTCGAGTCAAATAATTGCAAGTGGTAGTTTTTG3′. Transfections and treatments Transfections of various plasmids were done by the use of Polyethylenimine (PEI MAX 40000) (Polysciences, Inc.). In MCF7, transfection was performed at 70-80% confluency and the DNA:PEI ratio used for transfection was 1:3. In MDA-MB-231, reverse transfection was performed in which DNA: PEI mix was added before cell seeding. Tubacin (Sigma), a selective HDAC6 inhibitor, was used at 5 μM concentration for 5 hrs for optimum HDAC6 inhibition and an equal amount of DMSO was used as vehicle control. Immunohistochemistry (IHC) Paraffinized human breast cancer patient samples along with surrounding normal breast tissue were obtained from Ruby Hall Clinic, Pune, India. SMAR1, PKM1 and PKM2 expression were detected using standard immunohistochemical staining procedure. Briefly, after deparaffinization, endogenous peroxidase blockage and rehydration with decreasing concentrations of ethanol (100%, 95% and 70%), antigen retrieval was performed by heating the slides in 10 mM sodium citrate buffer (pH 6-7). Further samples were incubated with the antibody of SMAR1 (Bethyl) at 1: 100 dilution, PKM1 (CST) at 1: 100 dilution and PKM2 (CST) 1: 300 dilution overnight at 4ºC. After incubation with HRP-conjugated secondary antibodies for 1 hr at room temperature, sections were then treated in DAB (3,3'- Diaminobenzidine) for 15 mins to allow the development of brown precipitate corresponding to the sites of HRP-bound antibodies. Samples were washed and counterstained with Hematoxylin for nuclei staining. Tissue sections were observed at 20X magnification using a Nikon microscope (Eclipse E600) and images were acquired using Nikon’s NIS-elements imaging software. Dual reporter PKM minigene assay MCF7 cells were seeded onto a glass coverslip and after 24 hrs PKM minigene along with control or Flag- SMAR1 in a 1:1 ratio was introduced in these cells by PEI mediated transfection. After 48 hrs of transfection, cells were fixed with 4% paraformaldehyde for 10 mins at room temperature. Subsequently, cells were washed with 1X PBS thrice and the coverslips were mounted in fluoroshield media (Sigma) with DAPI. Cells were observed at 60X magnification using Nikon A1plus confocal microscope and images were acquired using Nikon’s NIS-elements imaging software. For eGFP and mCherry fluorescent intensity quantification, ImageJ software was used. Five independent regions of interest (ROIs) were selected per image covering cells expressing eGFP and mCherry and fluorescence intensity density was measured. Average relative fluorescence densities were calculated for all five ROIs per field and the fold Page 6/30 Page 6/30 change in eGFP/mCherry ratio was calculated. For statistical significance, three random fields per sample were selected and three independent experiments were performed. Mean ± SD fold change in eGFP/mCherry ratio was calculated for each sample and compared with the control sample. change in eGFP/mCherry ratio was calculated. For statistical significance, three random fields per sample were selected and three independent experiments were performed. Mean ± SD fold change in eGFP/mCherry ratio was calculated for each sample and compared with the control sample. Quantitative RT PCR Total RNA was extracted from cultured MCF7 cells and MDA-MB-231 cells by TRizolTM (Invitrogen) according to the manufacturer’s instruction. RNA was reverse transcribed by MMLV-RT (Invitrogen) as per the manufacturer’s instructions. Amplification reactions were prepared in triplicate using iQTaq SYBR green (Biorad) and amplification was performed on Eppendorf realplex 2.0 according to the manufacturer’s instruction. The average cycle thresholds from three independent biological replicate samples were calculated as described in Singh et al. with slight modifications32. Briefly, the average cycle thresholds from three independent biological replicate samples were normalized to housekeeping control gene 18S rRNA. Normalization was performed using 18S rRNA as a normalization control using the formula: [2^ (Ct control – Ct target)]. Along with control gene normalization, constitutive exon (exon 11) normalization was performed for PKM1 (Exon 8-9/9) and PKM2 (Exon 10-11/11) expression analysis. The student’s t-test was used to compare expression between two different groups. A list of primers used is given in Table S1. Antibody cross-linking, Co-Immunoprecipitation and sequential Co-Immunoprecipitation The majority of Immunoprecipitation (IP) experiments were done via covalently cross-linking the antibody with Protein G Dynabeads (Pierce) to avoid non-specific binding and contamination of immunoglobulin in the immunoprecipitated protein eluates. Approximately 1 µg of antibody was cross-linked with 10 µl of beads. Beads were washed thrice with ice-cold 1X PBS and further incubated with the desired antibody in IP buffer (1X PBS with 0.1% NP-40) containing protease inhibitor cocktail (Pierce). The antibody-bead mixture was incubated overnight at 4ºC. The unbound antibody was removed by three wash of the antibody-bead complex with IP buffer. Further, the antibody-bead complex was incubated with 500 μL of 10 mg/ml of Dimethyl pimelimidate (DMP) (Sigma) for 60 mins at room temperature with rotation. 50 μL of 1M Tris-Cl pH 8 was added to quench the reaction and incubated for 30 mins at room temperature with rotation. Unbound antibody was removed by washing it with 0.2 M Glycine pH 3 followed by three washes with IP buffer. The antibody-bead complex was further equilibrated with the IP buffer. These beads were either used immediately or stored at 4ºC for 2-3 days. Protein-bound to the antibody-bead complex was eluted at 95ºC by using SDS loading dye. For IP experiments, 500 µg of nuclear extracts were pre-cleared with control normal IgG (Sigma) bound dynabeads G and subsequently incubated for 12 hrs at 4ºC with dynabeads G crosslinked with the desired primary antibody. The protein associated bead complexes were washed thrice with IP buffer and then further eluted with SDS loading dye. The eluates were probed with indicated antibodies. For sequential IP experiments, nuclear extracts (1 mg) were immunoprecipitated first with 3 µg of anti- SMAR1 antibody. Before proceeding for the second IP, a minor fraction of the eluates was examined for the presence of HDAC6. Subsequently, the eluate was immunoprecipitated with 2 µg of anti-HDAC6. The final eluates were probed with anti-PTBP1 to evaluate the association. Western blotting Cells were incubated in TNN buffer [50 mM Tris-Cl pH 7.5, 5 mM EDTA, 0.5% NP40, 50 mM NaF, 1 mM DTT, 0.2 mM sodium orthovanadate, 0.5 mM PMSF, 150 mM NaCl and 1X Protease inhibitor cocktail (Thermo Scientific)] for cell lysis and lysates containing equal concentration of proteins were resolved Page 7/30 Page 7/30 using SDS-PAGE and transferred onto PVDF membrane. The membranes were incubated with primary antibodies such as anti-SMAR1 (Bethyl - A300-279A), Anti-PKM1 (CST- 7067), Anti-PKM2 (CST- 4053), Anti-hnRNP A1(CST- 8443), Anti-hnRNP A2 (Abcam- ab6102), Anti-PTBP1 (Thermo Scientific- 32-4800), Anti-HDAC6 (CST- 7558), Anti-HDAC1 (CST- 5356), Anti-GFP (Proteintech- 66002-1-Ig), Anti-mCherry (Proteintech- 26765-1-AP), Anti-β-actin (Sigma- A2228) and Anti-Flag tag (CST- 14793). This was followed by three washes and incubation with appropriate HRP conjugated secondary antibodies. Visualization was achieved with ECL substrate (Pierce) and exposure to X-ray films or imaging in Syngene G:BOX Chemi XRQ. UV-crosslinking and RNA immunoprecipitation (CLIP) CLIP experiment was done via covalently cross-linking the anti-PTBP1 antibody with Protein G Dynabeads (Thermo Scientific) as described in antibody cross-linking. Normal mouse IgG crosslinked beads were used as a negative control. Antibody cross-linked beads were further blocked for 1 hr at 4ºC with 100 nM yeast tRNA to avoid non-specific binding with RNA. Cells were washed with ice-cold 1X PBS, UV-irradiated (150 mJ/cm2) and harvested and lysed with lysis buffer (100 mM KCl, 5 mM MgCl2, 10 mM HEPES - pH 7.0, 0.5% NP40, 1 mM DTT, 100 units/ml RNase Out and Protease inhibitor cocktail) for 10 mins on ice. RNase A (1:1000 dilution) and DNase I (Invitrogen) was added to the lysate and incubated at 37ºC for 3 mins. 5% lysate was taken as input and TRizol LSTM (Invitrogen) was added for RNA extraction. An equal amount of protein (10 mg) was taken for control and SMAR1 knockdown sample and anti-PTBP1 and normal IgG conjugated beads were added for immunoprecipitation and incubated on rotation at 4ºC for 3 hrs. After two washes with wash buffer (50 mM Tris-HCl - pH 7.4, 150 mM NaCl, 1 mM MgCl2, 0.05% NP40) supplemented with RNase inhibitor, an aliquot (10%) of beads was kept as control of immunoprecipitation while the rest was treated with 30 μg of Proteinase K and incubated for 1 hr at 55ºC. RNA was then extracted by TRizol LSTM (Invitrogen) and RNA was reverse transcribed by MMLV-RT (Invitrogen) as per the manufacturer’s instructions. Immunoprecipitated fractions and 5% input were analyzed by quantitative real-time PCR in duplicate using iQTaq SYBR green (Biorad) and amplification was performed on Eppendorf realplex 2.0 and specific primers for PTBP1 binding site on Intron 8 of PKM were used (sequences mentioned in Table S1). Primer sequence for PKM Intron 8 PTBP1 binding site was obtained from Chen et al.13. The experiment was performed three times and normalization was performed to input using the formula: [2^ (Ct input – Ct immunoprecipitation)]. Fold enrichment was calculated relative to normal mouse IgG control. Resultant fold enrichment was further normalized with densitometry measurement of a western blot for immunoprecipitated PTBP1 protein samples. The student’s t-test was used to identify the significance between two different groups. Anti-Acetyl-Lysine Acetylation Assay For assessing the acetylation status of protein, 1 mg of nuclear extract was incubated with 2 μg of anti- acetyl-lysine antibody (CST- 9441) crosslinked with dynabeads G at 4ºC with slight mixing for 12 hrs. Page 8/30 Page 8/30 Immunocomplexes were washed thrice with 500 μL of IP buffer and further eluted at 95ºC with SDS loading dye. The eluates were loaded onto the SDS-PAGE and immunoblotted with indicated antibodies. Glucose (2-NBDG) uptake assay MCF7 cells were transfected with SMAR1 siRNA and control siRNA by Lipofectamine RNAiMaxTM (Ambion) according to the manufacturer’s protocol. After 24 hrs of transfection, media was removed and replenished with 10% FBS containing DMEM (Without glucose and Sodium Pyruvate) and incubated at 37ºC for 1 hr. 10 μM fluorescent d-glucose analog 2-[N-(7-nitrobenz-2-oxa-1,3-diazol-4-yl)amino]-2-deoxy- d-glucose (2-NBDG) (Invitrogen) was added to culture media and cells were incubated for 1 hr at 37ºC. The 2-NBDG uptake reaction was stopped by removing the incubation medium and the cells were washed with ice-cold 1X PBS. 1 μg/ml Propidium Iodide (PI) was added to distinguish the viable cell population. For each measurement, data from 10,000 single-cell events were collected using FACS Canto II (BD Bioscience). The percentage of 2-NBDG uptake was calculated from mean fluorescence intensity (MFI) compared to the control. Cell viability assay MCF7 cells were transfected with BORIS shRNA and shNT1 was used as a control in six-well culture plates. After the selection of transfected cells with puromycin, cells (4x103) were seeded in 96 well culture plates and were cultured for 24 hrs, 48 hrs, and 72 hrs. Cell growth was determined by measuring the conversion of MTT Tetrazolium salt (Sigma) to MTT formazan. In brief, 50 µl of MTT stock solution (5 mg/ml) was added to each well along with 50 µl of 10% FBS containing DMEM and incubated for 4 hrs. After the incubation time, formazan crystals formed in the cells were solubilized in Isopropanol. The cell viability was measured by Spectramax M5 (Molecular Devices) at an optical density of 570 nm. The percentage proliferation was calculated and cell viability at 0 hr for each sample was considered as 100%. Compared to 0 hr, the percentage of proliferation was calculated for other time points. Lactate assay MCF7 cells were transfected with respective shRNA and control shRNA. shRNA transfected cells were selected with puromycin and an equal number of cells were plated in a 6 well plate. After 24 hrs, the cells were replenished with 10% FBS containing high glucose DMEM (without sodium pyruvate). After 24 hrs of media replenishment, media was collected and the amount of lactate was calculated with the use of a Lactate Assay Kit (Abcam, ab65331) as per the manufacturer’s protocol. The amount of lactate present in each sample was normalized with the total amount of protein. The percentage of lactate production was calculated for each sample compared to the control. Glucose Assay MCF7 cells were transfected with respective shRNA and control shRNA. shRNA transfected cells were selected with puromycin and an equal number of cells were plated in a 6 well plate. After 24 hrs, the cells were replenished with 10% FBS containing high glucose DMEM (without sodium pyruvate). After 24 hrs of media replenishment, media was collected and the amount of glucose was calculated with the use of a Glucose Assay Kit (Abcam, ab65333) as per the manufacturer’s protocol by colorimetric method. The amount of glucose present was normalized with the total amount of protein. The glucose utilization was calculated by subtracting the glucose level of samples from that of cell-free media. The percentage of glucose utilization was calculated compared to the control. Page 9/30 Page 9/30 Transwell migration and invasion assay Transwell chamber: 24-well, 8.0-μm pore membranes (Corning USA) were used according to the manufacturer’s protocol. MCF7 cells were transfected with respective shRNAs and shNT1 was used as a control. After the selection of transfected cells with puromycin, 1x105 cells per well were seeded in the upper chamber in a serum-free medium, and DMEM with 5% FBS was added to the lower chamber as a chemoattractant at the same time. After incubation of 24 hrs at 37ºC, the cells remaining at the upper surface of the membrane were removed with cotton swabs, and the cells on the lower surface of the membrane are the migrated cells. After fixation with 4% paraformaldehyde and staining with 0.5% crystal violet solution, the cells were observed at 10X magnification using the Nikon microscope (Eclipse Ti2) and images were acquired using Nikon’s NIS-elements imaging software. The transwell invasion assay was carried out as described above, except that a transwell chamber with Matrigel (Corning, USA) was used and they were pre-incubated at 37ºC after hydration with 100 μL of serum-free medium for 2 hrs before the cells were seeded onto the membrane, followed by incubation of 48 hrs at 37ºC. Five fields were randomly captured and the number of migrating/invading cells were quantified manually with the help of Image J software. The percentage of migration/invasion was calculated for each sample compared to the control. Colony formation assay MCF7 cells were transfected with respective shRNAs and shNT1 was used as a control. The cells were selected with puromycin and 1x103 cells were seeded in the new 6-well plate. After 10 days, cells were Page 10/30 Page 10/30 fixed using methanol and acetic acid (3:1) for 5 mins. After fixation, cells were washed with 1X PBS thrice. After washing cells were stained with 0.05% crystal violet stain, images were taken and colonies formed were counted manually with help of ImageJ software. The percentage of colony formation was calculated for each sample compared to the control. Wound healing assay MCF7 cells were transfected with respective shRNAs and shNT1 was used as a control. The cells were selected with puromycin and 1.5x105 cells were seeded in each well of 12 well plate and allowed to grow at 37ºC to form a monolayer. Cells were synchronized by serum starvation for 12 hrs and scratch was introduced in the middle of the monolayer by sterile pipette tip, generating a cell-free area of approximately 1 mm in width and cell debris was removed by washing twice with 1X PBS. Three fields per well were imaged in the area where the wound was introduced at 0 hr 24 hrs. The area of the wound was measured by ImageJ for 0 hr and 24 hrs and the percentage of wound migration was calculated compared to that of the control. All images were taken at 10X magnification using the Nikon microscope (Eclipse Ti2) with help of Nikon’s NIS-elements imaging software. In vivo tumor generation Page 11/30 Page 11/30 All mice used in this experiment were bred at the animal resource facility of NCCS, Pune, India. Standard protocols approved and monitored by the Institutional Animal Ethical Committee were followed for this experiment. The MDA-MB-231 cells were transduced with SMAR1-adenovirus and control-adenovirus. 1 million cells were injected subcutaneously in 6-8 weeks old NOD/SCID mice. After 1 month of injection, mice were sacrificed and tumors were dissected. The volume and weight of tumors were measured and parts of the tumor were utilized for western blot and immunohistochemistry as described in the above section. Statistical analysis All statistical analysis was performed using Microsoft Excel and graphs were plotted using GraphPad Prism7. Data has been represented as mean ± SD. The student’s t-test was used to determine the statistical significance of the difference between the groups. The p-value of < 0.05 was considered significant. * p <0.05, p <0.01 and * p <0.001. Images were analyzed and quantified using ImageJ software. SMAR1 regulates PKM alternative splicing by promoting the incorporation of exon 9 and suppressing the incorporation of exon 10 SMAR1 is known to colocalize with one of the key splicing regulators SC35 in nuclear splicing speckles. It also interacts with snRNAs which are the core components of splicing machinery. Moreover, SMAR1 has been reported to regulate alternative splicing of CD44 variants and FAS ligand28. On the basis of an inverse correlation between SMAR1 and PKM2 expression as well as a positive correlation between SMAR1 and PKM1 observed in breast cancer, we hypothesized that SMAR1 might play a crucial role in regulating alternative splicing of the PKM gene. Based on the expression profile of SMAR1 in various breast cancer cell lines, MCF7 and MDA-MB-231 were used for further experiments. To investigate the role of SMAR1 in the regulation of PKM alternative splicing, PKM isoform expression was measured upon shRNA-mediated depletion of SMAR1 expression in MCF7 cells at the RNA level by performing quantitative RT-PCR and at the protein level by western blot. Knockdown of SMAR1 in MCF7 resulted in PKM1 downregulation and increased PKM2 expression at the transcript level (Fig. 2A). Relative fold change in SMAR1 expression upon shRNA-mediated SMAR1 knockdown in MCF7 has been represented in Figure S1A. Further analysis of PKM isoform expression in SMAR1 depleted cells at the protein level suggests diminished expression of PKM1 and upregulation of PKM2 which was further in coherence with RNA expression analysis (Fig. 2B). Moreover, PKM isoform expression was quantified upon ectopic expression of SMAR1 in LOH-containing MDA-MB-231 at the RNA level by qRT-PCR and at the protein level by western blot. Upon Flag-SMAR1 overexpression in MDA-MB-231, PKM1 was observed to be increased and PKM2 was downregulated at the transcript level (Fig. 2C). Flag-SMAR1 mediated overexpression in MDA-MB-231 at transcript level has been represented in Figure S1B. Moreover, PKM isoform expression at the protein level in SMAR1 overexpressed cells was in concordance with that of transcript level analysis (Fig. 2D). These results suggest that SMAR1 being a tumor suppressor protein, promotes the expression of PKM1 isoform over oncogenic isoform PKM2. To further validate the role of SMAR1 in PKM alternative splicing regulation, a dual reporter PKM minigene system has been generated. Schematic representation of dual reporter PKM minigene system has been described in Fig. 2E. In this dual reporter PKM minigene assay, incorporation of exon 10 leads to eGFP expression whereas incorporation of exon 9 leads to mCherry expression. SMAR1 exhibit a negative correlation with PKM2 and a positive correlation with PKM1 expression in breast cancer cells Dysregulation of various tumor suppressor proteins is one of the key features of various cancers33. SMAR1 being an important tumor suppressor protein is known to get downregulated in higher grades of breast and colon cancer21,23,24. LOH of SMAR1 locus (human chromosome 16q24) has been reported in various cancer25,26. Moreover, Cdc20 mediated proteasomal degradation of SMAR1 has been studied in breast cancer cells23. In contrast, upregulation of key oncogenes such as PKM2 gives metabolic as well as tumorigenic advantage to the cancer cells. The majority of normal differentiated cells have a lower expression of PKM2 and in cancer cells, there is an enhanced expression of PKM234. Expression of SMAR1, PKM2 and PKM1 was measured in breast cancer patient samples and compared with surrounding non-cancerous tissue by immunohistochemistry. IHC staining of SMAR1 revealed that its expression was diminished in tumor samples compared to that of normal control tissue (Fig. 1A). These results correspond with the earlier studies, confirming the downregulation of SMAR1 in breast cancer samples21,23. Moreover, IHC staining of PKM isoforms revealed that PKM2 expression was elevated and PKM1 expression was significantly low in tumor samples compared to that of normal control tissue. These results further comply with earlier reports suggesting a switch in PKM isoform expression between normal tissue and cancer tissue leading to higher expression of oncogenic isoform PKM2 and suppression of PKM13,32. Human breast cancer cells such as MDA-MB-231, MDA-MB-468 and T47D harbor LOH for SMAR1 locus, whereas MCF7 cells lack LOH for this locus25,27. LOH harboring cells have Page 12/30 significant downregulation of SMAR1 expression compared to that of MCF728. Expression profile of SMAR1, PKM1 and PKM2 in these breast cancer cell lines MCF7, MDA-MB-231, MDA-MB-468 and T47D were found to be correlating with that of the patient samples (Fig. 1B). These results suggest that SMAR1 has an inverse correlation with PKM2 expression whereas it has a direct correlation with PKM1 expression. This indicates that SMAR1 might be promoting the expression of PKM1 and suppressing the expression of oncogenic isoform PKM2 in the non-cancerous tissue. SMAR1 regulates PKM alternative splicing by promoting the incorporation of exon 9 and suppressing the incorporation of exon 10 Fluorescence imaging was done upon ectopic expression of SMAR1 along with the PKM minigene system by confocal Page 13/30 Page 13/30 microscopy to measure eGFP and mCherry expression. Further relative fluorescence intensity was calculated and the fold change of eGFP/mCherry ratio has been calculated. Expression of SMAR1, eGFP and mCherry was confirmed by western blot (Figure S1C). The fold difference in eGFP/mCherry ratio in SMAR1 overexpression condition was reduced by 2-fold as compared to that of control which further validates its role in PKM alternative splicing regulation (Fig. 2F). These results suggest that SMAR1 actively promotes the incorporation of exon 9 and the exclusion of exon 10 leading to higher expression of PKM1 and repression of PKM2 expression in breast cancer cells. SMAR1 interacts with PTBP1 and SMAR1-HDAC6 makes a triple complex with PTBP1 hnRNP A1, hnRNP A2 and PTBP1 are the three key regulators of PKM alternative splicing11,12. When these three proteins are present in higher concentration, they bind to intronic regions flanking exon 9 of PKM which leads to the inhibition of incorporation of exon 9 while promoting the incorporation of exon 10, thereby resulting in higher expression of PKM2 and diminished expression of PKM113. To delineate the detailed molecular mechanism of SMAR1 mediated regulation of PKM alternative splicing, expression of hnRNP A1, hnRNP A2 and PTBP1 was measured upon shRNA-mediated knockdown of SMAR1 in MCF7. It was observed that there was no change in the expression of these three splicing regulators upon SMAR1 knockdown (Fig. 3A). This observation eliminates the possibility of transcriptional repression of hnRNP A1, hnRNP A2 and PTBP1 by SMAR1. To determine the role of SMAR1 in the regulation of post- translational modification, the interaction of SMAR1 with these hnRNPs was checked by co- immunoprecipitation (co-IP) and observed that SMAR1 interacts with PTBP1 (Fig. 3B). Interaction of SMAR1 with PTBP1 was further confirmed by reverse co-IP of PTBP1 with SMAR1 (Fig. 3C). SMAR1 is known to regulate various cellular processes with the assistance of HDACs such as HDAC1 and HDAC620,22,28,35. To identify which HDAC is involved in SMAR1 mediated regulation of PKM alternative splicing, the interaction of PTBP1 with HDAC1 and HDAC6 was checked by co-IP which revealed that HDAC6 interacts with PTBP1 (Fig. 3C). Interaction of HDAC6 with PTBP1 and SMAR1 was further confirmed by reverse co-IP of HDAC6 (Fig. 3D). To further investigate the molecular interplay of SMAR1- HDAC6 interaction with PTBP1, sequential IP of SMAR1 and HDAC6 with PTBP1 was performed. Sequential IP experiment revealed that SMAR1-HDAC6 forms a triple complex (Fig. 3E). These results demonstrate that SMAR1 directly interacts with PTBP1 along with HDAC6 and the coexistence of SMAR1- HDAC6-PTBP1 as a ternary complex. This trimeric complex formation further indicates molecular dynamics between these proteins in SMAR1 mediated regulation of PKM alternative splicing. SMAR1 mediated regulation of PKM alternative splicing is HDAC6 dependent Page 14/30 To delineate the role of HDAC6 in SMAR1 mediated regulation of PKM alternative splicing, expression of PKM isoforms was checked upon shRNA-mediated knockdown of HDAC6 by Western blot. Upon depletion of HDAC6, there was a downregulation in PKM1 expression and an increase in PKM2 expression (Fig. 4A). This indicates the involvement of HDAC6 in PKM alternative splicing regulation. To further determine the role of HDAC6 in SMAR1 mediated regulation of PKM alternative splicing, SMAR1 overexpressed cells were treated with a specific HDAC6 inhibitor (Tubacin) and PKM isoform expression was checked by western blot. Expression of PKM1 was high and PKM2 was downregulated in the SMAR1 overexpression condition compared to that of control. However, upon Tubacin treatment in SMAR1 overexpressed cells, the expression of PKM1 was reduced and the expression of PKM2 was increased compared to that of only SMAR1 overexpressed cells (Fig. 4B). This confirms that SMAR1 mediated regulation of PKM alternative splicing is HDAC6 dependent process. Further validation of the role of HDAC6 in SMAR1 mediated regulation of PKM alternative splicing was done by dual reporter PKM minigene assay. Fold change in eGFP/mCherry ratio was measured upon SMAR1 overexpression along with Tubacin treatment in cells transfected with PKM minigene system compared to the control condition. Fold change in eGFP/mCherry ratio was decreased upon ectopic expression of SMAR1 compared to the control. However, upon tubacin treatment in SMAR1 overexpressed cells resulted in a further increase in the fold change of eGFP/mCherry ratio (Fig. 4C). These observations confirm the role of HDAC6 in SMAR1 mediated regulation of PKM alternative splicing. This demonstrates that HDAC6 is actively involved in PKM alternative splicing regulation orchestrated by SMAR1. SMAR1-HDAC6 deacetylates PTBP1 and modulates its affinity to PKM pre-mRNA Post-translational modifications such as phosphorylation, acetylation, ubiquitination, sumoylation of splicing factors are known to regulate various alternative splicing events36–39. SMAR1 is known to regulate the alternative splicing of CD44 variants by HDAC6 assisted deacetylation of Sam6828. Moreover, sirtuin-mediated deacetylation of hnRNP A1 has been reported to regulate alternative splicing of the PKM gene40. To further delineate the role of SMAR1-HDAC6 in the deacetylation of PTBP1, the acetylation status of PTBP1 was checked upon shRNA-mediated depletion of SMAR1 by immunoprecipitation with an anti-acetyl-lysine antibody. In the SMAR1 knockdown condition, there was an increase in the acetylation status of PTBP1 compared to control (Fig. 5A). This observation indicates that the SMAR1 level dictates the acetylation status of PTBP1 and it maintains PTBP1 in a deacetylated state. The acetylation status of PTBP1 was further estimated after SMAR1 overexpression along with Tubacin treatment. Ectopic expression of SMAR1 resulted in a decrease in acetylation status of PTBP1 but upon Tubacin treatment in SMAR1 overexpressed cells, the acetylation status of PTBP1 was restored (Fig. 5B). This suggests that SMAR1 maintains PTBP1 in the deacetylated state in HDAC6 dependent manner. Post-translational modifications such as phosphorylation, acetylation, ubiquitination, sumoylation of splicing factors are known to regulate various alternative splicing events36–39. SMAR1 is known to regulate the alternative splicing of CD44 variants by HDAC6 assisted deacetylation of Sam6828. Moreover, sirtuin-mediated deacetylation of hnRNP A1 has been reported to regulate alternative splicing of the PKM gene40. To further delineate the role of SMAR1-HDAC6 in the deacetylation of PTBP1, the acetylation status of PTBP1 was checked upon shRNA-mediated depletion of SMAR1 by immunoprecipitation with an anti-acetyl-lysine antibody. In the SMAR1 knockdown condition, there was an increase in the acetylation status of PTBP1 compared to control (Fig. 5A). This observation indicates that the SMAR1 level dictates the acetylation status of PTBP1 and it maintains PTBP1 in a deacetylated state. The acetylation status of PTBP1 was further estimated after SMAR1 overexpression along with Tubacin treatment. Ectopic expression of SMAR1 resulted in a decrease in acetylation status of PTBP1 but upon Tubacin treatment in SMAR1 overexpressed cells, the acetylation status of PTBP1 was restored (Fig. 5B). This suggests that SMAR1 maintains PTBP1 in the deacetylated state in HDAC6 dependent manner. SMAR1-HDAC6 deacetylates PTBP1 and modulates its affinity to PKM pre-mRNA Page 15/30 To inspect the effect of SMAR1-HDAC6 mediated deacetylation of PTBP1 on its affinity for PKM pre- mRNA, UV-crosslinking and RNA immunoprecipitation (CLIP) of PTBP1 was performed upon shRNA- mediated depletion of SMAR1 in MCF7. Enrichment of PTBP1 on Intron 8 of PKM pre-mRNA was enhanced by 2-fold in the case of SMAR1 knockdown condition as compared to that of control (Fig. 5C). Immunoprecipitation of PTBP1 in the CLIP experiment was confirmed by western blot (Fig. 5D). Increased enrichment of PTBP1 on PKM pre-mRNA in SMAR1 knockdown condition suggests that SMAR1-HDAC6 mediated deacetylation of PTBP1 leads to a reduction in its affinity for PKM pre-mRNA and ultimately modulates PKM alternative splicing. SMAR1 regulates the Warburg effect and breast cancer growth via regulation of PKM2 expression Higher expression of PKM2 compared to PKM1 is one of the key factors for cancer cells in achieving the metabolic advantage of the Warburg effect compared to normal cells3. Reduction in PKM2 expression might lead to inhibition of the Warburg effect and ultimately suppress the tumorigenic potential of cancer cells. SMAR1 mediated regulation of PKM alternative splicing leads to suppression of PKM2 and increased expression of PKM1. Based on this observation, we hypothesized that SMAR1 might be playing important role in the regulation of cancer cell metabolism. To determine the role of SMAR1 in the regulation of tumor metabolism, glucose utilization was estimated upon shRNA-mediated decrease of SMAR1 expression along with PKM2 in MCF7. Depletion of PKM2 by shRNA was confirmed by western blot (Figure S2A). A reduction in SMAR1 expression resulted in increased glucose utilization. Further depletion of PKM2 in SMAR1 knockdown cells resulted in decreased glucose utilization (Fig. 6A). Higher expression of PKM2 compared to PKM1 is one of the key factors for cancer cells in achieving the metabolic advantage of the Warburg effect compared to normal cells3. Reduction in PKM2 expression might lead to inhibition of the Warburg effect and ultimately suppress the tumorigenic potential of cancer cells. SMAR1 mediated regulation of PKM alternative splicing leads to suppression of PKM2 and increased expression of PKM1. Based on this observation, we hypothesized that SMAR1 might be playing important role in the regulation of cancer cell metabolism. To determine the role of SMAR1 in the regulation of tumor metabolism, glucose utilization was estimated upon shRNA-mediated decrease of SMAR1 expression along with PKM2 in MCF7. Depletion of PKM2 by shRNA was confirmed by western blot (Figure S2A). A reduction in SMAR1 expression resulted in increased glucose utilization. Further depletion of PKM2 in SMAR1 knockdown cells resulted in decreased glucose utilization (Fig. 6A). Expression of SMAR1 and PKM isoforms upon shRNA-mediated knockdown of SMAR1 along with PKM2 depletion was confirmed by western blot (Figure S2B). 2-NBDG mediated glucose uptake upon siRNA- mediated knockdown of SMAR1 validates its role in the regulation of glucose metabolism (Figure S2C). siRNA-mediated knockdown of SMAR1 in MCF7 was confirmed by western blot (Figure S2D). To further decipher the role of SMAR1 in the reversal of the Warburg effect, lactate production was measured in SMAR1 knockdown condition along with PKM2 depletion. SMAR1 depletion condition revealed enhanced lactate production compared to control. SMAR1 regulates the Warburg effect and breast cancer growth via regulation of PKM2 expression Moreover, PKM2 knockdown in SMAR1 depleted cells were having decreased lactate production compared to SMAR1 depleted cells (Fig. 6B). This suggests that SMAR1 mediated regulation of PKM alternative splicing inhibits the Warburg effect via suppression of PKM2. These observations reveal the important regulatory role of SMAR1 in keeping cellular metabolism in check via regulation of PKM isoform expression. p g ( g ) Expression of SMAR1 and PKM isoforms upon shRNA-mediated knockdown of SMAR1 along with PKM2 depletion was confirmed by western blot (Figure S2B). 2-NBDG mediated glucose uptake upon siRNA- mediated knockdown of SMAR1 validates its role in the regulation of glucose metabolism (Figure S2C). siRNA-mediated knockdown of SMAR1 in MCF7 was confirmed by western blot (Figure S2D). To further decipher the role of SMAR1 in the reversal of the Warburg effect, lactate production was measured in SMAR1 knockdown condition along with PKM2 depletion. SMAR1 depletion condition revealed enhanced lactate production compared to control. Moreover, PKM2 knockdown in SMAR1 depleted cells were having decreased lactate production compared to SMAR1 depleted cells (Fig. 6B). This suggests that SMAR1 mediated regulation of PKM alternative splicing inhibits the Warburg effect via suppression of PKM2. These observations reveal the important regulatory role of SMAR1 in keeping cellular metabolism in check via regulation of PKM isoform expression. PKM2 mediated increase in glucose uptake and lactate production provides a metabolic advantage to cancer cells leading to increased tumor growth3. For tumor cells, reprogramming of glucose metabolism and higher glucose utilization are essential for their proliferation and survival2,41. The proliferative potential of cancer cells upon shRNA-mediated depletion of SMAR1 along with PKM2 in MCF7 was assessed by proliferation assay and colony formation assay. In the SMAR1 knockdown condition Page 16/30 Page 16/30 proliferation rate and colony formation potential were increased which were further reduced due to PKM2 depletion (Fig. 6C and D). These results suggest that SMAR1 inhibits breast cancer growth via reversal of the Warburg effect by suppressing PKM2 expression. Increased lactate production by cancer cells creates acidic surroundings which contribute to the tumorigenic properties of cancer cells such as migration, invasion and metastasis42. Further, the tumorigenic potential of cancer cells upon depletion of SMAR1 along with PKM2 knockdown was measured by in vitro functional assays such as transwell migration, invasion and wound healing assays. SMAR1 knockdown resulted in an increase in migration, invasion and wound healing properties of MCF7. SMAR1 regulates in vivo tumor formation via regulation of PKM2 i PKM2 contributes to in vivo tumor generation and its progression via regulation of cancer cell metabolism3. To identify the role of SMAR1 in tumor formation, tumor xenografts of SMAR1 overexpressed cells were generated by injection of SMAR1-adenovirus transduced MDA-MB-231 cells in NOD/SCID mice. The tumor burden was compared among SMAR1 overexpressing MDA-MB-231 tumors to that of normal MDA-MB-231 control tumors. In SMAR1 overexpressing group, tumor burden was less compared to that of the control group (Fig. 7A). Moreover, in the SMAR1 overexpression condition, there was a significant reduction in the weight and volume of the tumor compared to control (Fig. 7B and C). These observations indicate that SMAR1 suppresses in vivo tumor generation. A quantitative expression of PKM1 and PKM2 in these xenograft tumor samples was checked by western blot. In adeno-SMAR1 over-expressed tumor samples expression of tumorigenic isoform PKM2 was reduced and PKM1 expression was increased compared to control (Fig. 7D). IHC staining of PKM isoforms in these in vivo tumor samples were observed to be correlating with western blot results suggesting higher expression of PKM1 isoform and depleted expression of PKM2 isoform in SMAR1 overexpressing tumors compared to that of control tumors (Fig. 7E). These observations substantiate that SMAR1 suppresses in vivo tumor formation via the regulation of PKM isoform expression. SMAR1 regulates the Warburg effect and breast cancer growth via regulation of PKM2 expression Knockdown of PKM2 in SMAR1 depleted cells showed a decrease in migration, invasion and wound healing ability of cancer cells (Fig. 6E-G). These observations demonstrate that SMAR1 inhibits the tumorigenic potential of cancer cells due to the inhibition of the cancer cell metabolism through the downregulation of PKM2. Discussion Here we describe the detailed mechanism of PKM alternative splicing regulation by one of the key tumor suppressor proteins, SMAR1 and its implications in inhibition of cancer cell metabolism and tumorigenesis. SMAR1 has been reported to co-localize with key splicing factor SC35 and it interacts with splicing regulator Sam6828. SMAR1 also contains the RS domain which facilitates its interactions with various snRNAs which are the core components of splicing machinery28. Furthermore, SMAR1 regulates alternative splicing of CD44 variants and FAS ligand28. This indicates the active involvement of SMAR1 in Page 17/30 Page 17/30 various alternative splicing events. Our results demonstrate that SMAR1 follows a positive correlation with PKM1 and a negative correlation with oncogenic isoform PKM2 in breast cancer patient samples as well as in breast cancer cell lines. Further experimental approaches have revealed that SMAR1 is actively involved in the regulation of PKM alternative splicing. SMAR1 promotes the inclusion of exon 9 and exclusion of exon 10 thus inhibiting the expression of PKM2 which is a tumorigenic isoform of the PKM gene. These observations were validated by dual chromatic PKM minigene assay, which further confirms the regulatory role of SMAR1 in dictating the outcome of PKM alternative splicing. The downregulation of SMAR1 in breast cancer is one of the reasons for enhanced PKM2 expression that leads to altered glucose metabolism and contributes to cancer cell growth. Nuclear matrix binding protein, SMAR1 is known to perform various biological functions with help of regulatory proteins such as HDACs20,22,28,35. Additionally, various HDACs have been reported to participate in the regulation of alternative splicing through modulation of the acetylation status of histone and non-histone proteins including splicing factors43. Earlier studies suggest that SMAR1 inhibits the expression of various genes such as Cyclin D1, BAX and PUMA by recruiting the repressor complex of HDAC1-mSin3a on the promoter and keeping it in a repressed state by epigenetic modulations20,22. Moreover, SMAR1 also plays a crucial role in deacetylating target proteins such as Ku70 and Sam68 with the help of HDAC628,35. This points to the crucial role of SMAR1 in bringing about the post-translational regulation of its target proteins and ultimately dictating their role in various molecular and biological processes. SMAR1-HDAC6 mediated deacetylation of Ku70 dictates cell fate upon exposure to ionizing radiation via regulation of DNA repair and apoptosis35. Discussion This demonstrates the tumor suppressor function of key nuclear matrix binding protein SMAR1 and its importance in the maintenance of metabolic equilibrium. Future studies on the role of SMAR1 in global alternative splicing regulation might give more insight into the active involvement of tumor suppressor proteins in cancer-associated alternative splicing regulation and its implication in various cellular processes. Highly proliferating cancer cells gain a metabolic advantage over normal cells via higher expression of PKM2 and exhibits peculiar metabolic properties known as the Warburg effect2. Glucose is a primary source of energy and anabolic demands for highly proliferating cancer cells2. Moreover, a higher amount of lactate production creates an acidic micro-environment in tumors which favors invasion and metastasis42. Our study demonstrates that SMAR1 mediated suppression of PKM2 isoform and upregulation of PKM1 isoform results in inhibition of cancer cell metabolism. SMAR1 inhibits glucose utilization and lactate production via regulation of PKM alternative splicing. Our results demonstrate that SMAR1 mediated reversal of the Warburg effect leads to inhibition of cancer cell proliferation, migration and invasion. Moreover, SMAR1 suppresses in vivo tumor formation via the regulation of PKM isoform expression. This demonstrates the tumor suppressor function of key nuclear matrix binding protein SMAR1 and its importance in the maintenance of metabolic equilibrium. Future studies on the role of SMAR1 in global alternative splicing regulation might give more insight into the active involvement of tumor suppressor proteins in cancer-associated alternative splicing regulation and its implication in various cellular processes. In higher grades of cancer, downregulation of SMAR1 expression has been correlated with an increase in tumorigenic potential21,23,24. SMAR1 expression and functions are also known to be modulated by various cancer-associated signaling pathways including JNK signaling, Wnt signaling, ERK-MAPK pathway23,24,28. SMAR1 gets highly dysregulated in Wnt signaling associated colorectal cancer (CRC)24. Moreover, activation of ERK-MAPK signaling leads to translocation of SMAR1 to the cytoplasm, resulting in increased CD44 variants alternative splicing and metastasis in breast cancer cells28. A study in the breast cancer model revealed that SMAR1 undergoes proteasomal degradation by Cdc20 in a JNK kinase-dependent manner23. Furthermore, microbial peptides that promote SMAR1 stabilization have been correlated with inhibition of Wnt/β-catenin activities in CRC24. A study involving stabilization of SMAR1 expression by treatment with isothiocyanate derivative has been co-related with anti-HIV activity44. Discussion Our findings reveal that SMAR1 knockdown promotes the Warburg effect and tumorigenic potential of breast cancer cells by modulating PKM isoform expression indicating an important regulatory role of tumor suppressor protein in cellular homeostasis. This indicates that future studies targeting cancer cell metabolism and cancer progression via stabilization of SMAR1 expression might lead to better therapeutic efficacy in cancer treatment. Discussion SMAR1 further inhibits metastasis via regulation of CD44 variants alternative splicing via HDAC6 mediated deacetylation of Sam68 in breast cancer cells28. Many post-translational modifications such as phosphorylation, acetylation, ubiquitination, and sumoylation of splicing regulators have been associated with the regulation of various alternative splicing events28,36−40. Previous studies involving PKM alternative splicing suggest that three important splicing inhibitors hnRNP A1, hnRNP A2 and PTBP1 regulate PKM alternative splicing11,12. Moreover, Sirtuin mediated deacetylation of hnRNP A1 has been reported to regulate alternative splicing of PKM gene40. This suggests that the acetylation status of splicing factors plays an important role in the regulation of alternative splicing events28,40. Herein, we showed that SMAR1 mediated PKM alternative splicing regulation is HDAC6 dependent and SMAR1-HDAC6 forms a triple complex with PTBP1 and maintains it in the deacetylated form. CLIP experiment of PTBP1 in SMAR1 depleted condition resulted in the increased binding of PTBP1 on Intron 8 of PKM pre-mRNA. This indicates that the binding affinity of PTBP1 is more on PKM pre-mRNA in absence of SMAR1. Based on this observation it can be further implied that the deacetylation of PTBP1 by SMAR1-HDAC6 results in the lower affinity of PTBP1 for PKM pre-mRNA and thus leading to the inclusion of exon 9 and exclusion of exon 10. This leads to higher expression of PKM1 and lower expression of tumorigenic isoform PKM2. Our findings suggest that SMAR1 brings about HDAC6 mediated deacetylation of PTBP1 and thus maintains PTBP1 in Page 18/30 deacetylated form. This deacetylated PTBP1 ultimately leads to the regulation of PKM alternative splicing promoting expression of PKM1 and repressing oncogenic isoform PKM2. Highly proliferating cancer cells gain a metabolic advantage over normal cells via higher expression of PKM2 and exhibits peculiar metabolic properties known as the Warburg effect2. Glucose is a primary source of energy and anabolic demands for highly proliferating cancer cells2. Moreover, a higher amount of lactate production creates an acidic micro-environment in tumors which favors invasion and metastasis42. Our study demonstrates that SMAR1 mediated suppression of PKM2 isoform and upregulation of PKM1 isoform results in inhibition of cancer cell metabolism. SMAR1 inhibits glucose utilization and lactate production via regulation of PKM alternative splicing. Our results demonstrate that SMAR1 mediated reversal of the Warburg effect leads to inhibition of cancer cell proliferation, migration and invasion. Moreover, SMAR1 suppresses in vivo tumor formation via the regulation of PKM isoform expression. Declarations Ethics approval All the experiments involving the usage of animals were conducted following CPCSEA guidelines approved by the Institutional Animal Ethics Committee. Conclusion Our study in the breast cancer model highlights that nuclear matrix binding protein SMAR1 regulates PKM alternative splicing and inhibits the expression of PKM2 and promotes the expression of PKM1. SMAR1 regulates PKM alternative splicing via deacetylation of a key alternative splicing inhibitor PTBP1 in an HDAC6 dependent manner. Further, in vitro enzymatic assays, functional assays and in vivo tumor Page 19/30 model suggest that SMAR1 inhibits tumor cell metabolism and tumorigenic properties of cancer cells by suppressing the expression of tumor-associated isoform PKM2 and promoting the expression of enzymatically more efficient PKM1 isoform. model suggest that SMAR1 inhibits tumor cell metabolism and tumorigenic properties of cancer cells by suppressing the expression of tumor-associated isoform PKM2 and promoting the expression of enzymatically more efficient PKM1 isoform. Availability of data and materials All data generated or analyzed during this study are included in this published article and its supplementary information files. Abbreviations ATP: Adenosine triphosphate Funding Page 20/30 The authors are thankful to the Council of Scientific & Industrial Research (CSIR), India; University Grants Commission (UGC), India; National Centre for Cell Science (NCCS), Pune, India and Department of Biotechnology (DBT), India for their fellowships. We would also like to acknowledge the DBT, India; Department of Science and Technology (DST), India and J. C. Bose fellowship-SERB (Science and Engineering Research Board), India to SC, for financial support. Acknowledgment We are thankful to the National Centre for Cell Science (NCCS), Pune, India for providing infrastructure and experimental facilities. We would like to acknowledge Dr. Marta Cortés-Cros for providing us sequence information of shPKM2. We would like to thank Dr. Adrian R. Krainer for providing the PKM minigene construct which we used to design dual reporter PKM minigene. Authors’ contributions AC and SC conceptualized, planned and designed the experiments. AC, AP, RP, VS, RN, PF performed experiments. AC wrote the manuscript. AC, AP, RP, VS, RN, PF, SmS, SaS, GCK and SC provided intellectual inputs and helped in preparing the manuscript. GCK helped in obtaining patient samples. SaS and SmS conceptualized the dual chromatic PKM minigene system and AC cloned the dual chromatic PKM minigene system. Conflict of interest The authors declare that there is no conflict of interest. References 1. O. Warburg. On the origin of cancer cells. Science 123, 309–314 (1956). 1. O. Warburg. On the origin of cancer cells. Science 123, 309–314 (1956). 1. O. Warburg. On the origin of cancer cells. Science 123, 309–314 (1956). 2. Heiden, M. G. V., Cantley, L. C. & Thompson, C. B. Understanding the warburg effect: The metabolic requirements of cell proliferation. Science 324, 1029–1033 (2009). 2. Heiden, M. G. V., Cantley, L. C. & Thompson, C. B. Understanding the warburg effect: The metabolic requirements of cell proliferation. Science 324, 1029–1033 (2009). 3. Christofk, H. R. et al. The M2 splice isoform of pyruvate kinase is important for cancer metabolism and tumour growth. Nature 452, 230–233 (2008). 4. Jurica, M. S. et al. The allosteric regulation of pyruvate kinase by fructose-1,6-bisphosphate. Structure 6, 195–210 (1998). 5. Noguchi, T., Inoue, H. & Tanaka, T. The M1- and M2-type isozymes of rat pyruvate kinase are produced from the same gene by alternative RNA splicing. J. Biol. Chem. 261, 13807–13812 (1986). 6. Takenaka, M. et al. Isolation and characterization of the human pyruvate kinase M gene. Eur. J. Biochem 198 101 106 (1991) 5. Noguchi, T., Inoue, H. & Tanaka, T. The M1- and M2-type isozymes of rat pyruvate kinase are produced from the same gene by alternative RNA splicing. J. Biol. Chem. 261, 13807–13812 (1986). 6. Takenaka, M. et al. Isolation and characterization of the human pyruvate kinase M gene. Eur. J. Biochem. 198, 101–106 (1991). Page 21/30 Page 21/30 7. Christofk, H. R., Vander Heiden, M. G., Wu, N., Asara, J. M. & Cantley, L. C. Pyruvate kinase M2 is a phosphotyrosine-binding protein. Nature 452, 181–186 (2008). 8. Israelsen, W. J. & Vander Heiden, M. G. Pyruvate kinase: Function, regulation and role in cancer. Semin. Cell Dev. Biol. 43, 43–51 (2015). 9. Wong, N., Ojo, D., Yan, J. & Tang, D. PKM2 contributes to cancer metabolism. Cancer Lett. 356, 184– 191 (2015). 10. Gao, X., Wang, H., Yang, J. J., Liu, X. & Liu, Z. R. Pyruvate Kinase M2 Regulates Gene Transcription by Acting as a Protein Kinase. Mol. Cell 45, 598–609 (2012). 11. Clower, C. V. et al. The alternative splicing repressors hnRNP A1/A2 and PTB influence pyruvate kinase isoform expression and cell metabolism. Proc. Natl. Acad. Sci. U. S. A. 107, 1894–1899 (2010). 12. David, C. J., Chen, M., Assanah, M., Canoll, P. & Manley, J. L. References HnRNP proteins controlled by c-Myc deregulate pyruvate kinase mRNA splicing in cancer. Nature 463, 364–368 (2010). 13. Chen, M., David, C. J. & Manley, J. L. Concentration-dependent control of pyruvate kinase M mutually exclusive splicing by hnRNP proteins. Nat. Struct. Mol. Biol. 19, 346–354 (2012). 14. Spoden, G. A. et al. Isotype-specific inhibitors of the glycolytic keyregulator pyruvate kinase subtype M2 moderately decelerate tumor cell proliferation. Int. J. Cancer 123, 312–321 (2008). 15. Berezney, R., Mortillaro, M. J., Ma, H., Wei, X. & Samarabandu, J. The Nuclear Matrix: A Structural Milieu for Genomic Function. Int. Rev. Cytol. 162, 1–12 (1996). 16. Zeitlin, S., Parent, A., Silverstein, S. & Efstratiadis, A. Pre-mRNA splicing and the nuclear matrix. Mol. Cell. Biol. 7, 111–120 (1987). 17. Chattopadhyay, S., Kaul, R., Charest, A., Housman, D. & Chen, J. SMAR1, a novel, alternatively spliced gene product, binds the scaffold/matrix-associated region at the T cell receptor β locus. Genomics 68, 93–96 (2000). 18. Kaul, R. et al. Direct interaction with and activation of p53 by SMAR1 retards cell-cycle progression at G2/M phase and delays tumor growth in mice. Int. J. cancer 103, 606–615 (2003). 19. Jalota-Badhwar, A. et al. SMAR1-derived P44 peptide retains its tumor suppressor function through modulation of p53. J. Biol. Chem. 282, 9902–9913 (2007). 20. Rampalli, S., Pavithra, L., Bhatt, A., Kundu, T. K. & Chattopadhyay, S. Tumor suppressor SMAR1 mediates cyclin D1 repression by recruitment of the SIN3/histone deacetylase 1 complex. Mol. Cell. Biol. 25, 8415–8429 (2005). 21. Singh, K. et al. p53 target gene SMAR1 is dysregulated in breast cancer: its role in cancer cell migration and invasion. PLoS One 2, e660 (2007). 21. Singh, K. et al. p53 target gene SMAR1 is dysregulated in breast cancer: its role in cancer cell migration and invasion. PLoS One 2, e660 (2007). 22. Sinha, S. et al. Coordinated regulation of p53 apoptotic targets BAX and PUMA by SMAR1 through an identical MAR element. EMBO J. 29, 830–842 (2010). 22. Sinha, S. et al. Coordinated regulation of p53 apoptotic targets BAX and PUMA by SMAR1 through an identical MAR element. EMBO J. 29, 830–842 (2010). 23. Paul, D. et al. Cdc20 directs proteasome-mediated degradation of the tumor suppressor SMAR1 in higher grades of cancer through the anaphase promoting complex. Cell Death Dis. 8, e2882 (2017). 23. Paul, D. et al. References Cdc20 directs proteasome-mediated degradation of the tumor suppressor SMAR1 in higher grades of cancer through the anaphase promoting complex. Cell Death Dis. 8, e2882 (2017). Page 22/30 Page 22/30 24. Taye, N., Alam, A., Ghorai, S. & Chatterji, D. G. SMAR1 inhibits Wnt / β-catenin signaling and prevents colorectal cancer progression. Oncotarget 9, 21322–21336 (2018). 25. Powell, J. A. et al. Sequencing, transcript identification, and quantitative gene expression profiling in the breast cancer loss of heterozygosity region 16q24.3 reveal three potential tumor-suppressor genes. Genomics 80, 303–310 (2002). 26. Birot, A. et al. Identification and molecular analysis of BANP. Gene 253, 189–196 (2000). 26. Birot, A. et al. Identification and molecular analysis of BANP. Gene 253, 189–196 (2000). 27. Callen, D. F. et al. Defining regions of loss of heterozygosity of 16q in breast cancer cell lines. Cancer Genet. Cytogenet. 133, 76–82 (2002). 27. Callen, D. F. et al. Defining regions of loss of heterozygosity of 16q in breast cancer cell lines. Cancer Genet. Cytogenet. 133, 76–82 (2002). 28. Nakka, K. K. et al. Nuclear matrix-associated protein SMAR1 regulates alternative splicing via HDAC6- mediated deacetylation of Sam68. Proc. Natl. Acad. Sci. 112, E3374–E3383 (2015). 28. Nakka, K. K. et al. Nuclear matrix-associated protein SMAR1 regulates alternative splicing via HDAC6- mediated deacetylation of Sam68. Proc. Natl. Acad. Sci. 112, E3374–E3383 (2015). 29. Mathai, J. et al. SMAR1 binds to T(C/G) repeat and inhibits tumor progression by regulating miR-371- 373 cluster. Sci. Rep. 6, 1–14 (2016). 29. Mathai, J. et al. SMAR1 binds to T(C/G) repeat and inhibits tumor progression by regulating miR-371- 373 cluster. Sci. Rep. 6, 1–14 (2016). 30. Cortés-Cros, M. et al. M2 isoform of pyruvate kinase is dispensable for tumor maintenance and growth. Proc. Natl. Acad. Sci. U. S. A. 110, 489–94 (2013). 30. Cortés-Cros, M. et al. M2 isoform of pyruvate kinase is dispensable for tumor maintenance and growth. Proc. Natl. Acad. Sci. U. S. A. 110, 489–94 (2013). 31. Wang, Z. et al. Exon-centric regulation of pyruvate kinase M alternative splicing via mutually exclusive exons. J. Mol. Cell Biol. 4, 79–87 (2012). 31. Wang, Z. et al. Exon-centric regulation of pyruvate kinase M alternative splicing via mutually exclusive exons. J. Mol. Cell Biol. 4, 79–87 (2012). 32. Singh, S. et al. Intragenic DNA methylation and BORIS-mediated cancer-specific splicing contribute to the Warburg effect. Proc. Natl. Acad. Sci. 114, 11440–11445 (2017). 32. Singh, S. et al. References Intragenic DNA methylation and BORIS-mediated cancer-specific splicing contribute to the Warburg effect. Proc. Natl. Acad. Sci. 114, 11440–11445 (2017). 33. Sherr, C. J. Principles of Tumor Suppression. Cell 116, 235–246 (2004). 34. Wong, N., De Melo, J. & Tang, D. PKM2, a central point of regulation in cancer metabolism. Int. J. Cell Biol. 2013, 242513 (2013). 34. Wong, N., De Melo, J. & Tang, D. PKM2, a central point of regulation in cancer metabolism. Int. J. Cell Biol. 2013, 242513 (2013). 35. Chaudhary, N., Nakka, K. K., Chavali, P. L., Chatterjee, S. & Chattopadhyay, S. SMAR1 coordinates HDAC6-induced deacetylation of Ku70 and dictates cell fate upon irradiation. Cell Death Dis. 5, 1–11 (2014). 35. Chaudhary, N., Nakka, K. K., Chavali, P. L., Chatterjee, S. & Chattopadhyay, S. SMAR1 coordinates HDAC6-induced deacetylation of Ku70 and dictates cell fate upon irradiation. Cell Death Dis. 5, 1–11 (2014). 36. Naro, C. & Sette, C. Phosphorylation-mediated regulation of alternative splicing in cancer. Int. J. Cell Biol. 2013, 151839 (2013). 36. Naro, C. & Sette, C. Phosphorylation-mediated regulation of alternative splicing in cancer. Int. J. Cell Biol. 2013, 151839 (2013). 37. Siam, A. et al. Regulation of alternative splicing by p300-mediated acetylation of splicing factors. RNA 25, 813–824 (2019). 38. Katzenberger, R. J., Marengo, M. S. & Wassarman, D. A. Control of alternative splicing by signal- dependent degradation of splicing-regulatory proteins. J. Biol. Chem. 284, 10737–10746 (2009). 39. Pozzi, B., Mammi, P., Bragado, L., Giono, L. E. & Srebrow, A. When SUMO met splicing. RNA Biol. 15, 689–695 (2018). 40. Yang, H. et al. Sirtuin-mediated deacetylation of hnRNP A1 suppresses glycolysis and growth in hepatocellular carcinoma. Oncogene 38, 4915–4931 (2019). 41. Hay, N. Reprogramming glucose metabolism in cancer: can it be exploited for cancer therapy? Nat. Rev. Cancer 16, 635–649 (2016). 41. Hay, N. Reprogramming glucose metabolism in cancer: can it be exploited for cancer therapy? Nat. Rev. Cancer 16, 635–649 (2016). Page 23/30 Page 23/30 42. de la Cruz-López, K. G., Castro-Muñoz, L. J., Reyes-Hernández, D. O., García-Carrancá, A. & Manzo- Merino, J. Lactate in the Regulation of Tumor Microenvironment and Therapeutic Approaches. Front Oncol. 9, 1143 (2019). 43. Rahhal, R. & Seto, E. Emerging roles of histone modifications and HDACs in RNA splicing. Nucleic Acids Res. 47, 4911–4926 (2019). 44. Trivedi, J. et al. A novel isothiocyanate derivative inhibits HIV-1 gene expression and replication by modulating the nuclear matrix associated protein SMAR1. References Antiviral Res. 173, 104648 (2020). 42. de la Cruz-López, K. G., Castro-Muñoz, L. J., Reyes-Hernández, D. O., García-Carrancá, A. & Manzo- Merino, J. Lactate in the Regulation of Tumor Microenvironment and Therapeutic Approaches. Front. Oncol. 9, 1143 (2019). Figures Page 24/30 Figure 1 Figure 1 Page 24/30 Page 24/30 Expression of SMAR1 and PKM isoforms in breast cancer. (A) Expression of SMAR1, PKM1 and PKM2 in paraffinized breast tumor samples compared to normal tissue by immunohistochemistry (n=3). (B) Expression profile of SMAR1, PKM1 and PKM2 in various breast cancer cell lines by western blot. Expression of SMAR1 and PKM isoforms in breast cancer. (A) Expression of SMAR1, PKM1 and PKM2 in paraffinized breast tumor samples compared to normal tissue by immunohistochemistry (n=3). (B) Expression profile of SMAR1, PKM1 and PKM2 in various breast cancer cell lines by western blot. Figure 2 SMAR1 regulates PKM alternative splicing. (A) qRT-PCR of PKM isoforms normalized to 18S rRNA and further normalized to constitutive PKM exon (Exon 11) upon shRNA-mediated knockdown of SMAR1 in MCF7 (n=3). (B) Expression of PKM isoforms upon shRNA-mediated knockdown of SMAR1 in MCF7 by western blot. (C) qRT-PCR of PKM isoforms normalized to 18S rRNA and further normalized to constitutive PKM exon (Exon 11) upon Flag-SMAR1 mediated overexpression of SMAR1 in MDA-MB-231 (n=3). (D) Expression of PKM isoforms upon Flag-SMAR1 mediated overexpression of SMAR1 in MDA- MB-231 by western blot. (E) Schematic representation of dual chromatic PKM minigene system. (F) Confocal microscopy to check the expression of eGFP and mCherry in MCF7 cells transfected with PKM minigene along with Flag-SMAR1. Relative fluorescence intensity of eGFP and mCherry was quantified with the help of ImageJ and fold change in eGFP/mCherry ratio was calculated for control and Flag- SMAR1 overexpression (n=3) Error bars show mean values ± SD Differences were considered Figure 2 Figure 2 SMAR1 regulates PKM alternative splicing. (A) qRT-PCR of PKM isoforms normalized to 18S rRNA and further normalized to constitutive PKM exon (Exon 11) upon shRNA-mediated knockdown of SMAR1 in MCF7 (n=3). (B) Expression of PKM isoforms upon shRNA-mediated knockdown of SMAR1 in MCF7 by western blot. (C) qRT-PCR of PKM isoforms normalized to 18S rRNA and further normalized to constitutive PKM exon (Exon 11) upon Flag-SMAR1 mediated overexpression of SMAR1 in MDA-MB-231 (n=3). (D) Expression of PKM isoforms upon Flag-SMAR1 mediated overexpression of SMAR1 in MDA- MB-231 by western blot. (E) Schematic representation of dual chromatic PKM minigene system. (F) Confocal microscopy to check the expression of eGFP and mCherry in MCF7 cells transfected with PKM minigene along with Flag-SMAR1. Relative fluorescence intensity of eGFP and mCherry was quantified with the help of ImageJ and fold change in eGFP/mCherry ratio was calculated for control and Flag- SMAR1 overexpression (n=3). Error bars show mean values ± SD. Differences were considered statistically significant with p < 0.05, p < 0.01 and *p < 0.001, ns non-significant difference (p > 0.05) Page 25/30 Figure 3 SMAR1 makes a triple complex with PTBP1 and HDAC6. (A) Expression of hnRNP A1, hnRNP A2 and PTBP1 upon shRNA-mediated knockdown of SMAR1 in MCF7. (B) Co-IP of SMAR1 with hnRNP A1, hnRNP A2, PTBP1 and HDAC6 in MCF7 suggest that SMAR1 interacts with PTBP1 and HDAC6. (C) Co-IP of PTBP1 with SMAR1, HDAC1 and HDAC6 in MCF7 suggest that PTBP1 interacts with SMAR1 and HDAC6. (D) Co-IP of HADC6 with hnRNP A1, hnRNP A2, PTBP1 and SMAR1 in MCF7 confirms that HDAC6 interacts with PTBP1 and SMAR1. (E) Sequential Co-IP of SMAR1 with HDAC6 and PTBP1 in MCF7 suggest that SMAR1-HDAC6 makes a triple complex with PTBP1. SMAR1 makes a triple complex with PTBP1 and HDAC6. (A) Expression of hnRNP A1, hnRNP A2 and PTBP1 upon shRNA-mediated knockdown of SMAR1 in MCF7. (B) Co-IP of SMAR1 with hnRNP A1, hnRNP A2, PTBP1 and HDAC6 in MCF7 suggest that SMAR1 interacts with PTBP1 and HDAC6. (C) Co-IP of PTBP1 with SMAR1, HDAC1 and HDAC6 in MCF7 suggest that PTBP1 interacts with SMAR1 and HDAC6. (D) Co-IP of HADC6 with hnRNP A1, hnRNP A2, PTBP1 and SMAR1 in MCF7 confirms that HDAC6 interacts with PTBP1 and SMAR1. (E) Sequential Co-IP of SMAR1 with HDAC6 and PTBP1 in MCF7 suggest that SMAR1-HDAC6 makes a triple complex with PTBP1. Page 26/30 Figure 4 SMAR1 mediated PKM alternative splicing is HDAC6 dependent. (A) Expression of PKM isoforms upon shRNA-mediated knockdown of HDAC6 in MCF7 by western blot. (B) Expression of PKM isoforms upon SMAR1 overexpression followed by Tubacin (5 μM) treatment in MCF7 by western blot. (C) Confocal microscopy to check the expression of eGFP and mCherry in MCF7 cells transfected with PKM minigene along with Flag-SMAR1. Relative fluorescence intensity of eGFP and mCherry was quantified with the help of ImageJ and fold change in eGFP/mCherry ratio was calculated for control, control + Tubacin (5 μM), Flag-SMAR1 and Flag-SMAR1 + Tubacin (5 μM). Fold change in eGFP/mCherry ratio for SMAR1 overexpression and Tubacin treatment validates the role of HDAC6 in SMAR1 mediated regulation of PKM alternative splicing (n=3). Error bars show mean values ± SD. Differences were considered statistically significant with p < 0.05, p < 0.01 and p < 0.001, ns non-significant difference (p > 0.05). Figure 4 Figure 4 SMAR1 mediated PKM alternative splicing is HDAC6 dependent. (A) Expression of PKM isoforms upon shRNA-mediated knockdown of HDAC6 in MCF7 by western blot. (B) Expression of PKM isoforms upon SMAR1 overexpression followed by Tubacin (5 μM) treatment in MCF7 by western blot. (C) Confocal microscopy to check the expression of eGFP and mCherry in MCF7 cells transfected with PKM minigene along with Flag-SMAR1. Relative fluorescence intensity of eGFP and mCherry was quantified with the help of ImageJ and fold change in eGFP/mCherry ratio was calculated for control, control + Tubacin (5 μM), Flag-SMAR1 and Flag-SMAR1 + Tubacin (5 μM). Fold change in eGFP/mCherry ratio for SMAR1 overexpression and Tubacin treatment validates the role of HDAC6 in SMAR1 mediated regulation of PKM alternative splicing (n=3). Error bars show mean values ± SD. Differences were considered statistically significant with p < 0.05, p < 0.01 and p < 0.001, ns non-significant difference (p > 0.05). SMAR1 mediated PKM alternative splicing is HDAC6 dependent. (A) Expression of PKM isoforms upon shRNA-mediated knockdown of HDAC6 in MCF7 by western blot. (B) Expression of PKM isoforms upon SMAR1 overexpression followed by Tubacin (5 μM) treatment in MCF7 by western blot. (C) Confocal microscopy to check the expression of eGFP and mCherry in MCF7 cells transfected with PKM minigene along with Flag-SMAR1. Relative fluorescence intensity of eGFP and mCherry was quantified with the help of ImageJ and fold change in eGFP/mCherry ratio was calculated for control, control + Tubacin (5 μM), Flag-SMAR1 and Flag-SMAR1 + Tubacin (5 μM). Fold change in eGFP/mCherry ratio for SMAR1 overexpression and Tubacin treatment validates the role of HDAC6 in SMAR1 mediated regulation of PKM alternative splicing (n=3). Error bars show mean values ± SD. Differences were considered statistically significant with p < 0.05, p < 0.01 and p < 0.001, ns non-significant difference (p > 0.05). Page 27/30 Page 27/30 Figure 5 SMAR1-HDAC6 deacetylates PTBP1 and inhibits its binding to PKM pre-mRNA. (A) The acetylation status of PTBP1 upon SMAR1 knockdown in MCF7 was checked by IP with an anti-Acetyl-Lysine antibody. (B) The acetylation status of PTBP1 upon SMAR1 overexpression followed by Tubacin (5 μM) treatment in MCF7 was checked by IP with the anti-Acetyl-Lysine antibody. (C) CLIP experiment of PTBP1 on Intron 8 of PKM pre-mRNA in SMAR1 knockdown condition (n=3). (D) IP of PTBP1 to confirm pull down in CLIP experiment upon SMAR1 knockdown in MCF7. Figure 4 Error bars show mean values ± SD. Differences were considered statistically significant with p < 0.05, p < 0.01 and p < 0.001, ns non-significant difference (p > 0.05). Figure 5 Figure 5 SMAR1-HDAC6 deacetylates PTBP1 and inhibits its binding to PKM pre-mRNA. (A) The acetylation status of PTBP1 upon SMAR1 knockdown in MCF7 was checked by IP with an anti-Acetyl-Lysine antibody. (B) The acetylation status of PTBP1 upon SMAR1 overexpression followed by Tubacin (5 μM) treatment in MCF7 was checked by IP with the anti-Acetyl-Lysine antibody. (C) CLIP experiment of PTBP1 on Intron 8 of PKM pre-mRNA in SMAR1 knockdown condition (n=3). (D) IP of PTBP1 to confirm pull down in CLIP experiment upon SMAR1 knockdown in MCF7. Error bars show mean values ± SD. Differences were considered statistically significant with p < 0.05, p < 0.01 and p < 0.001, ns non-significant difference (p > 0.05). Page 28/30 Figure 6 SMAR1 inhibits Warburg effect and tumorigenesis. (A) % Glucose utilization was measured in shRNA- mediated knockdown of SMAR1 along with PKM2 in MCF7 by enzymatic assay (n=3). (B) % Lactate formation was measured in shRNA-mediated knockdown of SMAR1 along with PKM2 in MCF7 by enzymatic assay (n=3). (C) % Cell proliferation was measured in shRNA-mediated knockdown of SMAR1 along with PKM2 in MCF7 by MTT assay (n=3). (D) % Colony formation was measured in shRNA- mediated knockdown of SMAR1 along with PKM2 in MCF7 (n=3). (E) % Transwell cell migration was measured in shRNA-mediated knockdown of SMAR1 along with PKM2 in MCF7 (n=3). (F) % Transwell cell invasion was measured in shRNA-mediated knockdown of SMAR1 along with PKM2 in MCF7 (n=3). (G) % Wound migration was measured in shRNA-mediated knockdown of SMAR1 along with PKM2 in MCF7 (n=3). Error bars show mean values ± SD. Differences were considered statistically significant with p < 0.05, p < 0.01 and p < 0.001, ns non-significant difference (p > 0.05). Figure 7 SMAR1 inhibits in vivo tumor formation. (A) Tumors generated in NOD/SCID mice upon injection of adeno-SMAR1 transduced MDA-MB-231 cells compared to control (n=4). (B and C) graphs representing the Tumor weight and Tumor volume. (D and E) Expression of PKM isoforms was checked in xenograft samples by western blot and immunohistochemistry. Error bars show mean values ± SD. Differences were considered statistically significant with p < 0.05, p < 0.01 and p < 0.001, ns non-significant difference (p > 0.05). SMAR1 inhibits in vivo tumor formation. (A) Tumors generated in NOD/SCID mice upon injection of adeno-SMAR1 transduced MDA-MB-231 cells compared to control (n=4). (B and C) graphs representing the Tumor weight and Tumor volume. (D and E) Expression of PKM isoforms was checked in xenograft samples by western blot and immunohistochemistry. Error bars show mean values ± SD. Differences were considered statistically significant with p < 0.05, p < 0.01 and p < 0.001, ns non-significant difference (p > 0.05). Figure 6 SMAR1 inhibits Warburg effect and tumorigenesis. (A) % Glucose utilization was measured in shRNA- mediated knockdown of SMAR1 along with PKM2 in MCF7 by enzymatic assay (n=3). (B) % Lactate formation was measured in shRNA-mediated knockdown of SMAR1 along with PKM2 in MCF7 by enzymatic assay (n=3). (C) % Cell proliferation was measured in shRNA-mediated knockdown of SMAR1 along with PKM2 in MCF7 by MTT assay (n=3). (D) % Colony formation was measured in shRNA- mediated knockdown of SMAR1 along with PKM2 in MCF7 (n=3). (E) % Transwell cell migration was measured in shRNA-mediated knockdown of SMAR1 along with PKM2 in MCF7 (n=3). (F) % Transwell cell invasion was measured in shRNA-mediated knockdown of SMAR1 along with PKM2 in MCF7 (n=3). (G) % Wound migration was measured in shRNA-mediated knockdown of SMAR1 along with PKM2 in MCF7 (n=3). Error bars show mean values ± SD. Differences were considered statistically significant with p < 0.05, p < 0.01 and p < 0.001, ns non-significant difference (p > 0.05). Page 29/30 Figure 7 SMAR1 inhibits in vivo tumor formation. (A) Tumors generated in NOD/SCID mice upon injection of adeno-SMAR1 transduced MDA-MB-231 cells compared to control (n=4). (B and C) graphs representing the Tumor weight and Tumor volume. (D and E) Expression of PKM isoforms was checked in xenograft samples by western blot and immunohistochemistry. Error bars show mean values ± SD. Differences were considered statistically significant with p < 0.05, p < 0.01 and *p < 0.001, ns non-significant difference (p > 0.05). Supplementary Files This is a list of supplementary files associated with this preprint. Click to download. Page 30/30 Page 30/30
https://openalex.org/W4232108061	https://acp.copernicus.org/articles/13/7631/2013/acp-13-7631-2013.pdf	English	null	Dependence of particle nucleation and growth on high molecular weight gas phase products during ozonolysis of α-pinene	null	2,013	cc-by	14,192	q Biogeosciences Open Access Dependence of particle nucleation and growth on high-molecular-weight gas-phase products during ozonolysis of α-pinene Climate of the Past Open Acces J. Zhao1,2, J. Ortega2, M. Chen1, P. H. McMurry1, and J. N. Smith2,3 1Department of Mechanical Engineering, University of Minnesota, Minneapolis, Minnesota, 55455, USA 2Atmospheric Chemistry Division, National Center for Atmospheric Research, Boulder, Colorado, 80307, USA 3Department of Applied Physics, University of Eastern Finland, Kuopio 70211, Finland Climate of the Past Open Acces J. Zhao1,2, J. Ortega2, M. Chen1, P. H. McMurry1, and J. N. Smith2,3 1Department of Mechanical Engineering, University of Minnesota, Minneapolis, Minnesota, 55455, USA 2Atmospheric Chemistry Division, National Center for Atmospheric Research, Boulder, Colorado, 80307, USA 3Department of Applied Physics, University of Eastern Finland, Kuopio 70211, Finland Correspondence to: J. Zhao (zhaoj@umn.edu) Correspondence to: J. Zhao (zhaoj@umn.edu) Received: 8 March 2013 – Published in Atmos. Chem. Phys. Discuss.: 8 April 2013 Revised: 23 June 2013 – Accepted: 5 July 2013 – Published: 8 August 2013 Received: 8 March 2013 – Published in Atmos. Chem. Phys. Discuss.: 8 April 2013 Revised: 23 June 2013 – Accepted: 5 July 2013 – Published: 8 August 2013 Abstract. We report the ﬁrst time-dependent measurements of high-molecular-weight (up to 700 amu) gas-phase oxida- tion products from α-pinene ozonolysis in an aerosol cham- ber under dry and low-NOx conditions. Measurements of products having mole fractions ranging from 10−14 to 10−11 were carried out with a chemical ionization mass spectrome- ter (the Cluster CIMS). Most products that were correlated with number concentrations of the smallest particles mea- sured (10–20 nm) had molecular weights in the 430–560 amu range. Those products are proposed to be likely responsi- ble for the initial nuclei formation and the early growth of the freshly nucleated particles based on their high molecu- lar weights and chemical identities, both of which suggest low-volatility compounds. Another group of oxidation prod- ucts in the lower mass range of 140–380 amu was well cor- related with particles larger than 20 nm. We postulate that those products contributed to the later growth of particles (i.e., larger than 20 nm in diameter). Although particle nucle- ation in this study was primarily due to condensation of ox- idation products from α-pinene ozonolysis, the involvement of residual sulfuric acid vapor in particle nucleation cannot be totally excluded. q Biogeosciences Open Access Dependence of particle nucleation and growth on high-molecular-weight gas-phase products during ozonolysis of α-pinene Geoscientific Instrumentation Methods and Data Systems Open Access Geoscientific Model Development Open Acc effect of aerosols on climate require better models for the chemical processes responsible for the nucleation and growth rates that control the chemical properties and concentration of CCN (Kuang et al., 2009; Merikanto et al., 2009; Yu and Luo, 2009; Kerminen et al., 2005; Laaksonen et al., 2005). Also, the assessment of potential human health risks linked to episodes of particulate matter pollution depends on the size- resolved aerosol chemical composition (Breitner et al., 2011; Franck et al., 2011). There is hence a need to understand the physical and chemical processes responsible for nanoparticle formation and growth in the atmosphere. Model Development cess Hydrology and Earth System Sciences Open Access Ocean Science Open Access Solid Earth Open Acces The last two decades have witnessed steady progress towards understanding the formation mechanisms of at- mospheric nanoparticles in the continental boundary layer (Chen et al., 2012; Zhang et al., 2012; Kirkby et al., 2011; Bzdek and Johnston, 2010; Kulmala et al., 2004). It is com- monly accepted that atmospheric nucleation is a multicom- ponent process that involves both inorganic and organic trace species. One unique species, gaseous sulfuric acid, is almost always found to be involved in the initial nuclei formation, and its concentrations are highly correlated with the forma- tion rates of newly formed nanoparticles. However, typical sulfuric acid concentrations (106–107 cm−3) are too low to account for the observed early rapid growth of atmospheric nanoparticles (Paasonen et al., 2010; Boy et al., 2008; Iida et al., 2008; Kuang et al., 2008; Riipinen et al., 2007; Si- hto e tal., 2006; Stolzenburg et al., 2005; Weber et al., 1996, 1997, 2001). Therefore, there must be other species that con- tribute to the early growth process, and different processes Atmos. Chem. Phys., 13, 7631–7644, 2013 www.atmos-chem-phys.net/13/7631/2013/ doi:10.5194/acp-13-7631-2013 © Author(s) 2013. CC Attribution 3.0 License. Atmos. Chem. Phys., 13, 7631–7644, 2013 www.atmos-chem-phys.net/13/7631/2013/ doi:10.5194/acp-13-7631-2013 © Author(s) 2013. CC Attribution 3.0 License. 1 Introduction The formation of nanoparticles by nucleation of atmospheric trace gases can be a signiﬁcant source of cloud condensa- tion nuclei (CCN). Accurate representations of the indirect Published by Copernicus Publications on behalf of the European Geosciences Union. (2012) showed that the chemical composition of nanoparticles generated from ozonolysis of α-pinene depends on particle size: 10 and 20 nm particles were enhanced in low-volatility carboxylic acids, while 40 nm particles showed higher concentrations of carbonyl-containing compounds and low-molecular-weight organic acids (Winkler et al., 2012). This ﬁnding reveals the important role of the Kelvin effect in the growth of biogenic nanoparticles. Viitanen et al. (2010) showed that gas-phase compounds with a mobility diameter of 1.4 nm, correspond- ing to a molecular weight of ∼355 amu, were responsible for particle nucleation from α-pinene ozonolysis in a cham- ber study using ion mobility spectrometry (Viitanen et al., 2011). While signiﬁcant progress has been made towards under- standing the formation mechanisms of atmospheric nanopar- ticles from sulfuric acid, much less is known about nanopar- ticles formed from other precursor gases such as biogenic organic compounds. A well-known example is the blue haze formation over forests, which has been attributed to aerosols formed from biogenic volatile organic compounds (BVOCs) (Went, 1960). Recent laboratory experiments showed that blue haze over forests can be enhanced by the interaction between organic acids of biogenic origin and sulfuric acid (Zhang et al., 2009). However, it is unclear whether BVOCs and/or their oxidation products over forests can form nuclei without the direct involvement of sulfuric acid vapor. Recently, Ehn et al. conducted chamber experiments of α- and β-pinene ozonolysis. Naturally charged negative ions (ions that are produced from collisions of air molecules with energetic particles, mainly galactic cosmic rays or radon) formed during the ozonolysis were measured with a high- resolution Atmospheric Pressure interface-Time Of Flight mass spectrometer (APi-TOF) (Ehn et al., 2012). Most of the ions were identiﬁed as clusters containing a negatively charged nitrate ion (NO− 3 ), the latter of which serves as the reagent ion that makes ion detection possible for the APi- TOF. Based on the ion spectra, the corresponding highly ox- idized multifunctional (HOM) neutral products were iden- tiﬁed with an O / C ratio of 0.7–1.3. The HOM products were proposed to be generated via formation of geminal di- ols and/or hydroperoxides in the gas phase rather than in the aerosol phase. However, Ehn et al. (2012) only measured the steady-state signals of the oxidation products; the time- dependent intensities of individual products that are related to the initial particle formation were not reported. Published by Copernicus Publications on behalf of the European Geosciences Union. Published by Copernicus Publications on behalf of the European Geosciences Union. 7632 J. Zhao et al.: Dependence of particle nucleation and growth may dominate in different atmospheric regimes. Laboratory experiments and quantum chemical studies showed that both atmospheric acidic (e.g., carboxylic acids) and basic com- pounds (ammonia and amines) can enhance nucleation rates at given sulfuric acid and water concentrations (Chen et al., 2012; DePalma et al., 2012; Kupiainen et al., 2012; Yu et al., 2012; Zollner et al., 2012; Erupe et al., 2011; Nadykto et al., 2011; Kirkby et al., 2011; Berndt et al., 2010; Metzger et al., 2010; Wang et al., 2010a; Zhang et al., 2009, 2004; Zhao et al., 2009; Kurten et al., 2008). Furthermore, recent chamber experiments and ﬁeld observations showed that, in sulfur-rich environments, nucleation can be explained by a simpliﬁed acid-base model through the formation of elec- trically neutral sulfuric acid and amine- and/or ammonia- containing clusters (Chen et al., 2012; Jiang et al., 2011). One recent study found that evaporation of sulfuric acid from dimers and trimers (clusters that respectively contain two and three sulfuric acid molecules plus unknown numbers of wa- ter and base molecules) limits nucleation, and that the nucle- ation rate (i.e., the rate at which stable clusters are produced) equals the tetramer formation rate (Chen et al., 2012). pinene ozonolysis have been identiﬁed in ambient secondary organic aerosols collected over forests (Hyder et al., 2012; Yasmeen et al., 2010; Claeys et al., 2009; Szmigielski et al., 2007; Herckes et al., 2006). Oxidation of α- and β-pinene has been studied more than that of other monoterpenes because of their ubiquity and their important roles in secondary organic aerosol formation. Most previous laboratory chamber studies focused either on the aerosol yields or on identiﬁcation of individual gas-phase products and their yields (Perraud et al., 2012; Chen et al., 2009, 2011; Lee et al., 2006; Berndt et al., 2003; Yu et al., 1999). In these studies, many multifunctional oxidation prod- ucts were identiﬁed in both gas and particulate phases. The chemical composition of laboratory-generated particles from ozonolysis of pinenes has been measured with off-line meth- ods, and oligomers were found to be the major constituents of secondary organic aerosols (Hall and Johnston, 2011, 2012a, b; Heaton et al., 2007). Winkler et al. J. Zhao et al.: Dependence of particle nucleation and growth A scanning mobility particle sizer (SMPS) mea- sured particle size distributions in the 10–350 nm size range with a ∼5 min measurement cycle. Typical experiment times were about 20 h from the start of ozone addition until several hours after the occurrence of a secondary particle formation event. 1. Instead of measuring naturally charged ions, the Clus- ter CIMS measured ions formed from chemical ioniza- tion of neutral oxidation products. The ion concentra- tions are thus much higher than those from naturally charged ions. Hence the Cluster CIMS has enough sen- sitivity to measure the oxidation products with a much shorter measurement cycle, which allows measurements of time-dependent proﬁles during particle nucleation. 2. The average residence time used in this study (de- termined by the chamber volume divided by the to- tal ﬂow rate) is much longer than that in Ehn et al. (2012) (∼300 min vs 50 min), which allows growth of particles to larger sizes and accumulation of sufﬁ- cient concentrations of particles to be measured. Time- dependent concentrations of both particles and gas- phase oxidation products were measured for the entire experiments (typically 20 h) in this study, while only the steady-state concentrations of gas-phase oxidation products were measured in Ehn et al. (2012). 3. Concentrations of measured oxidation products were estimated by directly applying the equation derived from chemical ionization (Zhao et al., 2010). The con- centrations were determined by the relative abundances of the ions resulting from ionization of the oxidation products and the reagent ions and thus avoid the high uncertainties caused by estimating the lifetimes of the product ions (Ehn et al., 2012). We report the ﬁrst time-dependent concentrations of gas- phase oxidation products during the entire process from ini- tial nuclei formation and growth to the occurrence of a later secondary nucleation event. The oxidation products were classiﬁed based on their characteristic temporal proﬁles and the correlations between the product abundances and the par- ticle concentrations. Gas-phase oxidation products that likely lead to the initial formation and growth of new particles were identiﬁed. The mechanisms of the particle nucleation and growth products are discussed. As a routine cleaning procedure, after each experiment high ozone levels (several ppm) were introduced into the chamber with the zero air ﬂow to oxidize residual organic compounds. This air was used to ﬂush the chamber and then vented into the laboratory exhaust. Published by Copernicus Publications on behalf of the European Geosciences Union. In addition, the concentrations of particles larger than 3 nm were found to be very low (< 10 cm−3) during the experiments. Hence no correlations between the HOM products and newly formed particles were reported in their study. While little is understood about processes responsible for nucleation and the early stages of growth from BVOCs, it is known that organic compounds are signiﬁcant or even dominant constituents of ambient submicron aerosols glob- ally (Jimenez et al., 2009; Zhang et al., 2007), and many of these compounds have biogenic origins. For example, prod- ucts of terpene oxidation were observed to be signiﬁcant components of ambient submicron particles at various loca- tions (Hyder et al., 2012; Wagener et al., 2011; Yasmeen et al., 2010; Szmigielski et al., 2007; Cahill et al., 2006; Her- ckes et al., 2006). In forested areas where biogenic emissions dominate, with minimal inﬂuences of anthropogenic pollu- tion, BVOCs are hypothesized to play a more dominant role than sulfuric acid vapor in the formation of initial nuclei (Paasonen et al., 2010). The monoterpenes α- and β-pinene are among the most abundant terpenes globally, second only to isoprene (Guenther et al., 1995). They react with atmo- spheric oxidants (e.g., OH radicals and ozone) to form a va- riety of oxidation products, some of which are sufﬁciently nonvolatile to form secondary organic aerosols. Some oxi- dation products such as pinic acid and cis-pinonic acid from Atmos. Chem. Phys., 13, 7631–7644, 2013 www.atmos-chem-phys.net/13/7631/2013/ www.atmos-chem-phys.net/13/7631/2013/ 7633 J. Zhao et al.: Dependence of particle nucleation and growth about 5 h. Ozone was introduced into the chamber by pass- ing one half of the total zero air ﬂow (15 lpm) through a 4.5 cm inner-diameter stainless steel tube containing a mer- cury lamp, which exposed the air to UV light. Ozone concen- trations were regulated by adjusting the length of the mer- cury lamp that was in contact with the air ﬂow. The vapor was added to the remaining zero air ﬂow from a gas cylinder containing 7 ppm α-pinene in nitrogen. The α-pinene cylin- der was prepared by introducing the liquid (Sigma-Aldrich, purity ≥99.5 %) into an electro-polished aluminum cylin- der and pressurizing with UHP nitrogen. The ﬁnal α-pinene concentration was veriﬁed using gas chromatography with ﬂame ionization detection (GC-FID) and compared with a NIST-certiﬁed benzene standard. The α-pinene ﬂow was ini- tiated at least 10 h before ozone was added in order to reach its steady-state concentration of about 5–20 ppb. No attempt to remove OH radicals from ozone–alkene reactions was made in these experiments. The OH radicals are formed through a complex process that involves collisional stabiliza- tion and subsequent thermal decomposition of Criegee in- termediates, and the competition of the decomposition with other reactions between the intermediates and reactants such as sulfur dioxide, water, aldehydes, and organic acids (Kroll et al., 2001). An OH yield of greater than 70 % has been demonstrated during the ozonolysis of α-pinene (Forester and Wells, 2011; Siese et al., 2001; Paulson et al., 1998). So both OH and ozone likely contributed to the formation of the reaction products detected in this study. In addition, aging by the OH radicals can signiﬁcantly increase the concentra- tion of ﬁrst-generation biogenic secondary organic aerosols as has recently been demonstrated by Donahue et al. (2012). Several instruments were used to monitor trace gases, gas- phase precursors and products, and particles. A home-built proton transfer reaction mass spectrometer (PTR-MS) moni- tored the concentration of α-pinene and a limited number of volatile gas-phase products (e.g., pinonaldehyde). Gas-phase species were also detected with the Cluster CIMS (Zhao et al., 2010). A scanning mobility particle sizer (SMPS) mea- sured particle size distributions in the 10–350 nm size range with a ∼5 min measurement cycle. Typical experiment times were about 20 h from the start of ozone addition until several hours after the occurrence of a secondary particle formation event. J. Zhao et al.: Dependence of particle nucleation and growth The background con- centrations of gas-phase precursors and pre-existing parti- cles met the following criteria prior to each experiment un- less otherwise speciﬁed: < 10 ppt for α-pinene, < 1 ppb for ozone, and < 2 cm−3 for particles larger than 10 nm. Con- centrations of other trace gases such as NOx and SO2 were not measured but were expected to be low (< 0.1 ppb) ac- cording to the zero-air-generator speciﬁcations. During one 2.1 Experimental setup The experiments were carried out in a 10 m3 continuous ﬂow reaction chamber shown in Fig. S1. These experiments were performed under dark conditions. Dry (RH < 1 %) zero air (model 737, Aadco Instruments Inc.) ﬂowed through the chamber at ∼30 lpm, resulting in a mean residence time of Atmos. Chem. Phys., 13, 7631–7644, 2013 www.atmos-chem-phys.net/13/7631/2013/ J. Zhao et al.: Dependence of particle nucleation and growth As a routine cleaning procedure after each experiment In this study, we measured gas-phase oxidation products from α-pinene ozonolysis using the Cluster CIMS (Zhao et al., 2010). There are three major differences between measurements in this study and those reported in Ehn et al. (2012). about 5 h. Ozone was introduced into the chamber by pass ing one half of the total zero air ﬂow (15 lpm) through a 4.5 cm inner-diameter stainless steel tube containing a mer- cury lamp, which exposed the air to UV light. Ozone concen- trations were regulated by adjusting the length of the mer- cury lamp that was in contact with the air ﬂow. The vapor was added to the remaining zero air ﬂow from a gas cylinder containing 7 ppm α-pinene in nitrogen. The α-pinene cylin- der was prepared by introducing the liquid (Sigma-Aldrich, purity ≥99.5 %) into an electro-polished aluminum cylin- der and pressurizing with UHP nitrogen. The ﬁnal α-pinene concentration was veriﬁed using gas chromatography with ﬂame ionization detection (GC-FID) and compared with a NIST-certiﬁed benzene standard. The α-pinene ﬂow was ini- tiated at least 10 h before ozone was added in order to reach its steady-state concentration of about 5–20 ppb. No attempt to remove OH radicals from ozone–alkene reactions was made in these experiments. The OH radicals are formed through a complex process that involves collisional stabiliza- tion and subsequent thermal decomposition of Criegee in- termediates, and the competition of the decomposition with other reactions between the intermediates and reactants such as sulfur dioxide, water, aldehydes, and organic acids (Kroll et al., 2001). An OH yield of greater than 70 % has been demonstrated during the ozonolysis of α-pinene (Forester and Wells, 2011; Siese et al., 2001; Paulson et al., 1998). So both OH and ozone likely contributed to the formation of the reaction products detected in this study. In addition, aging by the OH radicals can signiﬁcantly increase the concentra- tion of ﬁrst-generation biogenic secondary organic aerosols as has recently been demonstrated by Donahue et al. (2012). Several instruments were used to monitor trace gases, gas- phase precursors and products, and particles. A home-built proton transfer reaction mass spectrometer (PTR-MS) moni- tored the concentration of α-pinene and a limited number of volatile gas-phase products (e.g., pinonaldehyde). Gas-phase species were also detected with the Cluster CIMS (Zhao et al., 2010). 3.1 Species pertinent to particle nucleation and growth Three experiments (E1–E3) are discussed in this paper and their experimental conditions are summarized in Table S1. Most of the data presented below are from E1; similar results from E2 and E3 can be found in the Supplement. Ozone con- centrations increased linearly at a rate of 0.1–0.2 ppb min−1 during the ﬁrst three hours, and during this period an intense new particle formation event occurred, as shown in Fig. 1a and b for E1. Concentrations of the smallest measured parti- cles (10–20 nm diameter) began to increase within 1 h after ozone was ﬁrst introduced, and reached peak values of about 200–250 cm−3 after 0.5–1.5 h (Fig. 1b, point c, for E1; see also Fig. S2b for E2 and Fig. S3b for E3). Evidence of a second particle formation event beginning at about 8–10 h is also evident in all three experiments (e.g., Fig. 1b, point f). This probably occurs because the surface area of particles decreased (Fuchs surface area in Figs. 1c, S2c, and S3c), al- lowing the concentrations of condensable species to build up until new nuclei are formed and subsequently grow to de- tectable sizes (Kuang et al., 2010). Because no scavenger was added to remove OH radicals, we cannot rule out the possibility that reactions of OH radicals with products from α-pinene ozonolysis contributed to nucleation. Evidence for periodic bursts of nucleation has previously been observed in other studies (VanReken et al., 2006; McGraw and Saun- ders, 1984). In addition, at almost the same time, a distinct bimodal particle growth pattern is formed with a minimum number concentration observed at 150 nm diameter (Figs. 1a, S2a, and S3a). Similar observations have previously been re- ported for the ozonolysis of isoprene (Kamens et al., 1982). of the experiments, E1, background level exceeded the cri- teria presented above. Prior to that experiment, which is de- scribed in detail below, there were detectable levels of ox- idized VOCs and a moderately high (∼50 cm−3) concen- tration of 10–20 nm diameter particles. The source of those background oxidized VOCs and particles is not known. It is possible that they were formed from the residual products of the prior experiment. However, the amount of aerosol formed after the ozone was added greatly exceeded the amount of aerosol present when the ozone was added (point b in Fig. 1b). J. Zhao et al.: Dependence of particle nucleation and growth 7634 7634 CIMS inlet was modiﬁed to allow interfacing to the chamber. The inlet was shortened to 50 cm, which is half the length of the inlet used for ambient measurements. A sample ﬂow of 10 lpm was drawn from the chamber through a 1.25 cm outer- diameter Teﬂon tube. The N2 ﬂow rate through the ion source was set to 1 lpm to minimize sample ﬂow dilution. The reac- tion time for the reagent ions with the sampled products was estimated to be ∼0.2 s. For most experiments, we generated nitrate ions (the reagent ions) by introducing trace amount of nitric acid into the N2 ﬂow through the ion source to measure gas-phase oxidation products, similar to the detection scheme that was employed in previous ambient measurements (Zhao et al., 2010). We also carried out exploratory experiments to measure more gas-phase oxidation products by employing acetate, a less selective reagent ion (Veres et al., 2008, 2010). The Cluster CIMS acquires the spectra in 100–700 amu range using a quadrupole mass spectrometer, with a measurement cycle of about 10 min in length. Fig. 1. Characteristics of particle formation from α-pinene ozonoly- sis for E1: (a) contour plot of the particle size distribution measured with the SMPS in the 10–350 nm diameter range; (b) total num- ber concentration of particles between 10 and 20 nm. Points a–f denote the corresponding characteristic times as indicated below: a (t = −1.7 h), background levels; b (t = 0 h), addition of ozone; c (t = 0.5 h), reaching peak concentrations of Category I products for the ﬁrst particle formation event; d (t = 2.6 h), disappearance of al- most all 10–20 nm particles; e (t = 3.9 h), approaching steady-state concentration of Category I products; f (t = 12.9 h), reaching peak concentrations of Category I products for the second particle for- mation event (see text for the deﬁnition of Category I products). (c) Estimated time-dependent Fuchs surface area (in µm2 cm−3). 3.1 Species pertinent to particle nucleation and growth Although E1 was not ideal in this regard, it is rep- resentative of all experiments performed for which the back- ground criteria were met. Once the α-pinene concentration reached steady state, ozone was introduced into the cham- ber. In most experiments, the steady-state ozone concentra- tion was ∼50–100 ppb. The ozone was sampled from a port opposite the inlet using a UV photometric ozone analyzer (Model 49, Thermo Scientiﬁc Inc.). www.atmos-chem-phys.net/13/7631/2013/ 7634 Fig. 1. Characteristics of particle formation from α-pinene ozonoly- sis for E1: (a) contour plot of the particle size distribution measured with the SMPS in the 10–350 nm diameter range; (b) total num- ber concentration of particles between 10 and 20 nm. Points a–f denote the corresponding characteristic times as indicated below: a (t = −1.7 h), background levels; b (t = 0 h), addition of ozone; c (t = 0.5 h), reaching peak concentrations of Category I products for the ﬁrst particle formation event; d (t = 2.6 h), disappearance of al- most all 10–20 nm particles; e (t = 3.9 h), approaching steady-state concentration of Category I products; f (t = 12.9 h), reaching peak concentrations of Category I products for the second particle for- mation event (see text for the deﬁnition of Category I products). (c) Estimated time-dependent Fuchs surface area (in µm2 cm−3). www.atmos-chem-phys.net/13/7631/2013/ J. Zhao et al.: Dependence of particle nucleation and growth Color codes: red – Category I products; blue – Category II products; gray – Category III products (see text for the deﬁnition of the three categories). The correlations were per- formed between concentrations of products and the number concen- trations of particles for period t = 0–10 h in Fig. 1b. The horizontal dash lines indicate zero correlation coefﬁcients. ameter interval (point d), while concentrations of Category I compounds reach approximately steady-state levels (point e). As the experiment continues past the 6 h mark, Category I compounds appear to be anti-correlated with the Fuchs sur- face area (ref. Figs. 1c and 3a), increasing above the steady- state concentration at about 8 h, reaching a secondary maxi- mum at ∼13 h, and then decreasing during the second parti- cle formation event. The number concentration of 10–20 nm particles follows a similar trend: at about 8–10 h, their con- centration again starts to increase, reaching a secondary peak of about 30–40 cm−3 at ∼14 h. The formation rate for 10 nm particles during the second event, J10nm, was much lower than for the ﬁrst event (about 0.038 vs. 0.42 cm−3 s−1; see Supplement Sect. II for a discussion of particle formation and growth rates). If Category I compounds were responsi- ble for nucleation in both cases, their lower concentrations during the second event would have led to lower nucleation rates. As mentioned previously, the results from E2 and E3 were similar to those of E1. However, no obvious increase of concentrations of Category I compounds was found for E2 during the second particle formation event, probably due to a lower formation rate (see Supplement Sect. II) that requires a lower concentration (about 1.2 × 107 cm−3 for the steady- state concentration, Fig. S5). g Figure 3a shows a time series plot of the number con- centration of 10–20 nm particles and the total concentration of Category I compounds. The concentration of individual Category I compounds follows a similar trend. Estimated concentrations of the gas-phase products were obtained by applying the previously measured mass-dependent sensitiv- ities of the Cluster CIMS and assuming an ion-molecule rate constant of 2 × 10−9 cm3 s−1 and a reaction time of about 0.2 s (Zhao et al., 2010, ref. Fig. S4 for the mass- dependent sensitivities used to determine species concen- trations). As mentioned in Sect. J. Zhao et al.: Dependence of particle nucleation and growth 7635 Fig. 2. Coefﬁcients of correlation between gas-phase oxidation products and particles for E1: (a) with 10–20 nm particles; (b) with particles larger than 20 nm. Color codes: red – Category I products; blue – Category II products; gray – Category III products (see text for the deﬁnition of the three categories). The correlations were per- formed between concentrations of products and the number concen- trations of particles for period t = 0–10 h in Fig. 1b. The horizontal dash lines indicate zero correlation coefﬁcients. detected by the Cluster CIMS for which observed abun- dances are higher than their background levels. Figure 2b shows correlations for particles larger than 20 nm. The cor- relations are separated into three groups, which are shown in red (Category I), blue (Category II), and gray (Cate- gory III). The Category I compounds are mostly in the range of 490–630 amu, except for several in the ranges of 300–400 and 430–480 amu, and are well correlated with the smallest particles detected (10–20 nm) (correlation coefﬁcients ≥0.5 for most of Category I compounds except for 480, 530, 558 and 560 amu); we infer that these compounds may be as- sociated with particle nucleation. In the beginning of most experiments (in the absence of ozone), the chamber is typ- ically free of particles and hence near-zero particle surface area is expected. Without an aerosol source of condensation sink (i.e., not considering wall losses that are constant during the experiment), condensable products build up rapidly upon ozone addition. Saturation ratios of those species increase quickly and, above a certain limit, homogenous nucleation and rapid growth occur. For E1, the background concentra- tions of 10–20 nm particles are associated with detectable Category I products. Also a sharp increase of 10–20 nm par- ticle concentration was observed upon addition of ozone that clearly exceeded the concentration of any background particles (Fig. 1a). The Category II compounds are in the 140–380 amu range and are correlated with number con- centrations of particles larger than 20 nm (correlation coef- ﬁcients ≥0.5); we infer that these compounds may be as- sociated with particle growth. Category III compounds are not correlated with the concentrations of particles in either of those size ranges. Fig. 2. Coefﬁcients of correlation between gas-phase oxidation products and particles for E1: (a) with 10–20 nm particles; (b) with particles larger than 20 nm. 2.2 Cluster CIMS A detailed description of the Cluster CIMS along with il- lustrative measurements can be found in recent publications (Jiang et al., 2011; Zhao et al., 2010, 2011). Only the aspects relevant to the present study are described here. The Cluster Figure 2a shows correlations between the concentrations of 10–20 nm particles and the oxidation products that are Atmos. Chem. Phys., 13, 7631–7644, 2013 www.atmos-chem-phys.net/13/7631/2013/ J. Zhao et al.: Dependence of particle nucleation and growth 2.1, during E1 some Cat- egory I products were detected 1 h prior to ozone addition. Nevertheless, for this and all other experiments (ref. Fig. 1a, b; Figs. S2a, b; Fig. S3a, b in the Supplement), increases in the concentrations of 10–20 nm particles occurred about 0.5–1 h after the concentrations of Category I products were observed to increase. This time delay may be associated with the time required for nuclei to grow to 10 nm, the minimum size that was measured in this study. Once particle formation is observed, the Category I compounds and 10–20 nm parti- cles follow similar temporal proﬁles for about 2–3 h, reach- ing peak concentrations almost at the same time (Fig. 3a, point c). Subsequently, particle concentrations drop to near zero as nucleated particles grow beyond the 10–20 nm di- Figure 3b shows the time series plot of the total concen- tration of Category II compounds and the concentration of particles larger than 20 nm. Again, the concentration of indi- vidual Category II compounds follows a similar trend. These www.atmos-chem-phys.net/13/7631/2013/ J. Zhao et al.: Dependence of particle nucleation and growth 1b; (b) Category II products and particles larger than 20 nm; (c) Category III. Fig. 4. Time-dependent spectra of all category products between 100 and 700 amu measured with the Cluster CIMS for E1. (a–f) correspond to times a–f in Fig. 1b. Color codes: red – Category I products; blue – Category II products; gray – Category III products. Number 1–15 in Fig. 4c corresponds to different peak classiﬁcations (see text for details). time series are distinctly different from those in Fig. 3a. Con- centrations began to increase after ozone was introduced, and concentrations of both the > 20 nm particles and the Cat- egory II compounds reach peak values about 1 h after the 10–20 nm particles and Category I compounds reach their peak values, and decrease gradually thereafter. This suggests that these products are both ﬁrst of α-pinene ozonolysis and later-generation products formed through further OH radi- cal reactions. Figure 3c is a time series plot for a typical Category III compound. These compounds increase steadily throughout the 20 h experiment, which is likely due to the fact that their production rate is only slightly greater than their loss rate. Category III compounds are not correlated with particle concentrations in any size interval. Because they accumulate over time, we infer that Category III com- pounds are a mixture of ﬁrst- and later-generation products that are contributed from both α-pinene ozonolysis and fur- ther OH radical reactions. Category III compounds might also result from further heterogeneous oxidation of low- volatility compounds such as Category I compounds (Kroll et al., 2011). For example, particles might provide sufﬁcient surface area for heterogeneous reactions after 2 h when oxi- dant levels are high (Fig. 1c). and S3), are shown in these spectra. The background level of each peak was taken to be the average concentration from ﬁve consecutive scans about 2 h prior to ozone addition. Most peaks appear in two mass ranges: 200–400 amu and 490– 630 amu. Figure 4c shows that concentrations of Category I compounds reach peak values when the concentration of 10– 20 nm particles reaches the peak values (point c in Figs. 1b and 3a). Figure 4e shows the spectrum of neutral compound concentration vs. molecular weight at the point where Cate- gory I compounds drop to their lowest levels. J. Zhao et al.: Dependence of particle nucleation and growth 7636 Zhao et al.: Dependence of particle nucleation and growth Fig. 4. Time-dependent spectra of all category products between 100 and 700 amu measured with the Cluster CIMS for E1. (a–f) correspond to times a–f in Fig. 1b. Color codes: red – Category I products; blue – Category II products; gray – Category III products. Number 1–15 in Fig. 4c corresponds to different peak classiﬁcations (see text for details). Fig. 3. Time-dependent total concentrations of the three category (I–III) products measured with the Cluster CIMS, along with the total concentrations of two category particles (10–20 nm particles and particles larger than 20 nm measured with the SMPS for E1). (a) Category I products and 10–20 nm particles, points a–f corre- spond to times a–f in Fig. 1b; (b) Category II products and particles larger than 20 nm; (c) Category III. Fig. 4. Time-dependent spectra of all category products between 100 and 700 amu measured with the Cluster CIMS for E1. (a–f) correspond to times a–f in Fig. 1b. Color codes: red – Category I products; blue – Category II products; gray – Category III products. Number 1–15 in Fig. 4c corresponds to different peak classiﬁcations (see text for details). Fig. 3. Time-dependent total concentrations of the three category (I–III) products measured with the Cluster CIMS, along with the total concentrations of two category particles (10–20 nm particles and particles larger than 20 nm measured with the SMPS for E1). (a) Category I products and 10–20 nm particles, points a–f corre- spond to times a–f in Fig. 1b; (b) Category II products and particles larger than 20 nm; (c) Category III. Fig. 3. Time-dependent total concentrations of the three category (I–III) products measured with the Cluster CIMS, along with the total concentrations of two category particles (10–20 nm particles and particles larger than 20 nm measured with the SMPS for E1). (a) Category I products and 10–20 nm particles, points a–f corre- spond to times a–f in Fig. 1b; (b) Category II products and particles larger than 20 nm; (c) Category III. Fig. 3. Time-dependent total concentrations of the three category (I–III) products measured with the Cluster CIMS, along with the total concentrations of two category particles (10–20 nm particles and particles larger than 20 nm measured with the SMPS for E1). (a) Category I products and 10–20 nm particles, points a–f corre- spond to times a–f in Fig. www.atmos-chem-phys.net/13/7631/2013/ Atmos. Chem. Phys., 13, 7631–7644, 2013 J. Zhao et al.: Dependence of particle nucleation and growth 7637 Table 1. Summary of all measured Category I products for E1: ion (m/z), peak concentration of neutral species, cluster types, corresponding neutral identities, and correlation coefﬁcients with 10–20 nm particles. Table 1. Summary of all measured Category I products for E1: ion (m/z), peak concentration of neutral species, cluster types, corresponding neutral identities, and correlation coefﬁcients with 10–20 nm particles. g y p ( / ), p p , correlation coefﬁcients with 10–20 nm particles. J. Zhao et al.: Dependence of particle nucleation and growth Group Ion Peak conc.a Cluster Corresponding neutralc Correlation (m/z) (cm−3) typeb Mass(amu) Formula coefﬁcientd 1 328 4.8 × 105 1 266 C9H14O9 0.65 1 330 1.9 × 105 NA 0.50 2 340 2.0 × 106 1 278 C10H14O9 0.89 2 342 1.8 × 106 1 C9H12O10 0.87 or C10H16O9 3 358 1.9 × 106 1 296 C10H16O10 0.87 3 359 2.2 × 105 NA 0.53 4 372 1.1 × 106 1 310 C10H14O11 0.90 4 373 2.6 × 105 NA 0.63 4 374 6.3 × 105 1 312 C10H16O11 0.84 5 432 4.9 × 105 NA 0.64 6 460 1.1 × 106 NA 0.66 7 478 9.2 × 105 NA 0.78 7 480 6.5 × 105 1 418 C18H26O11 0.46 8 490 1.6 × 106 1 428 C14H20O15 0.69 8 492 1.7 × 106 NA 0.55 8 493 2.9 × 106 NA 0.84 8 494 1.9 × 107 1 432 C19H28O11 0.88 8 495 3.4 × 106 2 0.76 8 496 2.3 × 106 NA 0.61 8 498 3.3 × 106 NA 0.67 9 510 8.6 × 106 1 448 C20H32O11 0.70 9 511 2.5 × 106 2 0.72 9 512 2.1 × 106 NA 0.50 10 523 1.8 × 106 NA 0.66 10 524 1.0 × 107 1 462 C20H30O12 0.68 10 525 8.3 × 106 2 0.64 10 526 8.7 × 106 1 464 C19H28O13 0.74 10 530 3.0 × 106 1 468 C18H28O14 0.43 11 540 3.8 × 106 1 478 C20H30O13 0.52 11 541 2.4 × 106 2 0.67 11 542 1.0 × 107 1 480 C20H32O13 0.87 11 543 3.9 × 106 2 0.74 11 544 2.0 × 106 NA 0.56 12 556 6.4 × 106 1 494 C20H30O14 0.73 12 557 3.1 × 106 2 0.68 12 558 3.5 × 106 NA 0.48 12 559 1.7 × 106 2 0.57 12 560 2.0 × 106 NA 0.48 13 574 8.0 × 106 1 512 C20H32O15 0.78 13 575 2.6 × 106 2 0.56 14 588 1.3 × 107 1 526 C20H30O16 0.88 14 590 4.4 × 106 NA 0.73 15 620 1.4 × 107 1 558 C20H30O18 0.86 15 621 5.6 × 106 2 0.67 a concentration: background subtracted peak concentration (in cm−3) at time c in Fig. J. Zhao et al.: Dependence of particle nucleation and growth The Category I concentrations again pass through a second maximum when the concentration of 10–20 nm particles reaches the peak dur- ing the second particle formation event; the representative spectrum at this point is shown in Fig. 4f. Category I com- pounds can be further divided into 15 groups of peaks as shown in Fig. 4c. Each group contains 1 to 5 peaks as il- lustrated in Table 1. Ehn et al. (2012) used the APi-TOF to measure gas-phase species formed under the steady-state conditions during α-pinene ozonolysis, and identiﬁed the el- emental composition of certain products. As mentioned pre- viously, most of the detected ions in their chamber experi- ments have even m/z and were identiﬁed as clusters contain- ing an NO− 3 . In those experiments, NO− 3 formed in the cham- ber from reactions of residual HNO3 with other air ions such as O− 2 provided an in situ source of reagent ions. Ambient measurements in Hyyti¨al¨a showed that ions with odd masses were also present during the day when ambient sulfuric acid concentrations were high, and those ions were attributed to species that contain an HSO− 4 (Ehn et al., 2012). The Clus- ter CIMS did detect about 13 odd m/z ions (about one third Figure 4a–f show six spectra in which the estimated con- centrations of the neutral gas-phase compounds are plotted versus molecular weight over the 100–700 amu range dur- ing the period corresponding to points a–f in Fig. 1b. As ex- plained above, these compounds are classiﬁed as Category I (red), II (blue), and III (gray). Only peaks whose concen- trations were observably higher than background levels after addition of ozone, a total of 139 peaks (ref. Table 1, Table S2 Atmos. Chem. Phys., 13, 7631–7644, 2013 www.atmos-chem-phys.net/13/7631/2013/ www.atmos-chem-phys.net/13/7631/2013/ a concentration: background subtracted peak concentration (in cm−3) at time c in Fig. 1b; b Atmos. Chem. Phys., 13, 7631–7644, 2013 ( ); d correlation was performed with the concentration of 10–20 nm particles for time period t = 0-10 h in Fig. 1b (t = 0 was deﬁned as the time when ozone was added to the chamber). b cluster types: type 1 contains an NO− 3 (from Ehn et al., 2012); type 2 possibly contains an NO− 3 · HNO3; c from Ehn et al. (2012); 3.2 Contribution of residual sulfuric acid vapor to particle nucleation Figure 5 shows the time series plots of concentrations corre- sponding to the sulfuric acid monomer (160 amu, Fig. 5b), dimer (195 amu, Fig. 5c), trimer (293 amu, Fig. 5d), and tetramer (391 amu, Fig. 5e), along with the 10–20 nm and > 20 nm particles (Fig. 5a). The source of sulfuric acid was not known, but it was likely attributed to reactions of resid- ual SO2 with OH radicals that were generated during the ozonolysis. Residual SO2 was likely present in the zero air since the SO2 was not scrubbed before the house compressed air was delivered to the zero air generator. During the in- tense formation event of 10–20 nm particles, sulfuric acid monomer concentrations increased steadily, but its maximum level (about 5.5 × 105 cm−3 at about 4 h) occurred about 3.5 h after the peak concentration of the 10–20 nm particles. The concentration of the cluster corresponding to 293 amu varied between 5 × 104 and 3 × 105 cm−3. However, this ion is unlikely attributed to sulfuric acid trimers because con- centrations of ions associated with sulfuric acid dimers and tetramers were barely higher than their background levels. Concentrations of the measured sulfuric acid clusters con- taining 2–4 H2SO4 molecules were hence below the detec- tion limits (< 2 × 104 cm−3). g Four HOM compounds (C10H14O7, C10H14O9, C10H16O9, C10H16O11) in the Ehn et al. study were identiﬁed as the most abundant products, clustering with the reagent ion NO− 3 with peaks at 308, 340, 342, and 372 amu (Ehn et al., 2012). The Cluster CIMS measurements also show that during particle nucleation those products are among the most abundant. Based on their time-dependent proﬁles and their correlation with particles, one (308 amu) belongs to Category II and the rest (340, 342, and 372 amu) belong to Category I compounds. The intensities of Cate- gory I compounds in the 300–400 amu range are generally about twice those in the 490–630 amu range (corresponding to neutral masses of 430–560 amu); however, after mass- dependent sensitivities of the Cluster CIMS are applied (Fig. S4), the estimated concentrations of the latter are much higher than the former (about 20 times) because the Cluster CIMS sensitivities drop rapidly for ions larger than 400 amu (Zhao et al., 2010). The peak and steady-state Sulfuric acid concentrations were estimated to be below 2.0 × 105 cm−3 during intense particle nucleation. J. Zhao et al.: Dependence of particle nucleation and growth Because nitrate dimers were employed as reagent ions in chemical ionization and nitric acid concentrations were elevated in the ionization zone, those odd m/z ions were most likely from the addition of one more nitric acid (63 amu) to the even m/z clusters (containing an NO− 3 ), for example, 557 amu from 494 amu, 621 amu from 558 amu, etc. (Table 1). Our elemen- tal composition assignments in Table 1 are based on species identiﬁed by Ehn et al. using high-resolution mass spectrom- etry (Ehn et al., 2012). Table 1 also shows the correspond- ing neutral formulae and their molecular weights, as well as their correlation coefﬁcients with 10–20 nm particles. A few compounds having correlation coefﬁcients smaller than 0.5 (see column 7) are still included in Category I compounds because their time-dependent concentrations follow similar trends as those with correlation coefﬁcients greater than 0.5. J. Zhao et al.: Dependence of particle nucleation and growth J. Zhao et al.: Dependence of particle nucleation and growth 7638 Fig. 5. (a) Time-dependent concentrations of ions corresponding to sulfuric acid and its clusters containing up to 4 H2SO4, along with the concentrations of 10–20 nm particles and > 20 nm particles for E1 (b 160 amu, monomer; c 195 amu, dimer; d 293 amu, likely organic species other than trimer; e 391 amu, tetramer). concentrations for the sum of Category I compounds are estimated to be about 1.6 × 108 and 2.2 × 107 cm−3 respectively, predominately contributed from products in the 490–630 amu range (group 8–15 in Fig. 4c). Among them, there is one ion in each group that has much higher concentration than the adjacent ions except group 10, which has high concentrations in the adjacent three peaks (524, 525, 526 amu) (Table 1). The sum of the concentrations of those species comprises more than 70 % of the total Category I products. We estimate the overall uncertainty in concentrations of Category I compounds to be about a factor of 5, mainly inﬂuenced by the uncertainty of the mass-dependent sensitivities in the high mass range (i.e., 400–700 amu) and the estimated ion-molecule reaction time. Compared to ambient measurements, uncertainties associated with background identiﬁcation and subtraction are minor because of low background levels relative to the signals after ozone addition (Jiang et al., 2011). In compar- ison, a steady-state concentration of about 106–107 cm−3 for the neutral HOM products was estimated in Ehn et al. (2012), signiﬁcantly lower than those estimated in this study under similar precursor concentrations but a much lower humidity (less than 1 %) than that in Eh et al. (about 63 %). Fig. 5. (a) Time-dependent concentrations of ions corresponding to sulfuric acid and its clusters containing up to 4 H2SO4, along with the concentrations of 10–20 nm particles and > 20 nm particles for E1 (b 160 amu, monomer; c 195 amu, dimer; d 293 amu, likely organic species other than trimer; e 391 amu, tetramer). of the 44 ions in Table 1) in Category I products. J. Zhao et al.: Dependence of particle nucleation and growth 1b; b cluster types: type 1 contains an NO− 3 (from Ehn et al., 2012); type 2 possibly contains an NO− 3 · HNO3; c from Ehn et al. (2012); d correlation was performed with the concentration of 10–20 nm particles for time period t = 0-10 h in Fig. 1b (t = 0 was deﬁned as the time when ozone was added to the chamber). Atmos. Chem. Phys., 13, 7631–7644, 2013 www.atmos-chem-phys.net/13/7631/2013/ www.atmos-chem-phys.net/13/7631/2013/ J. Zhao et al.: Dependence of particle nucleation and growth volatilities (hence saturation vapor pressures) for oxidized organics containing up to 30 carbons with different func- tional groups. We did not detect signiﬁcant concentrations of carboxylic acid monomers in our gas-phase measure- ments using nitrate dimer ion as the reagent ion; instead, we observed highly oxidized high-molecular-weight com- pounds such as Category I products, predominantly in the 430–560 amu range. Some Category I products are possible dimers of products in a lower mass range (100–300 amu), forming via hydrogen bonding or covalent bonding that loses one water molecule, or other possible formation pathways, while some of them are high-molecular-weight compounds rather than dimers (Fig. S6). However, high mass resolu- tion is needed to conﬁrm the above statement. The current unit mass resolution of the Cluster CIMS cannot deﬁnitely identify the molecular composition of those compounds. The carboxylic acid monomers detected in Winkler et al. (2012) are likely fragments of high-molecular-weight compounds due to thermal decomposition of particulate species dur- ing resistive heating of particles in the TDCIMS measure- ments (Hall and Johnston, 2012b). The highly oxidized Cat- egory I products likely have even lower saturation vapor pres- sures (∼10−14–10−12 atm) than those estimated in Winkler et al. (2012) due to their higher molecular weights and higher oxygen contents. Hence they can partition into particles and help them grow to overcome the Kelvin effect during the early stages of particle growth, as indicated by their high correlations with the measured 10–20 nm particles. The to- tal peak concentration of Category I compounds is approxi- mately 1.6 × 108 cm−3, about a factor of 2 lower than what is required for growth of particles in the ﬁrst hour after de- tection of 10 nm particles (see Supplement Sect. III). Con- sidering the uncertainty of the concentration of Category I products and the required concentration estimated for the ob- served growth rate, it is possible that Category I products contribute to the rapid growth of particles in the ﬁrst hour after detection of 10 nm particles. Demonstrating this with certainty would require identifying the composition of the initial nucleated particles. or ambient measurements. However, if the available mea- sured nucleation rates are extrapolated down to sulfuric acid concentrations of ∼3 × 105 cm−3 observed in this study, an upper limit of about 0.3–0.5 cm−3 s−1 is obtained (Chen et al., 2012; Zollner et al., 2012). J. Zhao et al.: Dependence of particle nucleation and growth Hence they can partition into particles and help them grow to overcome the Kelvin effect during the early stages of particle growth, as indicated by their high correlations with the measured 10–20 nm particles. The to- tal peak concentration of Category I compounds is approxi- mately 1.6 × 108 cm−3, about a factor of 2 lower than what is required for growth of particles in the ﬁrst hour after de- tection of 10 nm particles (see Supplement Sect. III). Con- sidering the uncertainty of the concentration of Category I products and the required concentration estimated for the ob- served growth rate, it is possible that Category I products contribute to the rapid growth of particles in the ﬁrst hour after detection of 10 nm particles. Demonstrating this with certainty would require identifying the composition of the initial nucleated particles. Although the total concentration of Category I products is several orders of magnitude higher than that of the residual sulfuric acid vapor and the concentrations of sulfuric acid clusters (dimers, trimers, tetramers, etc.) were below the de- tection limit of the Cluster CIMS, the role of sulfuric acid in inﬂuencing particle nucleation cannot be totally ruled out. It is possible that particle nucleation is initiated by formation of sulfuric acid clusters and the subsequent rapid growth is due to condensation of the oxidation organic products. The growth rate enhancement factor, 0, is the ratio of the ob- served particle growth rate to the growth rate due to sulfuric acid vapor alone (Kuang et al., 2010). A 0 value of about 1600 was estimated for the ﬁrst hour after detection of 10 nm particles in the ﬁrst event shown in Fig. 1a, which demon- strates that organic products dominate particle growth. 3.2 Contribution of residual sulfuric acid vapor to particle nucleation We esti- mated an upper limit for sulfuric acid concentrations of about 3 × 105 cm−3, with a ±50 % uncertainty. At these concentra- tions, no nucleation has been reported either in laboratory www.atmos-chem-phys.net/13/7631/2013/ Atmos. Chem. Phys., 13, 7631–7644, 2013 www.atmos-chem-phys.net/13/7631/2013/ 7639 3.3 Mechanisms for particle growth Low-volatility carboxylic acids (C8–C10) have been iden- tiﬁed in the particulate phase as the major components of secondary organic aerosols formed from α-pinene ozonol- ysis (Yasmeen et al., 2010; Claeys et al., 2009), including terpenylic acid (172 amu), cis-pinonic acid (184 amu), and pinic acid (186 amu). Those carboxylic acids were found to be the major constituents of the 10 and 20 nm particles from α-pinene ozonolysis in a ﬂow reactor (Winkler et al., 2012). Winkler et al. estimated a saturation vapor pressure on the order of 10−11 atm or less for compounds responsible for growth of particles between 4-30 nm and 10−9 atm for compounds responsible for growth of particles larger than 30 nm, within the lower end of the previous measured sat- uration vapor pressures of C8–C18 monocarboxylic acids (10−3 ∼10−11 atm) (Cappa et al., 2008; Tao and McMurry, 1989). Donahue et al. (2011) predicted a broad range of The steady-state concentration of all Category II com- pounds is about 8 × 106 cm−3, three times lower than Cat- egory I compound concentrations (∼2.2 × 107 cm−3). The total steady-state concentration of both Category I and II compounds is at least one order of magnitude lower than re- quired for the observed particle growth rates of 28 nm h−1, indicating that they are probably not the dominant species responsible for the growth of larger (> 20 nm) particles. The measured peak concentration of sulfuric acid vapor is about 5 × 105 cm−3, several orders of magnitude lower than the minimum concentration required for particle growth rates of the ﬁrst or second events (see Supplement Sect. III, Ta- ble S5). We therefore conclude that particle growth is not signiﬁcantly inﬂuenced by sulfuric acid vapor in these ex- periments. Atmos. Chem. Phys., 13, 7631–7644, 2013 www.atmos-chem-phys.net/13/7631/2013/ J. Zhao et al.: Dependence of particle nucleation and growth We observed values of J10nm of about 0.42 and 0.038 cm−3 s−1, respectively, for the ﬁrst and second events from the SMPS measurements (Fig. 1a). The nucleation rates are not known because only particles larger than 10 nm were measured. The SMPS measurements showed an initial modal growth rate of about 36 nm h−1 for the ﬁrst hour (between 1 and 2 h in Fig. 1b) after detection of 10 nm particles. Subsequently, the modal growth rate de- creased until it reached a value of about 28 nm h−1. If the growth of initial nuclei (from 1 to 10 nm) follows similar trends, then a growth rate much larger than 36 nm h−1 can be expected for that size range. In addition, the condensation sink is low at the start of particle nucleation. It is hence rea- sonable to assume that the nucleation rate is close to J10nm (McMurry et al., 2005) and falls within the extrapolated up- per limit of the nucleation rate at the measured concentration of residual sulfuric acid vapor in the chamber. volatilities (hence saturation vapor pressures) for oxidized organics containing up to 30 carbons with different func- tional groups. We did not detect signiﬁcant concentrations of carboxylic acid monomers in our gas-phase measure- ments using nitrate dimer ion as the reagent ion; instead, we observed highly oxidized high-molecular-weight com- pounds such as Category I products, predominantly in the 430–560 amu range. Some Category I products are possible dimers of products in a lower mass range (100–300 amu), forming via hydrogen bonding or covalent bonding that loses one water molecule, or other possible formation pathways, while some of them are high-molecular-weight compounds rather than dimers (Fig. S6). However, high mass resolu- tion is needed to conﬁrm the above statement. The current unit mass resolution of the Cluster CIMS cannot deﬁnitely identify the molecular composition of those compounds. The carboxylic acid monomers detected in Winkler et al. (2012) are likely fragments of high-molecular-weight compounds due to thermal decomposition of particulate species dur- ing resistive heating of particles in the TDCIMS measure- ments (Hall and Johnston, 2012b). The highly oxidized Cat- egory I products likely have even lower saturation vapor pres- sures (∼10−14–10−12 atm) than those estimated in Winkler et al. (2012) due to their higher molecular weights and higher oxygen contents. Atmos. Chem. Phys., 13, 7631–7644, 2013 Supplementary material related to this article is available online at: http://www.atmos-chem-phys.net/13/ 7631/2013/acp-13-7631-2013-supplement.pdf. Supplementary material related to this article is available online at: http://www.atmos-chem-phys.net/13/ 7631/2013/acp-13-7631-2013-supplement.pdf. Acknowledgements. We thank National Science Foundation (NSF) award AGS 1068201 for supporting the chamber experiments performed at National Center for Atmospheric Research (NCAR). P.H. McMurray was supported by a Guggenheim Fellowship. The authors acknowledge James Greenburg for PTR-MS mea- surements of the gas-phase precursor and reaction products. The National Center for Atmospheric Research is sponsored by the National Science Foundation. Additional funding for this research was provided by the US Department of Energy grant no. DE-SC0006861. J.S. Smith acknowledges funding from the Finnish Academy grant no. 251007 and US NSF grant no. 0919317. Our experiments were performed with NOx and H2O con- centrations that are signiﬁcantly below typical ambient lev- els. Results from both chamber experiments (Eddingsaas et al., 2012; Lee at al., 2011) and model simulations (Lane et al., 2008) showed that the SOA yields are higher under low- NOx conditions than under high-NOx conditions because of the change in the product distributions. Low-volatility prod- ucts such as carboxylic acids were formed under low-NOx conditions, while more volatile products such as carbonyls, alcohols, and organic nitrates are generated under high-NOx conditions. Similarly, water vapor plays important roles in the product distributions of terpene oxidation and hence af- fects the aerosol mass yields (Jonsson et al., 2008; Fick et al., 2003). It is hence important to understand how humidity and NOx affect the chemical processes responsible for particle nucleation and growth. Edited by: N. M. Donahue Edited by: N. M. Donahue References Barsanti, K. C., McMurry, P. H., and Smith, J. N.: The potential con- tribution of organic salts to new particle growth, Atmos. Chem. Phys., 9, 2949–2957, doi:10.5194/acp-9-2949-2009, 2009. Berndt, T., Stratmann, F., Sipil¨a, M., Vanhanen, J., Pet¨aj¨a, T., Mikkil¨a, J., Gr¨uner, A., Spindler, G., Lee Mauldin III, R., Cur- tius, J., Kulmala, M., and Heintzenberg, J.: Laboratory study on new particle formation from the reaction OH + SO2: inﬂuence of experimental conditions, H2O vapour, NH3 and the amine tert-butylamine on the overall process, Atmos. Chem. Phys., 10, 7101–7116, doi:10.5194/acp-10-7101-2010, 2010. J. Zhao et al.: Dependence of particle nucleation and growth The nucleation-relevant products were found to be mainly in the high molecular weight range, with molecular weights in the 430–560 amu range. The growth-relevant products, which were well correlated with concentrations of particles larger than 20 nm, were in the 140–380 amu range. Measurements from this study suggest that for this chemical system, organic compounds were likely responsible for nucleation, although sulfuric acid vapor cannot be totally ruled out. The effects of NOx and humidity on the formation of Category I products warrant further studies. conditions. The time-dependent concentrations of gas-phase reaction products were measured with the Cluster CIMS us- ing nitrate dimers as reagent ions. The observable products were classiﬁed into two categories, which we found were correlated with particle nucleation (I) and growth (II). A third category (III) was not clearly correlated with either process, but could be lower-molecular-weight decomposition prod- ucts from the oxidation of Category I and II species. The nucleation-relevant products were found to be mainly in the high molecular weight range, with molecular weights in the 430–560 amu range. The growth-relevant products, which were well correlated with concentrations of particles larger than 20 nm, were in the 140–380 amu range. Measurements from this study suggest that for this chemical system, organic compounds were likely responsible for nucleation, although sulfuric acid vapor cannot be totally ruled out. The effects of NOx and humidity on the formation of Category I products warrant further studies. 4 Summary Boy, M., Karl, T., Turnipseed, A., Mauldin, R. L., Kosciuch, E., Greenberg, J., Rathbone, J., Smith, J., Held, A., Barsanti, K., Wehner, B., Bauer, S., Wiedensohler, A., Bonn, B., Kulmala, M., and Guenther, A.: New particle formation in the Front Range of Chamber experiments of α-pinene ozonolysis with ambient levels of precursors were performed under low-NOx and dry J. Zhao et al.: Dependence of particle nucleation and growth 7640 The nitrate dimer ion was used as the reagent ion for E1 and E2 in this study, and it has high selectivity for ioniza- tion of the reaction products. Oxidation products with less electronegativity might not be measured or might be detected with a lower sensitivity (e.g., category II and III) when the nitrate dimer ion is used as the reagent ion. An exploratory experiment (E3, Table S1) with acetate dimer ion as the reagent ion was performed to measure less electronegative compounds. A much higher total steady-state concentration (about 6–8 × 108 cm−3) of all products that arise from α- pinene ozonolysis was measured. The identiﬁcation of those products is not possible given the 1 amu resolution of the Cluster CIMS; however, it can be speculated that those prod- ucts might encompass compounds such as less electroneg- ative carboxylic acids, esters, or even carbonyls that are present in the gas phase with much higher concentrations and that might contribute more signiﬁcantly to particle growth, as observed in measurements by Winkler et al. (2012). Those compounds may contribute to particle growth by particulate- phase reactions such as acid-catalyzed reactions or polymer- ization (e.g., Wang et al., 2010b; Zhao et al., 2006, 2005; Hall and Johnston, 2012a; Hall and Johnston, 2011). Alter- natively, reactions between carboxylic acid and basic gases on or within freshly nucleated particles may also contribute to nanoparticle growth (Barsanti et al., 2009; Smith et al., 2010; Yli-Yuuti et al., 2013). It is likely that more oxidation products would be detected if acetate dimer ion were used as the reagent ion. Contributions of those products to the growth of larger (> 20 nm) particles will be studied in the future. Fu- ture studies will require simultaneous measurements in the gas and particle phases along with models that use such in- formation to explain the dependence of particle growth rate on gas-phase concentrations. conditions. The time-dependent concentrations of gas-phase reaction products were measured with the Cluster CIMS us- ing nitrate dimers as reagent ions. The observable products were classiﬁed into two categories, which we found were correlated with particle nucleation (I) and growth (II). A third category (III) was not clearly correlated with either process, but could be lower-molecular-weight decomposition prod- ucts from the oxidation of Category I and II species. J. Zhao et al.: Dependence of particle nucleation and growth 7641 the Colorado Rocky Mountains, Atmos. Chem. Phys., 8, 1577– 1590, doi:10.5194/acp-8-1577-2008, 2008. the Colorado Rocky Mountains, Atmos. Chem. Phys., 8, 1577– 1590, doi:10.5194/acp-8-1577-2008, 2008. Ehn, M., Kleist, E., Junninen, H., Pet¨aj¨a, T., L¨onn, G., Schobes- berger, S., Dal Maso, M., Trimborn, A., Kulmala, M., Worsnop, D. R., Wahner, A., Wildt, J., and Mentel, Th. F.: Gas phase formation of extremely oxidized pinene reaction products in chamber and ambient air, Atmos. Chem. Phys., 12, 5113–5127, doi:10.5194/acp-12-5113-2012, 2012. Breitner, S., Liu, L. Q., Cyrys, J., Bruske, I., Franck, U., Schlink, U., Leitte, A. M., Herbarth, O., Wiedensohler, A., Wehner, B., Hu, M., Pan, X. C., Wichmann, H. E., and Peters, A.: Sub-micrometer particulate air pollution and cardiovascular mortality in Beijing, China, Sci. Total Environ., 409, 5196–5204, 2011. Erupe, M. E., Viggiano, A. A., and Lee, S.-H.: The effect of trimethylamine on atmospheric nucleation involving H2SO4, At- mos. Chem. Phys., 11, 4767–4775, doi:10.5194/acp-11-4767- 2011, 2011. Bzdek, B. R. and Johnston, M. V.: New Particle Formation and Growth in the Troposphere, Anal. Chem., 82, 7871–7878, 2010. Cahill, T. M., Seaman, V. Y., Charles, M. J., Holzinger, R., and Goldstein, A. H.: Secondary organic aerosols formed from ox- idation of biogenic volatile organic compounds in the Sierra Nevada Mountains of California, J. Geophys. Res., 111, D16312, doi:10.1029/2006JD007178, 2006. Fick, J., Pommer, L., Nilsson, C., and Andersson, B.: Effect of OH radicals, relative humidity, and time on the composition of the products formed in the ozonolysis of alpha-pinene, Atmos. Env- iron., 37, 4087–4096, 2003. Forester, C. D. and Wells, J. R.: Hydroxyl radical yields from reac- tions of terpene mixtures with ozone, Indoor Air, 21, 400–409, doi:10.1111/j.1600-0668.2011.00718.x, 2011. Cappa, C. D., Lovejoy, E. R., and Ravishankara, A. R.: Evaporation rates and vapor pressures of the even-nunibered C8-C18 mono- carboxylic acids, J. Phys. Chem. A, 112, 3959–3964, 2008. Chen, M., Titcombe, M., Jiang, J. K., Jen, C., Kuang, C. A., Fischer, M. L., Eisele, F. L., Siepmann, J. I., Hanson, D. R., Zhao, J., and McMurry, P. H.: Acid-base chemical reaction model for nucle- ation rates in the polluted atmospheric boundary layer, P. Natl. Acad. Sci. USA, 109, 18713–18718, 2012. Franck, U., Odeh, S., Wiedensohler, A., Wehner, B., and Herbarth, O.: The effect of particle size on cardiovascular disorders – The smaller the worse, Sci. Total Environ., 409, 4217-4221, 2011. Guenther, A., Hewitt, C. www.atmos-chem-phys.net/13/7631/2013/ Atmos. Chem. Phys., 13, 7631–7644, 2013 www.atmos-chem-phys.net/13/7631/2013/ J. Zhao et al.: Dependence of particle nucleation and growth N., Erickson, D., Fall, R., Geron, C., Graedel, T., Harley, P., Klinger, L., Lerdau, M., McKay, W. A., Pierce, T., Scholes, B., Steinbrecher, R., Tallamraju, R., Tay- lor, J., and Zimmerman, P.: A global-model of natural volatile organic-compound emissions, J. Geophys. Res., 100, 8873– 8892, 1995. Chen, X. and Hopke, P. K.: Secondary organic aerosol from alpha- pinene ozonolysis in dynamic chamber system, Indoor Air, 19, 335–345, 2009. Chen, X., Hopke, P. K., and Carter, W. P. L.: Secondary organic aerosol from ozonolysis of biogenic volatile organic compounds: chamber studies of particle and reactive oxygen species forma- tion, Environ. Sci. Technol., 45, 276–282, 2011. Hall, W. A. and Johnston, M. V.: Oligomer content of alpha-pinene secondary organic aerosol, Aerosol Sci. Technol., 45, 37–45, 2011. Claeys, M., Iinuma, Y., Szmigielski, R., Surratt, J. D., Blockhuys, F., Van Alsenoy, C., Boge, O., Sierau, B., Gomez-Gonzalez, Y., Vermeylen, R., Van der Veken, P., Shahgholi, M., Chan, A. W. H., Herrmann, H., Seinfeld, J. H., and Maenhaut, W.: Terpenylic acid and related compounds from the oxidation of alpha-pinene: Im- plications for new particle formation and growth above forests, Environ. Sci. Technol., 43, 6976–6982, 2009. Hall, W. A. and Johnston, M. V.: Oligomer formation pathways in secondary organic aerosol from MS and MS/MS measurements with high mass accuracy and resolving power, J. Am. Soc. Mass Spectr., 23, 1097–1108, 2012a. Hall, W. A. and Johnston, M. V.: The thermal-stability of oligomers in alpha-pinene secondary organic aerosol, Aerosol Sci. Tech- nol., 46, 983–989, 2012b. Hao, L. Q., Romakkaniemi, S., Yli-Piril¨a, P., Joutsensaari, J., Ko- rtelainen, A., Kroll, J. H., Miettinen, P., Vaattovaara, P., Tiitta, P., Jaatinen, A., Kajos, M. K., Holopainen, J. K., Heijari, J., Rinne, J., Kulmala, M., Worsnop, D. R., Smith, J. N., and Laaksonen, A.: Mass yields of secondary organic aerosols from the oxida- tion of ?-pinene and real plant emissions, Atmos. Chem. Phys., 11, 1367–1378, doi:10.5194/acp-11-1367-2011, 2011. DePalma, J. W., Bzdek, B. R., Doren, D. J., and Johnston, M. V.: Structure and energetics of nanometer size clusters of sulfuric acid with ammonia and dimethylamine, J. Phys. Chem. A, 116, 1030–1040, 2012. Hao, L. Q., Romakkaniemi, S., Yli-Piril¨a, P., Joutsensaari, J., Ko- rtelainen, A., Kroll, J. H., Miettinen, P., Vaattovaara, P., Tiitta, P., Jaatinen, A., Kajos, M. K., Holopainen, J. K., Heijari, J., Rinne, J., Kulmala, M., Worsnop, D. R., Smith, J. J. Zhao et al.: Dependence of particle nucleation and growth Chem., 3, 133–139, 2011. dio:10.1029/2007JD009260, 2008. dio:10.1029/2007JD009260, 2008. Kuang, C., McMurry, P. H., McCormick, A. V., and Eisele, F. L.: Dependence of nucleation rates on sulfuric acid vapor concen- tration in diverse atmospheric locations, J. Geophys. Res., 113, D10209, doi:10.1029/2007JD009253, 2008. Jiang, J. K., Zhao, J., Chen, M. D., Eisele, F. L., Scheckman, J., Williams, B. J., Kuang, C. A., and McMurry, P. H.: First Mea- surements of Neutral Atmospheric Cluster and 1–2 nm Particle Number Size Distributions During Nucleation Events, Aerosol Sci. Technol., 45, II–V, doi:10.1080/02786826.2010.546817, 2011. Kuang, C., McMurry, P. H., and McCormick, A. V.: Determina- tion of cloud condensation nuclei production from measured new particle formation events, Geophys. Res. Lett., 36, L06812, doi:10.1029/2004GL022092, 2009. Jimenez, J. L., Canagaratna, M. R., Donahue, N. M., Prevot, A. S. H., Zhang, Q., Kroll, J. H., DeCarlo, P. F., Allan, J. D., Coe, H., Ng, N. L., Aiken, A. C., Docherty, K. S., Ulbrich, I. M., Grieshop, A. P., Robinson, A. L., Duplissy, J., Smith, J. D., Wil- son, K. R., Lanz, V. A., Hueglin, C., Sun, Y. L., Tian, J., Laak- sonen, A., Raatikainen, T., Rautiainen, J., Vaattovaara, P., Ehn, M., Kulmala, M., Tomlinson, J. M., Collins, D. R., Cubison, M. J., Dunlea, E. J., Huffman, J. A., Onasch, T. B., Alfarra, M. R., Williams, P. I., Bower, K., Kondo, Y., Schneider, J., Drewnick, F., Borrmann, S., Weimer, S., Demerjian, K., Salcedo, D., Cot- trell, L., Grifﬁn, R., Takami, A., Miyoshi, T., Hatakeyama, S., Shimono, A., Sun, J. Y., Zhang, Y. M., Dzepina, K., Kimmel, J. R., Sueper, D., Jayne, J. T., Herndon, S. C., Trimborn, A. M., Williams, L. R., Wood, E. C., Middlebrook, A. M., Kolb, C. E., Baltensperger, U., and Worsnop, D. R.: Evolution of Organic Aerosols in the Atmosphere, Science, 326, 1525–1529, 2009. Kuang, C., Riipinen, I., Sihto, S.-L., Kulmala, M., McCormick, A. V., and McMurry, P. H.: An improved criterion for new particle formation in diverse atmospheric environments, Atmos. Chem. Phys., 10, 8469–8480, doi:10.5194/acp-10-8469-2010, 2010. Kulmala, M., Vehkamaki, H., Petaja, T., Dal Maso, M., Lauri, A., Kerminen, V. M., Birmili, W., and McMurry, P. H.: Formation and growth rates of ultraﬁne atmospheric particles: a review of observations, J. Aerosol Sci., 35, 143–176, 2004. Kupiainen, O., Ortega, I. K., Kurt´en, T., and Vehkam¨aki, H.: Amine substitution into sulfuric acid – ammonia clusters, Atmos. Chem. Phys., 12, 3591–3599, doi:10.5194/acp-12-3591-2012, 2012. J. Zhao et al.: Dependence of particle nucleation and growth Kurt´en, T., Loukonen, V., Vehkam¨aki, H., and Kulmala, M.: Amines are likely to enhance neutral and ion-induced sulfuric acid-water nucleation in the atmosphere more effectively than ammonia, At- mos. Chem. Phys., 8, 4095–4103, doi:10.5194/acp-8-4095-2008, 2008. Jonsson, A. M., Hallquist, M., and Ljungstrom, E.: Inﬂuence of OH scavenger on the water effect on secondary organic aerosol for- mation from ozonolysis of limonene, Delta(3)-carene, and alpha- pinene, Environ. Sci. Technol., 42, 5938–5944, 2008. Laaksonen, A., Hamed, A., Joutsensaari, J., Hiltunen, L., Cavalli, F., Junkermann, W., Asmi, A., Fuzzi, S., and Facchini, M. C.: Cloud condensation nucleus production from nucleation events at a highly polluted region, Geophys. Res. Lett., 32, L06812, doi:10.1029/2004GL022092, 2005. Kamens, R. M., Gery, M. W., Jeffries, H. E., Jackson, M., and Cole, E. I.: Ozone-isoprene reactions-product formation and aerosol potential, Int. J. Chem. Kinet., 14, 955–975, 1982. Lane, T. E., Donahue, N. M., and Pandis, S. N.: Effect of NOx on secondary organic aerosol concentrations, Environ. Sci. Tech- nol., 42, 6022–6027, 2008. Kerminen, V. M., Lihavainen, H., Komppula, M., Viisanen, Y., and Kulmala, M.: Direct observational evidence linking atmospheric aerosol formation and cloud droplet activation, Geophys. Res. Lett., 32, L14803, doi:10.1029/2005GL023130, 2005. Lee, A., Goldstein, A. H., Keywood, M. D., Gao, S., Varutbangkul, V., Bahreini, R., Ng, N. L., Flagan, R. C., and Seinfeld, J. H.: Gas-phase products and secondary aerosol yields from the ozonolysis of ten different terpenes, J. Geophys. Res., 111, D17305, doi:10.1029/2006JD007050, 2006. Kirkby, J., Curtius, J., Almeida, J., Dunne, E., Duplissy, J., Ehrhart, S., Franchin, A., Gagne, S., Ickes, L., Kurten, A., Kupc, A., Met- zger, A., Riccobono, F., Rondo, L., Schobesberger, S., Tsagkoge- orgas, G., Wimmer, D., Amorim, A., Bianchi, F., Breitenlechner, M., David, A., Dommen, J., Downard, A., Ehn, M., Flagan, R. C., Haider, S., Hansel, A., Hauser, D., Jud, W., Junninen, H., Kreissl, F., Kvashin, A., Laaksonen, A., Lehtipalo, K., Lima, J., Love- joy, E. R., Makhmutov, V., Mathot, S., Mikkila, J., Minginette, P., Mogo, S., Nieminen, T., Onnela, A., Pereira, P., Petaja, T., Schnitzhofer, R., Seinfeld, J. H., Sipila, M., Stozhkov, Y., Strat- mann, F., Tome, A., Vanhanen, J., Viisanen, Y., Vrtala, A., Wag- ner, P. E., Walther, H., Weingartner, E., Wex, H., Winkler, P. M., Carslaw, K. S., Worsnop, D. R., Baltensperger, U., and Kulmala, M.: Role of sulphuric acid, ammonia and galactic cosmic rays in atmospheric aerosol nucleation, Nature, 476, 429–433, 2011. Lee, B. H., Pierce, J. J. Zhao et al.: Dependence of particle nucleation and growth N., and Laaksonen, A.: Mass yields of secondary organic aerosols from the oxida- tion of ?-pinene and real plant emissions, Atmos. Chem. Phys., 11, 1367–1378, doi:10.5194/acp-11-1367-2011, 2011. Donahue, N. M., Epstein, S. A., Pandis, S. N., and Robinson, A. L.: A two-dimensional volatility basis set: 1. organic-aerosol mixing thermodynamics, Atmos. Chem. Phys., 11, 3303–3318, doi:10.5194/acp-11-3303-2011, 2011. Heaton, K. J., Dreyfus, M. A., Wang, S., and Johnston, M. V.: Oligomers in the early stage of biogenic secondary organic aerosol formation and growth, Environ. Sci. Technol., 41, 6129– 6136, 2007. p Donahue, N. M., Henry, K. M., Mentel, T. F., Kiendler-Scharr, A., Spindler, C., Bohn, B., Brauers, T., Dorn, H. P., Fuchs, H., Till- mann, R., Wahner, A., Saathoff, H., Naumann, K. H., Mohler, O., Leisner, T., Muller, L., Reinnig, M. C., Hoffmann, T., Salo, K., Hallquist, M., Frosch, M., Bilde, M., Tritscher, T., Barmet, P., Praplan, A. P., DeCarlo, P. F., Dommen, J., Prevot, A. S. H., and Baltensperger, U.: Aging of biogenic secondary organic aerosol via gas-phase OH radical reactions, P. Natl. Acad. Sci. USA, 109, 13503–13508, 2012. Herckes, P., Engling, G., Kreidenweis, S. M., and Collett, J. L.: Particle size distributions of organic aerosol constituents during the 2002 Yosemite Aerosol Characterization Study, Environ. Sci. Technol., 40, 4554–4562, 2006. Hyder, M., Genberg, J., Sandahl, M., Swietlicki, E., and Jonsson, J. A.: Yearly trend of dicarboxylic acids in organic aerosols from south of Sweden and source attribution, Atmos. Environ., 57, 197–204, 2012. Eddingsaas, N. C., Loza, C. L., Yee, L. D., Chan, M., Schilling, K. A., Chhabra, P. S., Seinfeld, J. H., and Wennberg, P. O.: α-pinene photooxidation under controlled chemical conditions –Part 2: SOA yield and composition in low- and high-NOx environments, Atmos. Chem. Phys., 12, 7413–7427, doi:10.5194/acp-12-7413- 2012, 2012. Iida, K., Stolzenburg, M. R., McMurry, P. H., and Smith, J. N.: Estimating nanoparticle growth rates from size- dependent charged fractions: Analysis of new particle forma- tion events in Mexico City, J. Geophys. Res., 113, D05207, Atmos. Chem. Phys., 13, 7631–7644, 2013 www.atmos-chem-phys.net/13/7631/2013/ 7642 J. Zhao et al.: Dependence of particle nucleation and growth 7643 VanReken, T. M., Greenberg, J. P., Harley, P. C., Guenther, A. B., and Smith, J. N.: Direct measurement of particle formation and growth from the oxidation of biogenic emissions, Atmos. Chem. Phys., 6, 4403–4413, doi:10.5194/acp-6-4403-2006, 2006. Nadykto, A. B., Yu, F. Q., Jakovleva, M. V., Herb, J., and Xu, Y. S.: Amines in the Earth’s Atmosphere: A density functional theory study of the thermochemistry of pre-nucleation clusters, Entropy, 13, 554–569, 2011. Paasonen, P., Nieminen, T., Asmi, E., Manninen, H. E., Pet¨aj¨a, T., Plass-D¨ulmer, C., Flentje, H., Birmili, W., Wiedensohler, A., H˜orrak, U., Metzger, A., Hamed, A., Laaksonen, A., Facchini, M. C., Kerminen, V.-M., and Kulmala, M.: On the roles of sul- phuric acid and low-volatility organic vapours in the initial steps of atmospheric new particle formation, Atmos. Chem. Phys., 10, 11223–11242, doi:10.5194/acp-10-11223-2010, 2010. Veres, P., Roberts, J. M., Warneke, C., Welsh-Bon, D., Zahniser, M., Herndon, S., Fall, R., and de Gouw, J.: Development of negative- ion proton-transfer chemical-ionization mass spectrometry (NI- PT-CIMS) for the measurement of gas-phase organic acids in the atmosphere, Int. J. Mass Spectr., 274, 48–55, 2008. Veres, P., Roberts, J. M., Burling, I. R., Warneke, C., de Gouw, J., and Yokelson, R. J.: Measurements of gas-phase inorganic and organic acids from biomass ﬁres by negative-ion proton-transfer chemical-ionization mass spectrometry, J. Geophys. Res., 115, D23302, doi:10.1029/2010JD014033, 2010. Paulson, S. E., Chung, M., Sen, A. D., and Orzechowska, G.: Mea- surement of OH radical formation from the reaction of ozone with several biogenic alkenes, J. Geophys. Res., 103, 25533– 25539, 1998. Viitanen, A. K., Saukko, E., Virtanen, A., Yli-Pirila, P., Smith, J. N., Joutsensaari, J., and Makela, J. M.: Ion mobility distributions during the initial stages of new particle formation by the ozonoly- sis of alpha-pinene, Environ. Sci. Technol., 44, 8917–8923, 2010. Perraud, V., Bruns, E. A., Ezell, M. J., Johnson, S. N., Yu, Y., Alexander, M. L., Zelenyuk, A., Imre, D., Chang, W. L., Dab- dub, D., Pankow, J. F., and Finlayson-Pitts, B. J.: Nonequilibrium atmospheric secondary organic aerosol formation and growth, P. Natl. Acad. Sci. USA, 109, 2836–2841, 2012. Wagener, S., Langner, M., Hansen, U., Moriske, H. J., and Endlicher, W. R.: Spatial and seasonal variations of biogenic tracer compounds in ambient PM10 and PM1 samples in Berlin, Germany, Atmos. Environ., 47, 33–42, 2011. J. Zhao et al.: Dependence of particle nucleation and growth Riipinen, I., Sihto, S.-L., Kulmala, M., Arnold, F., Dal Maso, M., Birmili, W., Saarnio, K., Teinil¨a, K., Kerminen, V.-M., Laak- sonen, A., and Lehtinen, K. E. J.: Connections between atmo- spheric sulphuric acid and new particle formation during QUEST III–IV campaigns in Heidelberg and Hyyti¨al¨a, Atmos. Chem. Phys., 7, 1899–1914, doi:10.5194/acp-7-1899-2007, 2007. Wang, L., Lal, V., Khalizov, A. F., and Zhang, R. Y.: Heterogeneous chemistry of alkylamines with sulfuric acid: Implications for atmospheric formation of alkylaminium sulfates, Environ. Sci. Technol., 44, 2461–2465, 2010a. Wang, L., Khalizov, A.F., Zheng, J., Xu, W., Ma, Y., Lal, V., and Zhang, R.Y.: Atmospheric nanoparticles formed from het- erogeneous reactions of organics, Nat. Geosci., 3, 238–242, doi:10.1038/NGEO778, 2010b. Siese, M., Becker, K. H., Brockmann, K. J., Geiger, H., Hofzuma- haus, A., Holland, F., Mihelcic, D., and Wirtz, K.: Direct mea- surement of OH radicals from ozonolysis of selected alkenes: A EUPHORE simulation chamber study, Environ. Sci. Technol., 35, 4660–4667, 2001. Weber, R. J., Marti, J. J., McMurry, P. H., Eisele, F. L., Tanner, D. J., and Jefferson, A.: Measured atmospheric new particle forma- tion rates: Implications for nucleation mechanisms, Chem. Eng. Commun., 151, 53–64, 1996. Sihto, S.-L., Kulmala, M., Kerminen, V.-M., Dal Maso, M., Pet¨aj¨a, T., Riipinen, I., Korhonen, H., Arnold, F., Janson, R., Boy, M., Laaksonen, A., and Lehtinen, K. E. J.: Atmospheric sulphuric acid and aerosol formation: implications from atmospheric mea- surements for nucleation and early growth mechanisms, At- mos. Chem. Phys., 6, 4079–4091, doi:10.5194/acp-6-4079-2006, 2006. Weber, R. J., Marti, J. J., McMurry, P. H., Eisele, F. L., Tanner, D. J., and Jefferson, A.: Measurements of new particle formation and ultraﬁne particle growth rates at a clean continental site, J. Geophys. Res., 102, 4375–4385, 1997. Weber, R. J., Chen, G., Davis, D. D., Mauldin, R. L., Tanner, D. J., Eisele, F. L., Clarke, A. D., Thornton, D. C., and Bandy, A. R.: Measurements of enhanced H2SO4 and 3–4 nm particles near a frontal cloud during the First Aerosol Characterization Experi- ment (ACE 1), J. Geophys. Res., 106, 24107–24117, 2001. Smith, J. N., Barsanti, K. C., Friedli, H. R., Ehn, M., Kulmala, M., Collins, D. R., Scheckman, J. H., Williams, B. J., and McMurry, P. H.: Observations of aminium salts in atmospheric nanoparti- cles and possible climatic implications, P. Natl. Acad. Sci. USA, 107, 6634–6639, 2010. Went, F. W.: Blue hazes in the atmosphere, Nature, 187, 641–643, 1960. Stolzenburg, M. R., McMurry, P. J. Zhao et al.: Dependence of particle nucleation and growth R., Engelhart, G. J., and Pandis, S. N.: Volatil- ity of secondary organic aerosol from the ozonolysis of monoter- penes, Atmos. Environ., 45, 2443–2452, 2011. McGraw, R. and Saunders, J. H.: A Condensation Feedback Mech- anism for Oscillatory Nucleation and Growth, Aerosol Sci. Tech- nol., 3, 367–380, 1984. McMurry, P. H., Fink, M., Sakurai, H., Stolzenburg, M. R., Mauldin, R. L., Smith, J., Eisele, F., Moore, K., Sjostedt, S., Tanner, D., Huey, L. G., Nowak, J. B., Edgerton, E., and Voisin, D.: A criterion for new particle formation in the sulfur-rich Atlanta atmosphere, J. Geophys. Res., 110, D22S02, doi:10.1029/2005JD005901, 2005. Kroll, J. H., Sahay, S. R., Anderson, J. G., Demerjian, K. L., and Donahue, N. M.: Mechanism of HOx formation in the gas-phase ozone-alkene reaction. 2. Prompt versus thermal dissociation of carbonyl oxides to form OH, J. Phys. Chem. A, 105, 4446–4457, 2001. Merikanto, J., Spracklen, D. V., Mann, G. W., Pickering, S. J., and Carslaw, K. S.: Impact of nucleation on global CCN, At- mos. Chem. Phys., 9, 8601–8616, doi:10.5194/acp-9-8601-2009, 2009. Kroll, J. H., Donahue, N. M., Jimenez, J. L., Kessler, S. H., Cana- garatna, M. R., Wilson, K. R., Altieri, K. E., Mazzoleni, L. R., Wozniak, A. S., Bluhm, H., Mysak, E. R., Smith, J. D., Kolb, C. E., and Worsnop, D. R.: Carbon oxidation state as a metric for describing the chemistry of atmospheric organic aerosol, Nat. Metzger, A., Verheggen, B., Dommen, J., Duplissy, J., Prevot, A. S. H., Weingartner, E., Riipinen, I., Kulmala, M., Spracklen, D. V., Carslaw, K. S., and Baltensperger, U.: Evidence for the role of organics in aerosol particle formation under atmospheric condi- tions, P. Natl. Acad. Sci. USA, 107, 6646–6651, 2010. Atmos. Chem. Phys., 13, 7631–7644, 2013 www.atmos-chem-phys.net/13/7631/2013/ www.atmos-chem-phys.net/13/7631/2013/ J. Zhao et al.: Dependence of particle nucleation and growth H., Sakurai, H., Smith, J. N., Mauldin, R. L., Eisele, F. L., and Clement, C. F.: Growth rates of freshly nucleated atmospheric particles in Atlanta, J. Geophys. Res., 110, D22S05, doi:10.1029/2005JD005935, 2005. Winkler, P. M., Ortega, J., Karl, T., Cappellin, L., Friedli, H. R., Barsanti, K., McMurry, P. H., and Smith, J. N.: Iden- tiﬁcation of the biogenic compounds responsible for size- dependent nanoparticle growth, Geophys. Res. Lett., 39, L20815, doi:10.1029/2012GL053253, 2012. Szmigielski, R., Surratt, J. D., Gomez-Gonzalez, Y., Van der Veken, P., Kourtchev, I., Vermeylen, R., Blockhuys, F., Jaoui, M., Klein- dienst, T. E., Lewandowski, M., Offenberg, J. H., Edney, E. O., Seinfeld, J. H., Maenhaut, W., and Claeys, M.: 3-methyl- 1,2,3-butanetricarboxylic acid: An atmospheric tracer for terpene secondary organic aerosol, Geophys. Res. Lett., 34, L24811, doi:10.1029/2007GL031338, 2007. Yasmeen, F., Vermeylen, R., Szmigielski, R., Iinuma, Y., B¨oge, O., Herrmann, H., Maenhaut, W., and Claeys, M.: Terpenylic acid and related compounds: precursors for dimers in secondary or- ganic aerosol from the ozonolysis of α- and β-pinene, Atmos. Chem. Phys., 10, 9383–9392, doi:10.5194/acp-10-9383-2010, 2010. Tao, Y. and McMurry, P. H.: Vapor-pressure and surface free- energies of C14-C18 monocarboxylic acids and C5-dicarboxylic and C6-dicarboxylic acids, Environ. Sci. Technol., 23, 1519– 1523, 1989. Yli-Juuti, T., Barsanti, K., Hildebrandt Ruiz, L., Kieloaho, A.-J., Makkonen, U., Pet¨aj¨a, T., Ruuskanen, T., Kulmala, M., and Riip- inen, I.: Model for acid-base chemistry in nanoparticle growth Atmos. Chem. Phys., 13, 7631–7644, 2013 www.atmos-chem-phys.net/13/7631/2013/ J. Zhao et al.: Dependence of particle nucleation and growth 7644 (MABNAG), Atmos. Chem. Phys. Discuss., 13, 7175–7222, doi:10.5194/acpd-13-7175-2013, 2013. Zhang, R. Y., Khalizov, A., Wang, L., Hu, M., and Xu, W.: Nucle- ation and growth of nanoparticles in the atmosphere, Chem. Rev., 112, 1957–2011, 2012. Yu, F. and Luo, G.: Simulation of particle size distribution with a global aerosol model: contribution of nucleation to aerosol and CCN number concentrations, Atmos. Chem. Phys., 9, 7691– 7710, doi:10.5194/acp-9-7691-2009, 2009. Zhao, J., Levitt, N. P., and Zhang, R.: Heterogeneous chemistry of octanal and 2, 4-hexadienal with sulfuric acid, Geophys. Res. Lett., 32, L09802, doi:10.1029/2004GL022200, 2005. Zhao, J., Levitt, N. P., Zhang, R., and Chen, J.: Heterogeneous re- actions of methylglyoxal in acidic media: implication for sec- ondary organic aerosol formation, Environ. Sci. Technol., 40, 7682–7687, 2006. Yu, H., McGraw, R., and Lee, S. H.: Effects of amines on formation of sub-3 nm particles and their subsequent growth, Geophys. Res. Lett., 39, L02807, doi:10.1029/2011GL050099, 2012. Yu, J. Z., Cocker, D. R., Grifﬁn, R. J., Flagan, R. C., and Seinfeld, J. H.: Gas-phase ozone oxidation of monoterpenes: Gaseous and particulate products, J. Atmos. Chem., 34, 207–258, 1999. Zhao, J., Khalizov, A., Zhang, R. Y., and McGraw, R.: Hydrogen- Bonding Interaction in Molecular Complexes and Clusters of Aerosol Nucleation Precursors, J. Phys. Chem. A, 113, 80–689, 2009. Zhang, Q., Jimenez, J. L., Canagaratna, M. R., Allan, J. D., Coe, H., Ulbrich, I., Alfarra, M. R., Takami, A., Middlebrook, A. M., Sun, Y. L., Dzepina, K., Dunlea, E., Docherty, K., De- Carlo, P. F., Salcedo, D., Onasch, T., Jayne, J. T., Miyoshi, T., Shimono, A., Hatakeyama, S., Takegawa, N., Kondo, Y., Schneider, J., Drewnick, F., Borrmann, S., Weimer, S., Demer- jian, K., Williams, P., Bower, K., Bahreini, R., Cottrell, L., Grifﬁn, R. J., Rautiainen, J., Sun, J. Y., Zhang, Y. M., and Worsnop, D. R.: Ubiquity and dominance of oxygenated species in organic aerosols in anthropogenically-inﬂuenced Northern Hemisphere midlatitudes, Geophys. Res. Lett., 34, L13801, doi:10.1029/2007GL029979, 2007. Zhao, J., Eisele, F. L., Titcombe, M., Kuang, C. G., and McMurry, P. H.: Chemical ionization mass spectrometric measurements of atmospheric neutral clusters using the cluster-CIMS, J. Geophys. Res., 115, D08205, doi:10.1029/2009JD012606, 2010. Zhao, J., Smith, J. N., Eisele, F. L., Chen, M., Kuang, C., and McMurry, P. H.: Observation of neutral sulfuric acid-amine containing clusters in laboratory and ambient measurements, Atmos. Chem. Phys., 11, 10823–10836, doi:10.5194/acp-11- 10823-2011, 2011. Zollner, J. H., Glasoe, W. A., Panta, B., Carlson, K. J. Zhao et al.: Dependence of particle nucleation and growth K., McMurry, P. H., and Hanson, D. R.: Sulfuric acid nucleation: power de- pendencies, variation with relative humidity, and effect of bases, Atmos. Chem. Phys., 12, 4399–4411, doi:10.5194/acp-12-4399- 2012, 2012. Zhang, R. Y., Suh, I., Zhao, J., Zhang, D., Fortner, E. C., Tie, X. X., Molina, L. T., and Molina, M. J.: Atmospheric new particle formation enhanced by organic acids, Science, 304, 1487–1490, 2004. Zhang, R. Y., Wang, L., Khalizov, A. F., Zhao, J., Zheng, J., Mc- Graw, R. L., and Molina, L. T.: Formation of nanoparticles of blue haze enhanced by anthropogenic pollution, P. Natl. Acad. Sci. USA, 106, 17650–17654, 2009. Atmos. Chem. Phys., 13, 7631–7644, 2013 www.atmos-chem-phys.net/13/7631/2013/ www.atmos-chem-phys.net/13/7631/2013/
https://openalex.org/W4376614281	https://pubs.rsc.org/en/content/articlepdf/2023/cc/d3cc90158k	English	null	Correction: One-step mild preparation of tough and thermo-reversible poly(vinyl alcohol) hydrogels induced by small molecules	Chemical communications	2,023	cc-by	395	Chuang Dong,† Jiahua Zhou,† Dongjian Shi, Yufang Song, Xi Yu, Weifu Dong, Mingqing Chen* and Daisaku Kaneko* Correction for ‘One-step mild preparation of tough and thermo-reversible poly(vinyl alcohol) hydrogels induced by small molecules’ by Chuang Dong et al., Chem. Commun., 2021, 57, 3789–3792, https:// doi.org/10.1039/D1CC00578B. DOI: 10.1039/d3cc90158k rsc.li/chemcomm rsc.li/chemcomm The authors regret that a related reference was missing in the original article. On page 3791, right column, second paragraph, which reads: ‘‘This red shift of the -OH groups suggested the enhancement of the hydrogen bonding,36,37’’, a new reference should be added as reference 38 (shown as ref. 1 below). This work reported improved PVA hydrogels by the use of small molecules containing phenolic groups. which reads: ‘‘This red shift of the -OH groups suggested the enhancement of the hydrogen bonding,36,37’’, a new reference should be added as reference 38 (shown as ref. 1 below). This work reported improved PVA hydrogels by the use of small molecules containing phenolic groups. be added as reference 38 (shown as ref. 1 below). This work reported improved PVA hydrogels by the use of small molecules containing phenolic groups. g p g p The Royal Society of Chemistry apologises for these errors and any consequent inconvenience to authors and readers. CORRECTION View Article Online View Journal \| View Issue The Key Laboratory of Synthetic and Biological Colloids, Ministry of Education, School of Chemical and Material Engineering, Jiangnan University, Wuxi, Jiangsu 214122, China. E-mail: mqchen@jiangnan.edu.cn, daisaku@jiangnan.edu.cn † These authors contributed equally. Open Access Article. Published on 15 May 2023. Downloaded on 10/24/2024 4:56:21 This article is licensed under a Creative Commons Attribution 3.0 Unport Chuang Dong,† Jiahua Zhou,† Dongjian Shi, Yufang Song, Xi Yu, Weifu Dong, Mingqing Chen* and Daisaku Kaneko* ChemComm Open Access Article. Published on 15 May 2023. Downloaded on 10/24/2024 4:56:21 AM. This article is licensed under a Creative Commons Attribution 3.0 Unported Licence. Cite this: Chem. Commun., 2023, 59, 6422 Cite this: Chem. Commun., 2023, 59, 6422 References 1 E. M. Euti, A. Wolfel, M. L. Picchio, M. R. Romero, M. Martinelli, R. J. Minari and C. I. A. Igarzabal, Macromol. R 1 E. M. Euti, A. Wolfel, M. L. Picchio, M. R. Romero, M. Martinelli, R. J. Minari and C. I. A. Igarzabal, Macromol. Rapid Commun., 2019, 40, 1900217. 6422 \| Chem. Commun., 2023, 59, 6422 This journal is © The Royal Society of Chemistry 2023
https://openalex.org/W2993857569	https://www.iiste.org/Journals/index.php/JEDS/article/download/47992/49576	English	null	Money, Output and Inflation in Bangladesh: A Test of Cointegration and Causality	Journal of Economics and Sustainable Development	2,019	cc-by	6,870	iste.org iste.org www.iiste.org Journal of Economics and Sustainable Development ISSN 2222-1700 (Paper) ISSN 2222-2855 (Online) Vol.10, No.10, 2019 DOI: 10.7176/JESD Abstract This paper examines the relative impacts of money on economic growth and inflation in the economy of Bangladesh. It uses monthly data of Consumer Price Index, M2 multiplier, and industrial production index of Bangladesh from the Global Economic Monitor database of the World Bank. Using Johansen Co-integration and Vector Error Correction framework, the study finds that there is one cointegration relationship among these three variables. The Granger Causality results suggest that increased money multiplier has significant impacts on economic growth but non-significant impacts on inflation thus rejecting the neutrality of money in the Bangladesh economy over the study period. The results also suggest that the impact of money is unidirectional on economic growth. Keywords: neutrality of money, economic growth, cointegration, causality DOI: 10.7176/JESD/10-10-19 Publication date:May 31st 2019 DOI: 10.7176/JESD/10-10-19 Publication date:May 31st 2019 Publication date:May 31st 2019 Money, Output and Inflation in Bangladesh: A Test of Cointegration and Causality Md. Shahnewaz Khan Assistant Professor, Department of Economics, Jatiya Kabi Kazi Nazrul Islam University, Trishal, Mymensingh, Bangladesh Ab t t 1. Introduction The role of money in determining output and price occupies a central position in macroeconomic literature. There has been a great deal of controversy regarding the influence of monetary policy and the trade-offs associated with increased money supply on real economic growth and macroeconomic stabilization. The Classical view holds that a change in the money supply is associated with a corresponding change in the price level. According to Keynesian economics, expansionary monetary policy is an important tool to increase the effective demand by increasing spare capacity utilization. Monetarists, on the other hand, deny the role of policies and insist on the long run neutrality of money (Froyen 2005; Crowder 1998; Rangarajan & Arif 1990; Bannock et al. 2009). The prime focus of this paper is to examine the relative influence of money in raising the level of real output and inflation in Bangladesh. In other words, the objective is to examine if inflation is a monetary phenomena and if money is neutral in the long run in the economy of Bangladesh. 2.1 Context Monetarists view that the supply of money has predominant influence on the movement of nominal income. In the short run, it has substantial effect on the real economic variables like output and employment, although it has no impact on the real variables in the long run except raising the price level. They also view that the instability in the economy is caused by government policies. Thus they suggest a rule based monetary policy for attaining a stable economy. On the other hand, Keynesians propose for a discretionary monetary policy in order to attain high economic growth, particularly if the economy is characterized by less than full employment equilibrium (Froyen 2005; Crowder 1998; Bannok et al. 2009). Monetarist prediction about the role of money in the current economic activities has led to the development of a good number of empirical studies concerning the direction of causality between money and income. Using Post- War quarterly data (1947-1969) of money stock and current dollar GNP, Sims (1972) found unidirectional causality running from money to income. He argued that the usual practice of using money as exogenous in the distributed lag regression of income on money in the Quantity Theory framework is justified. The results of Sims support the monetarist view that the stock of money is the principal determinant of current economic activity in an economy. Using the same methodology developed by Sims (1972), Barth & Bennett (1974) conducted an empirical test of the causal link of money and income for the Canadian economy over the period 1957-1972. In contrast to Sims, they did not find any evidence of unidirectional causality running from money to income or from income to money. Rather, they found evidence of feedback relationship between money and income thus suggesting that money is not the principal agent in determining the real economic activity. Another study by Williams et al. (1976) also expressed doubt about the presence of unidirectional causality from money to income. Using UK data since 1958- I to 1971-III, they found some evidence of unidirectional causality from nominal income to money and also from money to prices. They suggested that the direction of causality is less clear-cut than which was suggested by Sims and that there could be a possibility of simultaneous causality rather than the unidirectional causality. 2.2 Empirical Studies – Bangladesh 2.2 Empirical Studies – Bangladesh To analyze the cause of inflation in Bangladesh, Taslim (1982) conducted a study in line with the monetarist- structuralist controversy. The study is perhaps the earliest in its genre in Bangladesh. He constructed three empirical models to test the cause of inflation in Bangladesh - a purely structuralist model which analyzed the role of agricultural and foreign exchange bottlenecks in creating inflation; a purely monetarist model which analysed the impact of monetary aggregate measures on inflation; and a hybrid model that combines both the structural constraints and money supply into the equation and then analyzed their respective roles to create inflation. He argued that neither the pure monetarist nor the pure structuralist model completely explain the inflation in Bangladesh and the hybrid model performs best, thus suggesting the relevance of both structural and monetary factors explaining inflation in Bangladesh. Jones & Sattar (1988) conducted a study using data from 1974-1985 to analyze the relationship between money, inflation and output in Bangladesh. They used three variables – (a) change in the money supply (M1 and M2); (b) change in the CPI (base year 1980) to represent inflation; and (c) change in the industrial production mix (base year 1980) to represent output. Using a Granger Causality framework, they suggested that the inflation in Bangladesh is not a purely monetary phenomenon and that the non-monetary factors have significant impacts on the inflation in Bangladesh. They also suggested that the money is not neutral in the economy of Bangladesh. Using quarterly data (1974-92), Chowdhury et al. (1995) claimed that the inflationary process in Bangladesh cannot be explained solely by the monetarist or structuralist explanation. Their multivariate VAR system included the following variables – M1 and M2 as the monetary aggregates, industrial production index as the output variable, consumer price index as the inflation measure, and the foreign exchange rate variable. They did not find any clear- cut cause and effect relationship between money supply and inflation in Bangladesh. They claimed that the monetary policy in Bangladesh has a significant unidirectional causal impacts on the real output. Mujeri et al. (2009) conducted another study to measure the inflationary pressures in Bangladesh. They used annualized data of real GDP, changes in the consumer price index, and broad money (M2) during 1980-2008 time periods. 2.1 Context 155 Journal of Economics and Sustainable Development www.iiste.org ISSN 2222-1700 (Paper) ISSN 2222-2855 (Online) DOI: 10.7176/JESD Vol.10, No.10, 2019 There are two main schools of thought that explain the causes of inflation in the developing countries. The structuralists argue that the prime cause of inflation in the developing countries is the structural constraints of their economies, such as agricultural weaknesses, persistent balance of payment deficits, and other structural rigidities (market imperfections, socio-economic and political constraints, etc.). In contrast, the monetarists view that inflation is essentially a monetary phenomenon. This view is based on the Quantity Theory of Money which suggests that money and inflation has a one-to-one correspondence (Taslim 1982; Bilquees 1988; Chowdhury et al. 1995). They view that inflation results not from the structural constraints of these countries but from the inappropriate monetary and fiscal policies like government deficit financing, expansionary credit policies, etc. (Taslim 1982). Rami (2010) examined the causal relationship between money, prices, and output in India using annualized data from 1951 to 2005. He applied the Granger Causality approach and claimed that the Indian case strongly validates the monetarist prediction that money plays an active role in determining nominal income and prices. Using quarterly data of real GDP and Consumer Price Index (1975:I-2003:IV), Kemal (2006) found that the inflation in Pakistan is a monetary phenomenon. Malik (2006) also found similar result in the case of Pakistan. 2.2 Empirical Studies – Bangladesh 3. Methodology and Data (2) where α1 indicates the short run impact to restore equilibrium and α2 indicates the speed of adjustment at which the equilibrium is restored. The idea of the ECM can be extended to a vector of multiple variables, known as the Vector Error Correction Model (VECM). In this paper, I have used multivariate Cointegration and Vector Error Correction Model (VECM) to test the long and short run relationships between money supply, economic growth, and inflation in Bangladesh. First, I check for the stationary of these three variables and check whether they are white noise (stationary) or unit root (non- stationary) process. Second, I test the existence of cointegration and the number of cointegrating relationships in the model. Third, I run the unrestricted and restricted vector error correction model. Finally, I use the Granger causality technique to see the direction of causality between money, growth, and inflation. 2.2 Empirical Studies – Bangladesh Consistent with the findings of the new open economy macroeconomics (NOEM) literature, the study found that monetary policy shocks has immediate effect on interest rate and exchange rate in Bangladesh; but it has effects on industrial production with a lag of over half a year and on inflation with a lag of more than one year. 3. Methodology and Data The study uses monthly data of Consumer Price Index (CPI), M2 multiplier and industrial production of Bangladesh during June1995 – July 2009 from the Global Economic Monitor database of the World Bank1. The CPI data is a widely used measure of inflation which is seasonality adjusted. The money (M2) multiplier is defined as the broad money to monetary base ratio, which indicates the ability of the commercial banks to create money. A high value of the index implies an increase of money supply and vice-versa. The industrial production data is seasonality adjusted at constant 2010 US $. The index is used as a measure of the real economic activity thus an increase in the index shows the growth of the real sector. I have used the log specifications of each variable. The three variables are defined as – ‘lcpi’ for log of CPI; ‘lM2’ for log of M2 multiplier; and ‘lipro’ for log of industrial production. The problem in dealing with the time series regression is that most time series are non-stationary (i.e., they exhibits time varying mean, variance, and covariance). In such cases, applying simple OLS leads to spurious regressions. Establishing valid relationship among different non-stationary time series requires the existence of cointegration among them. If two time series yt and xt are non-stationary but their linear combination is stationary, then we can say that these two series are cointegrated. Consider the following model – = + + … … … … . (1) where yt and xt represent two non-stationary time series variables and ut is white noise error term. Now if ut is stationary, then its linear combination (yt - α - βxt) will also be stationary; in other words, we can say that yt and xt are cointegrated. If there is a cointegrating relationship across variables, we can say that there is a valid long run relationship among them. Then their short run relationships (the deviation from the long run equilibrium) can be expressed in terms of an error correction model (Engle & Granger 1987; Gujarati et al. 2009). The Error Correction Model (ECM) of these two variables can be represented as follows – ∆ = + ∆ + ( − −) + … … … … … . 2.2 Empirical Studies – Bangladesh Their empirical work tested the forecasting performance of the P* model for the economy of Bangladesh which combines the long run determinants of price level based on the classical quantity theory of money and short run changes in inflation. They highlighted the significance of the non-monetary factors in explaining the inflation in Bangladesh. Using annualized data of money supply (M2) and income (nominal GDP) during 1974-2008, Hossain (2011) concluded that there is a stable long term (co-integrated) relationship between money and income in Bangladesh. Using an error correction approach, the study also found evidence of feedback relationship between these two variables in the short run. The study highlighted the need for undertaking monetary policy for achieving output goals in an economy like Bangladesh. Another study by Bhuiyan (2012) estimated the effects of monetary policy in Bangladesh using an open economy Bayesian structural VAR approach and using monthly data covering the time periods 1994-2009. They argued that some external variables like remittances, exports, oil prices, and non-fuel commodity prices influence the monetary policy and the real economy of Bangladesh despite the fact that Bangladesh is a small open economy. Their structural VAR approach identified a monetary policy function of Bangladesh and allowed simultaneous interactions of a number of domestic and foreign variables comprising the following variables – (i) monetary aggregate (M2); (ii) three months treasury bills rate; (iii) nominal exchange rate of Bangladeshi Taka against US 156 Journal of Economics and Sustainable Development www.iiste.org iste.org DOI: 10.7176/JESD dollar; (iv) monthly consumer price index; (v) industrial production index; (vi) remittances; (vii) exports; (viii) non-fuel commodity prices; and (ix) oil prices. Consistent with the findings of the new open economy macroeconomics (NOEM) literature, the study found that monetary policy shocks has immediate effect on interest rate and exchange rate in Bangladesh; but it has effects on industrial production with a lag of over half a year and on inflation with a lag of more than one year. dollar; (iv) monthly consumer price index; (v) industrial production index; (vi) remittances; (vii) exports; (viii) non-fuel commodity prices; and (ix) oil prices. 4. Unit Roots and Stationarity Figure 1 and 2 plots the data of the three variables over time. The plots of these variables show that they all have some general increasing trends thus indicating that their means might be different across time (non-stationary). 1 http://data.worldbank.org/data-catalog/global-economic-monitor, accessed on 01 May 2015 157 Journal of Economics and Sustainable Development ISSN 2222-1700 (Paper) ISSN 2222-2855 (Online) DOI: 10.7176/JESD Vol.10, No.10, 2019 Journal of Economics and Sustainable Development ISSN 2222-1700 (Paper) ISSN 2222-2855 (Online) DOI: 10.7176/JESD Vol.10, No.10, 2019 www.iiste.org 20.6 20.8 21 21.2 21.4 21.6 Log of industrial production 3.8 4 4.2 4.4 4.6 Log of CPI Jan-95 Jan-00 Jan-05 Jan-10 date Log of CPI Log of industrial production However, I have used two formal tests to examine the existence of unit roots – Augmented Dickey Fuller (ADF) and Phillips-Perron (PP) tests. The ADF test assumes autocorrelation of the error term of the random walk model and thus augments the Dickey Fuller (DF) equation by adding the lagged values. The ADF test thus takes the following regression – 1.4 1.5 1.6 1.7 1.8 1.9 Log of M2 3.8 4 4.2 4.4 4.6 Log of CPI Jan-95 Jan-00 Jan-05 Jan-10 date Log of CPI Log of M2 Figure 2: Plots of CPI and industrial production Figure 2: Plots of CPI and M2 multiplier 1.4 1.5 1.6 1.7 1.8 1.9 Log of M2 3.8 4 4.2 4.4 4.6 Log of CPI Jan-95 Jan-00 Jan-05 Jan-10 date Log of CPI Log of M2 Figure 2: Plots of CPI and M2 multiplier Figure 2: Plots of CPI and M2 multiplier Figure 2: Plots of CPI and industrial production 20.6 20.8 21 21.2 21.4 21.6 Log of industrial production 3.8 4 4.2 4.4 4.6 Log of CPI Jan-95 Jan-00 Jan-05 Jan-10 date Log of CPI Log of industrial production Figure 2: Plots of CPI and industrial production However, I have used two formal tests to examine the existence of unit roots – Augmented Dickey Fuller (ADF) and Phillips-Perron (PP) tests. The ADF test assumes autocorrelation of the error term of the random walk model and thus augments the Dickey Fuller (DF) equation by adding the lagged values. 4. Unit Roots and Stationarity The ADF test thus takes the following regression – ∆ = + + + ∆ + … … … … (3) where ∆Yt = Yt - Yt-1 and k=number of lags in the dependent variable, εt is white noise, β1 is drift and t is the trend term. The null hypothesis in the ADF test is that there is presence of unit root. In other words, if we cannot reject the null hypothesis that θ in the equation is zero, then there is evidence of unit root. The PP test uses nonparametric methods to account for the autocorrelation of the disturbance term. Both the ADF and PP tests can be conducted at different specifications – random walk; random walk with drifts, and random walk with drift and a deterministic trend (Gujarati et al. 2009; Greene 2006). I have conducted these tests by taking 12 lags and with and without trend specifications. The null hypothesis in both these tests is that there is unit root. The results of these tests are presented in Table 1. It shows that the ADF and PP statistics of all these variables in both specifications are lower than their critical values at 1%, 5% and 10% levels in their absolute terms, thus suggesting that these variables are non-stationary. Table 1: Results of ADF and PP tests (at 12 lags) Variable ADF statistic Phillips-Perron test statistic Decision Without trend With trend Without trend With trend lcpi 0.835 -1.095 1.894 -1.215 Non- stationary lM2 -1.458 -0.419 -1.633 -1.775 Non- stationary lipro 1.015 -1.893 -0.229 -7.214 Non- stationary Note: The ADF critical values are -3.49, -2.89, and -2.58 at 1%, 5%, and 10% levels are respectively (without trend). They are respectively -4.02, -3.44, and -3.14 (with trend). The critical values for the PP test are -3.49, -2.89, and -2.58 respectively at 1%, 5%, and 10% levels (without trend). They are respectively -4.02, -3.44, and -3.14 (with trend). stationary Note: The ADF critical values are -3.49, -2.89, and -2.58 at 1%, 5%, and 10% levels are respectively (without trend). They are respectively -4.02, -3.44, and -3.14 (with trend). The critical values for the PP test are -3.49, -2.89, and -2.58 respectively at 1%, 5%, and 10% levels (without trend). They are respectively -4.02, -3.44, and -3.14 (with trend). 1 See Stata User’s Guide: http://www.stata.com/manuals13/u.pdf 5. Cointegration g The Johansen technique has been used to test for the existence of cointegration. This is a widely used technique for testing cointegration in multivariate analysis (Johansen 1988). The first step in the Johansen procedure requires identifying the lag order of the underlying Vector Autoregressive model. Table 2 reports the result of VECM lag lengths. The suggested lag order is 2 according to Hannan and Quinn’s information criterion (HQIC) and Schwarz Bayesian information criterion (SBIC). However, according to Akaike’s final prediction error (FPE) and Akaike’s information criterion (AIC), the suggested lag is 3. I have followed the HQIC and SBIC lag order and thus choose a lag length of 2 for estimating the VECM. 158 Journal of Economics and Sustainable Development www.iiste.org ISSN 2222-1700 (Paper) ISSN 2222-2855 (Online) DOI: 10.7176/JESD Vol.10, No.10, 2019 DOI: 10.7176/JESD Table 2: Results of selection-order criteria lag LL LR df p FPE AIC HQIC SBIC 0 444.962 9.8e-07 -5.32485 -5.30202 -5.26861 1 1215.57 1541.2 9 0.000 1.0e-10 -14.5008 -14.4095 -14.2758 2 1244 56.861 9 0.000 8.0e-11 -14.7349 -14.5751* -14.3412* 3 1253.11 18.236* 9 0.033 8.0e-11* -14.7363* -14.508 -14.1739 4 1261.2 16.182 9 0.063 8.1e-11 -14.7254 -14.4286 -13.9942 Sample: 5 – 170; no of observations 166 The Johansen test involves the identification of the number of ranks (r) in the VECM. The null hypothesis is that there is no more than ‘r’ cointegrating relationship. Table 3 shows that the trace statistics are higher at r=0 (but lower at r =1) than their 5% critical values in all three specifications. It suggests that we can reject the null that there are no (r=0) cointegrating relationship but cannot reject the null that there are no more than one co-integrating relationship in the VECM. Thus it can be concluded that the maximum rank of the cointegrating VECM is 1 and that there is one cointegrating relationship between money supply, inflation, and industrial production in the VECM. Table 3: Result of Johansen test for cointegration Maximum rank Eigenvalue Trace statistic 5% critical value Constant and no trend specification 0 . 32.4019 29.68 1 0.15376 4.3543* 15.41 2 0.02540 0.0327 3.76 3 0.00019 3 21 No constant and no trend specification 0 . 42.5241 24.31 1 0.20765 3.4224* 12.53 2 0.02006 0.0187 3.84 3 0.00011 With trend and constant specification 0 . 5. Cointegration 39.9304 34.55 1 0.15385 11.8651* 18.17 2 0.06696 0.2221 3.74 3 0.00132 Sample: 3-170; No of observation=168; Lags=2 Table 3: Result of Johansen test for cointegration Sample: 3-170; No of observation=168; Lags=2 Sample: 3-170; No of observation=168; Lags=2 6. Vector Error Correction Model 6. Vector Error Correction Model 6. Vector Error Correction Model 6. Vector Error Correction Model ohansen approach considers the following equation as the VECM - The Johansen approach considers the following equation as the VECM - The Johansen approach considers the following equation as the VECM - ∆ = (՛ + ! + ") + Г∆ $ + + % + … … … … (4) ∆ = (՛ + ! + ") + Г∆ $ + + % + … … … … (4) where α and β are both kr rank matrices, yt is a k1 vector of variables; μ and ρ are r1 vectors of parameters and γ and Τ are k1 vectors of parameters. While the original VAR considers p lags in y, we have p-1 lags in ∆y. This approach considers several specifications of the VECM – (a) the unrestricted trend assumption puts no restrictions on the parameters thus assuming quadratic trends in the levels of the variables and that the cointegrating equations are stationary around time trends ; (b) the restricted trend assumption (Τ =0) holds that the levels of the variables show linear trends and that the cointegrating equations are trend stationary; (c) unrestricted constant assumes that the cointegrating equations are stationary around constant means but it also puts a linear time trend (i.e., Τ =0 and ρ=0) ; (d) restricted constant assumes that the cointegrating equations are stationary around constant means and that there are no linear time trend in the level of the variables (i.e., Τ =0 and ρ=0 and γ=0); and (e) no trend assumes that there are no trend and the cointegrating equations are stationary around zero means (i.e., Τ =0 and ρ=0, γ=0 and μ=0)1. 1 See Stata User’s Guide: http://www.stata.com/manuals13/u.pdf 159 Journal of Economics and Sustainable Development iste.org DOI: 10.7176/JESD I have carried out both unrestricted and restricted VECM. The result of the unrestricted model is presented in annex 1. It shows that the co-integrating relationship is significant at 1% level (chi sq=49.20). 6. Vector Error Correction Model The co-integrating relationships are estimated as follows - lcpi= - 13.89 - 0.14lM2 + 0.87 lipro ………………….(6) (0.039) (0.067) lcpi= - 13.89 - 0.14lM2 + 0.87 lipro ………………….(6) (0.039) (0.067) lcpi= - 13.89 - 0.14lM2 + 0.87 lipro ………………….(6) (0.039) (0.067) er se so ed nt he of he h, ed es g equations is correctly specified. The graph of the the circle (i.e, the values are <1; see Annex 3). It cified. -1 -.5 0 .5 1 Imaginary -1 -.5 0 .5 1 Real The VECM specification imposes 2 unit moduli Roots of the companion matrix % increase in the money multiplier in monthly CPI (or 1.7 % decrease ically significant at 5% level. It also industrial production is associated CPI, which is statistically significant sults support the findings of the redicts a non-significant impact of nt impact of output on inflation in the CM tells that every after one month, the long run equilibrium is restored e parameters of our VECM requires nd that the number of cointegrating equations is correctly specified. The graph of the that all the eigenvalues are inside the circle (i.e, the values are <1; see Annex 3). It able and the model is correctly specified. -1 -.5 0 .5 1 Imaginary -1 -.5 0 .5 1 Real The VECM specification imposes 2 unit moduli Roots of the companion matrix -1 -.5 0 .5 1 Imaginary -1 -.5 0 .5 1 Real The VECM specification imposes 2 unit moduli Roots of the companion matrix The result suggests that a 1% increase in the money multiplier associates a 0.14% decrease in monthly CPI (or 1.7 % decrease annually), which is not statistically significant at 5% level. It also suggests that 1% increase in industrial production is associated with 0.87 % increase in the CPI, which is statistically significant at 5% level. Thus the results support the findings of the unrestricted model which predicts a non-significant impact of money supply but a significant impact of output on inflation in the long run.The result of the VECM tells that every after one month, 1.3 % of the deviation from the long run equilibrium is restored (see Annex 2). However, the inference of the parameters of our VECM requires that the model is stationary and that the number of cointegrating equations is correctly specified. 6. Vector Error Correction Model It suggests the long run relationship between CPI, M2, and industrial production as follows: lcpi = - 17.32 - 0.10 lM2 + 1.03 lipro - 0.001t ………………….(5) (0.098) (0.176) (0.176) According to equation (5), we see that a 1% increase of the money multiplier is associated with a monthly decline of the CPI index by 0.10% (i.e., 1.2% decline annually) in the long run. It is somewhat puzzling that an increasing money supply is associated with decline in inflation; however, the result is not statistically significant at conventional level. We also see that a 1% increase in the industrial production is associated with an equivalent 1.03% increase in monthly CPI, and is statistically significant at 5% level. Thus it can be said that the impact of money supply is not significant but the impact of output growth is significant on inflation in the long run. The result of the VECM suggests that every after one month, 1.1 % of the deviation from the long run equilibrium is restored (see Annex 1). I have compared among different specifications of the restricted VECM based on the Likelihood Ratio (LR) statistics. While comparing between two different specifications, a low LR ratio implies that the new specification is better than the previous specification. Accordingly, I have found that the model with the unrestricted constant is a better specification (see results in Table 4). Table 4: Results of comparison between different specifications of the VECM Comparing models LR chi2 Prob > chi2 Decision ‘trend’ vs. ‘rtrend’ 3.82 0.1482 ‘rtrend’ is a better specification than ‘trend’ ‘rtrend’ vs. ‘constant’ 0.18 0.6685 ‘constant’ is better than ‘rtrend’ ‘constant’ vs. ‘rconstant’ 25.52 0.0000 ‘constant’ is better than ‘rconstant’ ‘constant’ vs. ‘none’ 26.60 0.0000 ‘constant’ is better specification than ‘none’ Notes: ‘trend’, ‘rtrend’, ‘constant’, ‘rconstant’, and ‘none’ stand for the ‘unrestricted trend’, ‘restricted trend’, ‘unrestricted constant’, ‘restricted constant’, and ‘no trend’ respectively. Table 4: Results of comparison between different specifications of the VECM Comparing models LR chi2 Prob > chi2 Decision I re-estimate the restricted VECM in terms of the unrestricted constant specification. The result of the model is shown in annex 2. The result shows that the co-integrating relationship is highly statistically significant at 1% level (chi sq= 1266.3). 7. Granger Causality Granger causality explains how much of the future values of a variable (say, yt) can be forecasted by the lagged values of another variable (xt), after controlling for the influence of the lagged values of the variable itself (Wooldridge 2009). The idea of the Granger causality is to look at the temporal order of events or the direction of the causality – whether x precedes y, or the other way round. The Granger Causality test involving two time series variables Yt and Xt requires estimation of the following pairs of regression – = ' ( + )) ( ) + … … … … … (7) ' = +' ( + ,)) ( ) + - … … … … . . (8) where ut and vt are assumed to be uncorrelated.. Now we can distinguish four possible cases from these equations – (a) unidirectional causality from X to Y when ∑ αi ≠ 0 and ∑ δj =0; (b) unidirectional causality from Y to X when∑ αi =0 and ∑ δj≠ 0; (c) feedback or bilateral causality when ∑ αi ≠ 0 and ∑ δj≠ 0; (d) independence when ∑ αi =0 and ∑ δj =0 (Granger 1969; Gujarati et al. 2009). The idea of the Granger Causality can be extended to a system of multivariate VAR model. I have applied the Granger causality test to see if - (a) money supply Granger causes inflation (or the other way round); (b) industrial production Granger causes inflation (or the opposite); and (c) money supply Granger causes growth (or the opposite). The result of the pair-wise Granger causality test is shown in Table 5. The Wald test statistics suggest that we cannot reject the null hypothesis that money supply does not Granger cause inflation. In contrast, we can reject the null that industrial production does not Granger cause inflation. The statistics also tells that money supply Granger causes growth in industrial production at 1% level of significance. Thus the result of the Granger Causality test lend support in favour of the findings of the VECM which shows a significant effect of output but a non-significant effect of money supply on inflation. It implies that the monetarist prediction that inflation is a monetary phenomenon does not hold true in Bangladesh. 6. Vector Error Correction Model The graph of the companion matrix show that that all the eigenvalues are inside the circle (i.e, the values are <1; see Annex 3). It suggests that the VECM is stable and the model is correctly specified. The VECM specification imposes 2 unit moduli 160 Journal of Economics and Sustainable Development Journal of Economics and Sustainable Development DOI: 10.7176/JESD ISSN 2222-1700 (Paper) ISSN 2222-2855 (Online) DOI: 10.7176/JESD Vol.10, No.10, 2019 ISSN 2222-1700 (Paper) ISSN 2222-2855 (Online) DOI: 10.7176/JESD Vol.10, No.10, 2019 Note: The result is computed at 12 lags 7. Granger Causality It seems reasonable in the context a developing country like Bangladesh where there are still lots of physical and human capital underutilized. Thus we can say that increased money supply causes economic growth and growth causes inflation in the Bangladesh economy. The Granger Causality result also indicates that inflation and industrial production do not have significant impact on money supply thereby rejecting the possibility of bi-directional causality running from real economic growth to money supply. Table 5: Result of the Granger Causality Wald test statistics chi2 df Prob> chi2 Dependent variable: lcpi Excluded lM2 7.1461 12 0.848 lipro 24.531 12 0.017 All 31.651 24 0.136 Dependent variable: lM2 Excluded lcpi 3.505 12 0.991 lipro 19.439 12 0.078 All 25.699 24 0.369 Dependent variable: lipro Excluded lcpi 20.496 12 0.058 lM2 34.981 12 0.000 All 50.483 24 0.001 Note: The result is computed at 12 lags Table 5: Result of the Granger Causality Wald test statistics 161 Journal of Economics and Sustainable Development www.iiste.org iiste.org DOI: 10.7176/JESD DOI: 10.7176/JESD ISSN 2222-1700 (Paper) ISSN 2222-2855 (Online) DOI: 10.7176/JESD Vol.10, No.10, 2019 ISSN 2222-1700 (Paper) ISSN 2222-2855 (Online) DOI: 10.7176/JESD Vol.10, No.10, 2019 8. Conclusion In this paper, I have used monthly data to analyze the long run relationship between money supply, inflation, and output growth of Bangladesh. Using the Johansen multivariate technique, I have found that there is one cointegrating relationship among these three variables. Using the Granger Causality technique, I have found that growth of money supply does not cause inflation in Bangladesh. I have found that increased money supply causes outputgrowth in Bangladesh, which rejects the monetarist view of the neutrality of money. However, my findings support another monetarist proposition that the causality is unidirectional from money to economic growth and not vice-versa. All these findings suggest the need for undertaking expansionary monetary policy for maximizing output growth in Bangladesh. References Bannock, G., R. E. Baxter & E. Davis (2009), The Penguin Dictionary of Economics, Penguin Books. Barth, J.R. & J.T. Bennett (1974), “The Role of Money in the Canadian Economy: An Empirical Test”, The Canadian Journal of Economics 7 (2), 306-311. Bhuiyan, R. (2012), “The Effects of Monetary Policy Shocks in Bangladesh: A Bayesian Structural VAR Approach”, International Economic Journal 26 (2), 301-316. Bilquees, F. (1988), “Inflation in Pakistan: Empirical Evidence on the Monetarist and Structuralist Hypothesis”, The Pakistan Development Review 27(2), 109-130. Chowdhury, A. R., M. Q. Dao & A. N. M. Wahid (1995), “Monetary Policy, Output, and Inflation in Bangladesh: A Dynamic Analysis”, Applied Economics Letters 2 (3), 51-55. Crowder, W.J. (1988), “The Long Run Link between Money, Growth, and Inflation”, Economic Inquiry 36, 229- 243. Engle, R. & C.W.J. Granger (1987), “Cointegration and Error Correction: Representation, Estimation and Testing”, Econometrica 55, 251-276. Froyen, R.T. (2005), Macroeconomics: Theories and Policies, 8th Edition, Pearson Education, Inc. Granger, C.W.J (1969), “Investigating Causal Relations by Econometric Models and Cross-Spectral Methods”, Econometrica 37(3), 424-438. Greene, W.H. (2006), Econometric Analysis , 5th Edition, Pearson. Gujarati, N. D., D. C Porter & S. Gunasekar (2009), Basic Econometrics, 5th Edition, New York: McGraw Hill Companies, Inc. Hossain, A. (2011), “Money-Income Causality in Bangladesh: An Error Correction Approach”, Bangladesh Development Studies 34 (1), 39-58. Johansen, S. (1988), “Statistical Analysis of Cointegrating Vectors”, Journal of Economic Dynamics and Control 12, 231-54. Jones, D. J. & Z. Sattar (1988), “Money, Inflation, Output and Causality: The Bangladesh Case, 1974-1985”, The Bangladesh Development Studies 16(1), 73-83. Kemal, M.A. (2006), “Is Inflation in Pakistan a Monetary Phenomenon?”, The Pakistan Development Review 45 (2), 213-220. Malik, W.S.(2006). “Money, Output, and Inflation: Evidence from Pakistan”, The Pakistan Development Review, 45(4), Part II, 1277-1286. Mujeri, M. K., M. Shahiduzzaman & M. E. Islam (2009), “Application of the P-Star Model for Measuring Inflationary Pressure in Bangladesh”, The Bangladesh Development Studies 32 (1), 1-22. 162 Journal of Economics and Sustainable Development ISSN 2222-1700 (Paper) ISSN 2222-2855 (Online) DOI: 10.7176/JESD Vol.10, No.10, 2019 www.iiste.org iiste.org DOI: 10.7176/JESD Rami, G.(2010), “Causality between Money, Prices and Output in India (1951-2005): A Granger Causality Approach”, Journal of Quantitative Economics 8 (2), 20-41. Rangarajan, C. & R.R. Arif (1990), “Money, Output and Prices: A Macro Econometric Model”, Economic and Political Weekly 25(16), 837-852. Sims, C.A. (1972), “Money, Income, and Causality”, The American Economic Review 62 (4), 540-552. Taslim, M. A. References of obs = 168 AIC = -14.63678 Log likelihood = 1246.489 HQIC = -14.50848 Det (Sigma_ml) = 7.21e-11 SBIC = -14.32066 Equation Parms RMSE R-sq chi2 P>chi2 D_lcpi 5 .00762 0.2944 68.0042 0.0000 D_lM2 5 .03246 0.2071 42.5776 0.0000 D_lipro 5 .03608 0.3196 76.5632 0.0000 Coef. Std.Err. z P>z [95% Conf. Interval] D_lcpi _ce1 L1. -0.013 0.018 -0.750 0.450 -0.048 0.021 lcpi LD. 0.149 0.078 1.900 0.057 -0.004 0.301 lM2 LD. 0.004 0.017 0.270 0.789 -0.028 0.037 lipro LD. -0.014 0.016 -0.920 0.357 -0.045 0.016 _cons 0.004 0.001 5.840 0.000 0.003 0.006 D_lM2 _ce1 L1. -0.046 0.075 -0.620 0.536 -0.194 0.101 lcpi LD. -0.004 0.332 -0.010 0.992 -0.654 0.647 lM2 LD. -0.431 0.070 -6.120 0.000 -0.569 -0.293 lipro LD. -0.117 0.067 -1.760 0.078 -0.248 0.013 _cons 0.004 0.003 1.150 0.249 -0.003 0.010 D_lipro _ce1 L1. 0.437 0.083 5.240 0.000 0.273 0.601 lcpi -0.516 0.369 -1.400 0.162 -1.240 0.207 Journal of Economics and Sustainable Development www.iiste.org ISSN 2222-1700 (Paper) ISSN 2222-2855 (Online) DOI: 10.7176/JESD Vol.10, No.10, 2019 Cointegrating equations Equation Parms chi2 P>chi2 _ce1 2 49.20194 0.0000 Identification: beta is exactly identified Johansen normalization restriction imposed beta Coef. Std. Err. z P>z [95% Conf. Interval] _ce1 lcpi 1 . . . . . lM2 0.0997 0.0975 1.0200 0.3060 -0.0913 0.2907 lipro -1.0288 0.1761 -5.8400 0.0000 -1.3740 -0.6836 _trend 0.0011 . . . . . _cons 17.3150 . . . . . Annex 2: Result of the Restricted VECM Annex 2: Result of the Restricted VECM Vector error-correction model Sample: 1995m8 - 2009m7 No. of obs = 168 AIC = -14.63678 Log likelihood = 1246.489 HQIC = -14.50848 Det (Sigma_ml) = 7.21e-11 SBIC = -14.32066 Equation Parms RMSE R-sq chi2 P>chi2 D_lcpi 5 .00762 0.2944 68.0042 0.0000 D_lM2 5 .03246 0.2071 42.5776 0.0000 D_lipro 5 .03608 0.3196 76.5632 0.0000 Coef. Std.Err. z P>z [95% Conf. Interval] D_lcpi _ce1 L1. -0.013 0.018 -0.750 0.450 -0.048 0.021 lcpi LD. 0.149 0.078 1.900 0.057 -0.004 0.301 lM2 LD. 0.004 0.017 0.270 0.789 -0.028 0.037 lipro LD. -0.014 0.016 -0.920 0.357 -0.045 0.016 _cons 0.004 0.001 5.840 0.000 0.003 0.006 D_lM2 _ce1 L1. -0.046 0.075 -0.620 0.536 -0.194 0.101 lcpi LD. -0.004 0.332 -0.010 0.992 -0.654 0.647 lM2 LD. -0.431 0.070 -6.120 0.000 -0.569 -0.293 lipro LD. -0.117 0.067 -1.760 0.078 -0.248 0.013 _cons 0.004 0.003 1.150 0.249 -0.003 0.010 D_lipro _ce1 L1. 0.437 0.083 5.240 0.000 0.273 0.601 lcpi LD. References (1982), “Inflation in Bangladesh: A Reexamination of the Structuralist-Monetarist Controversy” , The Bangladesh Development Studies 10(1), 23-52. Williams, D., C.A.E. Goodhart & D.H. Gowland (1976), “Money, Income, and Causality: The UK Experience”, The American Economic Review 66(3), 417-423. Wooldridge, M. J. (2009), “Introductory Econometrics: A Modern Approach”, 4th Edition, Cengage. Wooldridge, M. J. (2009), “Introductory Econometrics: A Modern Approach”, 4th Edition, Cengage. Annex 1: Result of the Unrestricted VECM Annex 1: Result of the Unrestricted VECM Annex 1: Result of the Unrestricted VECM Annex 1: Result of the Unrestricted VECM Vector error-correction model Sample: 1995m8 - 2009m7 No. of obs. = 168 AIC = -14.62488 Log likelihood = 1248.49 HQIC = -14.47395 Det(Sigma_ml) = 7.04e-11 SBIC = -14.25298 Equation Parms RMSE R-sq chi2 P>chi2 D_lcpi 6 .0076 0.3024 70.2411 0.0000 D_lM2 6 .0324 0.2139 44.0836 0.0000 D_lipro 6 .0361 0.3237 77.5329 0.0000 Coef. Std.Err. z P>z [95% Conf. Interval] D_lcpi _ce1 L1. -0.011 0.015 -0.730 0.466 -0.041 0.019 lcpi LD. 0.133 0.078 1.700 0.088 -0.020 0.286 lM2 LD. 0.006 0.017 0.350 0.727 -0.027 0.038 lipro LD. -0.015 0.016 -0.930 0.353 -0.046 0.016 _trend 0.000 0.000 1.460 0.145 0.000 0.000 _cons 0.003 0.001 2.370 0.018 0.001 0.005 D_lM2 _ce1 L1. -0.012 0.065 -0.180 0.860 -0.139 0.116 lcpi LD. 0.023 0.333 0.070 0.946 -0.629 0.675 lM2 LD. -0.442 0.070 -6.280 0.000 -0.580 -0.304 lipro LD. -0.103 0.067 -1.530 0.126 -0.235 0.029 _trend 0.000 0.000 -1.290 0.195 0.000 0.000 _cons 0.009 0.005 1.660 0.096 -0.002 0.019 D_lipro _ce1 L1. 0.384 0.072 5.300 0.000 0.242 0.526 lcpi LD. -0.490 0.370 -1.320 0.186 -1.215 0.236 lM2 LD. 0.161 0.078 2.050 0.040 0.007 0.315 lipro LD. -0.213 0.075 -2.850 0.004 -0.359 -0.066 _trend -1.52e-06 6.000 0.000 -0.030 0.979 0.000 _cons 0.001 0.006 0.130 0.896 -0.011 0.012 163 Journal of Economics and Sustainable Development www.iiste.org ISSN 2222-1700 (Paper) ISSN 2222-2855 (Online) DOI: 10.7176/JESD Vol.10, No.10, 2019 Cointegrating equations Equation Parms chi2 P>chi2 _ce1 2 49.20194 0.0000 Identification: beta is exactly identified Johansen normalization restriction imposed beta Coef. Std. Err. z P>z [95% Conf. Interval] _ce1 lcpi 1 . . . . . lM2 0.0997 0.0975 1.0200 0.3060 -0.0913 0.2907 lipro -1.0288 0.1761 -5.8400 0.0000 -1.3740 -0.6836 _trend 0.0011 . . . . . _cons 17.3150 . . . . . Annex 2: Result of the Restricted VECM Vector error-correction model Sample: 1995m8 - 2009m7 No. References -0.516 0.369 -1.400 0.162 -1.240 0.207 lM2 LD. 0.150 0.078 1.920 0.055 -0.003 0.304 lipro LD. -0.222 0.074 -3.000 0.003 -0.368 -0.077 _cons 0.001 0.004 0.150 0.882 -0.006 0.007 164 Journal of Economics and Sustainable Development www.iiste.org ISSN 2222-1700 (Paper) ISSN 2222-2855 (Online) DOI: 10.7176/JESD Vol.10, No.10, 2019 Cointegrating equations Equation Parms chi2 P>chi2 _ce1 2 1266.299 0.0000 Identification: beta is exactly identified Johansen normalization restriction imposed beta Coef. Std. Err. z P>z [95% Conf. Interval] _ce1 lcpi 1 . . . . . lM2 .1430 .0668 2.14 0.032 .0121 .2739 lipro -.8655 .0387 -22.34 0.000 -.9414 -.7896 _cons 13.8914 . . . . . Annex 3: Eigenvalue stability condition Eigenvalue Modulus 1 1 1 1 .6755747 .675575 -.3810964 + .138712i .405556 -.3810964 - .138712i .405556 .1831551 .183155 The VECM specification imposes 2 unit moduli Journal of Economics and Sustainable Development www.iiste.org ISSN 2222-1700 (Paper) ISSN 2222-2855 (Online) DOI: 10.7176/JESD Vol.10, No.10, 2019 Cointegrating equations Equation Parms chi2 P>chi2 _ce1 2 1266.299 0.0000 Identification: beta is exactly identified Johansen normalization restriction imposed beta Coef. Std. Err. z P>z [95% Conf. Interval] _ce1 lcpi 1 . . . . . lM2 .1430 .0668 2.14 0.032 .0121 .2739 lipro -.8655 .0387 -22.34 0.000 -.9414 -.7896 _cons 13.8914 . . . . . Annex 3: Eigenvalue stability condition Eigenvalue Modulus 1 1 1 1 .6755747 .675575 -.3810964 + .138712i .405556 -.3810964 - .138712i .405556 .1831551 .183155 The VECM specification imposes 2 unit moduli Annex 3: Eigenvalue stability condition Eigenvalue Modulus 1 1 1 1 .6755747 .675575 -.3810964 + .138712i .405556 -.3810964 - .138712i .405556 .1831551 .183155 The VECM specification imposes 2 unit moduli 165
https://openalex.org/W2969785912	https://www.econstor.eu/bitstream/10419/217866/1/sajbm-v15i3-1121.pdf	English	null	Never-buyers of consumer non-durables	South African journal of business management	1,984	cc-by	5,844	Provided in Cooperation with: University of Stellenbosch Business School (USB), Bellville, South Africa Suggested Citation: Firer, C. (1984) : Never-buyers of consumer non-durables, South African Journal of Business Management, ISSN 2078-5976, African Online Scientific Information Systems (AOSIS), Cape Town, Vol. 15, Iss. 3, pp. 155-161, https://doi.org/10.4102/sajbm.v15i3.1121 Terms of use: Documents in EconStor may be saved and copied for your personal and scholarly purposes. Die Dokumente auf EconStor dürfen zu eigenen wissenschaftlichen Zwecken und zum Privatgebrauch gespeichert und kopiert werden. You are not to copy documents for public or commercial purposes, to exhibit the documents publicly, to make them publicly available on the internet, or to distribute or otherwise use the documents in public. Sie dürfen die Dokumente nicht für öffentliche oder kommerzielle Zwecke vervielfältigen, öffentlich ausstellen, öffentlich zugänglich machen, vertreiben oder anderweitig nutzen. If the documents have been made available under an Open Content Licence (especially Creative Commons Licences), you may exercise further usage rights as specified in the indicated licence. Sofern die Verfasser die Dokumente unter Open-Content-Lizenzen (insbesondere CC-Lizenzen) zur Verfügung gestellt haben sollten, gelten abweichend von diesen Nutzungsbedingungen die in der dort genannten Lizenz gewährten Nutzungsrechte. https://creativecommons.org/licenses/by/4.0/ https://creativecommons.o Firer, C. Article Never-buyers of consumer non-durables South African Journal of Business Management Provided in Cooperation with: University of Stellenbosch Business School (USB), Bellville, South Africa Suggested Citation: Firer, C. (1984) : Never-buyers of consumer non-durables, South African Journal of Business Management, ISSN 2078-5976, African Online Scientific Information Systems (AOSIS), Cape Town, Vol. 15, Iss. 3, pp. 155-161, https://doi.org/10.4102/sajbm.v15i3.1121 Firer, C. Introduction In the field of consumer non-durables, an interesting segment of the market, known as 'never-buyers', has caused some dis- cussion in the literature. Ehrenberg (1972) claimed that, given a sufficiently long period of time, every consumer will eventually purchase a pro- duct at least once, and so the never-buyer does not exist. If a study shows a number of non-buyers in a period of time, Ehrenberg believes that, had the time of the study been longer, these non-buyers would not have been found. In fact, his Negative Binomial Distribution (NBD) theory which proposes that, provided the market is stationary frequently purchased consumer non-durables can be modelled using the NBD, (Equation Al) predicts this outcome. As time increases, the proportion of buyers of the product as predicted by the NBD tends to unity (Equation A2). (A brief description of the statistical distributions mentioned in the text is given in the Appendix.) Purely from a logical point of view one may seek to raise issue with Ehrenberg on the question of never-buyers. Non- smokers probably never buy cigarettes, non-drivers who do not own vehicles are unlikely to purchase petrol. Product classes can therefore be identified in which, by logical conside- rations alone, one would expect to find some never-buyers. Nevertheless, one should concede to Ehrenberg that such examples probably represent a small proportion of the con- sumer non-durables market. S. Afr. J. Bus. Mgmt. 1984, 15: 155- 161 In die artikel word die konsep van nooit-kopers van nle-duursame verbruikersgoedere bespreek. Ehrenbergse tradisionele Negatief· binomiaaldistribusie-benadering in verband met die vraag is voorgestel. Vorige ongepubliseerde werk uitgevoer by die Nagraadsebestuurskool, Universiteit van die Witwatersrand is bespreek, en hipoteses word voorgestel dat die waargenome groot nul-sel in die verkoops-frekwensiedistribusies miskien 'n gevolg mag wees van die bestaan van 'n groep nooit-kopers van die pro- duk, of as 'n gevolg van die opeenplasing van ten minste twee af- sonderlike koperpopulasies voorheen geidentifiseer as handels- naam-lojaal en handelsnaam-wlsselende gesinne. Die resultate van die navorsing wat gedoen is om die eerste hipotese te toets, word gegee. Twee sorgvuldig gekontroleerde stelsels data is gemodelleer met gebruik van die nulvermeerderde Negatief- binomiaal- en Sichel-distribusies. lndividuele handelsname in een datastesel (aankope van toiletseep) volg die voorspelling van die tra- disionele teorie - die proporsie van nie-verbruikers neem mettertyd af. C. Firer Graduate School of Business Administration, University of the Wit- watersrand, P.O. Box 31170, Braamfontein 2017, Republic of South Africa Never-buyers of consumer non-durables C. Firer Graduate School of Business Administration, University of the Witwatersrand, Johannesburg C. Firer Graduate School of Business Administration, University of the Witwatersrand, Johannesburg C e Graduate School of Business Administration, University of the Witwatersrand, Johannesburg In this article the concept of never-buyers of consumer non- durables is discussed. The traditional Negative Binomial Distribu- tion approach of Ehrenberg to the question is presented. Previous- ly unpublished work carried out at the Graduate School of Business Administration, University of the Witwatersrand, Is reviewed and hypotheses are put forward that the observed large zero cell in the purchase frequency distributions may be caused by the existence of a group of never-buyers of the product, or by the superimposition of at least two distinct buying populations, previously identified as brand-loyal and multibrand/brand-switchlng households. The results of the research aimed at testing the first hypothesis are presented here. Two carefully monitored data sets were modelled using zero-augmented Negative Binomial and Sichel distributions. The data were previously shown to exhibit the necessary mean households purchase/consumption stationarity. In- dividual brands in one data set (purchases of toilet soap) were shown to follow the predictions of the traditional theory - the proportion of non-buyers decreasing with time. In the second data set (consumption of packaged soup) the proportion of non- consumers of the brands fell towards zero as the length of the time period studied was increased, but at a rate faster than that predicted by the theory. The hypothesis of the existence of never- buyers/users of individual brands in these two product classes was therefore rejected. Introduction In die tweede datastelsel (verbruik van verpakte sop) het die proporsie van nie-verbruikers van die handelsname na nul gedaal met verlenging van die tydperk wat bestudeer is, maar teen 'n snelheid vinnlger as wat voorspel was deur die teorle. Die hipotese dat daar noolt-kopers/gebruikers van individuele handelsname in die twee produkklasse bestaan het, is dus verwerp. S.-Afr. Tydskr. Bedryfsl. 1984, 15: 155-161 Ehrenberg also claimed that his model is as applicable to an individual brand as it is to the product class as a whole. However, with respect to individual brands the situation may be quite different. Ehrenberg and other overseas writers have, with few published exceptions, only tested the NBD model in its univariate form. Where they found departures from the model, they ascribed them to effects which they labelled 'the shelving phenomenon' or 'the variance discrepency'. They believed this effect - an observed discontinuity in the frequen- cy of heavier purchases at or close to the number of units, equal to the number of weeks or period in the time period of the study- to be caused by a small sector of the market pur- chasing the product regularly. Sichel (1975) investigated this phenomenon using observa- tions on packaged-soup consumption. By their very nature, being consu.-nption-based, such data, according to Ehrenberg's theory, ought not to exhibit the effect described above. However, 'shelving' was observed in the data. C. Firer She also found that other statistical tests revealed signifi- cant departures from Ehrenberg's NBD model. As was subse- Accepted May 1984 S.-Afr. Tydskr. Bedryfsl. 1984, 15(3) 1S6 they were plotted in the fonn of frequency distributions, even- numbered purchase frequencies appeared high relative to the odd-numbered purchase frequencies, giving a zig-z.ag shaped distribution (Figure 1). This was attributed by Goodman to a consumer preference for purchasing the product in even- numbered lots. quently demonstrated by Zachos (l 977), although univariate distributions predicted by NBD theory may be accepted by the chi-square test, predictions in the bivariate fonn of the NBD often provide very unsatisfactory models. As a result, both of her own work and an earlier study by Tucker (1973), Sichel (1975) suggested that the true phen~ menon of repeat-consumption/buying may be represented by two distinctly different populations: 200 100 oL----~~~~~~:::::::c.-...:A::.. 0 2 4 6 8 10 12 14 No. OF OOTS PURCHASED Figure 1 Observed frequency distribution for Palmolive soap - eight months (N = 614) 200 100 oL----~~~~~~:::::::c.-...:A::.. 0 2 4 6 8 10 12 14 No. OF OOTS PURCHASED (a) A hard-core of stable, brand-loyal households which con- sume regularly in accordance with Ehrenberg's original model.' These therefore would be expected to follow the correlated NBD model (Equation A3). (a) A hard-core of stable, brand-loyal households which con- sume regularly in accordance with Ehrenberg's original model.' These therefore would be expected to follow the correlated NBD model (Equation A3). (b) 'A fair number of brand-switching, multibrand consum- ing households.' This group is statistically characterized by the uncorrelated bivariate NBD (Equation AS) in suc- cessive time periods. p She put forward the hypothesis that the observed bivariate distribution arose from the superimposition of these two populations. This hypothesis was supported by her work on the variation of market penetration, number of new and repeat consumers and correlation coefficients with increasing lengths of the observational period. She found that the observations usually lay between the predictions of the correlated and un- correlated models. Firer & Barnett (l 979) found similar results in the toilet soap market. Figure 1 Observed frequency distribution for Palmolive soap - eight months (N = 614) Joannides (l 976) fonnulated a binary model by combining the correlated and uncorrelated bivariate NBDs. He showed that his model could produce many of the observed shapes, patterns (including a shelO, and values of parameters which the original NBD theory could not. The two data sets analysed were: Soup consumption displayed a seasonal pattern. It dropped during spring and rose in autumn. Three separate stationary periods could thus be identified: Soup In the early 1970s the packaged-soup market in South Africa was dominated by the leading brand (here designated brand A) followed at some distance by the only other brand with a reasonable market share (designated brand B). Nine other brands made up the rest of the market. All brands were presented in unifonn packagings of four to six servings each. Joannides (1976) identified these populations as brand-loyal and multibrand, brand-switching households. However, the same distribution shape could arise from the superimposition of a buying population (which may, for a given time period contain non-buyers who will buy in a future period) and a group of never-buyers (whose distribution would consist of a certain frequency in the zero cell only). The data referred to as 'soup' in this paper were collected by Sichel (1975) as part of a South African National Dustbin Panel Survey. Housewives were provided with a bin into which they discarded empty packets of soup, the contents of which had been consumed by their families. Every twenty-eight days (i.e. at standard monthly intervals) the housewives were visited by interviewers who recorded the contents of their bins. The objective of the research reported in this article was to test these hypotheses by investigating the proportion of never- buyers in two frequently purchased South African consumer non-durables product classes. Accepted May 1984 For the purpose of testing a distribution fit to this data it was therefore necessary to group the cells. The z.ero cell was allowed to stand alone, the one-plus-two purchase frequency cells, the three-plus-four cells, etc., were grouped together. In practice, purchase/ consumption frequency distributions can be observed in which the size of the z.ero cell appears exces- sively large when compared to the distribution as a whole. This suggests that the distribution is not drawn from a single homogeneous population, but that at least two distinct popula- tions are involved. Data Infonnation relating to 512 households' consumption for sixteen twenty-eight day periods was collated. Table 1 Non-buyers of leading brands Table 1 Non-buyers of leading brands Table 1 Non-buyers of leading brands Proportions of non-buyers Brand over period studied Lux 0,450 Palmolive 0,492 Sunlight Mild 0,590 Breeze 0,651 Shield 0,694 Soup Brand A 0,203 Soup Brand B 0,309 According to Sichel (1980), individual brands need not neces- sarily follow the same distribution as that postulated for a total product field, and therefore Goodman's (1979) omission in not studying the zero-augmented NBD as a potential model for the individual brands is open to some criticism. This study therefore began by attempting to fit a zero- augmented NBD model to the individual brand data for Lux and Palmolive soaps. Earlier research (Firer & Barnett, 1979) had shown that for all time periods studied, the chi-square test rejected the hypothesis of an ordinary NBD model for the data grouped as suggested above. Parameter estimates for the model were made on ungrouped data. As can be seen from Table 2, the ungrouped data 'failed' the chi-square test. A detailed study of the tests revealed that the 'odd-even' purchase phenomenon was responsible for a substantial proportion of the total chi-square. The tests were thus repeated using grouped data. The zero-augmented Sichel distribution (-y = - 0,5) was then fitted to the same grouped data using the same tail-end grouping of cells of expected fre- quency less than 5,0. A visual inspection of the brand purchase frequency distribu- tions revealed certain interesting phenomena. First, in almost all cases a kink or a flattening out, and sometimes even a slight maximum, occurred at about the second or third cell. In the case of the toilet soap distributions, the problem of the con- sumers' favouring of even-numbered purchases (discussed above) may have been contributing towards this effect. Second- ly, in visually sweeping a smooth curve upwards through the data points, one would have expected such a curve to intercept the frequency axis at a much lower point than that actually recorded. An example from eacq data set is shown in Figures 1 and 2. From Table 2 it can be seen that, contrary to Goodman's expectations, the better fit was provided by the zero-augmented NBD model. An example of the fits obtained is shown in Figure 3. Systematic deviations of the two models were found. The Sichel distribution was first too low, then too high, then too low again, relative to the NBD. Soap The data referred to as 'soap' originate from the toilet soap product class. In South Africa this product class comprised more than thirty-five different brands, but was dominated by the two leading brands, Lux and Palmolive. Periods one through four (the first stationary winter period): - periods seven through ten (the stationary summer period); The data base used consisted of a sample of 614 White households' bimonthly purchases of toilet soap, over the period April to November 1978, details of which were collected by Market Research Africa using the consumer panel method. - periods thirteen through sixteen (the second stationary winter period). Similar mean consumption levels were found for the two winter periods. Goodman (1979) tested the data and concluded that pur- chasing rates over the eight-month period were stationary. The observations had one peculiarity worthy of mention. When This set of data differs fundamentally from the toilet soap observations in that it contains consumption and not purchase data. Thus idosyncrasies of individual households such as the 1so 100 so 6 8 10 12 14 No. Of OOTS Coo.HO Figure 2 Observed frequency distribution for combined brands A and B of soup - first four months (N = 614) S. Afr. J. Bus. Mgmt. 1984, 15(3) frequency of shopping and the number of units purchased on each shopping occasion are removed. Consumption should therefore produce a more uniformly distributed set of data. Table 1 Non-buyers of leading brands According to Sichel (1980) the nature of the deviations led one to the conclusion that decreas- ing the value of 'Y from - 0,5 to - 1,5 or - 2,5 would not significantly improve the fit. The hypothesis put forward is that these two effects are caus- ed by an 'excess' of zeros in the first cell of the distributions, and that this excess in fact comprises never-buyers/consumers of the brand. It was anticipated that the observations could best be model- led using a zero-augmented distribution (See (c) in Appendix). It was therefore concluded that, in the toilet soap field, in- dividual brand purchase frequencies are best modelled by the univariate zero-augmented NBD. Never-Buyers Never-buyers of the product field Never-buyers of the product field In the toilet soap product field, all of the 614 households in the sample had purchased some soap during the eight-month period of the survey. Thus no never-buyers of toilet soap ex- isted in the sample. However, in the packaged-soup market, after two winter periods, 20 of the 512 households had not yet consumed. Sichel (1975) concluded that, because there were so few non-consu- mers, these 20 households did not provide sufficient evidence of the existence of never-consumers. This conclusion is accepted by the author and the analysis below is thus devoted to indivi- dual brands. Never-buyers of individual brands: Never-buyers of individual brands: Figure 2 Observed frequency distribution for combined brands A and B of soup - first four months (N = 614) A study of the soap and soup data reveals that even after eight and sixteen months respectively, substantial proportions of the households in the samples had not purchased/ consumed the leading brands (Table 1). amined only the Sichel distribution when analysing the indivi- dual brands within the product field. Results Figure 4 is portions of n tions were de were made b observations In no case dicted by Eh the hypothes brands in th Analysis of In this analys period of the data did not Although t not correct, i portion of n seen from F Table 3 Soap: Estimates of proportion of never- buyers q derived from zero-augmented models the proportion of never-buyers in the sample, whereas the pro- portion of non-buyers as predicted by Ehrenberg's theory would tend to zero with lengthening time periods (Ehrenberg, 1972). Figure 4 is an example of the observed and expected pro- portions of non-buyers of Lux soap. The expected propor- tions were derived using Equation A4. The parameter estimates were made by the method of maximum likelihood using the observations of the first two-month period. In no case did the observed proportion lie above that pre- dicted by Ehrenberg's theory. As a result of these findings, the hypothesis of the existence of never-buyers for individual brands in the toilet soap market was rejected. Results Analysis of each of the data sets is discussed separately. One of the implications of the zero-augmented model is that estimates obtained for q, the proportion of never-buyers in the sample, should be independent of time. Table 3 shows the results obtained from the fitting of both the NBD and Sichel zero-augmented models. These estimates are certainly not constant. Analysis of the Soap Data Goodman (1979) found that the fit of the Sichel distribution (Sichel, 1971) with -y = -2,5 ((b) in Appendix) was generally superior to the NBD with respect to the toilet soap product class (all brands included). Results As a result of this finding, she ex- 1S8 Table 2 Soap: Results of chi-square tests for zero-augmented distributions NBD NBD Sichel ('y = - 0,5) Ungrouped Grouped Grouped Lax 1st 2 months P(x2 ~ 1,531 I 2)< =0,465 • a 1st 4 months P(x2 !l;: 16,379 I 5) < = 0,050 P(x2 ~ 2,023 I 2) = 0,364 P(x2~ 2,548 I 2)=0,275 6 months P(x2!l;:29,982 I 8)< =0,050 P(x2~5.747 I 3)=0,124 P(x2 ~ 7,791 I 3)=0,051 8 months P(x2 ~ 19,5(,6 I 11)=0.052 P(;x2~7,385 I 5)=0,194 P(x2 ~ 11,422 I 5)<0,050 Palmolive 1st 2 months P(x2 ~ 1,482 I 2)=0,477 a a 1st 4 months P(x2!l;:25,474 I 5)<0,050 P(x2!l;:2,536 I 1)=0,lll P(x2 ~ 2,916 I 1)=0,088 6 months P(x2!l;:28,420 I 7)<0,050 P(x2!l;:5,991 I 3)=0,112 P(x2 !l;: 8,398 I 3) < 0,050 8 months P(x2~21,031 I 9)<0.050 P(x2~9,181 I 4)=0,057 P(;x2!l;: 12,931 I 4)<0,050 'No degrees of flffllom left for chi-squatt tests Table 2 Soap: Results of chi-square tests for zero-augmented distributions l10 240 21) 11kl t; 150 z a ~120 a OOSERVED o ZEl{)-AUH:NTED NBD • ZERO-~NTED smn (y=-0.5) 'JO ZERO-AUMNTED NIil : 9=0.709 q=0.371 k=168 P(JCH 3855) =0.194 60 ZER>-AUMNTED S10£L: 8=0.814 q:O 406 p:1.87 (y=-0.5 a PRICH) Ptb11.422 l5l<O 05 J) 0 0 2 4 6 8 10 12 14 16 Nn CF LfflTS PLR:HASED and titted zero-augmented distributions for Lux soap - eight months (N = 614) l10 240 21) 11kl t; 150 z a ~120 a OOSERVED o ZEl{)-AUH:NTED NBD • ZERO-~NTED smn (y=-0.5) 'JO ZERO-AUMNTED NIil : 9=0.709 q=0.371 k=168 P(JCH 3855) =0.194 60 ZER>-AUMNTED S10£L: 8=0.814 q:O 406 p:1.87 (y=-0.5 a PRICH) Ptb11.422 l5l<O 05 J) 0 0 2 4 6 8 10 12 14 16 Nn CF LfflTS PLR:HASED Fl&are 3 Observed and titted zero-augmented distributions for Lux soap - eight months (N = 614) Fl&are 3 Observed and titted zero-augmented distributions for Lux soap - eight months (N = 614) Table 3 Soap: Estimates of proportion of never- buyers q derived from zero-augmented models Lux 1st 2 months 1st 4 months 6 months 8 months Palmolive 1st 2 months 1st 4 months 6 months 8 months Proportion of never-buyers NBD Siebel 0,634 0,635 0,498 0,515 0,410 0,440 0,37) 0,407 0,686 0,686 0,474 0,503 0,450 0,485 0,376 0,430 If never-buyers do exist, one would expect the observed pro- portion of non-buyers of a product to eventually level out at the proportio portion of n would tend t 1972). Analysis of the Soup Data In this analysis the non-consumer over the entire sixteen-month period of the survey were considered, despite the fact that the data did not exhibit stationarity over the period studied. Although the use of the sixteen-month period is theoretically not correct, it does at least give some guidance as to the pro- portion of non-consumers over a lengthy period. As will be seen from Figure S, the observed point plotted at sixteen If never-buyers do exist, one would expect the observed pro- portion of non-buyers of a product to eventually level out at S. Afr. J. Bus. Mgmt. 1984, 1S(3) 1,0 --o-- OBSERVED - EXPECTED 0.4 +-----,--"'T"""----,.-----r--- 0 2 4 6 8 LEt{iTH OF OBSERVATIONAL PERIOD I 11'.JNTHS I Figure 4 Proportion of non-buyers of Lux soap as a function of length of observational period (N = 614) 1,0 ---o--- OBSERVED ......,.__ ~XPECTED 159 0,0 +--,---.,..--.....----,--,--,---,---,--- 0 2 4 6 8 1l 12 14 16 LEt«iTH Of OBSER\ATIONAL PERIOO !l'llNTHSl Figure 5 Proportion of non-consumers of brand A soup as a function of length of observational period (N = S 12) S. Afr. J. Bus. Mgmt. 1984, 1S(3) 1,0 --o-- OBSERVED - EXPECTED 0.4 +-----,--"'T"""----,.-----r--- 0 2 4 6 8 LEt{iTH OF OBSERVATIONAL PERIOD I 11'.JNTHS I Figure 4 Proportion of non-buyers of Lux soap as a function of length of observational period (N = 614) S. Afr. J. Bus. Mgmt. 1984, 1S(3) S. Afr. J. Bus. Mgmt. 1984, 1S(3) 1,0 --o-- OBSERVED - EXPECTED 0.4 +-----,--"'T"""----,.-----r--- 0 2 4 6 8 LEt{iTH OF OBSERVATIONAL PERIOD I 11'.JNTHS I 1,0 ---o--- OBSERVED ......,.__ ~XPECTED 159 0,0 +--,---.,..--.....----,--,--,---,---,--- 0 2 4 6 8 1l 12 14 16 LEt«iTH Of OBSER\ATIONAL PERIOO !l'llNTHSl Figure 5 Proportion of non-consumers of brand A soup as a function of length of observational period (N = S 12) 159 1,0 ---o--- OBSERVED ......,.__ ~XPECTED 159 0,0 +--,---.,..--.....----,--,--,---,---,--- 0 2 4 6 8 1l 12 14 16 LEt«iTH Of OBSER\ATIONAL PERIOO !l'llNTHSl Figure 5 Proportion of non-consumers of brand A soup as a function of length of observational period (N = S 12) Figure 4 Proportion of non-buyers of Lux soap as a function of length of observational period (N = 614) theory. This theory does not provide for the existence of never- buyers. A similar calculation was made for the individual brands in the soup product field. months is well within the general sweep of the curve. Analysis of the Soup Data The consumption frequency distributions of the individual brands were not modelled except for the case of the combined brands A and B over the eight stationary winter months. The results of the chi-square tests on the two zero-augmented models were: Therefore, on the basis of the available evidence in two pro- duct fields, the never-buyer/consumer hypothesis was rejected. The Binary Model of Joannides (1976) should thus be con- sidered as the explanation for the observed distribution shapes. He showed that the marginal distributions obtained by sum- ming a correlated NBD, representing loyal repeat-buyers, and an uncorrelated NBD, representing brand-switching/multi- brand buyers, displayed high zero cell frequencies as well as points of inflection or secondary modes close to the zero cell. NBD Sichel P<x2 ~ 2S, 732 I 27) = 0,531 P(x2 ~33,304 I 26)=0,166 P<x2 ~ 2S, 732 I 27) = 0,531 P(x2 ~33,304 I 26)=0,166 The NBD model, which provided the better fit, gave an estimate of zero for the proportion of never-consumers. Because of this result, the fitting of models to the individual brands was omitted, and the graphs of observed and expected (via Ehrenberg's NBD theory) proportions of non-consumers of brand A, brand B and brands A and B combined were im- mediately studied. Thus the results obtained, together with a consideration of the Joannides model, lead one to the conclusion that perhaps brand-disloyal buyers/ consumers are better represented by a distribution of frequencies rather than the subset thereof (zero cell frequencies only) implied by the never-buyer hypothesis. These results have important implications for the marketers of consumer non-durables. Instead of devoting time and ef- fort in persuading never-buyers to use their products (in a generic sense), they would be strongly advised to undertake advertising campaigns aimed at wooing existing users of other brands of the product into trying their particular brand. The parameter estimates for brands A and B were based on averages of each of the first four months' estimates, calculated using the method of moments. For the combined brands the maximum likelihood parameter estimates for the first month were used. It was found that the observed pro- portions of non-consumers approached zero more rapidly than the expected proportions as predicted by the Ehrenberg theory. An example of the observed and expected proportions of non- consumers is shown in Figure 5. References Ehrenberg, A.S.C. 1972. Repeat buying: theory and applications. Amsterdam: North Holland Publishing Co. Analysis of the Soup Data Finally, this study highlighted the need to model the field of consumer non-durable purchase patterns from a multibrand perspective, rather than by looking at individual brands in isola- tion. Research in this area has been undertaken and will be reported on in the future. On the basis of this evidence, one must therefore also re- ject the hypothesis of the existence of never-consumers of in- dividual brands in the two data sets studied. (a) The NBD Repeat-buying Model (i) Univariate Case The model assumes that an individual household's purchases are characterized by a Poisson distribution. In order to study market behaviour as a whole, the individual household's distributions arc compounded using a 'Y distribution for the individual average purchase frequencies per unit time (Sichel, 1971). The result is the Negative Binomial Distribution (NBD): Ax I fJ = [kl(k + tm)lk rex + k) ltml(k + tm)]X (Al) r(k) r(x+ I) where xis the number of purchases in time t, 0< m < co ,0< k < oo , and x = 0,1,2, ... The proportion of buyers of the product in time t is given by b1 = 1 - (I +tmlkF" (A2) where a buyer is defined as one who buys at least one unit during time period t. (ii) The Bivariate Correlated NBD The bivariate model is obtained by mixing the product of two univariate Poisson distributions with the 'Y distribution, to form the bivariate correlated NBD model. In its symmetric form (where successive purchase periods are of equal length) it can be written Jtx I fJ = (l+Umlk)-1c r (x+y+k) [tml(k+Um)f+" (A3) ,Y r(k) r(x+ 1) re,,+ I) where k>O, m >0, k(fJ = k, m(t)= tm, x=0,1,2, ... , y=0,1,2, ... The market penetration (proportion of buyers of the product in time fJ is given by c"t = l -(1 + tmlk)-k (A4) (iii) The Uncorrelated Bivariate NBD Sichel (1980) derived the uncorrelated bivariate NBD in an attempt to model the behaviour of product-disloyal consumers (brand- switchers and multibrand buyers) whose consumptions are uncor- related between two time periods. The uncorrelated NBD is obtain- ed from the product of two univariate NBDs. It can be written in the form ,t, (r) represents a family of discrete distributions. Many of the bet- ter known discrete distributions such as the NBD, Poisson, geometric, etc., are special or limiting forms of 1/J(.r) (Sichel, 1974). Making gamma negative allows one to generate an entirely new set of discrete distributions. The distributions are mathematically dif- ficult to handle unless 'Y is a half-integer. -y could therefore be set, a priori, to - O,S and values of - l ,S or - 2,S would be used in order to try and improve the fit if 'Y = - O,S is not satisfactory. Two parameters, a and 0, have therefore to be estimated. Concluslons • r+} l r = 0,1,2, ... (A7) Sichel, H.S. 1980. Private Communication. _ (vf=e)l (a0/2)' K (a) 1/J(.r) - K (a(v' I - 0))r! • r+} l Sichel, N. 197S. Problem areas in repeat consumption models. Un- published M.B.A. dissertation, University of the Witwatersrand, Johannesburg. Tucker, A.E. 1973. The repeat purchase patterns of non-durable con- sumer goods in selected households. Unpublished M.B.A. disser- tation, University of the Witwatersrand, Johannesburg. r = 0,1,2, ... (A7) r = 0,1,2, ... (A7) (A7) It is derived by mixing the Poisson distribution with the flexible mixing distribution Zachos, G. 1977. Non-durable consumption patterns of individual households. Unpublished M.B.A. dissertation, University of the Witwatersrand, Johannesburg. ft>.) = j2v'{r=e)/ a0P • >. ~ - 1 • exp! _ ( _!_) >. _ a 2 e (AS) 2K (av'l -0) 0 4 } ' (AS) Appendix A brief description of the statistical distributions mentioned in the text is given here. In this distribution - oo < -y < oo , O < 0 < I and a >0 are the three parameters and K_r (.) is the modified Bessel function of the second kind or order (.). (a) The NBD Repeat-buying Model If the proportion of observed frequenci~ in the zero cell is large (general- ly above S00fo), the method of equating mean and zero cell propor- tions is fairly efficient (Sichel, 1973). If this is not the case, a maxi- mum likelihood method is available (Sichel, 1971). However, this method is very complicated, so when the i.ero cell is small, the method of moments is generally used (Sichel, 1974). (A2) (c) Concluslons Firer, C. 1980. The development of new mathematical models for multibrand buying. Unpublished M.B.A. dissertation, University of the Witwatersrand, Johannesburg. For the two data samples available, no evidence could be found to support the hypothesis that the apparent excessive size of the zero cell in the purchase/ consumption frequency distribu- tion was due to the superimposition of a buying/consuming population and a group of never-buyers/consumers. Firer, C. & Barnett, A.E. 1979. Testing the NBD repeat buying model for exposure periods of varying lengths. Unpublished Report. p Goodman, M.A. 1979. Multibrand purchasing behaviour - theory and applications. Unpublished M.B.A. dissertation, University of the Witwatersrand, Johannesburg. The q parameters of the z.ero-augmented NBD (which it was hoped would provide an estimate of the proportion of never- buyers of toilet soap) were not constant over the various time periods studied. In addition, with lengthening time periods, the observed proportion of non-buyers decreased at almost ex- actly the same rate as that predicted by the Ehrenberg NBD Joannides, J. 1976. A new distribution for repeat consumption. Un- published M.B.A. dissertation, University of the Witwatersrand, Johannesburg. g Morrison, D.G. 1969. Conditional trend analysis: a model that allows S.-Afr. Tydskr. Bedryfsl. 1984, 1S(3) 160 k>O, m>O, k(t)=tk, m(t)=tm, x=0,1,2, ... , y=0,1,2, ... for non-users. J. Market. Res., VI, 342-346. for non-users. J. Market. Res., VI, 342-346. Sichel, H.S. 1971. On a family of discrete distributions particularly Sichel, H.S. 1971. On a family of discrete distributions particularly suited to represent long-tailed frequency data. In Proceedings of the Third Symposium on Mathematical Statistics (N.F. Laubscher, In the case of this distribution, the proportion of buyers is ed.). S.A. C.S.l.R. (Pretoria), Sl-97. ed.). S.A. C.S.l.R. (Pretoria), Sl-97. ,/'r = I - (1 + ml k) - '" (A6) ,/'r = I - (1 + ml k) - '" (A6) Sichel, H.S. 1973. Statistical valuation of diamondiferous deposits . ,/'r = I - (1 + ml k) - '" The Sichel distribution (A6) (A6) Journal of the South African Institute of Mining and Metallergy, 23S-239. The Sichel distribution The Sichel distribution The Sichel distribution (b) Sichel, H.S. 1974. On a distribution representing sentence length in written prose. Journal of the Royal Statistical Society, 137(1), 2S-34. This distribution was also derived by Sichel in 1971. It can be writ- ten in the form _ (vf=e)l (a0/2)' K (a) 1/J(.r) - K (a(v' I - 0))r! Zero-augmented distributions A large number of observed purchase/consumption frequency distributions have an (what appears to the eye), excessively large mo cell. Irrespective of the true cause of this effect, one requires to be able to model such distributions. In order to do so, it is assumed that this cell is made up of two separate groups. The situation is illustrated graphically in Figure Al· An implication of this model is that, if a hard core of never-buyen does exist, the estimate of IJ should be independent of time. The values of IJ for different brands within a product field need not, of course, be similar. course, be similar. ~r ~q 1i / / / / / .... .... .... ( 1-q )do (0) -- 0 1 2 3 i. S r-.,_ OF I..HITS PIR:HASEO / CON9.Jt£0 fla11n Al A graphical representation of never-buyers ~r ~q 1i / / / / / .... .... .... ( 1-q )do (0) -- 0 1 2 3 i. S r-.,_ OF I..HITS PIR:HASEO / CON9.Jt£0 fla11n Al A graphical representation of never-buyers Sichel (1980) proposed a similar modd based on the Sichel distnbu- tion. Here, too, three parameters require estimation: a and 0 of the Sichel, and lj the proportion of never-buyers. The techniques of estimation of the parameters for both distribu- tions are discussed by Firer (1980). The techniques of estimation of the parameters for both distribu- tions are discussed by Firer (1980). Sichel (1980) proposed a similar modd based on the Sichel distnbu- tion. Here, too, three parameters require estimation: a and 0 of the Sichel, and lj the proportion of never-buyers. S. Afr. J. Bus. Mgmt. 1984, 15(3) 161 Morrison (1969) showed that the existence of never-buyers would introduce a systematic bias into the NBD model. He therefore pro- posed a model of the form given above. In this model three parameters require estimation: the k and m parameters of the NBD and lj, the proportion of never-buyers. Sichel (1980) proposed a similar modd based on the Sichel distnbu- tion. Here, too, three parameters require estimation: a and 0 of the Sichel, and lj the proportion of never-buyers. The techniques of estimation of the parameters for both distribu- tions are discussed by Firer (1980).
W3176345680.txt	https://www.researchsquare.com/article/rs-544454/latest.pdf	en		Industrial ecological efficiency of cities in the Yellow River Basin in the background of China’s economic transformation: spatial-temporal characteristics and influencing factors	Environmental science and pollution research international	2,021	cc-by	10,098	Industrial Ecological E ciency of Cities in the Yellow River Basin in the Background of China's Economic Transformation: Spatial-Temporal Characteristics and In uencing Factors Chengzhen Song Shandong Normal University Guanwen Yin Shandong Normal University Zhilin Lu Shandong Normal University Yanbin Chen (  ychen@sdnu.edu.cn ) Shandong Normal University https://orcid.org/0000-0001-6259-595X Research Article Keywords: Industrial ecological e ciency, Super-e ciency DEA model, Economic transformation, Yellow River Basin Posted Date: June 25th, 2021 DOI: https://doi.org/10.21203/rs.3.rs-544454/v1 License:   This work is licensed under a Creative Commons Attribution 4.0 International License. Read Full License Version of Record: A version of this preprint was published at Environmental Science and Pollution Research on August 17th, 2021. See the published version at https://doi.org/10.1007/s11356-021-159642. Industrial ecological efficiency of cities in the Yellow River Basin in the background of China's economic transformation: spatial-temporal characteristics and influencing factors Chengzhen Song Postgraduate student College of Geography and Environment Shandong Normal University Jinan 250358, China Email: songchengzhensdnu@163.com Guanwen Yin Associate professor College of Geography and Environment Shandong Normal University Jinan 250358, China Email: yingw@sdnu.edu.cn Zhilin Lu Postgraduate student College of Geography and Environment Shandong Normal University Jinan 250358, China Email: 496260672@qq.com Yanbin Chen Lecturer College of Geography and Environment Shandong Normal University Jinan 250358, China Email: ychen@sdnu.edu.cn (Corresponding author) 1 1 Industrial ecological efficiency of cities in the Yellow River Basin 2 in the background of China's economic transformation: spatial- 3 temporal characteristics and influencing factors 4 Abstract: At present, China's economic development has entered a "new normal." Exploring 5 Industrial ecological efficiency (IEE) in the background of economic transformation is of great 6 significance to promote China's industrial transformation and upgrading and achieving high-quality 7 economic development. Based on the Super-Efficiency DEA model, this study evaluated the IEE of 8 cities in the Yellow River Basin from 2008 to 2017. Exploratory spatial data analysis methods were 9 used to explore the spatial-temporal evolutionary characteristics, and a panel regression model was 10 established to explore the influencing factors of IEE. The research results showed that: The IEE in 11 the Yellow River Basin exhibited an elongated S-shaped evolutionary trend from 2008 to 2017, and 12 the mean IEE of cities presented a trend whereby Yellow River Basin’s regions could be ranked in 13 the following order: lower reaches > middle reaches > upper reaches. There was significant spatial 14 autocorrelation of the IEE in the Yellow River Basin, and the hot and cold spots showed an obvious 15 "spatial clubs" phenomenon. The results of panel regression show that the influence factors of IEE in 16 the Yellow River Basin showed spatial heterogeneity in their effect. 17 Keywords: Industrial ecological efficiency; Super-efficiency DEA model; Economic transformation; 18 Yellow River Basin 19 1. Introduction 20 Since China’s economic reform and opening up, it has relied on the traditional extensive industrial 21 development pattern of “high resource input, high energy consumption, and high pollution emissions,” 22 enabling the economy to achieve high-speed growth in a short period of time (Wang et al. 2016). However, 2 23 the traditional industrial development model has also led to problems such as excessive resource 24 consumption, serious environmental pollution, and low production efficiency (Yu et al. 2015, Yu et al. 25 2018). At present, the Chinese economy has shifted from a high-speed growth stage to a high-quality 26 development stage. Achieving the coordinated development of industrial economic growth, resource 27 conservation, and environmental protection has become the only way to promote China's ecological 28 civilization construction and high-quality economic development. 29 In the early 1990s, Schaltegger andSturm (1990) proposed the concept of eco-efficiency, or the ratio 30 between the increase in economic output and the increase in environmental impact, to measure the 31 environmental performance of economic development. The basic idea of eco-efficiency is to create 32 greater economic benefits with lower resource consumption and environmental cost. IEE not only reflects 33 the eco-efficiency of economic growth, but it also expresses the efficiency of industrial production under 34 environmental constraints. It is an effective means to characterize the relationship between industrial 35 development and the ecological environment (Lin &Zhu 2020, Zhou et al. 2018). IEE can take into 36 account the relationship between industrial economic material exchange and energy conversion and 37 comprehensively evaluates the overall status of industrial economy, energy consumption, and 38 environmental benefits. It is of great significance for transforming the traditional industrial development 39 model, promoting green industrial development, and supporting the construction of an ecological 40 civilization in China. 41 Scholars have conducted research on IEE mainly from the following aspects: First, at the regional 42 scale, recent research of IEE has mainly focused on spatiotemporal changes in different countries and 43 provinces. For example, Marques et al. (2019) explored the IEE of 11 EU countries from 1997 to 2015. 44 Zhou et al. (2020) studied the eco-efficiency of 31 industrial sectors in China. Zhang et al. (2017a) 3 45 analyzed the IEE of 30 provinces (autonomous regions and municipalities) in China from 2005 to 2013. 46 Second, in terms of methodology, scholars usually adopt the traditional DEA model (Shah et al. 2020, 47 Wu et al. 2015, Zhang et al. 2008) for quantitative evaluation of the IEE. This method has strong 48 objectivity and effectively avoids the interference of human factors, but when the value of each effective 49 DMU calculated by the DEA model is 1, it is impossible to compare and distinguish effective DMUs. 50 Finally, in terms of influencing factors, the Panel Tobit model and other methods are mostly used to 51 explore the level of economic development (Haoran et al. 2019, Liu et al. 2020a), technological progress 52 (Chen et al. 2019), foreign investment (Zhao et al. 2018), environmental regulation (Liu et al. 2020b), 53 industrial structure (Wencong et al. 2018, Zhou et al. 2019), and other factors on IEE. 54 However, there have been relatively few studies of the effects of economic transformation on IEE. 55 Globalization, marketization, and decentralization, as the three important forces of economic 56 transformation, play an important role in reconstructing the spatial pattern of China's regional economy. 57 Since its reform and opening up, China's industrialization process has also been deeply affected by 58 economic transformation. Therefore, it is of great practical significance to explore IEE in the background 59 of economic transformation. In addition, ecological protection and high-quality economic development 60 in the Yellow River Basin was established as a major national strategy in September 2019 (Wang et al. 61 2021). However, the ecological environment of the Yellow River Basin is extremely fragile, and the 62 heavy chemical industry is the development pillar of many provinces in the basin. The continuous 63 development of heavy chemical industry has caused serious pollution in the ecological environment of 64 the Yellow River Basin (Wang et al. 2020). Therefore, it is of great significance to explore the impact 65 mechanism of IEE in the background of economic transformation to promote the ecological environment 66 protection and high-quality development in the Yellow River Basin. 4 67 On this basis, this study used the Super-Efficiency DEA model to measure the IEE of the Yellow 68 River Basin from 2008 to 2017 and reveal its spatial distribution characteristics. The panel regression 69 model was used to analyze the influencing factors of the IEE of the Yellow River Basin in the background 70 of economic transformation. Compared with other studies, this study was advanced in the two aspects 71 that follow. (1) Based on the relationship between economic transformation and IEE, a theoretical 72 analysis framework of the impact of economic transformation on IEE was constructed. (2) The Super- 73 Efficiency DEA model was introduced to evaluate the IEE, which addresses the problem that it is 74 impossible to compare multiple effective DMUs when the calculation results of the traditional DEA 75 model show that multiple effective DMUs are 1. 76 The other parts of this paper were arranged as follows. The second section describes the theoretical 77 framework of economic transformation and IEE. The third section is the overview of the research area, 78 model methods, and data sources. The fourth section is the research results and discussion. The fifth 79 section is the summary of the research results. 80 2. Economic transformation and IEE 81 China's economy has experienced a transformation from planned economy to market economy, from 82 closed development to opening up, and into the process of economic globalization. Economic 83 transformation generally refers to the process of globalization, marketization, and decentralization. In 84 recent years, experts and scholars have gradually realized the important role of economic transformation 85 and have explored regional economic development (Jiya et al. 2020), industrial pollution (Zhiqing et al. 86 2018), and land use (Huang et al. 2015, Liu et al. 2018) from the perspective of economic transformation. 87 Therefore, this study attempted to explore the impact mechanism of globalization, marketization, and 88 decentralization on the IEE of cities in the Yellow River Basin from the perspective of economic 5 89 transformation. 90 Globalization plays a vital role in promoting China's industrialization, and its impact on IEE is 91 mainly characterized by foreign direct investment (FDI) (Pao &Tsai 2011). FDI may have a positive or 92 negative impact on IEE. On the one hand, FDI may have a positive impact on the IEE of developing 93 countries (Feng 2014); the "pollution halo" hypothesis posits that foreign-funded enterprises may bring 94 advanced equipment and technology to developing countries and promote the industrial technology 95 upgrading of developing countries through the technology spillover effect so as to reduce energy 96 consumption and pollution emissions (Jiang et al. 2019). On the other hand, FDI may have a negative 97 impact on the IEE of developing countries (Lan et al. 2012, Liu et al. 2017b); the "pollution haven" 98 hypothesis holds that the migration of foreign-funded enterprises from developed countries to developing 99 countries with less environmental regulation not only promotes the economic growth of developing 100 countries, but it also aggravates the local resource and energy consumption and leads to an increase in 101 pollutant emissions and the deterioration of the ecological environment (Bao et al. 2015). 102 The effect of marketization on IEE is mainly reflected in promoting the redistribution of elements 103 among industrial production sectors and promoting the reform of the market system. Marketization of 104 the allocation of resources can break barriers to the flow of resources, promote the cross-regional and 105 cross-industry flow of resources, and improve the efficiency of the allocation of production factors, which 106 is conducive to reducing the cost of obtaining production factors, improving the survival ability of 107 enterprises, and improving the production efficiency of enterprises (Campos &Horváth 2012, Merlevede 108 2003). Moreover, Yi et al. (2015) believe that marketization can promote reform of the market system, 109 which is conducive to attracting foreign-funded enterprises, introducing advanced industrial technology 110 and management experience, promoting industrial enterprises' production efficiency, and reducing 6 111 industrial pollution emissions. 112 Decentralization is also an important feature of China's economic transformation. China's 113 decentralization process endows local governments with more independent development rights and also 114 makes local governments bear the responsibility of regional economic development (Yin et al. 2018). At 115 present, there are mainly three views on the effect of decentralization on IEE. One view holds that 116 decentralization is conducive to the improvement of IEE; decentralization refers to the transfer of 117 economic and financial power from the central government to the local governments, which is conducive 118 to improving the allocation efficiency of local resources and reducing industrial pollution emissions 119 (Chen &Chang 2020, Mu 2018). Another view is that decentralization is not conducive to improving IEE. 120 In order to improve the performance in promoting local economic development, local governments 121 relaxed the regulatory environment to gain more space for economic development, thus promoting the 122 increase of industrial pollution emissions (Zhang et al. 2017b). The last point of view is that the 123 relationship between decentralization and IEE is non-linear. Under fiscal decentralization, in order to 124 improve political performance, local governments must not only promote local economic development, 125 but also improve the quality of the local ecological environment (Yang et al. 2020). Some scholars have 126 found a nonlinear relationship between decentralization and pollution emissions through research (Cheng 127 et al. 2019, Liu et al. 2017a). 128 This study was based on the theoretical research of economic transformation and took into account 129 the regional economic development level and industrial development status of the Yellow River Basin. 130 In addition, based on the existing research results (Guan &Xu 2016, HA et al. 2020, Ren et al. 2020), and 131 by adding basic factors such as industrial structure, economic development level, scientific and 132 technological innovation, and population and industrial agglomeration, this study constructed a 7 133 theoretical analysis framework of IEE in the background of economic transformation (Fig. 1). Economic transformation Globalization Marketization Decentralization Foreign direct investment Proportion of non-stateowned economy Proportion of local fiscal expenditure Economic development Factor resource allocation Economic development Skill improved Market system reform Environmental protection Industrial ecological efficiency Basic factors Industrial structure The level of economic development Technological innovation Industrial Cluster 134 135 Fig. 1. Theoretical analysis framework of Economic transformation and IEE. 136 3. Research setting and methodology 137 3.1.Study area 138 The Yellow River Basin originates in the northern foot of Bayankala Mountain, flows eastward 139 through Qinghai, Sichuan, Gansu, Ningxia, Inner Mongolia, Shaanxi, Shanxi, Henan, and Shandong, and 140 then flows into the Bohai Sea from Kenli County of Shandong Province. In order to ensure the continuity 141 of the region and learn from the definition of the Yellow River Basin in the study of physical geography 142 and social economy by relevant scholars (Zhang &Miao 2020), we considered that Sichuan province 143 belongs to the Yangtze River Basin, and the East Four Leagues of Inner Mongolia belong to Northeast 144 China, so they were not included. Therefore, in this study, the remaining 8 provinces and 91 cities 145 (prefectures and leagues) were selected as research objects (Fig. 2). Referring to the Yellow River 8 146 Yearbook, combined with the natural environment and socio-economic characteristics of the provinces 147 in the Yellow River Basin, this study defined the Qinghai Province, Gansu Province, Inner Mongolia 148 Autonomous Region, and Ningxia Hui Autonomous Region as the upper reaches, Shaanxi Province and 149 Shanxi Province as the middle reaches, and Henan Province and Shandong Province as the lower reaches. 150 151 Fig. 2. Scope of the Yellow River Basin. 152 3.2.Methodology 153 (1) Super-efficiency DEA model 154 In the 1970s, American scholars Charnes, Cooper, and Rhodes first proposed the DEA model. This 155 model is an analysis model based on "relative efficiency evaluation," which can better eliminate the 156 interference of human factors when evaluating the efficiency of DMUs (Charnes et al. 1979), However, 157 in the model calculation results, when the value of each effective DMU calculated by the DEA model is 158 1, it is impossible to compare and distinguish effective DMUs. In order to solve this problem, Andersen 159 andPetersen (1993) proposed a Super-efficiency DEA model which was based on the traditional DEA 160 model. This model makes up the defect of the traditional DEA model, which can make the effective DMU 161 efficiency value greater than 1, and can compare multiple effective DMUs. 162 In view of the advantages of the Super-efficiency DEA model for efficiency evaluation, this study 9 163 used the Super-efficiency DEA model to evaluate the IEE of cities in the Yellow River Basin. The Super- 164 efficiency DEA model is as follows: 165 Min   n    X j  j  s  X 0  i 1, jn1 s.t. X j  j  s   Y0   i 1, j 1   0, J  1,2,  k  1, k  j  s   0, s   0 166 where  is the value of IEE; X and Y are input variables and output variables, respectively;  is the 167 combination proportion of effective DMUs to represent the return of scale of DMUs; 168 variables and residual variables, respectively. When  >1, the IEE is optimal; when  <1, the IEE is not 169 optimal. 170 (2) Exploratory spatial data analysis (ESDA) , (1) s and s  are slack 171 The ESDA mainly includes global spatial autocorrelation (Moran &P.A.P 1948) and local spatial 172 autocorrelation (Getis &Ord 1992). Global spatial autocorrelation is a generalization of the spatial 173 dependence of economic phenomena or attributes with adjacency in a certain space. The most commonly 174 used correlation index is global Moran ' sI . n 175 I n nwij ( xi  x)  (x j  x) i 1 j 1 n n  n  wij   (xi  x)2    i 1 j1  i1 , (2) n is the number of units in the analysis space; x i and xj are 176 where I is the global Moran index; 177 the observed values of space unit i and j, respectively; x is the mean value of observation values of all 178 space units; and w ij is the spatial weight matrix of spatial elements i and j . 179 Hot spot analysis is a judgment method of local spatial autocorrelation, which can directly reflect 180 the agglomeration of high and low values in a certain area and the degree of agglomeration (Chen et al. 10 181 2021). In this study, the Getis  OrdGi * index reflecting local spatial autocorrelation was used to measure 182 the local spatial agglomeration of IEE in the Yellow River Basin. The specific formula is as follows: n W ( d ) X ij Gi  * 183 j 1 j 1 (i  j )，Z (G )  * i n X j Gi  E (Gi* ) j Var(Gi* ) , (3) * 184 where W ij is the weight of space, and is 1 if the space is adjacent or 0 if the space is not adjacent. E ( G i ) 185 and Var (Gi ) are the mathematical expectation and variance of Gi , respectively. If the value of 186 Z ( G i* ) is positive and significant, the surrounding values of the spatial unit i are high, and it is a hot 187 spot in the region; if the value of Z ( G i ) is negative and significant, the unit is a cold spot in the region. * * * 188 189 3.3. Dependent and independent variables 190 Based on the core idea of "low input and high output" and referring to relevant research results (J. 191 et al. 2009, Tong et al. 2020), combined with the industrial development status of various regions in the 192 Yellow River Basin, the industrial power consumption and industrial wastewater discharge were taken 193 as the input index of resource consumption, the industrial SO2 emission and industrial dust emission were 194 taken as the input index of environmental pollution, and the total industrial output value was taken as the 195 output index. Descriptive statistical results of the input and output indicator data are shown in Table 1. 196 Table 1. Descriptive statistics of input and output data. Indicators Variables Unit Mean Std. Dev. Min Max Input Industrial 100 million 62.36 77.41 0.13 991.61 electricity KW·h 10 kilo-tons 5424.54 5057.39 99 33007 ton 63699.62 51176.44 1016 321133 consumption Industrial wastewater discharge Industrial SO2 emissions 11 Industrial ton 49558.46 228183.33 450 5168812 Industrial 100 million 2,709.07 3,161.44 8.36 16,811.83 output yuan smoke and dust emissions Output 197 198 In terms of model construction, based on the perspective of economic transformation, this study 199 identified IEE as the dependent variable, and globalization, marketization, and decentralization as the 200 core explanatory variables. In addition, by referring to the relevant research literature (Guan &Xu 2016, 201 HA et al. 2020, Ren et al. 2020) combined with the development status of the Yellow River Basin and 202 the availability of data, the five indicators of industrial structure, economic development level, scientific 203 and technological innovation, population agglomeration, and industrial agglomeration were selected as 204 the control variables. The panel regression model was constructed to explore the driving factors of IEE 205 in the Yellow River Basin in the background of economic transformation, and the relevant variables and 206 their meanings are shown in Table 2. In order to avoid the influence of heteroscedasticity on the 207 regression results, the index data were logarithmically processed, and the descriptive statistical results of 208 the indicator data are shown in Table 3. 209 Table 2. Explanation of urban IEE and impact indicators. Variables Indicators Definition Codes Dependent variable IEE IEE value ep Explanatory variable Globalization FDI/urban population glo Marketization Non-state-owned economy / Total mar industrial output value Decentralization Urban per capita fiscal dec expenditure/National per capita fiscal expenditure Control variable Industrial structure The proportion of secondary industry in str GDP The level of economic GDP per capita pgdp Technological The proportion of science and technology tec innovation expenditure to local fiscal expenditure development 12 Population City population/area pop Industrial output value/area agg agglomeration Industrial Cluster 210 211 Table 3. Descriptive statistics of variables. Indicators Observation Mean Std. Dev. Min Max lnep 800 0.182 0.196 0.006 1.263 lnglo 800 -1.183 1.815 -6.604 2.550 lnmar 800 1.234 1.571 -5.464 4.599 lndec 800 1.320 0.485 0.078 3.170 lnstr 800 -0.315 1.529 -4.577 2.918 lnpgdp 800 3.895 0.273 2.608 4.471 lntec 800 2.347 0.068 2.103 2.502 lnpop 800 7.042 1.662 1.826 9.987 lnagg 800 1.692 0.230 0.468 1.976 212 213 3.4. Data Sources 214 The research objects of this study were 91 cities in the Yellow River Basin. However, because of 215 the inaccessibility of some cities' data, 80 cities in the Yellow River Basin were selected as the research 216 samples. The data of the aforementioned variables were obtained from the 2009–2018 Statistical 217 Yearbook of Chinese Cities, the Statistical Yearbook of Shandong, Statistical Yearbook of Henan, the 218 Statistical Yearbook of Shanxi, the Statistical Yearbook of Shaanxi, Statistical Yearbook of Inner 219 Mongolia, the Statistical Yearbook of Ningxia, Statistical Yearbook of Gansu, the Qinghai Statistical 220 Yearbook, and the statistical yearbook data of some prefecture-level cities in the study area from 2009 to 221 2018. 222 4. Results and discussion 223 4.1. The temporal evolution characteristics of IEE 224 Owing to the large regional differences in the upper, middle, and lower reaches of the Yellow River 225 Basin, the IEE of the whole basin and different regions were studied individually. The evolution results 13 226 of the mean IEE are shown in Fig. 3. From 2008 to 2017, the mean IEE of the Yellow River Basin showed 227 an elongated S-shaped evolutionary trends. The mean IEE of the whole region rose from 0.102 in 2008 228 to 0.316 in 2016, when it reached its highest value, and fell to 0.291 in 2017, indicating that the IEE of 229 the Yellow River Basin began to deteriorate after a long period of improvement. In the past few years, 230 local industrial restructuring in the Yellow River Basin has hampered the development of local enterprises, 231 resulting in a sharp drop in the production efficiency of local enterprises in a short period of time. 232 Regional differences of IEE were as follows: lower reaches > middle reaches > upper reaches. 233 Among them, the mean IEE of the upper reaches was always lower than that of the whole basin, and the 234 mean IEE of the middle reaches showed an evolution trend of "first rising and then declining," which 235 was consistent with the evolution curve of the whole basin. The mean IEE of the lower reaches was 236 always higher than the mean IEE of the whole basin. Moreover, the mean IEE of the lower reaches had 237 been steadily rising before 2015, jumped in 2016, and then stabilized in 2017. In summary, there were 238 obvious regional differences in the IEE of the Yellow River Basin, which may have been closely related 239 to the regional differences in the level of economic development. The higher the level of economic 240 development, the higher the IEE. Therefore, it is important to promote the high-quality development of 241 the Yellow River Basin to improve the level of economic development, promote industrial transformation 242 and upgrading, and reduce industrial pollution emissions. 14 243 244 Fig. 3. The mean IEE in the Yellow River Basin from 2008 to 2017. 245 In order to further analyze the temporal evolution characteristics of regional differences of IEE in 246 the Yellow River Basin, four years (2008, 2011, 2014, and 2017) were selected, and the kernel density 247 estimation method was used to analyze the regional differences of IEE in the Yellow River Basin in 248 different periods (Fig. 4). From 2008 to 2017, the kernel density curve of the whole basin showed the 249 characteristics of a long-tail rising peak, the peak gradually decreased, and the peak type changed from 250 narrow peak to wide peak. The change interval increased continuously, which indicated that the overall 251 level of IEE in the Yellow River Basin was improving, but the regional gap was widening. In the middle 252 and lower reaches of the region, the kernel density curve showed that the main peak gradually shifted to 253 the right, the peak decreased continuously, and the peak changed from narrow peak to wide peak, which 254 indicated that the level of IEE in the middle and lower reaches of the region was increasing while the 255 regional gap was widening. The kernel density curve of the upper reaches showed that the peak value 256 gradually decreased and the right tailing phenomenon gradually weakened, which indicated that the 257 overall level of IEE in the upper reaches was decreasing while the regional gap was gradually decreasing. 15 258 259 260 Fig. 4. Regional differences in IEE from 2008 to 2017. 4.2.The spatial evolution characteristics of IEE 261 The spatial visualization results of IEE in the Yellow River Basin in 2008, 2011, 2014, and 2017 are 262 shown in Fig. 5. The higher IEE zone in the Yellow River Basin showed the spatial distribution 263 characteristics of “scattered Mosaic.” The overall level of IEE presented a spatial differentiation pattern, 264 in which lower reaches were higher than middle reaches and higher than upper reaches. 265 Specifically, in 2008, the higher IEE zone included Yantai and Weihai in the lower reaches and 266 Ordos and Qingyang in the upper reaches. In 2011, the higher IEE zone included Qingdao, Yantai, Weihai, 267 and Dongying in the lower reaches, Luliang in the middle reaches, and Qingyang in the upper reaches. 268 In 2014, the higher IEE zone included Qingdao, Yantai, Weihai, and Dongying in the lower reaches, and 269 Weinan and Shangluo were added in the middle reaches. In 2017, the higher IEE zone was transformed 270 into six cities in the lower reaches, including Qingdao, Weihai, Dongying, Puyang, Zhoukou, and Luohe, 271 and Ankang City was added in the middle reaches. The trend surface analysis was used to identify the 272 geographic trends of the IEE in the Yellow River Basin from 2008 to 2017. The trend surface analysis 16 273 results are shown in Fig. 6. The IEE of the Yellow River Basin in the east-west direction evolved from 274 "high at both ends and low in the middle" in 2008 to "high in the east and low in the west" in 2017. In 275 the north-south direction, the IEE changed from the inverted "U" pattern in 2008 to the gradient spatial 276 pattern of "high in the south and low in the north" in 2017, and the curvature of the curve decreased 277 significantly, demonstrating that the IEE of the Yellow River Basin improved significantly during the 278 research period, and the southeast region became the high-value area of IEE, while the northwest region 279 became the weak area of IEE. 280 The IEE in the lower reaches of the Yellow River Basin was relatively stable and at a high level, 281 while the IEE in the middle and upper reaches of the Yellow River Basin was relatively low. Qingdao, 282 Yantai, Weihai, and other coastal cities in the lower reaches had an early start in industrial development 283 and a high level of economic development. On the one hand, the good industrial base has strongly 284 promoted the development of regional economy. On the other hand, the higher level of economic 285 development has forced the upgrading of regional industrial technology, which is conducive to the higher 286 level of IEE in the lower reaches. Shanxi and Shaanxi provinces in the middle reaches are both large in 287 coal resources, and the traditional industrial development model of "high consumption and high 288 pollution" has been dominant for a long time, resulting in low IEE. The ecological environment in the 289 upper reaches is very fragile. Owing to the late start of industrial development, industrial technology and 290 management experience are relatively backward, and the level of IEE in this region is far lower than that 291 in the middle and lower reaches. 17 292 293 Fig. 5. The IEE of the Yellow River Basin in 2008, 2011, 2014 and 2017. 294 295 296 Fig. 6. The overall trend of IEE in 2008 and 2017. 4.3.Spatial correlation analysis of IEE 297 The global Moran's I index calculated by GeoDa software is shown in Table 4. The global Moran's 298 I of the IEE in the Yellow River Basin from 2008 to 2017 were all positive and passed the significance 299 test at the level of 1%, indicating that there was a significant and positive spatial autocorrelation of the 300 IEE in the Yellow River Basin during the research period. At the same time, the global Moran's I index 301 presented an evolutionary characteristic of "rising first and then decreasing," reflecting that the IEE of 302 the Yellow River Basin has gone through a process of increasing spatial agglomeration and then 18 303 decreasing spatial agglomeration. 304 Table 4. Global Moran’s I Index of IEE in the Yellow River Basin. Year Moran’s I Z-score P-value 2008 0.2716 5.2292 0.0020 2009 0.3170 6.2012 0.0020 2010 0.2884 5.1526 0.0020 2011 0.2892 5.0120 0.0020 2012 0.2494 4.4847 0.0020 2013 0.3667 6.4440 0.0020 2014 0.3279 5.2251 0.0020 2015 0.2639 4.3143 0.0020 2016 0.3348 6.2003 0.0020 2017 0.3234 5.5456 0.0020 305 306 In order to further explore the local spatial agglomeration of IEE in the Yellow River Basin, the 307 Getis-Ord Gi* index of IEE of cities in the Yellow River Basin was measured by ArcGIS software, and 308 the spatial distribution map of cold and hot spots was created (Fig. 7). There was an obvious "spatial 309 club" phenomenon in the distribution of cold and hot spots of urban IEE in the Yellow River Basin. The 310 hot spots were mainly concentrated in the middle and lower reaches of the Yellow River Basin, while the 311 cold spots were mainly concentrated in middle and upper reaches. Specifically, in 2008, Qingdao, Yantai, 312 Weihai, Weifang, Dongying, and Yulin were hot spots of IEE, while Zhongwei and Baiyin were cold 313 spots of IEE. In 2011, Zibo City, Yan 'an City, and Tongchuan City were added as hot spots on the basis 314 of 2008 IEE hot spots, while Wuwei City, Lanzhou City, Dingxi City, and Yinchuan City were added as 315 cold spots on the basis of 2008 IEE cold spots. In 2014, the hot spots of IEE evolved into 11 cities (i.e., 316 Qingdao, Yantai, Weihai, Weifang, Dongying, Tongchuan, Weinan, Shangluo, Xi'an, Xianyang, and 317 Ankang), while the cold spots evolved into 10 cities (Shuozhou, Xinzhou, Taiyuan, Yangquan, Jinzhong, 318 Changzhi, Yinchuan, Wuzhong, Zhongwei, and Wuwei). In 2017, IEE hot spots continued to increase to 319 19 cities, including Qingdao, Yantai, Weihai, Zhengzhou, and Zhoukou, and cold spots increased to 11 320 cities, including Shuozhou, Xinzhou, Taiyuan, Yangquan, Zhongwei, and Wuzhong. From the evolution 19 321 of the cold and hot spots of IEE in the Yellow River Basin from 2008 to 2017, most of the cities were in 322 the state of random distribution, reflecting that the connection degree of IEE among the cities in the 323 Yellow River Basin was not high during the research period. The number of hot spots was increasing and 324 mainly concentrated in the lower reaches, while the cold spots were concentrated in the middle and upper 325 reaches. Therefore, the key to improving the IEE of the Yellow River Basin is to strengthen the interaction 326 between the hot spots of IEE in the lower reaches and other regions and give full play to the radiation- 327 driven role of the region. 328 329 330 Fig. 7. Spatial evolution map of cold and hot spots of IEE in the Yellow River Basin. 4.4. The Influence mechanism of IEE in the background of economic transformation 331 In this study, Stata software was used to analyze the influencing factors of IEE in the Yellow River 332 Basin from 2008 to 2017. First, the fixed effect model and random effect model were used to 333 comprehensively evaluate the index factors, and then the optimal explanatory model was selected by 334 Hausmann test. Based on the Hausmann test and regression results, the fixed-effect model was found to 20 335 be suitable for influencing factor analysis in the whole basin and the middle reaches, while the random 336 effect model was found to be suitable for influencing factor analysis in the upper and lower reaches. The 337 regression results for the influencing factors are presented in Table 5. 338 (1) Globalization had a positive effect on the IEE of the lower reaches in the research period, which 339 was consistent with the research conclusions of Wang andChen (2014), but it was not significant for the 340 upper reaches, middle reaches and the whole basin. Specifically, the impact coefficient of globalization 341 on the IEE in the lower reaches of the Yellow River Basin was 0.1038, and it passed the test at the 342 significance level of 5%, indicating that every 1% increase in per capita FDI caused the IEE in the lower 343 reaches of the Yellow River Basin to increase by 0.1038%. Shandong Province, in the lower reaches, is 344 located in the eastern coastal economic zone, with a relatively high degree of globalization and strong 345 attraction to foreign-funded enterprises, which is conducive to promoting the economic development of 346 the region; Henan Province, in the lower reaches, has a large area, with the east adjacent to the coastal 347 economic zone and the south adjacent to the Yangtze River Economic Zone, which promotes the diffusion 348 of foreign-funded enterprises to the region. Thus, FDI could promote the economic growth, industrial 349 technology progress, and industrial structure adjustment in the lower reaches, improve the level of 350 industrial technology and industrial production efficiency, and promote the improvement of IEE in the 351 lower reaches, which confirmed the "halo effect" of FDI. The effect of globalization on the IEE of the 352 whole basin, middle reaches, and upper reaches was not significant. The middle and upper reaches of the 353 Yellow River Basin are located in the central and western inland areas of China. The degree of 354 globalization is low, the level of economic development is relatively slow, and the attraction to foreign- 355 funded enterprises is relatively insufficient, which makes the investment of foreign-funded enterprises 356 relatively insufficient. Moreover, owing to the fragile ecological environment in the middle and upper 21 357 reaches of the Yellow River Basin, relevant government departments give more management and 358 constraints to foreign-invested enterprises when introducing foreign capital, which further increases the 359 difficulty of attracting foreign capital and makes the impact of globalization on IEE insignificant. In 360 addition, the impact of globalization on the IEE of the Yellow River Basin was not significant in the 361 middle and upper reaches, so the impact of globalization on the IEE of the whole basin was not significant. 362 (2) Marketization had a positive effect on the IEE of the lower reaches during the research period, 363 which was consistent with the research conclusion of Yi et al. (2015), while the effect on the middle 364 reaches, upper reaches, and the whole basin was not significant. Specifically, the impact coefficient of 365 marketization on the IEE of the lower reaches was 0.0331, and it passed the significance test at the level 366 of 10%. This showed that the degree of marketization had a positive impact on the IEE of the lower 367 reaches. Every 1% increase in the proportion of non-state-owned economy caused the IEE of the lower 368 reaches to increase by 0.0331%. Shandong and Henan provinces in the lower reaches are located in the 369 eastern coastal economic zone and its adjacent regions respectively. The degree of globalization and 370 marketization in these regions is relatively high, which is conducive to promoting the market system 371 reform in this region, enhancing its attraction to foreign enterprises, and thus promoting regional 372 industrial and technological progress. Marketization is conducive to promoting the cross-regional and 373 cross-industry flow of factor resources in the lower reaches, attracting and retaining foreign-invested 374 enterprises, promoting the adjustment and optimization of industrial structure in the lower reaches, thus 375 promoting the improvement of IEE in the region. The effect of marketization on the IEE of the middle 376 reaches, upper reaches, and the whole basin was not significant. The middle and upper reaches of the 377 Yellow River Basin, owing to their depth in China's inland regions, are relatively less open to the outside 378 world, less non-state-owned enterprises, and less globalized and marketized than the lower reaches. 22 379 Moreover, Shanxi and Shaanxi provinces, in the middle reaches of the Yellow River Basin, are both big 380 coal resources provinces. In the process of industrial development, they have a strong dependence on 381 coal resources, a lack of attraction to foreign-funded enterprises, and a low degree of marketization. The 382 upper reaches are an important ecological environment protection area in China, and the government has 383 strong environmental supervision on local enterprises, which is not conducive to the introduction of 384 foreign-funded enterprises. Therefore, the proportion of non-state-owned enterprises was low. In addition, 385 compared with the Yangtze River Basin, the water flow of the Yellow River Basin is small and unstable, 386 the navigation capacity is poor, and the Yellow River Basin lacks gateway cities and hub city clusters 387 like Shanghai and the Yangtze River Delta, so it is difficult to form the Yellow River economic belt, 388 which is not conducive to regional globalization and marketization. Therefore, the effect of marketization 389 on the IEE of the whole basin was not significant. 390 (3) Decentralization had a negative effect on the IEE of the whole basin and the middle reaches 391 during the research period, which was consistent with the research conclusion of Zhang et al. (2017b), 392 and had a positive promoting effect on the upper reaches, which was consistent with some previous 393 research conclusions (Chen &Chang 2020, Mu 2018). However, the effect on the lower reaches was not 394 significant. Specifically, the impact coefficient of decentralization on the IEE of the whole basin was - 395 0.1495, and it passed the significance test at the level of 5%. Every 1% increase in the degree of 396 decentralization would reduce the IEE of the whole basin by 0.1495%. The impact coefficient of 397 decentralization on the IEE of the middle reaches was -0.3972, and it passed the significance test at the 398 level of 5%. Every 1% increase in the degree of decentralization would reduce the IEE of the middle 399 reaches by 0.3972%. The degree of globalization and marketization in the middle reaches of the Yellow 400 River Basin is relatively low, and the degree of decentralization is relatively high. In the process of 23 401 decentralization, local governments often take measures to promote the rapid economic growth of the 402 region, such as increasing enterprise taxes and relaxing the regulatory environment, which leads to the 403 increase of enterprise production costs and pollution emissions, and then reduces the IEE. The impact 404 coefficient of decentralization on the IEE of the upper reaches was 0.0678, and it passed the significance 405 test at the level of 5%. Every 1% increase in the degree of decentralization would cause the IEE of the 406 upper reaches to increase by 0.0678%. The ecological environment in the upper reaches of the Yellow 407 River Basin is very fragile, and the local government has a high degree of financial autonomy, which is 408 beneficial for the local government to strengthen the control of regional ecological environment and to 409 promote the protection of ecological environment in the upper reaches of the Yellow River Basin, which 410 is consistent with the results of some previous studies (Chen &Chang 2020, Hao et al. 2021). However, 411 decentralization is conducive to improving the allocation efficiency of local resources, strengthening 412 exchanges and cooperation with other regions, selectively introducing advanced industries and 413 technologies, and contributing to the ecological environment protection and high-quality development of 414 the upper reaches. The effect of decentralization on the IEE in the lower reaches was not significant, 415 which may be because of the relatively high degree of globalization and marketization in the lower 416 reaches, the relatively high level of economic development, and relatively weak control of local 417 enterprises by the local government. 418 (4) Control variable analysis. The industrial structure had a significant negative effect on the IEE of 419 the upper, middle, and lower reaches as well as the whole basin. This was in contrast to the fact that most 420 areas of the Yellow River Basin were dominated by the development of heavy and chemical industry 421 during the research period. A large number of environmental pollutants are discharged in the process of 422 industrial production, thus reducing the IEE of the whole basin. The level of economic development had 24 423 a significant positive effect on the IEE of the lower reaches and the whole basin, which is consistent with 424 the results of some previous studies (Liu et al. 2020a). The level of economic development would 425 promote the adjustment of regional industrial structure and the progress of industrial technology and then 426 promote the improvement of IEE. However, the level of economic development in the middle and upper 427 reaches of the region is relatively backward, so the impact on IEE was not significant. Scientific and 428 technological innovation has a significant positive role in promoting the IEE of the lower reaches, which 429 is consistent with the research conclusion of Chen et al. (2020). Scientific and technological innovation 430 helps to promote the upgrading of industrial technology, improve the production efficiency of industrial 431 enterprises, and promote the improvement of IEE. Population agglomeration plays a negative role in 432 promoting the IEE of the lower reaches and the whole basin. There are many industrial enterprises in the 433 area of population agglomeration; however, most of the industries in the Yellow River Basin are still 434 dominated by the secondary industry, and the level of industrial technology and management experience 435 is relatively backward, which makes the production efficiency of most industrial enterprises low. 436 Industrial agglomeration has a significant positive effect on the IEE of the lower reaches, middle reaches, 437 and the whole basin, which is consistent with the research conclusion of Shen andPeng (2020). Industrial 438 agglomeration can effectively reduce the overall pollutant emission in the Yellow River Basin by its 439 pollution control scale effect. However, industrial agglomeration can promote industrial technology 440 upgrading through technology spillover effect, and then promote the improvement of IEE. 441 Table 5. Panel regression results of factors affecting IEE in the Yellow River Basin. All Lower Middle Upper Variable fe re fe re fe re fe re -0.1826 -0.1918 -0.3359 0.4431 2.4489 0.7252 -0.0658 (- -0.0989 (- (-0.18) (-0.22) (-0.11) (0.22) (0.50) (0.30) 0.13) 0.21) -0.0168 -0.0388* 0.0634 0.1038 0.0848* 0.0050 -0.0134 (- -0.0119 (- (-0.98) (-2.57) (1.34) (2.66) (1.77) (0.12) 1.52) 1.52) Cons lnglo 25 0.0036 -0.0023 0.0072 0.0331* 0.0056 -0.0064 (- -0.0019 (- -0.0001 (- (0.28) (-0.24) (0.14) (1.25) (0.18) 0.25) 0.33) 0.01) -0.1495 0.0252 -0.2052 (- -0.0611 -0.3972 -0.1040 (- 0.0594 0.0678 (-2.51) (0.61) 1.42) (-0.54) (-2.67) 0.80) (2.00) (2.66) -0.0273* -0.0417* -0.0987 -0.1310*** -0.0946* -0.0286 (- -0.0154* -0.0127* (-1.77) (-2.83) (-2.36) (-4.01) (-1.91) 0.65) (-1.97) (-1.69) 0.0518* 0.0971** 0.2492* 0.1272*** 0.1108 0.0351 0.0109 0.0082 (1.00) (2.07) (1.68) (3.76) (0.92) (0.31) (0.38) (0.30) 0.1936 0.2503 0.9937* 0.6069* 0.6727 0.1361 0.0399 0.0855 lnmar lndec lnstr lnpgdp lntec (0.54) (0.72) (1.15) (1.87) (0.87) (0.17) (0.19) (0.44) -0.8187 -0.3459* -1.4285 -1.1636* -2.8287 -0.9936 -0.0522 -0.0748 (-2.55) (-2.78) (-1.10) (-2.70) (-1.07) (-1.58) (-0.39) (-0.94) 0.2384* 0.0975* 0.2408* 0.2038* 0.2928* 0.1620 0.0082 -0.0020 (6.96) (4.61) (4.57) (4.96) (3.63) (2.11) (0.33) (-0.11) R2 0.3860 0.3230 0.6081 0.5982 0.5100 0.4550 0.3147 0.3080 F 18.23 - 18.23 - 7.15 - 3.85 - lnpop lnagg 442 443 represents random effects model. 444 5. Conclusions and policy suggestions 445 5.1.Conclusions Note: , , and refer to the 1%, 5%, and 10% significance levels, respectively; fe represents fixed effect model; re 446 This study put forward a theoretical framework for the impact of economic transformation on IEE 447 based on the Super-efficiency DEA model to measure the IEE of the Yellow River basin from 2008 to 448 2017. The ESDA method was used to explore the spatial and temporal evolution of IEE, and the panel 449 regression model was used to study the main driving factors of IEE. The main conclusions were as 450 follows: 451 (1) From 2008 to 2017, the IEE of the Yellow River Basin showed an elongated S-shaped temporal 452 evolution characteristic. During the study period, the regional differences of IEE of cities in the Yellow 453 River Basin were significant. The IEE of the lower reaches was the highest, followed by the middle 454 reaches, and the upper reaches were the lowest. The results of kernel density estimation indicate that the 455 overall level of IEE in the Yellow River Basin is improving, but the gap between regions is widening. 456 (2) The IEE in the Yellow River Basin had significant global and local spatial autocorrelation, and 26 457 the distribution of cold spots and hot spots showed obvious "space club" phenomenon, respectively. Thus, 458 the southeast region of the Yellow River Basin is the high-value area of IEE, while the northwest region 459 is the low-value area. 460 (3) The influence factors of IEE in the Yellow River Basin showed spatial heterogeneity in their 461 effect. Globalization and marketization had a positive effect on the IEE in the lower reaches, but had no 462 significant effect in the middle and upper reaches and the whole basin. Decentralization had a negative 463 effect on the whole basin and the middle reaches, a positive effect on the upper reaches, and an 464 insignificant effect on the lower reaches. Decentralization had a negative effect on the whole basin and 465 the middle reaches, a positive effect on the upper reaches, and an insignificant effect on the lower reaches. 466 In addition, industrial structure, economic development level, scientific and technological innovation, 467 population agglomeration, industrial agglomeration also had different effects on the IEE of the Yellow 468 River Basin. 469 5.2. Policy suggestions 470 In view of the regional differences of IEE in the Yellow River Basin, different industrial 471 development policies and environmental control measures should be formulated according to local 472 conditions. Based on the research conclusions, we make the suggestions that follow. (1) The lower 473 reaches of the Yellow River Basin should make full use of the development opportunities brought by 474 globalization and marketization, accelerate the adjustment of industrial structure, improve the market 475 system and mechanism, transform the model of industrial development, improve the efficiency of 476 industrial production, and reduce energy consumption and pollution emission. (2) The middle reaches 477 should adjust the industrial structure in a timely manner, reduce excessive dependence on coal and other 478 resource-based industries, actively seek new driving forces for industrial development, strengthen 27 479 exchanges and cooperation with the developed areas along the eastern coast, and promote the upgrading 480 of regional industrial structure and high-quality economic development. (3) The ecological environment 481 of the upper reaches is very fragile. The local government should give full play to the role of macro- 482 control in the development of regional industries. On the basis of protecting the local ecological 483 environment, it should selectively introduce advanced industries and technologies from the eastern 484 developed regions, strengthen the industrial and technological exchanges with eastern developed regions, 485 and promote ecological environment protection and high-quality development of the upper reaches. 486 487 488 Declarations 489 Availability of data and materials The datasets used and/or analysed during the current study are 490 available from the corresponding author on reasonable request. 491 Ethics approval and consent to participate Not applicable. 492 Consent for publication Not applicable. 493 Competing interests The authors declare no competing interests. 494 Funding his study is supported by the National Science Foundation of China(Grant No. 41701177). 495 Author contribution All authors contributed to the study conception and design. Methodology, Software 496 and Writing-Original draft preparation were performed by Chengzhen Song. Conceptualization, 497 Methodology, Writing- Review & Editing, and Funding Acquisition were performed by Guanwen Yin. 498 Data curation and Software were performed by Zhilin Lu. Conceptualization, Writing-Review & Editing, 499 Methodology, Supervision, and Funding acquisition were performed by Yanbin Chen. All authors read 500 and approved the final manuscript. 501 References 502 503 504 505 506 Andersen P, Petersen NC (1993): A Procedure for Ranking Units in Data Envelopment Analysis. Management Science 39, 1261-1264. Bao J, Yang X, Zhao Z, Wang Z, Yu C, Li X (2015): The Spatial-Temporal Characteristics of Air Pollution in China from 2001–2014. Int J Environ Res Public Health 12, 15875-15887. DOI: https://doi.org/10.3390/ijerph121215029. 28 507 508 509 510 511 512 513 514 515 516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 534 535 536 537 538 539 540 541 542 543 544 545 546 547 548 549 550 Campos NF, Horváth R (2012): Reform redux: Measurement, determinants and growth implicati ons. European Journal of Political Economy 28, 227-237. DOI: https://doi.org/10.1016/j. ejpoleco.2011.12.002. Charnes A, Cooper WW, Rhodes E (1979): Measuring the efficiency of decision-making units. European Journal of Operational Research 3, 339. Chen F, Zhao T, Wang J (2019): The evaluation of energy–environmental efficiency of China's industrial sector: based on Super-SBM model. Clean Technologies and Environmental Policy 21, 1-18. DOI: https://doi.org/10.1007/s10098-019-01713-0. Chen W, Si W, Chen ZM (2020): How technological innovations affect urban eco-efficiency in China: A prefecture-level panel data analysis. Journal of Cleaner Production 270, 122479. DOI: https://doi.org/10.1016/j.jclepro.2020.122479. Chen X, Chang CP (2020): Fiscal decentralization, environmental regulation, and pollution: a s patial investigation. Environmental Science and Pollution Research 27. DOI: https://doi. org/10.1007/s11356-020-09522-5. Chen Y, Yin G, Liu K (2021): Regional differences in the industrial water use efficiency of China: The spatial spillover effect and relevant factors. Resources, Conservation and Recycling 167, 105239. DOI: https://doi.org/10.1016/j.resconrec.2020.105239. Cheng S, Fan W, Chen J, Meng F, Yang Z (2019): The impact of fiscal decentralization on CO2 emissions in China. Energy 192, 116685. DOI: https://doi.org/10.1016/j.energy.2019.116685. Feng L (2014): Does Foreign Direct Investment Harm the Host Country's Environment? Evidence from China. Current Topics in Management. Getis A, Ord JK (1992): The Analysis of Spatial Association by Use of Distance Statistics. Geographical Analysis 24. Guan W, Xu S (2016): Study of spatial patterns and spatial effects of energy eco-efficiency in China. Journal of Geographical Sciences 26, 1362-1376. DOI: https://doi.org/10.1007/s1 1442-016-1332-x. HA, Lin, TU, Jianjun, YANG, Jianping, XU, Chunhai, PANG, Jiaxing (2020): Regional eco-efficiency evaluation and spatial pattern analysis of the Yangtze River Economic Zone. Journal of Geographical Sciences v.30, 78-100. DOI: https://doi.org/10.1007/s11442-020-1773-0. Hao Y, Gai Z, Yan G, Wu H, Irfan M (2021): The spatial spillover effect and nonlinear relati onship analysis between environmental decentralization, government corruption and air pollution: Evidence from China. Science of The Total Environment 763, 144183. DOI: https://doi.org/10.1016/j.scitotenv.2020.144183. Haoran, Zhao, Sen, Guo, Huiru, Zhao (2019): Provincial energy efficiency of China quantified by three-stage data envelopment analysis. Energy. DOI: https://doi.org/10.1016/j.energy.2 018.10.063. Huang Z, Wei YD, He C, Li H (2015): Urban land expansion under economic transition in C hina: A multi-level modeling analysis. Habitat International 47, 69-82. DOI: https://doi. org/10.1016/j.habitatint.2015.01.007. J., Quariguasi, Frota, Neto, and, G., Walther, and, J., Bloemhof (2009): A methodology for ass essing eco-efficiency in logistics networks. European Journal of Operational Research. DOI: https://doi.org/10.1016/j.ejor.2007.06.056. Jiang L, He S, Zhong Z, Zhou H, He L (2019): Revisiting environmental kuznets curve for carbon dioxide emissions: The role of trade. Structural change and economic dynamics 50, 245-257. DOI: 29 551 552 553 554 555 556 557 558 559 560 561 562 563 564 565 566 567 568 569 570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585 586 587 588 589 590 591 592 593 594 https://doi.org/10.1016/j.strueco.2019.07.004. Jiya AN, Sama MC, Ouedraogo I (2020): Infrastructure, trade openness and economic transformation in Common Market for Eastern and Southern Africa member countries. Social Sciences & Humanities Open. DOI: https://doi.org/10.1016/j.ssaho.2020.100072. Lan J, Kakinaka M, Huang X (2012): Foreign Direct Investment, Human Capital and Environ mental Pollution in China. Environmental and Resource Economics 51, 255-275. DOI: https://doi.org/10.1007/s10640-011-9498-2. Lin B, Zhu J (2020): Chinese electricity demand and electricity consumption efficiency: Do the structural changes matter? Applied Energy 262, 114505. DOI: https://doi.org/10.1016/j. apenergy.2020.114505. Liu L, Ding D, He J (2017a): Fiscal Decentralization, Economic Growth, and Haze Pollution Decoupling Effects: A Simple Model and Evidence from China. Computational Economics. DOI: https://doi.org/10.1007/s10614-017-9700-x. Liu Q, Wang S, Li B, Zhang W (2020a): Dynamics, differences, influencing factors of eco-efficiency in China: A spatiotemporal perspective analysis. Journal of Environmental Management 264, 110442. DOI: https://doi.org/10.1016/j.jenvman.2020.110442. Liu Y, Hao Y, Gao Y (2017b): The environmental consequences of domestic and foreign invest ment: Evidence from China. Energy Policy 108, 271-280. DOI: https://doi.org/10.1016/j. enpol.2017.05.055. Liu Y, Li J, Yang Y (2018): Strategic adjustment of land use policy under the economic transformation. Land Use Policy 74, 5-14. DOI: https://doi.org/10.1016/j.landusepol.2017.07.005. Liu Y, Zhu J, Li EY, Meng Z, Song Y (2020b): Environmental regulation, green technological innovation, and eco-efficiency: The case of Yangtze river economic belt in China. Tech nological Forecasting and Social Change 155, 119993. DOI: https://doi.org/10.1016/j.tec hfore.2020.119993. Marques AC, Fuinhas JA, Tomas C (2019): Energy efficiency and sustainable growth in industrial sectors in European Union countries: A nonlinear ARDL approach. Journal of Cleaner Production 239, 118045.1-118045.12. DOI: https://doi.org/10.1016/j.jclepro.2019.118045. Merlevede B (2003): Reform reversals and output growth in transition economies. Economics of Transition 11, 649-669. DOI: https://doi.org/10.1111/j.0967-0750.2003.00165.x. Moran, P.A.P (1948): The interpretation of statistical maps. J. R. Stat. Soc. Ser B–Stat. Methodol 10, 243–251. Mu R (2018): Bounded Rationality in the Developmental Trajectory of Environmental Target Policy in China, 1972–2016. Sustainability 10, 199. DOI: https://doi.org/10.3390/su10010199. Pao H-T, Tsai C-M (2011): Multivariate Granger causality between CO2 emissions, energy consumption, FDI (foreign direct investment) and GDP (gross domestic product): Evidence from a panel of BRIC (Brazil, Russian Federation, India, and China) countries. Energy 36, 685-693. DOI: https://doi.org/10.1016/j.energy.2010.09.041. Ren Y, Fang C, Li G (2020): Spatio-temporal characteristics and influential factors of eco-efficiency in Chinese prefecture-level cities: A spatial panel econometric analysis. Journal of Cleaner Production 260. DOI: https://doi.org/10.1016/j.jclepro.2020.120787. Schaltegger S, Sturm A (1990): Ecological rationality: approaches to design of ecology-oriented management instruments. Die Unternehm 44, 273-290. Shah IH, Dong L, Park HS (2020): Tracking urban sustainability transition: An eco-efficiency analysis 30 595 596 597 598 599 600 601 602 603 604 605 606 607 608 609 610 611 612 613 614 615 616 617 618 619 620 621 622 623 624 625 626 627 628 629 630 631 632 633 634 635 636 637 638 on eco-industrial development in Ulsan, Korea. Journal of Cleaner Production 262, 121286. DOI: https://doi.org/10.1016/j.jclepro.2020.121286. Shen N, Peng H (2020): Can industrial agglomeration achieve the emission-reduction effect? SocioEconomic Planning Sciences, 100867. DOI: https://doi.org/10.1016/j.seps.2020.100867. Tong Y, Zhou H, Jiang L (2020): Exploring the transition effects of foreign direct investment on the ecoefficiency of Chinese cities: Based on multi-source data and panel smooth transition regression models. Ecological Indicators. DOI: https://doi.org/10.1016/j.ecolind.2020.107073. Wang DT, Chen WY (2014): Foreign direct investment, institutional development, and environmental externalities: Evidence from China. Journal of Environmental Management 135, 81–90. DOI: https://doi.org/10.1016/j.jenvman.2014.01.013. Wang L, Zhang F, Fu W, Tan Q, Chen J (2020): Analysis of temporal and spatial differences and influencing factors of energy eco-efficiency in energy-rich area of the Yellow River Basin. Physics and Chemistry of the Earth. DOI: https://doi.org/10.1016/j.pce.2020.102976. Wang S, Zhou C, Li G, Feng K (2016): CO2 , economic growth, and energy consumption in China's provinces: Investigating the spatiotemporal and econometric characteristics of C hina's CO2 emissions. Ecological Indicators 69, 184-195. DOI: https://doi.org/10.1016/j.e colind.2016.04.022. Wang Z, Liu B, Wang L, Shao Q (2021): Measurement and Temporal & Spatial Variation of Urban ecoefficiency in the Yellow River Basin. Physics and Chemistry of the Earth Parts A/B/C, 102981. DOI: https://doi.org/10.1016/j.pce.2021.102981. Wencong Y, Yanpeng C, Zhifeng Y, Qiangqiang R, Zhi D (2018): Structural optimization for i ndustrial sectors to achieve the targets of energy intensity mitigation in the urban clust er of the Pearl River Delta. Ecological Indicators 95, 673-686. DOI: https://doi.org/10.1 016/j.ecolind.2018.08.009. Wu J, Zhu Q, Chu J, Liang L (2015): Two-Stage Network Structures with Undesirable Intermediate Outputs Reused: A DEA Based Approach. Computational Economics 46, 455-477. DOI: https://doi.org/10.1007/s10614-015-9498-3. Yang Y, Tang D, Zhang P (2020): Effects of fiscal decentralization on carbon emissions in China. International Journal of Energy Sector Management 14, 213-228. Yi J, Chen Y, Wang C, Kafouros M (2015): Spillover Effects of Foreign Direct Investment: H ow do Region-Specific Institutions Matter? Management International Review 55, 539-5 61. DOI: https://doi.org/10.1007/s11575-014-0235-2. Yin G, Liu Y, Wang F (2018): Emerging Chinese new towns: Local government-directed capita l switching in inland China. Cities 79, 102-112. DOI: https://doi.org/10.1016/j.cities.201 8.02.026. Yu C, Li H, Jia X, Li Q (2015): Improving resource utilization efficiency in China's mineral resourcebased cities: A case study of Chengde, Hebei province. Resources, Conservation and Recycling 94, 1-10. DOI: https://doi.org/10.1016/j.resconrec.2014.10.013. Yu Y, Huang J, Zhang N (2018): Industrial eco-efficiency, regional disparity, and spatial conver gence of China's regions. Journal of Cleaner Production 204, 872-887. DOI: https://doi. org/10.1016/j.jclepro.2018.09.054. Zhang B, Bi J, Fan Z, Yuan Z, Ge J (2008): Eco-efficiency analysis of industrial system in C hina: A data envelopment analysis approach. Ecological Economics 68, 306-316. DOI： https://doi.org/10.1016/j.ecolecon.2008.03.009. 31 639 640 641 642 643 644 645 646 647 648 649 650 651 652 653 654 655 656 657 658 659 660 661 662 Zhang B, Miao C (2020): The evolution and driving forces of the spatial-temporal pattern of land use in the Yellow River Basin. Resources Science 42, 460-473.(in Chinese) Zhang J, Liu Y, Chang Y, Zhang L (2017a): Industrial eco-efficiency in China: A provincial quantification using three-stage data envelopment analysis. Journal of Cleaner Production 143, 238-249. DOI： https://doi.org/10.1016/j.jclepro.2016.12.123. Zhang K, Zhang ZY, Liang QM (2017b): An empirical analysis of the green paradox in China: From the perspective of fiscal decentralization. Energy Policy 103, 203-211. DOI：htt ps://doi.org/10.1016/j.enpol.2017.01.023. Zhao, Changhong, Zhang, Haonan, Zeng, Yurong, Li, Fengyun, Liu, Yuanxin (2018): Total-Factor Energy Efficiency in BRI Countries: An Estimation Based on Three-Stage DEA Model. Sustainability. DOI: https://doi.org/10.3390/su10010278. Zhiqing L, Yong G, Hancheng D, Jeffrey W, Yang X, Rui W, Wei Y, Zhongjue Y (2018): Regional impacts of launching national carbon emissions trading market: A case study of Shanghai. Applied Energy 230, 232-240. DOI：https://doi.org/10.1016/j.apenergy.2018.08.117. Zhou C, Shi C, Wang S, Zhang G (2018): Estimation of eco-efficiency and its influencing factors in Guangdong province based on Super-SBM and panel regression models. Ecological Indicators 86, 67-80. DOI：https://doi.org/10.1016/j.ecolind.2017.12.011. Zhou Y, Kong Y, Sha J, Wang H (2019): The role of industrial structure upgrades in eco-efficiency evolution: Spatial correlation and spillover effects. Science of the Total Environment 687. DOI: https://doi.org/10.1016/j.scitotenv.2019.06.182. Zhou Y, Liu Z, Liu S, Chen M, Zhang X, Wang Y (2020): Analysis of industrial eco-efficiency and its influencing factors in China. Clean Technologies and Environmental Policy 22, 2023-2038. DOI: https://doi.org/10.1007/s10098-020-01943-7. 32
https://openalex.org/W4381682313	https://journals.ufs.ac.za/index.php/trp/article/download/7116/4737	English	null	A review of some of the criteria used in land-demarcation processes	Town and regional planning/Stads- en streekbeplanninq	2,023	cc-by	8,629	Abstract In recent years, settlements have sprawled beyond the urban growth boundaries, due to a number of factors, including mobility, technology, and urban blight. As a result, governments globally have opted to restructure their local administrative boundaries (municipalities) to be more accommodating to the unplanned growth, while setting a clear limit to the extent to which the urban region could grow. Without a clear administrative boundary, urban problems such as civil conflict, administrative duplication, political corruption, lack of service delivery, and environmental degradation become more prevalent. In order to understand the driving forces behind administrative delimitation, this article reviews how local administrative boundaries can be delineated from a theoretical stance. It further unpacks various criteria to contextualise how boundaries could be demarcated and their resultant structure. The article reveals that, from a theoretical stance, the method of demarcating administrative urban boundaries is not apparent, since a variety of factors influence open systems. Consequently, this article provides awareness of the challenges of demarcating local administrative boundaries, with both administrative and policy implications. Administratively, it sheds light on criteria that can influence boundary demarcation. In terms of policy, it demonstrates that the demarcation of boundaries is a huge challenge that requires further research and action. administratiewe grens word stedelike probleme soos burgerlike konflik, administratiewe duplisering, politieke korrupsie, gebrek aan dienslewering en omgewingsagteruitgang meer algemeen. Ten einde die dryfkragte agter administratiewe afbakening te verstaan, poog hierdie artikel om te verstaan hoe plaaslike administratiewe grense vanuit ’n teoretiese standpunt afgebaken kan word. Die wetgewende kriteria word verder ontleed en die teoretiese rasionaal van afbakening word gekontekstualiseer. Die artikel het bevind dat vanuit ’n teoretiese standpunt die metodes om administratiewe stedelike grense af te baken nie duidelik is nie, aangesien dit oop sisteme is wat deur ’n verskeidenheid faktore beïnvloed word. Gevolglik beklemtoon hierdie artikel die uitdagings van toepassing op die afbakening van plaaslike administratiewe grense, wat beide administratiewe en beleid- implikasies inhou. Administratief werp dit lig op kriteria wat grensafbakening kan beïnvloed en in terme van beleid demonstreer dit dat die afbakening van grense uitdagend is en verdere navorsing verg. How to cite: Jeeva, Z., Cilliers, J. & Gumbo, T. 2023. A review of some of the criteria used in land demarcation processes. Town and Regional Planning, no. 82, pp. 52-61. © Creative Commons With Attribution (CC-BY) Published by the UFS http://journals.ufs.ac.za/index.php/trp TLHAHLOBO EA E MENG EA MEKHOA E SEBELISOANG HO AROLA MEELI EA LEFATS’E Lilemong tsa morao tjena, libaka tsa bolulo li nametse ka nģ’ane ho meeli ea kholo ea litoropo, ele litla-morao tsa mabaka a ‘maloa, a kenyelelitseng ho falla, theknoloji le ho senyeha ha metse litoropong. Ka lebaka leo, mebuso lefatšeng ka bophara e khethile ho hlophisa bocha meeli ea eona ea tsamaiso ea lehae (masepala) hore e lumellane le kholo e sa reroang, ha e ntse e beha moeli o hlakileng oa hore na tikoloho ea litoropo e ka hola hakae. Ntle le moeli o hlakileng oa tsamaiso, mathata a litoropong a kang likhohlano tsa sechaba, phetiso ea tsamaiso, bobolu ba lipolotiki, khaello ea phano ea litšebeletso, le ho senyeha ha tikoloho li tsoa taolong. E le ho utloisisa matla a susumetsang karolo ea meeli ea tsamaiso, sengoloa sena se hlahloba hore na meeli ea tsamaiso ea lehae e ka hlalosoa joang ho latela maikutlo a khopolo-taba. E tsoela pele ho manolla mekhoa e fapaneng ea ho hlalosa hore Keywords: local administration, demarcation, local boundaries, functional linkages Review article DOI: https://doi.org/10.38140/trp.v82i.7116 Received: February 2022 Peer reviewed and revised: March 2022 Published: June 2023 *The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article A review of some of the criteria used in land- demarcation processes Zaakirah Jeeva, Juanee Cilliers & Trynos Gumbo Town and Regional Planning 2023 (82):52 61 \| ISSN 1012 280 \| e ISSN 2415 0495 Dr Zaakirah I. Jeeva, Unit for Environmental Sciences and Management, North-West University, Potchefstroom, South Africa. Phone: 0725702956, email: Zaakirah.Jeeva@gmail.com, ORCID: https://orcid.org/0000-0001-5327-859X Prof. Juanee Cilliers, School of Built Environment, University of Technology, Sydney, Australia. Email: Juanee.cilliers@uts.edu.au, ORCID: https:// orcid.org/0000-0002-8581-6302 Prof. Trynos Gumbo, Department of Urban and Regional Planning, University of Johannesburg, South Africa. Phone: +27 (0)11 559 6062, email: tgumbo@uj.ac.za, ORCID: https://orcid.org/0000-0003-3617-4996 Zaakirah Jeeva, Juanee Cilliers & Trynos Gumbo Review article DOI: https://doi.org/10.38140/trp.v82i.7116 Review article DOI: https://doi.org/10.38140/trp.v82i.7116 Dr Zaakirah I. Jeeva, Unit for Environmental Sciences and Management, North-West University, Potchefstroom, South Africa. Phone: 0725702956, email: Zaakirah.Jeeva@gmail.com, ORCID: https://orcid.org/0000-0001-5327-859X Prof. Juanee Cilliers, School of Built Environment, University of Technology, Sydney, Australia. Email: Juanee.cilliers@uts.edu.au, ORCID: https:// orcid.org/0000-0002-8581-6302 Prof. Trynos Gumbo, Department of Urban and Regional Planning, University of Johannesburg, South Africa. Phone: +27 (0)11 559 6062, email: Town and Regional Planning 2023 (82):52-61 \| ISSN 1012-280 \| e-ISSN 2415-0495 ’N OORSIG VAN SOMMIGE KRITERIA WAT IN GRONDAFBAKENINGSPROSESSE GEBRUIK WORD Nedersettings het die afgelope jare buite die stedelike groeigrense uitgebrei, as gevolg van ’n aantal faktore, insluitende mobiliteit, tegnologie en die verval van stedelike areas. Gevolglik het regerings regoor die wêreld begin om hul plaaslike administratiewe grense (munisipaliteite) te herstruktureer om meer tegemoetkomend te wees vir die onbeplande groei, maar terselfdertyd ’n duidelike beperking te stel tot die mate waarbinne die totale stedelike streek kan groei, want sonder ’n duidelike 52 Jeeva, Cilliers & Gumbo 2023 Town and Regional Planning (82):52-61 administrative management of the area (Zhou et al., 2016: 1). which can extend as far as 300 km (Webster & Muller, 2004: 282). This raises the question: How should sprawling settlements be contained via local government administrative boundaries? administrative management of the area (Zhou et al., 2016: 1). na meeli e ka aroloa joang le sebopeho sa eona. Sengoliloeng se senola hore, ho latela maikutlo a khopolo-taba, mokhoa oa ho beha meeli ea tsamaiso ea litoropo ha o bonahale, kaha lintlha tse fapaneng li susumetsa litsamaiso tse se nang taolo. Qetllong, sengoloa sena se fana ka tlhokomeliso ea liqholotso tsa ho arola meeli ea tsamaiso ea lehae, ka litlamorao tsa tsamaiso le maano. Tsamaisong, sengoloa se fana ka leseli holim’a mekhoa e ka susumetsang ho seha meeli. Mabapi le maano, se bontša hore ho seha meeli ke phephetso e kholo e hlokang lipatlisiso le mehato e meng e tla fokotsa litla-morao. According to Zhang (2016: 1), urban sprawl occurs due to three reasons, namely natural population growth, urbanisation, and decentralisation. Natural population growth is when the birth rate in the region exceeds the death rate and settlements grow naturally beyond their boundaries, transforming villages to towns; urbanisation is rapid development caused by the in-migration of individuals from rural areas into urban areas transforming towns to cities, and decentralisation is the movement of individuals out of urban areas into the peripheral rural areas, transforming cities into metropolitan regions, conurbations, or polycentric regions (Sporna & Krzysztofik, 2020: 2). The latter (regions) are larger than the traditional settlements, as they are created through the interdependent functional relationship between settlements (Vasanen, 2012: 1; Sinclair- Smith, 2015: 132). The functional region is more abstract than the traditionally, demographically, and economically classified administrative settlement, in that it encompasses the functional linkages of various communities between where they live, work, shop, access healthcare, and recreate (Vasanen 2012: 2; Ramutsindela, 2013: 42). 1. INTRODUCTION Human beings have a need to name and categorise objects, concepts, and places to make sense of their surroundings, in order for them to function more efficiently (Fulkerson, Waxman & Seymour, n.d.:1). Politicians alike demarcate their territory into smaller administrative regions such as electoral wards, municipalities, districts, and provinces, in order to manage civil administration and provide a platform for the redistribution of resources. It is anticipated that subdivision provides insight into regional disparities and assists government in their long-term plans of promoting spatial equality, economic efficiency, sustainability, and social cohesiveness at various levels within a country (Feleke, 2018: 130; Zhou et al., 2016: 1). For the purposes of this article, only the method of demarcating municipalities (basic administrative boundaries), which are the lowest tier of local government, is analysed, and not ward and district boundaries (Drupal, 2021). Human beings have a need to name and categorise objects, concepts, and places to make sense of their surroundings, in order for them to function more efficiently (Fulkerson, Waxman & Seymour, n.d.:1). A qualitative research approach was employed primarily through the application of desktop research. Relevant materials used in this review consisted of articles, theses, reports, and other documents obtained from the North-West University library database and the internet. As an initial search, relevant documents were identified through the application of a database keyword search for terms associated with the delimitation of local administrative boundaries, delimitation of municipalities, demarcation of boundaries, motivation for spatial structural reform, categorisation of municipalities, and local boundaries formation. The primary search was performed between 13 May 2020 and 15 December 2020, with supplementary supporting information sourced between 2 February 2021 and 29 February 2022. Preference was given to more recent literature to reflect on current research on, and approaches to the topic. To expand on the relevant literature for review, several of the references cited in the initial texts were also evaluated for their relevance to this topic. Rather recent research has been conducted on the delimitation of these municipal regions. Currently, limited comprehensive studies can be referenced on the criteria and the However, Bennet (1989) has an alternative view. He believes that the administrative boundary should be extended to coincide with the functional boundary. In his opinion, one cannot restrict sprawl. ’N OORSIG VAN SOMMIGE KRITERIA WAT IN GRONDAFBAKENINGSPROSESSE GEBRUIK WORD Friedmann (1978) suggests that the growth and development of the functional region should be contained through the delimitation of a formal administrative boundary. In this way, he hoped to contain sprawl through compact development 2. METHODS AND REVIEW APPROACH This review seeks to explore the criteria that guide the delimitation of boundaries and how it is implemented in municipal formation. First, the delimitation of boundaries was reviewed to provide insights into the current state of administrative boundaries. Secondly, the different administrative delimitation rationales were introduced and reviewed from an urban planning perspective to motivate the criterion for local municipal boundaries and their different outcomes. 3.1 Delimiting local government boundaries According to the OECD (2020), delimiting local government boundaries can be done either intensively or extensively. These can be observed in a study conducted in Paris and Rome by the OECD in 2016 (see Figure 1). On the other hand, some municipalities such as Rome, which are monocentric structures, choose to have one large extensive administrative region, which encompasses all the peripheral settlements. The entire region is managed by a single local government under the leadership of a single mayor who takes instructions from the national government (Beauregard, 1991: 190). In this regard, national governments pass laws on how the region should be demarcated and administrated, according to their political value sets, and the administration is carried out by government officials on the local level, in a top-down manner (Beauregard, 1991: 191). However, even in this case, development and management become difficult as the region becomes too large to administer efficiently (Sikander, 2015: 172). Accordingly, the question as to how the boundaries around these administrative regions could be demarcated to allow for the efficient and effective management of the contained settlements remains open (OECD, 2020: 11). In the above case, the administrative boundary around Paris would be intensive or under-bounded, since it only consists of the main urban area, while the functional region sprawls beyond it into alternative administrative regions. Some could argue that this makes the development and management of the functional area rather difficult in economic, social, administrative, and political terms (Webster & Muller, 2004: 283; Feleke, 2018: 127-129; Sikander, 2015: 171). However, in the Paris case, there are 20 smaller local government administrative regions called ‘municipal arrondissements’ that form part of a larger metropolitan commune entity. Consequently, each arrondissement entity serves the area that it is demarcated to serve, under the leadership of its own mayor and council. It is responsible for managing the community facilities and registering the population in its area. However, each arrondissement is under the management of the Paris council and Commune Mayor. In this case, the local government territory Due to the rareness of relevant literature, and the exploratory and evaluative nature of the present research endeavour, the review relied heavily on grey literature. 3. KEY THEMES effects of local boundary extension and settlement incorporation (Zhou et al., 2016: 2; Feleke, 2018: 152). The initial search for relevant academic material on the delimitation of local boundaries found that most of the studies focused on the technical methodology of how boundaries can be delimited and the relevant outcomes. This may be because there is no standard way of delimiting boundaries globally and the outcome of the various methods brings forth different structures. Consequently, research output on the criteria is scarce. effects of local boundary extension and settlement incorporation (Zhou et al., 2016: 2; Feleke, 2018: 152). The initial search for relevant academic material on the delimitation of local boundaries found that most of the studies focused on the technical methodology of how boundaries can be delimited and the relevant outcomes. This may be because there is no standard way of delimiting boundaries globally and the outcome of the various methods brings forth different structures. Consequently, research output on the criteria is scarce. is more intensively developed into smaller administrative areas which allows for more efficient and coordinated management. However, the entire Metropolitan is managed through mutual cooperation. This kind of management is rather costly and tedious to manage. 1. INTRODUCTION The extension of the administrative boundary to incorporate many functionally related settlements would allow for the more efficient administrative management of the region as an entity. However, the functional region cannot be easily noticed and it could include a number of demographically categorised settlements and peri-urban zones, Historically, the categorisation of local administrations was based on the economic activities within the regions and they were delimited as either ‘urban’ or ‘rural’ areas. From a town-planning perspective, urban administrative regions included cities and towns, while rural areas encompassed villages and hamlets (UN, 2012: 4). However, the interdependent relationship between these administrative regions has resulted in many expanding beyond their delimited boundaries, thus challenging the 53 Jeeva, Cilliers & Gumbo 2023 Town and Regional Planning (82):52-61 3.1 Delimiting local government boundaries This documentation primarily consists of reports and policy documents from both governmental and non- governmental agencies, which were identified and sourced either through internet search terms, or because they were referenced or referred to in the referenced journal articles. It is important to note that this review utilises sources with varying degrees of credibility, and this may influence the final accuracy and validity of the discussion. Furthermore, the examples discussed were chosen on the basis that information was easy to access. Thus, educated assumptions had to be made regarding which data to use. The review approach included a thematic analysis to identify criteria used to delimit local boundaries. Subsequently, the main criterion identified from the literature review was coded and grouped into six criteria. In the discussion section, the literature on each coded criterion was evaluated to determine its main rationale, the outcome of the boundary delineation (large or small municipalities), and the criterion’s shortcomings, challenges and outcomes, and their potential for effective and efficient implementation. Figure 1: Different administrative boundaries Source: OECD, 2016: 2 Figure 1: Different administrative boundaries Source: OECD, 2016: 2 54 Jeeva, Cilliers & Gumbo 2023 Town and Regional Planning (82):52-61 Subsequently, a bottom-up planning approach, which allows for public participation in matters of policy, was introduced towards the late 1980s (Sikander, 2015: 173; UL Haque, 2013: 6). This allowed individuals to communicate on where they would like the administrative boundaries to extend to, with the aim of creating more efficient, effective, integrated, and equitable societies and administrative regions (Sikander, 2015: 174-176). However, the public engagement led to a perpetual altering of processes and goals, which, in turn, compromised efficiency and increased financial cost (Zhou et al., 2016: 2; ul Haque, 2013: 8; Cilliers, 2009: 943). as optimal operations, financial viability, cohesiveness, equality, or functionality are too often used to motivate the delimitation processes (Barlow & Wastl-Walter, 2017: 20; Dampegama et al., 2019: 197; Feleke, 2018: 128). this criterion has been used in a legislative context in countries such as the UK, Denmark, Australia, Canada, and the USA to motivate administrative delimitation based on its ability to create more efficient, accessible, and viable administrative regions (Glaeser, 2011: 1-5). In the South African case, municipal boundaries needed to be restructured post-1994, from being based on race to being more inclusive, integrated, and cohesive. 3.1 Delimiting local government boundaries The challenge, however, came in identifying the correct criteria to address the administrative objectives of a particular country and how this translates into creating efficient municipal regions – in many cases trade-off has to be made. In terms of application, national governments legally incorporated smaller settlements that are closely related, in order to create a single larger administrative region with the hope that it would reduce the fixed costs of infrastructure and decrease administrative duplication, while containing sprawl. However, research suggests that the relationship between size, performance, and agglomeration are inconclusive (Soja, 2015), since it is largely unknown how population size decreases the costs of providing services, given that the factors and circumstances of each municipality differ. Furthermore, the delimitation of the boundary in terms of distance from the centre or the size of the municipality or the required density within the municipality is unknown for it to benefit from agglomeration advantages or economics of scale (Giuliano et al., 2019: 379). As a result, during the economic recession of the 2000s, many national governments, globally, sought to find ways to save costs and become more efficient while engaging with the public. This brought about a renewed interest in the local administrative structure within their countries (OECD, 2016: 5). During the post-2000 period, many governments such as the USA, Canada, Australia, Turkey, Denmark, and South Africa legally incorporated settlements to form larger administrative regions, with the hope of containing sprawl and decreasing administrative cost, by encouraging higher density development through infill development (Dollery & Robotti, 2008: 10; Feleke, 2018: 141). In light of this renewed interest, some countries followed international election standards provided by regional and nongovernmental organisations (including the OSCE, the European Commission for Democracy Through Law, the Commonwealth Secretariat, and the Electoral Institute of Southern Africa [EISA]), while other governments created their own criteria on how their local administrative regions should be delimited. Research reveals that, even though definition and terms of administrative categories vary depending on the preferences of countries, most of the administrative regions are delimited based on the natural differentiation of space and the relevant spatial interaction (Halas et al., 2016: 2). Terms such 3.2 Rationales behind administrative delimitation from an urban planning perspective The delimitation of local municipal boundaries is a significant task considering that they contain multi- ethnic, multi-religion, and multicultural communities residing in regions that are developed on different levels (Dampegama et al., 2019: 198). Upon review, the criteria found in the literature that guide municipal delimitation have common themes such as population concentrations, functional linkages, cohesiveness, or integration (Handley, 2007: 61-62). With the common principles that underlie municipal delimitations is the hope of creating regions that would ensure optimal usage of resources, decreased expenditure, allow efficient administration, equitable voting capacity, and minimise social conflict while still being transparent in application (Mulindwa, 2020: 3-10). The following criteria are considered for delimiting local administrative boundaries, based on their outcomes and implications. However, there is a recognised trade-off between the size of a local authority and the quality of life (Glaeser, 2011: 1-5). Smaller administrative areas with high densities are well known to have higher administrative costs associated with congestion, pollution, social ills, and longer commuting times (OECD, 2020. On the other hand, larger administrations with lower population densities are costly to manage and many do not have the ability to attract investment (OECD, 2020). Subsequently, public choice theorists disagree with the agglomeration economics argument to delimit administrative regions, since they believe that smaller independent administrative regions promote competitive behaviour amidst themselves. This necessitates greater efficiency and quality of services than a single large administration (Combes, Duranton & Gobillon, 2011: 257). 3.2.2 Functional linkages Commuting patterns between settlements are common attributes used globally to delimit local administrative regions (Halas et al., 2016: 2). This is due to the strong socio-economic interrelationship between residential suburbs, peripheral settlements, and commercial centres. However, defining the spheres of influence and the range of the functional activity space involves identifying the settlement typology and measuring commuting patterns between the various settlements. This is challenging, considering that traffic is in constant flux (Martínez-Bernabes, Coombes & Casado-Díaz, 2020: 740- 743). Furthermore, to date there is no consensus regarding the intensity of the activity or the traffic required for it to form an administrative entity or the method that could be used to determine it (OECD, 2020). Consequently, central governments have to revise their boundaries on an annual or semi-annual basis to keep them updated if they are based on functional linkages, which is costly (Bennett, 1989). In Denmark, municipal areas were delimited based on access to health services. The Danish government condensed 1 386 boroughs in 1970 to create 275 municipalities in 2007. However, the cost of providing health services has been higher than expected. Thus, it could be said that delimitation has had limited success (Olejaz et al., 2012). 3.2.3 Economic functions This criterion shares obvious links with the agglomeration economics approach in that a delimited local boundary would have to contain a minimum number of people for it to be economically efficient and effective (Craythorne, 1998: 6; Nijkamp, 2013: 4). In this regard, the less important the economic functions on offer in the centre, the lower the population density required and vice versa (Christaller, 1966; Losch, 1962). This methodology is more objective than the functional linkages and agglomeration, as it enables consistent application (Slack & Bird, 2010: 5). For example, the OECD (2022: 1) proposes that urban areas can be delimited and categorised into four types according to population size: 3.2.1 Agglomeration Agglomeration economics consists of leveraging the benefits associated with the clustering of economic activities in close proximity with the hope that it will decrease the cost of infrastructure, administration, and production (Giuliano, Kang & Yuan, 2019: 377). In recent years, 55 Jeeva, Cilliers & Gumbo 2023 Town and Regional Planning (82):52-61 In summary, the agglomeration of scale criterion has been used to support both small and large territories. Small territories with higher densities are delimited to promote economics of scale and reduce administrative costs. Large territories are delimited to counteract sprawl and promote efficiency in terms of administration (Glaeser, 2011: 5). The motivation depends on the legislative stance of the country. For example, Australia reduced its municipalities from 869 in 1980 to 565 in 2014, based on the economics of scale or agglomeration criteria. However, the findings revealed no correlation between city size and performance (Dollery & Byrnes 2008: 5). that the region could be too large to administer properly, resulting in the region becoming non-profitable, as in the case of Canada. In Canada, municipalities were decreased from 851 municipalities in 1998 to 444 in 2009, based on mergers and cooperation related to functional linkages, equality, and socio- geographic relationships. However, amalgamation failed to solve any of the problems the municipalities faced prior to amalgamating (Slack & Bird, 2010: 8). which to determine boundaries or administrative classification, since they do not necessarily reflect the values of the local governments and their political objectives (Cameron, 2006: 80). Furthermore, regions are fluid and will need to be re-examined at least every five to 10 years, depending on the rate of demographic growth, rate of urbanisation, and economic development, which is costly (Giraut & Maharaj, 2002: 26). Moreover, there are many different economic activities and basic services on offer within a municipality, all of which differ in terms of the required threshold to be profitable. Which service or economic function would be given preference to determine the density required within an administrative region? As much as criteria motivate functional linkages as a basis for the legal incorporation of settlements, it is more challenging to implement them in practice and there is currently no consensus on how this can be implemented or measured (Halas et al., 2016: 3; OECD, 2020: 14). 3.2.4 Socio-demographic This delimitation criterion seeks to balance the need for speedy and sustainable delivery of services with the concern of addressing inequalities. From a socio- demographic perspective, it is argued that a more equitable distribution of services and taxes could be achieved, by merging a city with its poorer surrounding rural areas through consolidation, incorporation, or amalgamation (Cameron, 2001: 6; Glaeser, 2011: 2). This approach is similar to that adopted by the South African democratic government, which sought to integrate the previously segregated racial groups who had different tax bases (Mabin & Smit, 1997). • Small urban areas, with a population of below 200 000; • Medium-sized urban areas, with a population of between 200 000 and 500 000; • Metropolitan areas, with a population of between 500 000 and 1.5 million, and Many governments who use this approach opt for over-bounded local boundaries that extend beyond the range of normal activities of the majority of the people, thus allowing for future growth and expansion (see Rome in Figure 1) (Ramutsindela, 2013: 54). However, the problem with this over-bounded approach is However, researchers have reservations about using this criterion to create administrative regions, in order to reduce inequality. As municipalities could risk losing their more affluent ratepayers, who would object to paying higher taxes, the • Large metropolitan areas, with a population of 1.5 million or more. Nonetheless, while population number and population density are useful, on their own they do not provide a clear basis on 56 Jeeva, Cilliers & Gumbo 2023 Town and Regional Planning (82):52-61 problems (Fonseca, Lukosch & Brazier, 2019: 231-240). region could instead attract less affluent citizens who do not have funds to pay taxes (Cameron, 2006: 78; Geyer, Geyer & Du Plessis, 2015: 1-2). In addition, if rich and poor areas are consolidated, the tax base has to be divided over a larger area. The question then is: Do you develop the underdeveloped regions with the municipal funds, or should the funds be used to maintain aged infrastructure? In the case of South Africa, this has contributed to the deterioration of local municipalities. in the area that pay taxes (as the economic approach discussed in 3.2.3), or would it require a certain number of employed citizens who pay taxes and rates? This is unclear. 3.2.5 Financial viability Equality of voting strength is a recurrent principle found in delimitation reports. In many cases, this comes in the form of political motivations for local administrative boundary adjustments. However, government is not always candid and the motivation comes in the form of quoting ‘efficiency, sustainability, equity, viability, inclusive and integration’. Each of these principles is important but cannot be objectively measured or determined, since it differs according to context (Dampegama et al., 2019: 197). This makes the motivation and rationale elusive and easy to manipulate, in order to reach alternative political objectives. Financial viability is defined as a municipality’s ability to meet its expenditure commitments and cost obligations from its own income stream, including revenues, delivery of services, property taxes, and transfers (Harrison & Todes, 2016: 1-5). In other words, if an administrative area is not financially viable, self-sufficient, or self-reliant, it will not be able to afford its expenses from its own revenue and will depend on funding from national government (Khumelo & Ncube, 2016: 6-8). Thus, it needs to be an optimal size for it to have a tax base that will enable it to be profitable and financially viable (Rondinelli, McCullough & Johnson, 2012: 57-60). Optimal, in this sense, refers to the urban size or administrative region that would offer the greatest return. This is where the total benefit (average revenue and marginal revenue) is higher than the total cost (average cost and marginal cost) (Camagni, Diappi & Leonardi, 1986: 150). The problem, however, is: How would the boundary be delimited to ensure viability? Would it be based on the number of businesses In South Africa, financial viability was a delimiting criterion between 2011 and 2016. The Municipal Demarcation Board claimed that, by merging financially non-viable municipalities with financially viable municipalities, they would be self-sufficient (Thupane, 2015). However, in 2022, at least one third of the South African municipalities were dysfunctional. This brings into question the success of the mergers (Jeeva & Cilliers, 2021: 83). 3.2.4 Socio-demographic In light of this, some countries believe that smaller administrative units offer a more socially homogeneous composition and a greater sense of community, as in the apartheid model of the South African government (Skaburskis, 1992: 161). Alternatively, ethnic lines could be exchanged with national lines to create a large municipality with a heterogeneous population to incorporate a nationalist mindset – similar to the democratic government of South Africa. However, this heterogeneity could and does give rise to cultural conflict and feelings of under-representation on the part of the minority populations (Fonseca et al., 2019: 231-245; Cameron, 2001: 9). Thus, a legal incorporation based on cohesiveness depends on the political value preferences of the national government. How big or small these administrative regions need to be is also left open and differs according to context. According to Khumelo and Ncube (2016: 5), there is no apparent connection between financial viability and delimitation, since many factors, including poor management, political and administrative issues, poor investment decisions, low revenue bases, unemployment, high dependency, and poverty levels within its borders, can cause a municipality to become financially non-viable (Rondinelli et al., 2012: 57-60). Hence, legal incorporation will not necessarily make weaker municipalities financially viable, self-sufficient, or self-reliant, but could result in the contrary if internal issues are not addressed appropriately (Cameron & Meligrana, 2010: 10). In addition, it is not known whether larger municipalities would be more financially viable or whether smaller municipalities would be better. This also ties in with the agglomeration criterion. Furthermore, in this delimitation approach, municipalities could also become too large if rural areas are located too far from urban areas. Nevertheless, these scenarios are influenced by the politics of the consolidated municipalities and the preferences of the national government (Cameron, 2001: 5). This approach is believed to bring about spatial equity. However, there is no set way of delimiting the region and there are limited results to prove it to date. 3.2.6 Cohesiveness or nation building There is growing evidence that social infrastructure, as opposed to physical infrastructure, creates a sense of community (Aldrich & Kyota, 2017). The reason for this is that social infrastructure creates a form of collective identity that requires common history, culture, geography, as well as social and economic conditions. This is, however, a complex social construct, due to different societies having different geographies, political representations, economics, and According to Stabingis (2015: 3), one of the telling signs that politicians are seeking influence over administrative regions to secure maximum votes is when these regions are often larger than others and do not offer much equity to the citizens. This guarantees the ruling power within that region more parliamentary seats. However, this criterion is not acknowledged and not openly discussed, making 57 Jeeva, Cilliers & Gumbo 2023 Town and Regional Planning (82):52-61 it challenging to measure and contest (Cameron, 2001: 8; Barlow & Wastl-Walter, 2017: 50). qualitative data and cannot really be measured) (see Table 1). the desires of the central government (see section 3.1) (Beauregard, 1991: 191). Since the 1980s, a more liberal bottom-up approach was adopted to allow communities to have input on how boundaries should be delineated (Sikander, 2015: 173). However, due to the multicultural background and various motivating factors, the delineation based on public opinion was challenging. As a result, in recent years, there has been a further shift in the search for a middle ground between top-down and bottom-up planning to create more efficient and administratively effective, as well as context-sensitive regions. Consequently, many governments employed independent delimitation commissions to delimit and standardise their local administrative regions, in categories based on legislation and public consultation. It was anticipated that the standardisation would help compare, plan, and implement a strategic socio-economic policy. From Table 1, it can be noted that criteria guiding delimitation are currently subjective and context specific. As a result, the outcomes are inconsistent and lack transparency and can be easily manipulated in the method of delineation. For instance, the motivation to create compact administrative regions, by increasing density and clustering services and infrastructure, comes about strongly in the agglomeration economics, socio- geographic and financial viability motivations. This is why the functional, economic, and cohesive approaches appear to support larger and more integrated regions, in order to increase the market area and decrease administrative costs. 3.2.6 Cohesiveness or nation building However, all these criteria are subjective and applied according to the importance provided by central government. Consequently, the subjectivity is open to political manipulation which could easily be packaged differently, since principles such as integration, equity, and equality cannot be measured spatially. From the above review, it can be concluded that there is currently no well-defined, precise, best single model for the delimitation of local administrative boundaries. Many criteria, some of which are opposing, are used to motivate the delimitation of local boundaries. The review also highlights that the criterion can be defined and measured in different ways. As a result, the outcome differs according to context and the country implementing it. If anything, the review highlights the complexity and the lack of objectivity in delimiting local boundaries, making it difficult to apply consistently and easy to manipulate. 4. DISCUSSION The delimitation of administrative regions leads to the categorisation of administrative units, which are important for strategic planning and policy implementation. The approach to delimiting administrative regions was previously based on a top-down modernist approach, solely based on The review identified criteria that guide delimitation globally post-2000, in terms of their objectivity (it can be measured and is transparent) and subjectivity (based on Table 1: Criteria on boundary demarcation and their associated shortcomings Criteria Motivation Challenge of application Outcome Literature Country Economics of scale / Agglomeration Clustering settlements to decrease administrative costs and basic service duplication to increase profits. No proof to date that larger administrative entities are more profitable than smaller ones, or if smaller regions decrease service costs and promote more competition. Subjective application Giuliano et al., 2019 Dollery and Byrnes, 2008 Glaeser, 2011 Australia Functional approach Minimal distance required to make the provision of a service profitable between centres. Unknown what the maximum distance between settlements is required to draw a boundary. Internet can offer services without being dependant on immediate population. Could be objective but no threshold determined to date on what the size of the region needs to be or the distance between settlements. Hanes and Wikström, 2010 Canada Economic approach Minimal threshold required to make the provision of a service profitable in the region. The viability of economic services is different. Hence, it is difficult to prove which size would be best. Could be objective but the threshold for each service is different to be provided efficiently. Olejaz et al., 2012 Denmark Financial viability To decrease administrative costs and increase profits by delimiting areas optimally. Different services require different thresholds – which service would be used to delimit boundaries to determine optimal city size and profitability. Subjective Slack and Bird, 2010 Thupane, 2015 Canada South Africa Cohesiveness Cluster settlements that have socio-geographic or cultural linkages to form one identity. Unknown to date where or how boundary lines are to be demarcated. Subjective Mabin and Smit, 1997 South Africa Political motivation Political motivation of boundaries to ensure integration, equality, equity, and sustainability. Subjective criteria that are easily manipulated based on context and definition. Subjective Stabingis, 2015 Not easy to clearly state or prove. 4. DISCUSSION Source: Authors Source: Authors 58 Jeeva, Cilliers & Gumbo 2023 Town and Regional Planning (82):52-61 Furthermore, even though all the above-mentioned criteria are important for administrative delineation, the simultaneous implementation of all delimitation criteria and principles is not always possible. There must be trade-offs and negotiations between numerous independent participants such as businesses, lobbyists, politicians, real-estate developers, and the public (Glushko, 2016: 52). Accordingly, the outcome of the delimited region could mean smaller or larger municipalities. BENNETT, R. 1989. Territory and administration in Europe. London, UK: Pinter. BENNETT, R. 1989. Territory and administration in Europe. London, UK: Pinter. administrative units more effective and optimal (OECD, 2020: 12). Subsequently, it could also be argued that the criteria discussed above are too focused on the effective and efficient governance of municipalities and do not take note of sustainability or quality of life. In light of this, the 2030 Agenda for Sustainable Development (UN 2015) includes several indicators that could be collected for cities or for rural and urban areas to ensure sustainable delimitation and categorisation. However, the data required to apply contemporary methodologies are not always collected across countries. CAMAGNI, R., DIAPPI, L. & LEONARDI, G. 1986. Urban growth and decline in a hierarchical system. A supply-oriented dynamic approach. Regional Science and Urban Economics, 16(1), pp.145-160. DOI:10.1016/0166-0462(86)90017-7 CAMERON, R. 2001. The upliftment of South African local government? Local Government Studies, 27(3), pp. 97-118. DOI:10.1080/714004109 Moreover, the delimited administrative regions need to be categorised, in order to provide an understanding of the region. Since the delimitation is based on a trade- off between many choices, including the nature of resources, the purpose of the delimitation, the effort put into the delimitation, the intricacy of the domain, and the capabilities of the people doing the delimitation, the result is biased in one way or another and largely inconsistent. Globally, many administrative boundaries have had to be re-delimited since the early 2000s to either separate areas that have fused or consolidate areas that were functionally related, in order to combat sprawl, promote smart growth, efficient management, and enforce authority over a region. Considering the contemporary relevance of the exercise and the impact of administrative delineation in a country, it would have been expected that a clear method would be prescribed, at least from a theoretical perspective, on how local boundaries could be demarcated. REFERENCES ALDRICH, D.P. & KYOTA, E. 2017. Creating community resilience through elder-led physical and social infrastructure. Disaster Medicine and Public Health Preparedness, 11(1), pp.120-126. DOI:10.1017/ dmp.2016.206 CRAYTHORNE, D.L. 1998. Municipal administration: A handbook. 4th edition. Cape Town, South Africa: Juta Academic. DAMPEGAMA, K.P., HALANAYAKE, N.T. & RANAWAKA, C. 2019. Delimitation of administrative boundaries: History, importance and using Geographic Information System for improved delimitation of administrative boundaries. In: Proceedings of the 12th International Research Conference, 11-12 September, Kotelawala Defence University (KDU), Ratmalana, Sri Lanka, pp. 197-208. ACKNOWLEDGEMENTS This work is based on research supported by the National Institute for The Humanities and Social Sciences. COMBES, P.-P., DURANTON, G. & GOBILLON, L. 2011. The identification of agglomeration economies. Journal of Economic Geography, 11(2), pp. 253‑266. DOI: 10.1093/jeg/lbq038 CRAYTHORNE, D.L. 1998. Municipal administration: A handbook. 4th edition. Cape Town, South Africa: Juta Academic. COMBES, P.-P., DURANTON, G. & GOBILLON, L. 2011. The identification of agglomeration economies. Journal of Economic Geography, 11(2), pp. 253‑266. DOI: 10.1093/jeg/lbq038 4. DISCUSSION However, this study found that there are not many theories or technical methodologies that could be referred to, in order to guide delimitation. CAMERON, R. 2006. Local government boundary reorganization. In: Pillay, U., Tomlinson, R. & Toit, D. (Eds). Democracy and delivery: Urban policy in South Africa. Cape Town: HSRC Press, pp. 76-106. This article recommends that further research be conducted on how administrative regions could be delimited based on the quality of life in administrative regions to be more socially conscious and sustainable. It further reveals that the application of classic theories in delimiting regions is subjective and that the outcome differs according to context. This article calls for more objective criteria to be developed, in order to ensure that delimited regions are used not only to increase profits, but also to promote sustainability and offer a good quality of life to all within the administrative regions. CAMERON, R.G. & MELIGRANA, J. 2010. Criteria for determining municipal boundaries and categorisation of metropolitan municipalities. MDB (Bid No. A03/2010). Pretoria: MDB Publication. CHRISTALLER, W. 1933. (1966) Central places in Southern Germany. Translated by C. Baskin. Englewood Cliffs, NJ: Prentice Hall. CILLIERS, J.E. 2009. The Urban Development Boundary as a planning tool for sustainable urban form: The South African situation. International Journal of Humanities and Social Sciences, 3(6), pp. 942-947. DOI: 10.5281/zenodo.1078478 CHRISTALLER, W. 1933. (1966) Central places in Southern Germany. Translated by C. Baskin. Englewood Cliffs, NJ: Prentice Hall. CILLIERS, J.E. 2009. The Urban Development Boundary as a planning tool for sustainable urban form: The South African situation. International Journal of Humanities and Social Sciences, 3(6), pp. 942-947. DOI: 10.5281/zenodo.1078478 5. CONCLUSIONS & CASADO-DÍAZ, J.M. 2020. Functional regions for policy: A statistical ‘toolbox’ providing evidence for decisions between alternative geographies. Applied Spatial Analysis and Policy, 13(3), pp. 739-758. DOI: 10.1007/s12061-019-09326-2 England: Edward Elgar Publishing. https://doi.org/10.4337/9781781956687 DRUPAL. 2021. Administrative boundary hierarchy. [Online]. Available at: <https://www.drupal.org/ node/2063033> [Accessed: 12 May 2022]. GLUSHKO, R.J. 2016. The discipline of organizing: Professional edition. GLUSHKO, R.J. 2016. The discipline of organizing: Professional edition. 4th edition. Sebastopol, CA, USA.: O’Reilly Media, Inc. MULINDWA, P. 2020. Interstate border conflicts and their effects on region- building and integration of the east African community. African Journal of Governance and Development, 9(2), pp. 599-620. DOI: 9971-1-10-20201223 NIJKAMP, P. 2013. The universal law of gravitation and the death of distance. Romanian Journal of Regional Science, 7(2), pp. 1-10. MULINDWA, P. 2020. Interstate border conflicts and their effects on region- building and integration of the east African community. African Journal of Governance and Development, 9(2), pp. 599-620. DOI: 9971-1-10-20201223 HALAS, M., MARTIN, E., PAVEL, K. & MARTIN, T. 2016. Efficiency of delineation of administrative regions: A functional region approach. In: Proceedings of the 19th International Colloquium on Regional Sciences, 15‑17 June, Cejkovice, Czech Republic, pp. 565-572. 10.5817/ CZ.MUNI.P210-8273-2016-72 FELEKE, A.T. 2018. The dilemma of determining urban administrative boundary for land administration: The case of Shire Inda Silase. International Journal of Social Sciences Perspectives, 2(2), pp. 126-160. DOI: 10.33094/7.2017.2018.22.126.160 FELEKE, A.T. 2018. The dilemma of determining urban administrative boundary for land administration: The case of Shire Inda Silase. The case of Shire Inda Silase. International Journal of Social Sciences Perspectives, 2(2), pp. 126-160. DOI: 10.33094/7.2017.2018.22.126.160 NIJKAMP, P. 2013. The universal law of gravitation and the death of distance. Romanian Journal of Regional Science, 7(2), pp. 1-10. FONSECA, X., LUKOSCH, S. & BRAZIER, F. 2019. Social cohesion revisited: A new definition and how to characterize it. Innovation, 32(2), pp. 1-23. DOI: 10.1080/13511610.2018.1497480 HANDLEY, L. 2007. Challenging the norms and standards of election administration: Boundary delimitation. IFES White Paper funded by USAID, pp. 59-77. OECD (Organization for Economic Cooperation and Development). 2016. Comparison of administrative boundaries and the urban extent of cities. [Online]. Available at: <https:// www.oecd-ilibrary.org/sites/ba8c7dc7- en/index.html?itemId=/content/ component/ba8c7dc7-en> [Accessed: 10 January 2022]. HANES, N. & WIKSTRÖM, M. 2010. Amalgamation impacts on local growth: Are voluntary amalgamations more efficient than compulsory amalgamations? Canadian Journal of Regional Economics, 33(1), pp. 57-70. FRIEDMANN, J. 1978. The spatial organization of power in the development of urban systems. In: Bourne, L.S. 5. CONCLUSIONS DAMPEGAMA, K.P., HALANAYAKE, N.T. & RANAWAKA, C. 2019. Delimitation of administrative boundaries: History, importance and using Geographic Information System for improved delimitation of administrative boundaries. In: Proceedings of the 12th International Research Conference, 11-12 September, Kotelawala Defence University (KDU), Ratmalana, Sri Lanka, pp. 197-208. BARLOW, M. & WASTL-WALTER, D. 2017. New challenges in local and regional administration. London, UK: Routledge. DOI:10.4324/9781315248110 BEAUREGARD, R.A. 1991. Without a net: Modernist planning and the postmodern abyss. Journal of Planning Education and Research, 10(3), pp.189-194. DOI:10.1177/0739456x9101000305 BARLOW, M. & WASTL-WALTER, D. 2017. New challenges in local and regional administration. London, UK: Routledge. BARLOW, M. & WASTL-WALTER, D. 2017. New challenges in local and regional administration. London, UK: Routledge. DOI 10 4324/9781315248110 This article showed that the principled approach based on subjective criteria does not necessarily mean that it is transparent or that it is consistently applied. This results in there being no way of comparing and contrasting administrative units globally. This makes it difficult to determine what is efficient and what can be applied to make the BEAUREGARD, R.A. 1991. Without a net: Modernist planning and the postmodern abyss. Journal of Planning Education and Research, 10(3), pp.189-194. DOI:10.1177/0739456x9101000305 59 Jeeva, Cilliers & Gumbo 2023 Town and Regional Planning (82):52-61 DOLLERY, B. & BYRNES, J. 2008. Local government amalgamation: A conceptual analysis population size and scale economies in municipal service provision. Australasian Journal of Regional Studies, 14(2), pp. 167-175. DOLLERY, B. & BYRNES, J. 2008. Local government amalgamation: A conceptual analysis population size and scale economies in municipal service provision. Australasian Journal of Regional Studies, 14(2), pp. 167-175. MABIN, A. & SMIT, D. 1997. Reconstructing South Africa’s cities? The making of urban planning 1900-2000. Planning Perspectives, 12(2), pp. 193-223. DOI: 10.1080/026654397364726 GIULIANO, G., KANG, S. & YUAN, Q. 2019. Agglomeration economies and evolving urban form. The Annals of Regional Science, 63(3), pp. 377-398. DOI: 10.1007/s00168-019-00957-4 MABIN, A. & SMIT, D. 1997. Reconstructing South Africa’s cities? The making of urban planning 1900-2000. Planning Perspectives, 12(2), pp. 193-223. DOI: 10.1080/026654397364726 GLAESER, E.L. 2011. Introduction to “Agglomeration Economics”. In: Glaeser, E.L. (Ed.). Agglomeration economics. Chicago, Illinois: The University of Chicago Press, pp. 1-14. https://doi.org/10.7208/ chicago/9780226297927.003.0001 DOLLERY, B.E. & ROBOTTI, L. (Eds). 2008. The theory and practice of local government reform. Cheltenham, England: Edward Elgar Publishing. https://doi.org/10.4337/9781781956687 DOLLERY, B.E. & ROBOTTI, L. (Eds). 2008. The theory and practice of local government reform. Cheltenham, MARTÍNEZ-BERNABEU, L., COOMBES, M. KHUMELO, B. & NCUBE, M. 2016. KHUMELO, B. & NCUBE, M. 2016. Financial and functional viability and sustainability of municipalities: Beyond the demarcation instrument. Paper presented at the MDB Conference on Demarcation and Spatial Transformation. Financial and Fiscal Commission, Pretoria, 23 June LÖSCH, A. 1962. Die räumliche KHUMELO, B. & NCUBE, M. 2016. Financial and functional viability and sustainability of municipalities: Beyond the demarcation instrument. Paper presented at the MDB Conference on Demarcation and Spatial RUDKJOBING, A., BIRK, O., KRASNIK. A. & HERNANDEZ- QUEVEDO, C. 2012. Denmark: Health system review. Health System Transit, 4(2), pp. 1-192. PMID: 22575801. RAMUTSINDELA, M.F. 2013. Experienced regions and borders: The challenge for transactional approaches. Regional Studies, 47(1), pp. 43-54. https://doi.org/10.1080/00343404.2011 .618121 RAMUTSINDELA, M.F. 2013. Experienced regions and borders: The challenge for transactional approaches. Regional Studies, 47(1), pp. 43-54. https://doi.org/10.1080/00343404.2011 .618121 5. CONCLUSIONS & Simmons, J.W. (Eds). Systems of cities: Readings on structure, growth, and policy. New York: Oxford University Press, pp. 328-339. FRIEDMANN, J. 1978. The spatial organization of power in the development of urban systems. In: OECD (Organization for Economic Cooperation and Development). 2020. Why delineate functional areas in all territories? In: OECD Territorial reviews – Delineating functional areas in all territories, pp. 11-18. https://doi. org/10.1787/07970966-en HARRISON, P. & TODES, A. 2016. Spatial transformations in a “loosening state”: South Africa in a comparative perspective. Geoforum, 61, pp. 148-161. https://doi.org/10.1016/j. geoforum.2015.03.003 FULKERSON, A.L., WAXMAN, S.R. & SEYMOUR, J.M. n.d. Linking object names and object categories: Words (but not tones) facilitate object categorization in 6- and 12-month- olds, Www.bu.edu. [Online]. Available at: <https://www.bu.edu/bucld/ files/2011/05/30-FulkersonBUCLD2005. pdf> [Accessed: 8 August 2022]. OECD (Organization for Economic Cooperation and Development). 2022. Urban population by city size (indicator). OECD iLibrary. DOI: 10.1787/b4332f92-en JEEVA, Z. & CILLIERS, J. 2021. An explorative approach to the evolving municipal landscape of South Africa: 1993-2020. Town and Regional Planning, 78, pp. 81-91. https://doi. org/10.18820/2415-0495/trp78i1.6 GEYER, H.S., Jr, GEYER, H.S. & DU PLESSIS, D. 2015. Primary cities in sub-Saharan Africa: Quasars, loose connections, and black holes. International Planning Studies, 20(1-2), pp. 39-51. DOI: 10.1080/13563475.2014.942494 GEYER, H.S., Jr, GEYER, H.S. & DU PLESSIS, D. 2015. Primary cities in sub-Saharan Africa: Quasars, loose connections, and black holes. International Planning Studies, 20(1-2), pp. 39-51. DOI: 10.1080/13563475.2014.942494 OLEJAZ, M., NIELSEN, J., RUDKJOBING, A., BIRK, O., KRASNIK. A. & HERNANDEZ- QUEVEDO, C. 2012. Denmark: Health system review. Health System Transit, 4(2), pp. 1-192. PMID: 22575801. GIRAUT, F. & MAHARAJ, GIRAUT, F. & MAHARAJ, B. 2002. Contested terrains: Cities and hinterlands in post- apartheid boundary delimitations. GeoJournal, 57(1/2), pp. 39-51. DOI: 10.1023/a:1026018819206. GIRAUT, F. & MAHARAJ, B. 2002. Contested terrains: Cities and hinterlands in post- apartheid boundary delimitations. GeoJournal, 57(1/2), pp. 39-51. DOI: 10.1023/a:1026018819206. LÖSCH, A. 1962. Die räumliche Ordnung der Wirtschaft. 3rd edition. Berlin, Germany: De Gruyter Oldenbourg. LÖSCH, A. 1962. Die räumliche Ordnung der Wirtschaft. 3rd edition. Berlin, Germany: De Gruyter Oldenbourg. 60 Jeeva, Cilliers & Gumbo 2023 Town and Regional Planning (82):52-61 Jeeva, Cilliers & Gumbo 2023 Town and Regional Planning (82):52-61 UL HAQUE, A. 2013. Theoretical perspective of local government: Literature review. Munich Personal RePEC Archive. MPRA Paper No. 45868. [Online]. Available at: <https:// mpra.ub.uni-muenchen.de/45868/> [Accessed: 2 August 2022]. RONDINELLI, D.A., MCCULLOUGH, J.S. & JOHNSON, R.W. 2012. Analysing decentralization policies in developing countries: A political- economy framework. Development and Change, 20(1), pp. 57-87. https:// doi.org/10.1111/j.1467-7660.1989. tb00340.x UN (United Nations). 2012. Chapter III: Towards sustainable cities. [Online]. Available at: <http://www.un.org/ en/development/desa/policy/wess/ wess_current/wess2013/Chapter3.pdf> [Accessed: 2 February 2017]. SIKANDER, T. 2015. A theoretical framework of local government. SIKANDER, T. 2015. A theoretical framework of local government. International Journal of Humanities and Social Science, 5(6), pp. 171-178. International Journal of Humanities and Social Science, 5(6), pp. 171-178. SINCLAIR-SMITH, K. 2015. Polycentric development in the Cape Town city-region: Empirical assessment and consideration of spatial policy implications. Development Southern Africa, 32(2), pp. 131-150. DOI: 10.1080/0376835x.2014.984378 UN (United Nations). 2015. 2030 Agenda for Sustainable Development [Online]. Available at: <https://www. undp.org/sustainable-development- goals> [Accessed: 2 February 2017]. UN (United Nations). 2015. 2030 Agenda for Sustainable Development [Online]. Available at: <https://www. undp.org/sustainable-development- goals> [Accessed: 2 February 2017]. VASANEN, A. 2012. Functional polycentricity: Examining metropolitan spatial structure through the connectivity of urban sub-centres. Urban Studies, 49(16), pp. 3627-3644. DOI: 10.1177/0042098012447000 SKABURSKIS, A. 1992. Goals for restructuring local government SKABURSKIS, A. 1992. Goals for restructuring local government boundaries: Canadian lessons. Environment and Planning C: Government and Policy, 10(2), pp. 159‑172. https://doi.org/10.1068/ c100159 WEBSTER, D. & MULLER, L. 2004. Peri-urbanization: Zones of rural-urban transition. In: Sassen, S. (Ed.). Human settlement development, volume 1. Chapter 13. Paris, France, UNESCO: Eolss Publishers, pp. 281-290. SLACK, E. & BIRD, R. 2010. Merging municipalities: Is bigger better? [Online]. Available at: <https://tspace.library.utoronto. ca/bitstream/1807/81253/1/imfg_ paper_14_slack_bird_Feb_27_2013. pdf> [Accessed: 5 May 2017]. ZHANG, X.Q. 2016. The trends, promises and challenges of urbanisation in the world. Habitat International, 54, pp. 241-252. DOI: 10.1016/j.habitatint.2015.11.018 ZHANG, X.Q. 2016. GIRAUT, F. & MAHARAJ, The trends, promises and challenges of urbanisation in the world. Habitat International, 54, pp. 241-252. DOI: 10.1016/j.habitatint.2015.11.018 ZHOU, R., ZHANG, H., XIN-YUE, Y., WANG, X. & SU, H. 2016. The delimitation of urban growth boundaries using the CLUE-S land-use change model: Study on Xinzhuang town, Changshu city, China. Sustainability, 8(11), article number 1182. DOI: 10.3390/su8111182 SOJA, E. 2015. Accentuate the regional. International Journal of Urban and Regional Research, 39(2), pp. 372‑381. doi. org/10.1111/1468-2427.12176 ZHOU, R., ZHANG, H., XIN-YUE, Y., WANG, X. & SU, H. 2016. The delimitation of urban growth boundaries using the CLUE-S land-use change model: Study on Xinzhuang town, Changshu city, China. Sustainability, 8(11), article number 1182. DOI: 10.3390/su8111182 SPORNA, T. & KRZYSZTOFIK, R. 2020. Inner suburbanisation – Background of the phenomenon in a polycentric, post-socialist and post- industrial region. Example from the Katowice conurbation, Poland. Cities, 104, article number 102789. DOI: 10.1016/j.cities.2020.102789 STABINGIS, R. 2015. Case study: Equal voting power and boundary delimitation in Lithuania. [PPT]. [Online]. Available at: <https:// aceeeo.org/sites/default/files/ case_study_equal_voting_power_and_ boundary_delimitation_in_lithuania.pdf> [Accessed: 5 July 2018]. THUPANE, J. 2015. SA municipal boundaries changed ahead of 2016 elections. [Online]. Available at: <https://businesstech.co.za/news/ government/97069/sa-municipal- boundaries-changed-aheadof-2016- elections/> [Accessed: 5 July 2018]. 61
https://openalex.org/W2515848441	https://europepmc.org/articles/pmc5129240?pdf=render	English	null	Displaced populations due to humanitarian emergencies and its impact on global eradication and elimination of vaccine-preventable diseases	Conflict and health	2,016	cc-by	2,403	© 2016 The Author(s). Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated. Abstract Populations affected by humanitarian emergencies may require unique strategies to ensure access to life-saving vaccines and attain sufficiently high population immunity to interrupt virus circulation. Vaccination strategies among displaced populations should not be an afterthought and must be part of the vaccine-preventable disease eradication and elimination initiatives from the start. Keywords: Vaccines, Refugees, Poliomyelitis, Measles, Rubella, Disease eradication, Internally displaced persons, Humanitarian emergency, Immunizations, Outbreaks, Civil conflicts, Displacement agent as a result of deliberate efforts; whereas elimin- ation is the reduction of the incidence of a specified dis- ease to zero in a defined geographic area [3]. Among the eradication and elimination initiatives that are currently in place, perhaps the two most prominent and relevant to emergency settings are the Global Polio Eradication Initiative (GPEI) and the Measles and Rubella Initiative (MRI) [4, 5]. The GPEI launched at the World Health Assembly in 1988 and is a joint partnership between United Nations Children’s Fund (UNICEF), Rotary International, the United States Centers for Disease Control and Prevention (CDC), and the World Health Organization (WHO). The MRI began in 2001 and is a partnership between the American Red Cross, CDC, UNICEF, United Nations Foundation and WHO. With the respective goals of certified global eradication of polio by 2018 [4] and the elimination of measles and rubella in at least five WHO regions by 2020 [5], these programs are among the most promising and ambitious public health endeavors of our time. According to the UN High Commissioner for Refugees (UNHCR), there were 59.5 million forcibly displaced persons worldwide by the end of 2014 due to persecu- tion, conflict, generalized violence, or human rights violations, and this number is only expected to increase with numerous ongoing global conflicts [1]. These for- cibly displaced populations include both refugees who cross international boarders in their escape from con- flict, and internally displaced persons (IDPs) who flee conflict but stay within the borders of their own country. The unique challenges of emergency settings often inter- fere with routine health services and prevent access to rec- ommended vaccinations. This disruption of immunization services increases the number of susceptible individuals and creates a population at particularly high risk for vaccine-preventable diseases (VPDs) targeted for eradica- tion and elimination [2]. There are a number of VPD eradication and elimin- ation initiatives taking place worldwide. Open Access * Correspondence: elam@cdc.gov 1Emergency Response and Recovery Branch; Division of Global Health Protection, Center for Global Health; Centers for Disease Control and Prevention, 1600 Clifton Road NE, Atlanta, GA 30309, USA Full list of author information is available at the end of the article Displaced populations due to humanitarian emergencies and its impact on global eradication and elimination of vaccine- preventable diseases Eugene Lam1, Michael Diaz2, Allen Gidraf Kahindo Maina3 and Muireann Brennan1 Lam et al. Conflict and Health (2016) 10:27 DOI 10.1186/s13031-016-0094-5 Lam et al. Conflict and Health (2016) 10:27 DOI 10.1186/s13031-016-0094-5 Abstract Eradication is defined as the permanent reduction to zero of the worldwide incidence of infection caused by a specific Global VPD eradication and elimination initiatives are resource intensive in terms of finance and human re- sources. Experiences with GPEI and MRI have shown that populations displaced due to humanitarian emer- gencies can increase the risk of VPD outbreaks and thereby adding to the resources needed to meet VPD Correspondence: elam@cdc.gov 1Emergency Response and Recovery Branch; Division of Global Health Protection, Center for Global Health; Centers for Disease Control and Prevention, 1600 Clifton Road NE, Atlanta, GA 30309, USA Full list of author information is available at the end of the article Lam et al. Conflict and Health (2016) 10:27 Page 2 of 3 Page 2 of 3 eradication and elimination goals [6–10]. Although most outbreaks of polio and measles worldwide do not occur among displaced persons, recent outbreaks among refu- gees, IDPs, and conflict-affected populations suggest that these vulnerable groups are at risk of transmission of VPDs. The breakdown of immunization services in Syria due to the protracted civil war has caused multiple out- breaks of VPDs, both within the country and the region. The reemergence of polio in Syria for the first time in over 15 years caused concern of re-introduction of polio to neighboring countries into areas with suboptimal vaccination [6]. Outbreaks of measles also increased substantially across Syria and its neighbors including Turkey, Jordan, and Lebanon [10]. routine immunization to prevent repeated large-scale outbreaks of VPDs [17]. Roles and responsibilities of dif- ferent agencies regarding various aspects of the vaccin- ation programs should be clearly allocated among local and international agencies to capitalize on operational strengths of each partner. Operational guidelines for VPD eradication and elimination efforts should be adapted to the evolving epidemiology of both the disease and affected populations in order to address the unique challenges of reaching displaced populations in humani- tarian emergencies [18–20]. Unless plans are made upfront to include displaced populations in national and regional VPD control ef- forts, global eradication and elimination goals may not be achieved within the targeted time frame. Vulnerable populations affected by humanitarian emergencies may require unique strategies to ensure that these popula- tions have access to life-saving vaccines and attain suffi- ciently high population immunity to interrupt virus circulation. Consent for publication Not applicable. Consent for publication Not applicable. Availability of data and materials Availability of data and materials Not applicable. Funding Various guidance documents for vaccination efforts in acute humanitarian emergencies are currently available [13, 14] Traditional vaccination strategies among dis- placed populations have focused on camp settings in- cluding vaccination of new arrivals, utilization of transit centers to provide health services including vaccinations, and mass vaccination campaigns [15]. As more displaced persons have moved out of camps and into urban settings [16], different strategies to increase vaccination coverage among non-camp populations are warranted given the unique challenges in locating such populations. To ensure that health services are available to displaced populations who are not in camp setting but integrated into host communities, key partners should provide ef- fective social mobilization and communication to facili- tate access to vaccination and other health services along with appropriate language and translation support. Protracted emergencies highlight the need to reinforce Abstract The Global Vaccine Action Plan (GVAP) 2011–2020 highlights, in its third strategic objective, the importance of ensuring equitable access of vaccines to all people including those in humanitarian crises and conflict zones [21]. GVAP calls for effective communica- tion and establishment of vaccine stockpiles to prevent and respond to VPDs among displaced populations in humanitarian emergencies. If vaccines are to be equit- ably accessible to all, deliberate efforts to include the most marginalized communities are a must. For this rea- son, vaccination strategies among displaced populations should not be an afterthought and must be part of the eradication and elimination strategies from the start. y J [ ] The spread of VPDs among displaced populations and susceptible host communities has not been limited to the Middle East. A 2013 outbreak of polio in refugee camps along the Kenya Somalia border spread into the surrounding communities resulting in reactive vaccin- ation efforts in both the camps and non-camp popula- tions [7]. Outbreaks of measles in Dadaab refugee camp in Kenya and Dollo Ado refugee camp in Ethiopia in 2010 and 2011 were linked to the arrival of new refugees from Somalia that was experiencing a measles outbreak during that time [8, 9]. Furthermore, a joint statement by WHO, UNHCR, and UNICEF on refugees, asylum- seekers and migrants entering the WHO European Region has called for the integration of refugee and mi- grant health into broader national health interventions to ensure equitable and non-discriminatory access to health services, including vaccinations [11]. Due to the dynamic movement and unique characteristics of dis- placed populations, specific vaccination strategies for such populations at risk of VPD circulation should be examined and documented in light of global efforts for eradication and elimination [12]. Acknowledgements Not applicable. Acknowledgements Not applicable. Acknowledgements Not applicable. Funding US CDC. Authors’ contributions EL d MD d f d h EL and MD drafted the manuscript and were major contributors in writing the manuscript. AGKM and MB reviewed and revised the manuscript. All authors read and approved the final manuscript. Competing interests Competing interests The authors declare that they have no competing interests. Consent for publication Not applicable. References 1. UN High Commissioner for Refugees (UNHCR), UNHCR Mid-Year Trends 2015. 2015. Available at: http://www.refworld.org/docid/568fbb8f4.html . Accessed 12 Mar 2016. Accessed 12 Mar 2016. 2. Connolly MA, Gayer M, Ryan MJ, Salama P, Spiegel P, Heymann DL. Communicable diseases in complex emergencies: impact and challenges. Lancet. 2004;364(9449):1974–83. 3. Dowdle WR. The principles of disease elimination and eradication. Bull World Health Organ. 1998;76 Suppl 2:22–5. 4. Global Polio Eradication Initiative, Polio Eradication & Endgame Strategic Plan 2013–2018. 2013. Available at: http://www.polioeradication.org/Portals/0/ Document/Resources/StrategyWork/PEESP_EN_US.pdf. Accessed 12 Mar 2016. 5. World Health Organization, Global Measles and Rubella Strategic Plan 2012– 2020. 2012. Available at: http://apps.who.int/iris/bitstream/10665/44855/1/ 9789241503396_eng.pdf. Accessed 13 Mar 2016. 6. Eichner M, Brockmann SO. Polio emergence in Syria and Israel endangers Europe. Lancet. 2013;382(9907):1777. 7. Combined use of inactivated and oral poliovirus vaccines in a large-scale campaign in refugee camps and host communities - Kenya, December 2013. Wkly Epidemiol Rec, 2014;89(12): 127–32 8. Navarro-Colorado C, Mahamud A, Burton A, Haskew C, Maina GK, Wagacha JB, et al. Measles outbreak response among adolescent and adult Somali refugees displaced by famine in Kenya and Ethiopia, 2011. J Infect Dis. 2014; 210(12):1863–70. 8. Navarro-Colorado C, Mahamud A, Burton A, Haskew C, Maina GK, Wagacha JB, et al. Measles outbreak response among adolescent and adult Somali refugees displaced by famine in Kenya and Ethiopia, 2011. J Infect Dis. 2014; 210(12):1863–70. 9. Polonsky JA, Ronsse A, Ciglenecki I, Rull M, Porten K. High levels of mortality, malnutrition, and measles, among recently-displaced Somali refugees in Dagahaley camp, Dadaab refugee camp complex, Kenya, 2011. Confl Health. 2013;7(1):1. doi:10.1186/1752-1505-7-1. 10. Sharara SL, Kanj SS. War and infectious diseases: challenges of the Syrian civil War. PLoS Pathog. 2014;10(11), e1004438. 10. Sharara SL, Kanj SS. War and infectious diseases: challenges of the Syrian civil War. PLoS Pathog. 2014;10(11), e1004438. 11. World Health Organization, UN Children's Fund, and UN High Commissioner for Refugees, WHO-UNHCR-UNICEF Joint statement on general principles of vaccination of refugees, asylum-seekers and migrants in the WHO European Region. 2015. Available at:http://reliefweb.int/sites/reliefweb.int/files/ resources/EuropeVaccinationPosition_WHO-UNHCR-UNICEFNov.pdf. Accessed 4/9/2016. 11. World Health Organization, UN Children's Fund, and UN High Commissioner for Refugees, WHO-UNHCR-UNICEF Joint statement on general principles of vaccination of refugees, asylum-seekers and migrants in the WHO European Region. 2015. Available at:http://reliefweb.int/sites/reliefweb.int/files/ resources/EuropeVaccinationPosition_WHO-UNHCR-UNICEFNov.pdf. Accessed 4/9/2016. 12. Minetti A, Bopp C, Fermon F, François G, Grais RF, Grout L, et al. Measles outbreak response immunization is context-specific: insight from the recent experience of Médecins Sans Frontières. measlesrubellainitiative.org/wp-content/uploads/2013/06/WHO-AFRO- Measles-Fieldguide-April-2011.pdf. Accessed on 8/3/2016. 20. World Health Organization, Measles Elimination Field Guide. Regional Office for the Western Pacific (WPRO). 2013. Available at: http://www.wpro.who. int/immunization/documents/measles_elimination_field_guide_2013.pdf. Accessed on 8/3/2016. 21. World Health Organization, Global vaccine action plan 2011–2020. 2013. Available at: http://www.who.int/immunization/global_vaccine_action_plan/ en/. Accessed 16 Mar 2016. Author details 1 1Emergency Response and Recovery Branch; Division of Global Health Protection, Center for Global Health; Centers for Disease Control and Prevention, 1600 Clifton Road NE, Atlanta, GA 30309, USA. 2Department of Global Epidemiology; Rollins School of Public Health, Emory University, Atlanta, GA, USA. 3United Nations High Commissioner for Refugees (UNHCR), Geneva, Switzerland. Received: 26 July 2016 Accepted: 10 August 2016 Published: 30 November 2016 Lam et al. Conflict and Health (2016) 10:27 Lam et al. Conflict and Health (2016) 10:27 Disclaimer The findings and conclusions in this report are those of the authors and do not necessarily represent the official position of the Centers for Disease Control and Prevention. Page 3 of 3 Page 3 of 3 Page 3 of 3 Page 3 of 3 References PLoS Med. 2013;10(11), e1001544. 13. Sphere Project, Sphere Handbook: Humanitarian Charter and Minimum Standards in Disaster Response, 2011. 2011. Available at: http://www. refworld.org/docid/4ed8ae592.html. Accessed on 13 Mar 2016. 13. Sphere Project, Sphere Handbook: Humanitarian Charter and Minimum Standards in Disaster Response, 2011. 2011. Available at: http://www. refworld.org/docid/4ed8ae592.html. Accessed on 13 Mar 2016. 14. World Health Organization, SAGE working group on vaccination in humanitarian emergencies : vaccination in acute humanitarian emergencies: a framework for decision making. 2013. Available at: http://www.who.int/ iris/handle/10665/92462. Accessed 13 Mar 2016. 14. World Health Organization, SAGE working group on vaccination in humanitarian emergencies : vaccination in acute humanitarian emergencies: a framework for decision making. 2013. Available at: http://www.who.int/ iris/handle/10665/92462. Accessed 13 Mar 2016. Submit your next manuscript to BioMed Central and we will help you at every step: Submit your next manuscript to BioMed Central and we will help you at every step: 15. Kamugisha C, Cairns KL, Akim C. An outbreak of measles in Tanzanian refugee camps. J Infect Dis. 2003;187 Suppl 1:S58–62. 15. Kamugisha C, Cairns KL, Akim C. An outbreak of measles in Tanzanian refugee camps. J Infect Dis. 2003;187 Suppl 1:S58–62. • We accept pre-submission inquiries • Our selector tool helps you to ﬁnd the most relevant journal • We provide round the clock customer support • Convenient online submission • Thorough peer review • Inclusion in PubMed and all major indexing services • Maximum visibility for your research Submit your manuscript at www.biomedcentral.com/submit and we will help you at every step: 16. Spiegel PB, et al. Health-care needs of people affected by conflict: future trends and changing frameworks. Lancet. 2010;375(9711):341–5. 16. Spiegel PB, et al. Health-care needs of people affected by conflict: future trends and changing frameworks. Lancet. 2010;375(9711):341–5. 17. Mancini S, Coldiron ME, Ronsse A, Ilunga BK, Porten K, Grais RF. Description of a large measles epidemic in Democratic Republic of Congo, 2010–2013. Confl Heal. 2014;8:9. 18. World Health Organization, Planning and Implementing High Quality Supplementary Immunization Activities for Injectable Vaccines using an example of Measles and Rubella Vaccines. A Field Guide. 2016. Available at: http://www.who.int/immunization/diseases/measles/SIA-Field-Guide-revised. pdf. Accessed on 8/3/2016. 19. World Health Organization, Measles SIAs Planning and Implementation Field Guide. Regional Office for Africa (AFRO). 2010. Available at: http://www.
https://openalex.org/W3108150818	https://zenodo.org/records/4300995/files/60%2021105%201570677108%20edit1.pdf	English	null	A simple levelized cost of electricity for EV charging with PV and battery energy storage system: Thailand case study	International Journal of Power Electronics and Drive Systems/International Journal of Electrical and Computer Engineering	2,020	cc-by	5,230	International Journal of Power Electronics and Drive System (IJPEDS) Vol. 11, No. 4, December 2020, pp. 2223~2230 ISSN: 2088-8694, DOI: 10.11591/ijpeds.v11.i4.pp2223-2230 International Journal of Power Electronics and Drive System (IJPEDS) Vol. 11, No. 4, December 2020, pp. 2223~2230 ISSN: 2088-8694, DOI: 10.11591/ijpeds.v11.i4.pp2223-2230 International Journal of Power Electronics and Drive System (IJPEDS) Vol. 11, No. 4, December 2020, pp. 2223~2230 ISSN: 2088-8694, DOI: 10.11591/ijpeds.v11.i4.pp2223-2230  2223 NOMENCLATURE NOMENCLATURE sLCOEx Simple levelized cost of electricity for x technology CRF Capital recovery factor CF Capacity factor i Interest rate, % n Year of the project O&M Operation & Maintenance Pc Charging price Econ Converter efficiency A simple levelized cost of electricity for EV charging with PV and battery energy storage system: Thailand case study Aree Wangsupphaphol, Surachai Chaitusaney Department of Electrical Engineering, Chulalongkorn University, Pathumwan, Bangkok, Thailand ABSTRACT This paper proposes the calculation of the simple levelized cost of electricity of PV and battery energy storage system for supporting the investment decision of the EV hybrid charging station. The paper introduces the problems of EV charging against the grid power system. Thus, the hybrid charging for EV is suggested. The study provides an architecture of the hybrid EV charging station along with the factor impacting the EV infrastructure for acknowledgment. The cost elements of the station are presented to address the benefits of the investment. Besides, the profit is mainly from the margin of the electricity price, therefore, this study compares the electricity cost of PV and PV equipped with a battery with the commercial on-peak electricity tariff. The results show that the charging cost contributed by PV alone has the lowest amount throughout the study period year 2020 – 2030. In contrast, the hybrid charging cost contributed by PV and battery has a higher value than the on-peak tariff during 2020 – 2025 but it is lower afterward. The result supports the feasibility of charging an EV by solar power and the hybrid power system in the future. Journal homepage: http://ijpeds.iaescore.com Keywords: Electric vehicle Power systems Renewable energy resources Energy storage This is an open access article under the CC BY-SA license. This is an open access article under the CC BY-SA license. Corresponding Author: Surachai Chaitusaney Department of Electrical Engineering Chulalongkorn University Pathumwan, Bangkok 10330, Thailand Email: Surachai.C@chula.ac.th 1. INTRODUCTION Electric vehicle (EV) is becoming an effective transportation mode, thus the power supply adequacy and quality are the main concern to prevent the power supply outage [1]. The energy management strategy, therefore, plays a crucial role in maximizing energy utilization [2–7]. In [8], the impact of EV load on the power system in Thailand has been studied and it is found out that the traditional charging period of EV is from 6 PM – 7 AM. The EV load generates high power ramps in the early morning and evening. Besides, in Journal homepage: http://ijpeds.iaescore.com  2224 2224  ISSN: 2088-8694 case of incorporating photovoltaic (PV), the low power can be occurred in the daytime by the reverse power flow generated by PV; this power profile is the so-called “duck curve” [9-11]. case of incorporating photovoltaic (PV), the low power can be occurred in the daytime by the reverse power flow generated by PV; this power profile is the so-called “duck curve” [9-11]. A study of practical solutions to mitigate EV load impacts against the distribution power shows that the splitting transformer to two units and relocation closing to the EV loads can reduce the impact of the voltage drop and peak load demand [1, 10]. However, this requires a high investment from utilities. A simple but effective method is the incentive-based charging period using time-of-use (TOU) tariff; the method provides the capability of flattening power and voltage. Though the TOU tariff has not been deployed in many utilities and the charging regime causes the inconvenience against the EV owner [12]. This scenario has also happened the same in Thailand. pp Incorporating with PV distributed energy resources for charging EV, the effective socio-economic charging solution can be obtained [13]. The scheme contributes to the low charging price in the daytime, however, without EV load, it may trouble the power system. Thus, the battery energy storage system (BESS) is applied to store the surplus energy and reuse when required; this is an optimized solution in the real-world of EV applications [14]. The composition of PV and BESS for EV hybrid charging becomes an effective solution to maintain power quality, deploy the renewable energy source, and contribute zero-emission electric mobility [15]. In Thailand, there is a need to study the feasibility of the EV hybrid charging station. 𝑃𝑟𝑜𝑓𝑖𝑡 = 𝐼𝑛𝑐𝑜𝑚𝑒 – 𝐼𝑛𝑣𝑒𝑠𝑡𝑚𝑒𝑛𝑡 – 𝑂𝑝𝑒𝑟𝑎𝑡𝑖𝑜𝑛 & 𝑀𝑎𝑖𝑛𝑡𝑒𝑛𝑎𝑛𝑐𝑒 (1) 𝑃𝑟𝑜𝑓𝑖𝑡 = 𝐼𝑛𝑐𝑜𝑚𝑒 – 𝐼𝑛𝑣𝑒𝑠𝑡𝑚𝑒𝑛𝑡 – 𝑂𝑝𝑒𝑟𝑎𝑡𝑖𝑜𝑛 & 𝑀𝑎𝑖𝑛𝑡𝑒𝑛𝑎𝑛𝑐𝑒 (1) (1) The (2) shows the income calculation which is derived from charging demand (kWh) and charging price in which the margin is added to the electricity price (baht/kWh). 𝐼𝑛𝑐𝑜𝑚𝑒 = 𝐷𝑒𝑚𝑎𝑛𝑑 𝐸𝑛𝑒𝑟𝑔𝑦 𝑥 𝐸𝑉 𝐶ℎ𝑎𝑟𝑔𝑖𝑛𝑔 𝑝𝑟𝑖𝑐𝑒 (2) (2) The EV charger subsidized by Thai government is required to promote the EV charging station, and it is deducted from the construction and charger system cost to build up the investment cost as in (3). 𝐼𝑛𝑣𝑒𝑠𝑡𝑚𝑒𝑛𝑡= 𝐶𝑜𝑛𝑠𝑡𝑟𝑢𝑐𝑡𝑖𝑜𝑛 𝑎𝑛𝑑 𝑙𝑎𝑛𝑑 𝐶𝑜𝑠𝑡𝑠+ 𝐶ℎ𝑎𝑟𝑔𝑒𝑟 𝑆𝑦𝑠𝑡𝑒𝑚 𝐶𝑜𝑠𝑡𝑠− (𝑈𝑛𝑖𝑡 𝐶ℎ𝑎𝑟𝑔𝑒𝑟 𝑆𝑦𝑠𝑡𝑒𝑚 𝑆𝑢𝑏𝑠𝑖𝑑𝑦 𝑥 𝑁𝑜. 𝑜𝑓 𝐶ℎ𝑎𝑟𝑔𝑒𝑟) (3 𝐼𝑛𝑣𝑒𝑠𝑡𝑚𝑒𝑛𝑡= 𝐶𝑜𝑛𝑠𝑡𝑟𝑢𝑐𝑡𝑖𝑜𝑛 𝑎𝑛𝑑 𝑙𝑎𝑛𝑑 𝐶𝑜𝑠𝑡𝑠+ 𝐶ℎ𝑎𝑟𝑔𝑒𝑟 𝑆𝑦𝑠𝑡𝑒𝑚 𝐶𝑜𝑠𝑡𝑠− (𝑈𝑛𝑖𝑡 𝐶ℎ𝑎𝑟𝑔𝑒𝑟 𝑆𝑦𝑠𝑡𝑒𝑚 𝑆𝑢𝑏𝑠𝑖𝑑𝑦 𝑥 𝑁𝑜. 𝑜𝑓 𝐶ℎ𝑎𝑟𝑔𝑒𝑟) (3) (3) The operation and maintenance (O&M) cost is developed from the amount of electricity sold, operational cost (supposed to be subsidized by Thai government), land rental cost, and maintenance cost as shown in (4). 𝑂&𝑀 𝐶𝑜𝑠𝑡 = (𝐷𝑒𝑚𝑎𝑛𝑑 𝐸𝑛𝑒𝑟𝑔𝑦 𝑥 𝐸𝑙𝑒𝑐𝑡𝑟𝑖𝑐𝑖𝑡𝑦 𝑃𝑟𝑖𝑐𝑒) + (𝑂𝑝𝑒𝑟𝑎𝑡𝑖𝑜𝑛 𝐶𝑜𝑠𝑡 − 𝑂𝑝𝑒𝑟𝑎𝑡𝑖𝑜𝑛 𝑆𝑢𝑏𝑠𝑖𝑑𝑦) + 𝐿𝑎𝑛𝑑 𝑅𝑒𝑛𝑡 + 𝑀𝑎𝑖𝑛𝑡𝑒𝑛𝑎𝑛𝑐𝑒 𝐶𝑜𝑠𝑡 (4) (4) Int J Pow Elec & Dri Syst ISSN: 2088-8694 The direct factors are mainly based on the financial approach. The charging demand, the subsidy from the government, construction, maintenance, electric tariff, and number of EV are the main concerns of the EV charging infrastructure development. The indirect factor is based on the EV owner perspective; power train and energy source performance, charging technology, support policies, and human factors. The relations of the factors are shown along with their colors highlighted either in/out of the group of factors, for example, the charging demand can be influenced by the customer behavior because the driver may charge the EV accordingly to his convenient time. Therefore, the incentive-based charging period may be ineffective. The psychological factor of the driver is related to the charging location, the number of EVs in the market, technology of the EV system, and charging technic. The improvements of the technology factors mitigate the driver anxiety effectively, i.e. the EV can be conveniently charged until full within a short period or has a long driving range by the improvement of charging and battery technology [16, 18]. Regarding the direct factor that can be evaluated monetarily, the cost elements of the EV charging station in Thailand are suggested as shown in Figure 3. Figure 3. Monetary structure of EV charging station Figure 3. Monetary structure of EV charging station The profit composes of any financial elements; income, investment, and operation and maintenance, and they are obtained by (1). The profit composes of any financial elements; income, investment, and operation and maintenance, and they are obtained by (1). 1. INTRODUCTION 4, December 2020 : 2223 – 2230  2225 Int J Pow Elec & Dri Syst 1. INTRODUCTION Therefore, this paper proposes the contemporary key elements for the feasibility study of EV charging station as shown in Figure 1. In section 2, it gives the factor impacting the EV infrastructure with cost elements of the charging station. The architecture of the EV hybrid charging station is presented in section 3. As the electricity price is the important factor designated the possibility of implementation, the simple levelized cost of electricity (sLCOE) calculation is proposed in section 4. The cost projection of the PV module and Li-ion battery in China during the year 2020-2030 is used [12]. The engineering, procurement, construction (EPC) cost of PV and battery from suppliers in Thailand year 2020 is compared with China’s cost to find the gain and establish Thailand’s cost projection during 2020-2030 (see Table 1). After that, the sLCOE of the PV and hybrid charging is made and compared with the commercial on-peak electricity tariff. The results are used for supporting the investment decision as suggested in section 5. Figure 1. Key elements for the EV hybrid charging station feasibility study Factor impact the EV infrastructure Hybrid charging station architecture Cost elements of charging station Simple calculation of electricity cost Comparison of charging costs Figure 1. Key elements for the EV hybrid charging station feasibility study 2. FACTOR IMPACTING THE EV CHARGING INFRASTRUCTURE 2. FACTOR IMPACTING THE EV CHARGING INFRASTRUCTURE The penetration of EV is reinforced by the charging network [16]. Besides, the charging infrastructure is influenced by the direct and indirect factors as shown in Figure 2 [17]. 2. FACTOR IMPACTING THE EV CHARGING INFRASTRUCTURE The penetration of EV is reinforced by the charging network [16]. Besides, the charging infrastructure is influenced by the direct and indirect factors as shown in Figure 2 [17]. 2. FACTOR IMPACTING THE EV CHARGING INFRASTRUCTURE The penetration of EV is reinforced by the charging network [16]. Besides, the charging infrastructure is influenced by the direct and indirect factors as shown in Figure 2 [17]. The penetration of EV is reinforced by the charging network [16]. Besides, the charging tructure is influenced by the direct and indirect factors as shown in Figure 2 [17]. Figure 2. The factors influencing of EV charging infrastructure Figure 2. The factors influencing of EV charging infrastructure Int J Pow Elec & Dri Syst, Vol. 11, No. 4. SIMPLE LEVELIZED COST OF ELECTRICITY 4. SIMPLE LEVELIZED COST OF ELECTRICITY The sLCOE calculation identifies the commercializing of a power generation technology through its lifetime. The fundamental definition of sLCOE is expressed in (5) [23]. 4. SIMPLE LEVELIZED COST OF ELECTRICITY The sLCOE calculation identifies the commercializing of a power generation technology through its lifetime. The fundamental definition of sLCOE is expressed in (5) [23]. The sLCOE calculation identifies the commercializing of a power generation technology throu lifetime. The fundamental definition of sLCOE is expressed in (5) [23]. 3. HYBRID CHARGING STATION ARCHITECTURE The hybrid charging station structure in this study encompasses the power from the grid, PV, and BESS which manages via a hybrid inverter as is shown in Figure 4. Because of the Li-ion battery is the suitable technical solution together with commercial availability as suggested in IEEE standard of energy storage, it is the assured choice for this study [19]. Based on the hybrid charging system, the maximum grid power can be decreased by 45% during EV charging and the system can offset the EV daily charging load completely [15, 20]. With the character of the hybrid inverter, the power from the grid and BESS can flow bi- directionally between the EV load while the solar power is controlled to supply only [20]. The PV can be simple levelized cost of electricity for EV charging with PV and battery energy … (Aree Wangsupphapho 2226  2226  ISSN: 2088-8694 installed at any available space i.e. building/charging station rooftop or empty land. Without EV load, the surplus solar power is supplied to the BESS for reuse. At the EV charging station, DC fast charger and AC charger level 2 is normally used as the standard chargers for EV [21, 22]. To find the feasibility to implement the hybrid charging system, the electricity price generated by any technologies is evaluated using sLCOE. installed at any available space i.e. building/charging station rooftop or empty land. Without EV load, the surplus solar power is supplied to the BESS for reuse. At the EV charging station, DC fast charger and AC charger level 2 is normally used as the standard chargers for EV [21, 22]. To find the feasibility to implement the hybrid charging system, the electricity price generated by any technologies is evaluated using sLCOE. Figure 4. The architecture of an EV charging station incorporating with PV and BESS Figure 4. The architecture of an EV charging station incorporating with PV and BESS 𝑠𝐿𝐶𝑂𝐸= 𝑇𝑜𝑡𝑎𝑙 𝐿𝑖𝑓𝑒𝑡𝑖𝑚𝑒 𝐶𝑜𝑠𝑡 𝑇𝑜𝑡𝑎𝑙 𝐿𝑖𝑓𝑒𝑡𝑖𝑚𝑒 𝐸𝑛𝑒𝑟𝑔𝑦 𝑃𝑟𝑜𝑑𝑢𝑐𝑡𝑖𝑜𝑛 (5) The simple levelized cost of electricity of the PV system, sLCOEPV, can be obtained by ( The simple levelized cost of electricity of the PV system, sLCOEPV, can be obtained by (6). (6) It accounts for the combination of capital costs (Ccap,PV), fixed operation & maintenance (O&Mfix) of rooftop PV [24], variable O&M (O&Mvary) [25], the capacity factor of PV (CFPV) which is the ratio of the annual generating power to its capacity. For CFPV, it is selected based on the small size of the residential solar rooftop, 0.16, as is studied in [26]. Also, the calculation includes the capital recovery factor (CRFPV) which is the constant annuity to the present value of the annual receiving for a given year (n) along with the interest rate (i) as shown in (7). It should be noted that the number of the year using for CRF calculation for PV and BESS is 20 and 10 respectively as per its normal usage and lifetime. To replace a daily consumption 7.5 kWh of an EV supplied by the power grid, the solar peak power could be well managed by a smart charger to charge an EV efficiently. Thus, the high-cost grid power purchased may not be required [12]. However, in the real world, the charging demand and intensity are random whereby solar energy may not be ready or enough. 𝐶𝑅𝐹= 𝑖(1+𝑖)𝑛 (1+𝑖)𝑛−1 (7) To capture the surplus solar energy when EV is away, the battery is utilized i.e. saving the energy on a weekend. Based on this assumption, the battery will be charged by PV for 6.4 hours/week for 15 kWh that is used by an EV on a weekend [12]. Thus, the simple levelized cost of electricity of BESS, sLCOEBES, can To capture the surplus solar energy when EV is away, the battery is utilized i.e. saving the energy on a weekend. Based on this assumption, the battery will be charged by PV for 6.4 hours/week for 15 kWh that is used by an EV on a weekend [12]. Thus, the simple levelized cost of electricity of BESS, sLCOEBES, can Int J Pow Elec & Dri Syst, Vol. 11, No. 4, December 2020 : 2223 – 2230  2227 2227 ISSN: 2088-8694  Int J Pow Elec & Dri Syst be obtained by (8). The first term presents the discounted of capital investment and running O&M yearly cost. 𝑠𝐿𝐶𝑂𝐸= 𝑇𝑜𝑡𝑎𝑙 𝐿𝑖𝑓𝑒𝑡𝑖𝑚𝑒 𝐶𝑜𝑠𝑡 𝑇𝑜𝑡𝑎𝑙 𝐿𝑖𝑓𝑒𝑡𝑖𝑚𝑒 𝐸𝑛𝑒𝑟𝑔𝑦 𝑃𝑟𝑜𝑑𝑢𝑐𝑡𝑖𝑜𝑛 They are derived from the battery capital cost, Ccap,bat, considering the capital recovery factor of BESS, CRFBES, with yearly O&M (O&Myear) over the yearly discharge hour. The discharge hour is assumed the same as charging time 6.4 hours/week and this is calculated by using a capacity factor of BESS, CFBES, as of 0.038. Because the battery is charged by solar power, sLCOEPV is used for the charging price (Pc), and it is deducted with the cost of energy conversion loss where the energy conversion efficiency (Econ) is considered. This second term is to segregate the conversion loss cost out to find the only electricity cost contributed by the battery. 𝑠𝐿𝐶𝑂𝐸𝐵𝐸𝑆= (𝐶𝑐𝑎𝑝,𝐵𝑎𝑡𝑥 𝐶𝑅𝐹𝐵𝐸𝑆) + 𝑂&𝑀𝑦𝑒𝑎𝑟 8760 𝑥 𝐶𝐹𝐵𝐸𝑆 −(𝑃𝑐 𝑥 (1 −𝐸𝑐𝑜𝑛)) (8) (8) The evaluation of sLCOEPV and sLCOEBES during 2020-2030 is calculated based on its CRF aforesaid. The interest rate of 4.7% for Thailand’s government project is applied. The capital cost, operation, and maintenance parameter, and other parameters are declared (with references) in Table 2. The purpose of the calculation is to find the sLCOE of the PV and hybrid charging to compare with the commercial on-peak electricity tariff. The complex calculation such as future replacement or degradation costs is not included in this preliminary study. From the cost projection of EV charging as depicted in Figure 5, one can observe that the sLCOEPV is the cheapest cost among others and it has a little decrease during year 2020-2030. In contrast, the sLCOE of hybrid system, PV equipped with BESS, is the highest cost between 2020 and 2025, but it is lower than the on-peak tariff afterward. In [27], we found that the sLCOE of residential PV-battery system since 2020 has the same agreement as in our study, however, it has a small different of the competed year that the price of hybrid charging is lower than the grid tariff. The discrepany is mainly from the different interest rate where we use the interest rate for a government project, 4.7%, but the afromentioned study used the discount rate for a residential scale, 2.89%. Figure 5. Cost projection of EV charging Figure 5. Cost projection of EV charging APPENDIX A Evaluation results of sLCOE for hybrid charging Parameter of sLCOEpv 2020 2021 2022 2023 2024 2025 2026 2027 2028 2029 2030 % interest / 100 0.047 0.047 0.047 0.047 0.047 0.047 0.047 0.047 0.047 0.047 0.047 npv (yr) 20 20 20 20 20 20 20 20 20 20 20 CRFpv 0.078 0.078 0.078 0.078 0.078 0.078 0.078 0.078 0.078 0.078 0.078 Cpv ($/kW) 617.4 562.8 516.6 483.0 453.6 424.2 403.2 386.4 369.6 352.8 340.2 O&Mfix ($/kW-yr) 10 10 10 10 10 10 10 10 10 10 10 O&Mvary ($/kWh) 0.002 0.002 0.002 0.002 0.002 0.002 0.002 0.002 0.002 0.002 0.002 CFpv % 0.16 0.16 0.16 0.16 0.16 0.16 0.16 0.16 0.16 0.16 0.16 sLCOEpv ($/kWh) 0.045 0.042 0.039 0.037 0.035 0.034 0.032 0.031 0.030 0.030 0.029 Parameter of sLCOEBES % interest / 100 0.047 0.047 0.047 0.047 0.047 0.047 0.047 0.047 0.047 0.047 0.047 nBES (yr) 10 10 10 10 10 10 10 10 10 10 10 CRFBES 0.128 0.128 0.128 0.128 0.128 0.128 0.128 0.128 0.128 0.128 0.128 CBat ($/kW) 428 365 311 269 235 210 189 172 155 139 122 O&Myear ($/kW-yr) 6.9 6.9 6.9 6.9 6.9 6.9 6.9 6.9 6.9 6.9 6.9 CFBat % 0.038 0.038 0.038 0.038 0.038 0.038 0.038 0.038 0.038 0.038 0.038 Energy conversion efficiency 0.94 0.94 0.94 0.94 0.94 0.94 0.94 0.94 0.94 0.94 0.94 sLCOEBES ($/kWh) 0.180 0.157 0.136 0.120 0.108 0.098 0.090 0.084 0.078 0.071 0.065 Total sLCOE for PV+BES ($/kWh) 0.225 0.198 0.175 0.157 0.143 0.132 0.123 0.115 0.108 0.101 0.094 REFERENCES [1] S. Alshahrani, M. Khalid, and M. Almuhaini, “Electric Vehicles Beyond Energy Storage and Modern Power Networks: Challenges and Applications ” IEEE Access vol 7 pp 99031 99064 2019 doi: Table 1. APPENDIX B N Table 2. Evaluation results of sLCOE for hybrid charging Parameter of sLCOEpv 2020 2021 2022 2023 2024 2025 2026 2027 2028 2029 2030 % interest / 100 0.047 0.047 0.047 0.047 0.047 0.047 0.047 0.047 0.047 0.047 0.047 npv (yr) 20 20 20 20 20 20 20 20 20 20 20 CRFpv 0.078 0.078 0.078 0.078 0.078 0.078 0.078 0.078 0.078 0.078 0.078 Cpv ($/kW) 617.4 562.8 516.6 483.0 453.6 424.2 403.2 386.4 369.6 352.8 340.2 O&Mfix ($/kW-yr) 10 10 10 10 10 10 10 10 10 10 10 O&Mvary ($/kWh) 0.002 0.002 0.002 0.002 0.002 0.002 0.002 0.002 0.002 0.002 0.002 CFpv % 0.16 0.16 0.16 0.16 0.16 0.16 0.16 0.16 0.16 0.16 0.16 sLCOEpv ($/kWh) 0.045 0.042 0.039 0.037 0.035 0.034 0.032 0.031 0.030 0.030 0.029 Parameter of sLCOEBES % interest / 100 0.047 0.047 0.047 0.047 0.047 0.047 0.047 0.047 0.047 0.047 0.047 nBES (yr) 10 10 10 10 10 10 10 10 10 10 10 CRFBES 0.128 0.128 0.128 0.128 0.128 0.128 0.128 0.128 0.128 0.128 0.128 CBat ($/kW) 428 365 311 269 235 210 189 172 155 139 122 O&Myear ($/kW-yr) 6.9 6.9 6.9 6.9 6.9 6.9 6.9 6.9 6.9 6.9 6.9 CFBat % 0.038 0.038 0.038 0.038 0.038 0.038 0.038 0.038 0.038 0.038 0.038 Energy conversion efficiency 0.94 0.94 0.94 0.94 0.94 0.94 0.94 0.94 0.94 0.94 0.94 sLCOEBES ($/kWh) 0.180 0.157 0.136 0.120 0.108 0.098 0.090 0.084 0.078 0.071 0.065 Total sLCOE for PV+BES ($/kWh) 0.225 0.198 0.175 0.157 0.143 0.132 0.123 0.115 0.108 0.101 0.094 5. CONCLUSION This study concludes the basic knowledge for simple levelized cost of electricity calculation of PV and battery energy storage systems. The benefits of renewable energy resources and storage that mitigate the problem of EV charging against the grid power are presented. Moreover, the architecture of the hybrid charging station is illustrated. The factor that impacts the EV infrastructure is declared to seek out the cost element. Thus, the sLCOE of the power generations has been addressed for observation its suitability. Finally, the electricity cost of PV and hybrid systems which composes of PV and storage are evaluated with the on-peak tariff. Considering only the sLCOE, PV is the cheapest solution of power generation but, in the real-world condition, other factors may cause the application difficulty, i.e. installation area, available charging period which depends on the solar intensity, etc. To enhance the PV application, the battery storage is studied for its feasibility to implement. Based on the parameters used for the battery and PV in this study, the sLCOE of hybrid charging will be acknowledged after 2025 compared with the on-peak tariff. Therefore, the grid and solar power could be a worth solution for EV charging presently. simple levelized cost of electricity for EV charging with PV and battery energy … (Aree Wangsupphapho ISSN: 2088-8694  2228 APPENDIX A Cost projection of PV and Li-ion battery of Thailand based on China’s evaluations Year China's annual PV module production (GW) China's PV cost projection (USD/kW) China's estimated PV cost including EPC (USD/kW) Thailand cost, 50% higher than China’s EPC (USD/kW) China's annual Li- ion battery productio n (GWh) China's Li- ion battery cost projection (USD/kWh) China's Estimated Li-ion battery including EPC (USD/kW) Thaiand cost 50% higher than China’s EPC (USD/kW) 2020 100 205.8 411.6 617.4 129.1 142.8 285.6 428.4 2021 100 187.6 375.2 562.8 199.6 121.8 243.6 365.4 2022 100 172.2 344.4 516.6 275.1 103.6 207.2 310.8 2023 100 161 322 483 355.6 89.6 179.2 268.8 2024 100 151.2 302.4 453.6 441.2 78.4 156.8 235.2 2025 100 141.4 282.8 424.2 525.4 70 140 210 2026 100 134.4 268.8 403.2 614.2 63 126 189 2027 100 128.8 257.6 386.4 707.7 57.4 114.8 172.2 2028 100 123.2 246.4 369.6 142.1 51.8 103.6 155.4 2029 100 117.6 235.2 352.8 217.7 46.2 92.4 138.6 2030 100 113.4 226.8 340.2 297.4 40.6 81.2 121.8 APPENDIX B APPENDIX B Int J Pow Elec & Dri Syst, Vol. 11, No. 4, December 2020 : 2223 – 2230 ACKNOWLEDGEMENTS ACKNOWLEDGEMENTS The research is supported by Ratchadaphisek Somphot Fund for Postdoctoral Fellowship, Chulalongkorn University. The research is supported by Ratchadaphisek Somphot Fund for Postdoctoral Fellowship, longkorn University. The research is supported by Ratchadaphisek Somphot Fund for Postdoctoral Fellowship, Chulalongkorn University. APPENDIX A Table 1. Cost projection of PV and Li-ion battery of Thailand based on China’s evaluations Year China's annual PV module production (GW) China's PV cost projection (USD/kW) China's estimated PV cost including EPC (USD/kW) Thailand cost, 50% higher than China’s EPC (USD/kW) China's annual Li- ion battery productio n (GWh) China's Li- ion battery cost projection (USD/kWh) China's Estimated Li-ion battery including EPC (USD/kW) Thaiand cost 50% higher than China’s EPC (USD/kW) 2020 100 205.8 411.6 617.4 129.1 142.8 285.6 428.4 2021 100 187.6 375.2 562.8 199.6 121.8 243.6 365.4 2022 100 172.2 344.4 516.6 275.1 103.6 207.2 310.8 2023 100 161 322 483 355.6 89.6 179.2 268.8 2024 100 151.2 302.4 453.6 441.2 78.4 156.8 235.2 2025 100 141.4 282.8 424.2 525.4 70 140 210 2026 100 134.4 268.8 403.2 614.2 63 126 189 2027 100 128.8 257.6 386.4 707.7 57.4 114.8 172.2 2028 100 123.2 246.4 369.6 142.1 51.8 103.6 155.4 2029 100 117.6 235.2 352.8 217.7 46.2 92.4 138.6 2030 100 113.4 226.8 340.2 297.4 40.6 81.2 121.8 APPENDIX B Table 2. REFERENCES Santoso, M. P. Cloud, and M. Waclawiak, “Determining Time-of-Use Schedules for Electric Vehicle Loads : A Practical Perspective,” IEEE Power Energy Technol. Syst. J. Receiv., vol. 2, no. 1, 2015. p , gy y [11] Q. Hoarau and Y. Perez, “Interactions between electric mobility and photovoltaic generation: A review,” Renew. Sustain. Energy Rev., vol. 94, no. February, pp. 510–522, 2018, doi: 10.1016/j.rser.2018.06.039. [12] J. Liu and C. Zhong, “An economic evaluation of the coordination between electric vehicle storage and distributed renewable energy,” Energy, vol. 186, Nov. 2019, doi: 10.1016/j.energy.2019.07.151. [13] D. Gielen, F. Boshell, D. Saygin, M. D. Bazilian, N. Wagner, and R. Gorini, “The role of renewable energy in the global energy transformation,” Energy Strateg. Rev., vol. 24, no. June 2018, pp. 38–50, 2019, doi: 10.1016/j.esr.2019.01.006. 4] X. Wu, X. Hu, X. Yin, C. Zhang, and S. Qian, “Optimal battery sizing of smart home via convex programming Energy, vol. 140, pp. 444–453, 2017, doi: 10.1016/j.energy.2017.08.097. [14] X. Wu, X. Hu, X. Yin, C. Zhang, and S. Qian, “Optimal battery sizing of smart home via convex programming,” Energy, vol. 140, pp. 444–453, 2017, doi: 10.1016/j.energy.2017.08.097. [15] L Novoa and J Brouwer “Dynamics of an integrated solar photovoltaic and battery storage nanogrid for electric [15] L. Novoa and J. Brouwer, “Dynamics of an integrated solar photovoltaic and battery storage nanogrid for electric vehicle charging,” J. Power Sources, vol. 399, no. March, pp. 166–178, 2018, doi: 10.1016/j.jpowsour.2018.07.092. [16] L. Shi, Y. Hao, S. Lv, L. Cipcigan, and J. Liang, “A comprehensive charging network planning scheme for promoting EV charging infrastructure considering the Chicken-Eggs dilemma,” Res. Transp. Econ., no. March, p. 100837, 2020, doi: 10.1016/j.retrec.2020.100837. 7] Q. Zhang et al., “Factors influencing the economics of public charging infrastructures for EV – A review,” Rene Sustain. Energy Rev., vol. 94, no. June, pp. 500–509, 2018, doi: 10.1016/j.rser.2018.06.022. [18] A. Tomaszewska et al., “Lithium-ion battery fast charging: A review,” eTransportation, vol. 1, p. 100011, 2019, doi: 10.1016/j.etran.2019.100011. j [19] E. Society and T. Society, IEEE Draft Guide for the Interoperability of Energy Storage Systems Integrated with the Electric Power Infrastructure. 2014. [20] O. Elma, “A dynamic charging strategy with hybrid fast charging station for electric vehicles,” Energy, vol. 202, p. 117680, 2020, doi: 10.1016/j.energy.2020.117680. [21] J. Martínez-Lao, F. G. Montoya, M. G. Montoya, and F. Manzano-Agugliaro, “Electric vehicles in Spain: An overview of charging systems,” Renewable and Sustainable Energy Reviews, vol. 77. pp. REFERENCES REFERENCES [1] S. Alshahrani, M. Khalid, and M. Almuhaini, “Electric Vehicles Beyond Energy Storage and Modern Power Networks: Challenges and Applications,” IEEE Access, vol. 7, pp. 99031–99064, 2019, doi: 10.1109/access.2019.2928639. [1] S. Alshahrani, M. Khalid, and M. Almuhaini, “Electric Vehicles Beyond Energy Storage and Modern Power Networks: Challenges and Applications,” IEEE Access, vol. 7, pp. 99031–99064, 2019, doi: 10.1109/access.2019.2928639. [2] A. Wangsupphaphol, N. Rumzi, and N. Idris, “Student Research Highlight : Acceleration-Based Design of Electric Vehicle Auxiliary Energy Source,” IEEE Aerosp. Electron. Syst. Mag., no. 10, pp. 32–35, 2016, doi: 10.1109/MAES.2016.140011. [2] A. Wangsupphaphol, N. Rumzi, and N. Idris, “Student Research Highlight : Acceleration-Based Design of Electric Vehicle Auxiliary Energy Source,” IEEE Aerosp. Electron. Syst. Mag., no. 10, pp. 32–35, 2016, doi: 10.1109/MAES.2016.140011. [3] A. Wangsupphaphol et al., “Deceleration-based Design Auxiliary Energy Source for Electric Vehicle Applications, version 2 Deceleration-based Design Auxiliary Energy Source for Electric Vehicle Applications.” [Online]. Available: https://engine.lib.uwaterloo.ca/ojs-. g j [4] A. Wangsupphaphol, N. Rumzi, N. Idris, A. Jusoh, N. D. Muhamad, and S. Chamchuen, “Acceleration- Int J Pow Elec & Dri Syst, Vol. 11, No. 4, December 2020 : 2223 – 2230  2229 Int J Pow Elec & Dri Syst ISSN: 2088-8694 Control Strategy and Design for Hybrid Electric Vehicle Auxiliary Energy Source,” ECTI Trans. Comput. Inf. Technol., vol. 9, no. 1, pp. 83–92, 2015. pp [5] A. Wangsupphaphol, N. R. N. Idris, A. Jusoh, N. D. Muhamad, and L. W. Yao, “The Energy Management Control Strategy for Electric Vehicle Applications,” Green Energy Sustain. Dev. (ICUE), 2014 Int. Conf. Util. Exhib., no. March, pp. 1–5, 2014, [Online]. Available: http://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=6828980. [6] A. Wangsupphaphol, N. R. N. Idris, A. Jusoh, N. D. Muhamad, and I. M. Alsofyani, “Acceleration-based design auxiliary power source for Electric Vehicle applications,” 2014 11th Int. Conf. Electr. Eng. Comput. Telecommun. Inf. Technol., no. March, pp. 1–6, 2014, doi: 10.1109/ECTICon.2014.6839873. [7] A. Wangsupphaphol, N. R. N. Idris, A. Jusoh, N. D. Muhamad, and I. M. Alsofyani, “Energy and power control strategy for battery electric vehicle with supercapacitors,” 2014, doi: 10.1109/CENCON.2014.6967469. [8] W. Wangjiraniran, J. Pongthanaisawan, S. Junlakarn, and D. Phadungsri, “Scenario Analysis of Disruptive Technology Penetration on the Energy System in Thailand,” Energy Procedia, vol. 142, pp. 2661–2668, 2017, doi: 10.1016/j.egypro.2017.12.208. j gyp J. Romero Agüero, E. Takayesu, D. Novosel, and R. Masiello, “Grid modernization: challenges and opportunities,” Electr. J., vol. 30, no. 4, pp. 1–6, 2017, doi: 10.1016/j.tej.2017.03.008. [10] A. Dubey, S. A simple levelized cost of electricity for EV charging with PV and battery energy … (Aree Wangsupphaphol) REFERENCES 970–983, 2017, doi: 10.1016/j.rser.2016.11.239. j [22] R. H. Ashique, Z. Salam, M. J. Bin Abdul Aziz, and A. R. Bhatti, “Integrated photovoltaic-grid dc fast charging system for electric vehicle: A review of the architecture and control,” Renewable and Sustainable Energy Reviews, vol. 69, no. October 2016. Elsevier, pp. 1243–1257, 2017, doi: 10.1016/j.rser.2016.11.245. [23] T. T. D. Tran and A. D. Smith, “Incorporating performance-based global sensitivity and uncertainty analysis into LCOE calculations for emerging renewable energy technologies,” Appl. Energy, vol. 216, no. February, pp. 157– 171, 2018, doi: 10.1016/j.apenergy.2018.02.024. j p gy [24] “Levelized Cost of Energy Calculator,” 2020. https://www.nrel.gov/analysis/tech-lcoe.html. [24] Levelized Cost of Energy Calculator, 2020. https://www.nrel.gov/analysis/tech lcoe.html. [25] M. H. Mostafa, S. H. E. Abdel Aleem, S. G. Ali, Z. M. Ali, and A. Y. Abdelaziz, “Techno-economic assessment of energy storage systems using annualized life cycle cost of storage (LCCOS) and levelized cost of energy (LCOE) metrics,” J. Energy Storage, vol. 29, no. December 2019, p. 101345, 2020, doi: 10.1016/j.est.2020.101345. [26] S. Tongsopit, S. Junlakarn, W. Wibulpolprasert, A. Chaianong, P. Kokchang, and N. V. Hoang, “The economics of solar PV self-consumption in Thailand,” Renew. Energy, vol. 138, pp. 395–408, 2019, doi: 10.1016/j.renene.2019.01.087. j 7] A. Chaianong, A. Bangviwat, C. Menke, B. Breitschopf, and W. Eichhammer, “Customer economics of residenti PV–battery systems in Thailand,” Renew. Energy, vol. 146, pp. 297–308, 2020, doi: 10.1016/j.renene.2019.06.15 simple levelized cost of electricity for EV charging with PV and battery energy … (Aree Wangsupphapho  2230 ISSN: 2088-8694    Int J Pow Elec & Dri Syst, Vol. 11, No. 4, December 2020 : 2223 – 2230 BIOGRAPHIES OF AUTHORS BIOGRAPHIES OF AUTHORS Aree Wangsupphaphol was born in Bangkok, Thailand, in 1975. He received a Degree in Engineering and a Masters in Engineering in electrical engineering from King Mongkut’s Institute of Technology, Lad- krabang, Thailand, in 1999 and 2007, and the Ph.D. degree from Universiti Teknologi Malaysia, Malaysia in 2019. He is currently a Post-Doctoral Research Fellow with Chulalongkorn University. His research interests are mainly in energy management system, electric vehicle applications, power system, and control of power electronics systems. Surachai Chaitusaney (M’07) received the B.Eng. and M.Eng. degrees in electrical engineering from Chulalongkorn University, Thailand, in 2000 and 2002, respectively, and the Ph.D. degree from the University of Tokyo, Japan, in 2007. From 2002 to 2004, he worked at the Energy Research Institute and Center of Excellence in Power Technology, Chulalongkorn University, where he is currently the Head of the Power Systems Research Laboratory. His special fields of interest include renewable energy and power systems planning, operation, and reliability. Int J Pow Elec & Dri Syst, Vol. 11, No. 4, December 2020 : 2223 – 2230
https://openalex.org/W2765775345	https://ruj.uj.edu.pl/xmlui/bitstream/handle/item/42199/popiolek_wieczorkowski_zarzadzanie_prywatnoscia_w_internecie_2017.pdf?sequence=1&isAllowed=y	Polish	null	Online privacy management- selected methodological issues	Ekonomiczne Problemy Usług	2,017	cc-by-sa	3,472	Ekonomiczne Problemy Usług nr 1/2017 (126), t. 2 ISSN: 1896-382X \| www.wnus.edu.pl/epu DOI: 10.18276/epu.2017.126/2-23 \| strony: 231–239 ISSN: 1896-382X \| www.wnus.edu.pl/epu DOI: 10.18276/epu.2017.126/2-23 \| strony: 231–239 Malwina Popiołek Uniwersytet Jagielloński Wydział Zarządzania i Komunikacji Społecznej Katedra Zarządzania, Ekonomiki Mediów i Reklamy e-mail: malwinapopiolek@gmail.com Malwina Popiołek Uniwersytet Jagielloński Wydział Zarządzania i Komunikacji Społecznej Katedra Zarządzania, Ekonomiki Mediów i Reklamy e-mail: malwinapopiolek@gmail.com Jędrzej Wieczorkowski Szkoła Główna Handlowa w Warszawie Instytut Informatyki i Gospodarki Cyfrowej Kolegium Analiz Ekonomicznych e-mail: jedrzej.wieczorkowski@sgh.waw.pl Jędrzej Wieczorkowski Szkoła Główna Handlowa w Warszawie Instytut Informatyki i Gospodarki Cyfrowej Kolegium Analiz Ekonomicznych e-mail: jedrzej.wieczorkowski@sgh.waw.pl Kody JEL: B40, B49, D83, O33 Kody JEL: B40, B49, D83, O33 Słowa kluczowe: prywatność, cyberbezpieczeństwo, technologie informacyjno-komunikacyjne Streszczenie. W artykule podjęto próbę uchwycenia i opisania najważniejszych problemów metodologicznych związanych z badaniem zarządzania prywatnością w internecie. Ukazano różnorodność teoretycznych kontekstów zagadnienia prywatności online, a także możliwe podej- ścia badawcze w różnych naukach: informatyce i naukach technicznych, prawnych, społecznych. Pokazano możliwe sposoby pozyskiwania danych (metody ilościowe i jakościowe), a także ro- dzaje badań z podziałem ze względu na cel – eksploracja, opis, wyjaśnienie. Przedstawiono także nowe metody badania prywatności, np. netnografię. Wprowadzenie Problem prywatności w ostatnich latach staje się coraz częstszym tematem dysku- sji zarówno w publikacjach naukowych, jak i w prasie popularnej. Wiąże się to przede wszystkim z rozwojem technologii informacyjno-komunikacyjnych (ICT) i zastosowa- niem jej w codziennym życiu. Wykorzystywanie internetu skutkuje ciągłym pozosta- wianiem śladów swojej aktywności, które jako dość dobrze ustrukturyzowane dane, mogą być łatwo wykorzystywane do różnych celów. Są to dane prywatne, często oso- Zarządzanie prywatnością w internecie – wybrane problemy metodologiczne 232 bowe, w tym wrażliwe, różnorodne informacje o zainteresowaniach, kontaktach z in- nymi osobami, codziennej aktywności itp. Nad takimi danymi aktywny internauta nie ma pełnej kontroli. Coraz częściej spotkać można się z opiniami, że przeciętny użytkownik technologii informacyjno-komunikacyjnych ma nierówne szanse i przegraną pozycję względem pod- miotów działających w branży technologicznej. Często bowiem nie jest on nawet dosta- tecznie świadomy zagrożeń dla prywatności, towarzyszących korzystaniu z różnych usług online (Kuczyński, 2009). Przykładem może być chociażby popularny na całym świecie serwis społecznościowy Facebook, z którego korzystanie, z różnych względów, postrze- gane jest jako konieczność, a jego polityka prywatności budzi liczne kontrowersje (zob. Papińska-Kacperek, Polańska, 2016; Giermak, Popiołek, 2014). Rodzi się pytanie o zrozumienie zmieniającego się zakresu prywatności, o stopień akceptacji tej sytuacji oraz o pożądany poziom prywatności. Zagadnienie wydaje się szczególnie interesujące wśród młodych osób, często określanych mianem cyfrowych tubylców (Prensky, 2001) wychowanych w świecie wszechobecnej technologii informa- cyjno-komunikacyjnej. Niniejszy artykuł stanowi próbę uchwycenia najważniejszych problemów metodologicznych, z jakimi można się zetknąć podejmując próby badawcze w obszarze zarządzania prywatnością online. Za cel autorzy powzięli pokazanie różno- rodności podejść do tego zagadnienia, przedstawienie możliwych sposobów badania pry- watności w internecie, a dzięki temu także teoretyczne wprowadzenie do własnych badań. Dokonano przeglądu literatury odnoszącej się do zagadnienia prywatności online, wyróż- niając możliwe warianty badawcze. Wyciągnięto także pierwsze wnioski z dotychczaso- wych, własnych badań nad zarządzaniem prywatnością w sieci. 1. Zagadnienie prywatności – różnorodność ujęć teoretycznych Pojęcie prywatności jest oczywiście znacznie starsze od współczesnych technolo- gii informacyjno-komunikacyjnych (ICT), lecz w porównaniu z innymi prawami czło- wieka jest stosunkowo młode (zob. Pryciak, 2010). Prywatność jest ściśle powiązana z poziomem oraz stylem życia. Inaczej postrzega się prywatność przykładowo w obec- nej europejskiej kulturze dużych zamożnych miast, inaczej – w kulturach tradycyjnych ubogich wiejskich społeczności z wielopokoleniowymi rodzinami, jeszcze inaczej była postrzegana kilka wieków temu. I można postawić tezę, że w ostatnich kilku lat podej- ście do prywatności wśród osób z szerokim dostępem do technologii ICT także gwał- townie się zmienia. Widać jednak, że ewolucja prywatności jest rzeczą w pełni natural- ną., a podejście zależy od kultury, zmienia się wraz z zamożnością społeczeństwa i postępem technicznym. Koncepcję prywatności opracowaną przez Warrena i Brandeisa (1890), którą określono jako prawo do bycia pozostawionym w spokoju (right to be let alone), można traktować jako pierwszą dość dojrzałą jej teorię. Tak jak obecnie jako zagrożenie pry- watności traktowane są technologie ICT, wówczas podchodzono tak do fotografii. Tym Malwina Popiołek, Jędrzej Wieczorkowski 233 niemniej jako element ponadczasowy tej teorii należy uznać wydzielenie obszarów życia – prywatnego i publicznego. Już wówczas uznano, że niedopuszczalne jest ujaw- nianie życia prywatnego, jeśli nie wiąże się to z interesem publicznym. W kolejnych latach podejście do prywatności ewoluowało, jej ochrona stawała się coraz silniejsza i stopniowo prywatność stała się istotnym elementem praw człowieka i godności ludz- kiej. Trudno o jednoznaczną definicję prywatności, ze względu na jej wieloaspektowy charakter: m.in. prawny, filozoficzny, socjologiczny, technologiczny. Przykładowo można wyróżnić trzy podstawowe metody traktowania prywatności: relacyjna (iteracyjny) zwią- zana z kontrolą kontaktów społecznych; informacyjna związana z zasobem i charakterem przekazywanych informacji oraz przestrzenna (fizyczna) dotyczący fizycznej dostępności do osoby (Dopierała, 2013). Z punktu widzenia zachowania prywatności w internecie istotne są dwa pierwsze podejścia. W tym kontekście można rozumieć prywatność jako kontrolę przepływu informacji prywatnej – sprawowanie kontroli nad informacjami doty- czącymi jednostki, a w szczególności nad tym, w jakim zakresie dana informacja jest udostępniana innym (Westin, 1967). Kontrolę dostępu do takich informacji należy rozpa- trywać zawsze w określonym kontekście społecznym – które informacje, komu, kiedy i jakiej sytuacji mogą zostać przekazane (Nisseubaum, 2004). Zakładając że użytkownicy internetu mają potrzebę zachowania pewnego poziomu prywatności, zapewne podejmują jakieś działania z tym związane. Można więc mówić o praktykach zarządzania prywatnością. Przykładowo A. 1. Zagadnienie prywatności – różnorodność ujęć teoretycznych Westin (1996) wyróżnił posta- wy: fundamentalistyczną (najsilniejsza ochrona kosztem rezygnacji z korzyści), beztroska (najsłabsza ochrona), pragmatyczną (każdorazowa ocena korzyści i kosztów udostępnie- nia prywatnych informacji). Należy jednak pamiętać, że np. serwisy społecznościowe są komercyjnymi przedsięwzięciami, a ich podstawowy model biznesowy zakłada udostęp- nienie platformy w zamian za dostęp do spersonalizowanych informacji współtworzonych przez społeczność (Polańska, Wassilew, 2015). Jednocześnie niejednokrotnie korzystanie z takich serwisów przestaje być wyborem, a staje się koniecznością ze względu na sytu- ację zawodową lub kontakty towarzyskie. Bardzo uciążliwe jest także wykorzystywanie internetu jako źródła informacji bez pozostawiania śladów w postaci danych prywatnych. Pytaniem więc pozostaje jakie postawy przyjmują obecnie młodzi ludzie wobec prywat- ności w internecie i jak badać te postawy? 2. Badanie prywatności w internecie – perspektywy badawcze i podejścia metodologiczne 1 Pojęcie cyberbezpieczeństwa najczęściej kojarzone jest z tzw. bezpieczeństwem telein- formatycznym. W anglojęzycznym nazewnictwie występują dwa terminy: computer security i cyber security, przy czym pierwsze dotyczy głównie systemów informatycznych, drugie odnosi się do bezpieczeństwa w cyberprzestrzeni w sensie bardziej ogólnym. 2. Badanie prywatności w internecie – perspektywy badawcze i podejścia metodologiczne Zainteresowanie problematyką zarządzania prywatnością w internecie systematycznie wzrasta. Można przypuszczać, że wraz z rozwojem i upowszechnieniem dostępu do ICT trend ten będzie się utrzymywał. Wśród publikacji naukowych polskich autorów poświęco- nych temu zagadnieniu pojawiają się głównie prace eksploracyjne (zob. Surma, 2013; Koło- Zarządzanie prywatnością w internecie – wybrane problemy metodologiczne 234 dziejczyk, 2014; Wieczorkowski, Pawełoszek, 2016); rzadko znaleźć można monografie, w których autorzy opisywaliby zagadnienie kompleksowo (zob. Dopierała, 2013). Zagadnienie zarządzania prywatnością w internecie jest dość problematycznym obsza- rem refleksji naukowo-badawczej. Wynika to przede wszystkim z dużej transdyscyplinarno- ści Wiele problemów badawczych które mieszczą się w ramach szeroko rozumianej reflek Zagadnienie zarządzania prywatnością w internecie jest dość problematycznym obsza- rem refleksji naukowo-badawczej. Wynika to przede wszystkim z dużej transdyscyplinarno- ści. Wiele problemów badawczych, które mieszczą się w ramach szeroko rozumianej reflek- sji nad społeczeństwem informacyjnym ma charakter inter- lub transdyscyplinarny, co może być postrzegane jako pewnego rodzaju bariera metodologiczna (zob. Sienkiewicz, 2016). Refleksja nad zarządzaniem prywatnością online rozwija się równolegle w ramach wielu dyscyplin, w tym przede wszystkim: 1. W naukach informatycznych i technicznych – pojęcie prywatności wiąże się głównie z metodami ochrony danych w przestrzeni cyfrowej, a więc zagad- nieniem bezpieczeństwa teleinformatycznego, rozumianego jako ochrona systemów teleinformatycznych przed włamaniami, kradzieżą czy też nie- uprawnionym wykorzystaniem danych (Klimek, 2015, s. 96; Malko, Lis, 2016, s. 182–185). 2. W naukach prawnych – dotyczy prawnych aspektów związanych z ochroną i udostępnianiem informacji online. W tym ujęciu uwaga skupia się zwykle na analizowaniu praktyk podmiotów oferujących swoje usługi w sieci i ich ocenie pod kątem zgodności z obowiązującym stanem prawnym. Szeroko analizowana jest także kwestia zasadności wykorzystywania prywatnych da- nych pozyskanych online przez różne organy państwowe, np. dla potrzeb za- pewnienia szeroko rozumianego bezpieczeństwa (Świeboda, 2013). 3. W naukach społecznych – bada się szerokie spektrum przyczyn i konsekwencji związanych z przemianami w obszarze prywatności, a także wpływem interne- tu i ICT na rodzaj i charakter tych przemian. Samo zarządzanie prywatnością rozumiane jest najczęściej jako jedna z e-umiejętności czy e-kompetencji i ana- lizowane z perspektywy użytkowników internetu. Dominują tu przede wszyst- kim badania socjologiczne (Mazurek, Zając, Rakocy, 2007), psychologiczne (Sanecka, 2014), pedagogiczne (Czopek, 2015), politologiczne (Lakomy, 2013), a także medioznawcze. Mamy zatem do czynienia z dużą różnorodnością perspektyw badawczych, jak również wielorakim definiowaniem pojęcia prywatności online i sytuowaniem jej raz w obszarze cyberbezpieczeństwa59F1, innym zaś razem w obszarze e-kompetencji cyfro- wych. 2. Badanie prywatności w internecie – perspektywy badawcze i podejścia metodologiczne Z metodologicznego punktu widzenia należałoby wyodrębnić dwa główne podej- ścia do badania prywatności online – podejście ilościowe i jakościowe. Pierwsze ma przede wszystkim na celu stwierdzenie czy jakieś zjawiska występują lub nie, Malwina Popiołek, Jędrzej Wieczorkowski 235 a także określenie skali ich występowania (często na poziomie makro). Dominować będą w nim badania kwestionariuszowe w formie ankiet przeprowadzane na dużych reprezentatywnych próbach populacji. Przykładem mogą być analizy zrobione na zlece- nie różnych instytucji, np. Generalnego Inspektoratu Ochrony Danych Osobowych (Raport z badań…). W badaniach jakościowych nie analizuje się skali zjawisk, a ich złożoność, przy- czyny i ewentualne konsekwencje (zob. Flick, 2012; Silverman, 2012). Wybór metod jest tu dość szeroki, począwszy od obserwacji i metod etnograficznych, poprzez indy- widualne wywiady pogłębione, aż do zogniskowanych wywiadów grupowych. W tego typu badaniach dobór próby ma często charakter celowy lub arbitralny, a jej struktura nie musi być reprezentatywna (Babbie, 2008, s. 212). Z względu na cel prowadzonych badań można mówić o trzech głównych przyczy- nach poznawczych badania prywatności w internecie (Babbie, 2008, s. 107–110): 1. Badania eksploracyjne – ich celem jest dokonanie wstępnego rozeznania w określonym obszarze refleksji. Podejście eksploracyjne jest częste, zwłaszcza wówczas, gdy przedmiot badań jest nowym zjawiskiem. Zagadnienie zarządza- nia prywatnością online niewątpliwie takim zjawiskiem jest, toteż w literaturze dominują właśnie prace o charakterze eksploracyjnym (np. prezentujące wyniki badań pilotażowych. Badania te mogą wprawdzie zawierać zarówno elementy opisu, jak i wyjaśnienia zjawisk, stanowią one jednak wstęp do dalszych pogłę- bionych badań, a nie próbę uzyskania ostatecznych odpowiedzi (Babbie, 2008, s. 107). Praktyka badawcza pokazuje, że w przypadku eksploracji poszczegól- nych zagadnień stosowane są różne warianty metod, zarówno ilościowe, jak i jakościowe, przy czym w obydwu przypadkach wyniki raczej nie dają się eks- trapolować na ogół populacji. 2. 2. Badania deskryptywne (opisowe) – ich celem jest dokonanie szczegółowej charakterystyki badanego przedmiotu. Szuka się zatem odpowiedzi na pyta- nia: co, ile, kiedy, gdzie? Dominuje podejście ilościowe. Aby dokonać opisu tego jak zmieniają się preferencje w zakresie ustawień prywatności, np. ze względu na wiek, płeć, czy wykształcenie badanej populacji, sięgamy właśnie po badania ilościowe. 2. Badania deskryptywne (opisowe) – ich celem jest dokonanie szczegółowej charakterystyki badanego przedmiotu. Szuka się zatem odpowiedzi na pyta- nia: co, ile, kiedy, gdzie? Dominuje podejście ilościowe. Aby dokonać opisu tego jak zmieniają się preferencje w zakresie ustawień prywatności, np. ze względu na wiek, płeć, czy wykształcenie badanej populacji, sięgamy właśnie po badania ilościowe. 3. 2. Badanie prywatności w internecie – perspektywy badawcze i podejścia metodologiczne Badania eksplanacyjne (wyjaśniające) – ich celem jest udzielenie odpowiedzi na pytania dotyczące przyczyn zjawisk, a także próba zrozumienia ich złożo- ności. Dominować będzie w nich podejście jakościowe, a więc np. obserwa- cja, badania terenowe, wywiady, czy podejście etnograficzne (np. netnogra- fia). Wywiady stanowią bogate źródło informacji na temat motywów jakimi kierują się użytkownicy usług online udostępniający swoje prywatne dane w różnych sytuacjach. Przede wszystkim jednak wywiady indywidualne czy zogniskowane wywiady grupowe pozwalają zorientować się na ile prywat- ność jest dla badanych istotna, jaka jest ich świadomość w zakresie zagrożeń związanych z korzystaniem z usług online, a także czy postrzegają oni prze- 3. Badania eksplanacyjne (wyjaśniające) – ich celem jest udzielenie odpowiedzi na pytania dotyczące przyczyn zjawisk, a także próba zrozumienia ich złożo- ności. Dominować będzie w nich podejście jakościowe, a więc np. obserwa- cja, badania terenowe, wywiady, czy podejście etnograficzne (np. netnogra- fia). Wywiady stanowią bogate źródło informacji na temat motywów jakimi kierują się użytkownicy usług online udostępniający swoje prywatne dane w różnych sytuacjach. Przede wszystkim jednak wywiady indywidualne czy zogniskowane wywiady grupowe pozwalają zorientować się na ile prywat- ność jest dla badanych istotna, jaka jest ich świadomość w zakresie zagrożeń związanych z korzystaniem z usług online, a także czy postrzegają oni prze- Zarządzanie prywatnością w internecie – wybrane problemy metodologiczne 236 strzeń wirtualną jako przestrzeń prywatną czy publiczną. W badaniu prywat- ności ciekawym sposobem gromadzenia danych może być także nowa metoda jaką jest netnografia (zob. Kozintes, 2012). 3. Netnografia w badaniu prywatności online Netnografia, jak pisze D. Jemielniak (2013), to metoda badań wirtualnych oparta na bezpośrednim zastosowaniu zespołu technik i metod stosowanych w repertuarze antropolo- gicznym. Początkowo miała charakter pomocniczy i polegała na uzupełnianiu danych służą- cych triangulacji, np. przez wzbogacenie informacji z wywiadów o kontekst stron interne- towych dla danej zbiorowości. Współcześnie etnografie wirtualne dotyczą przede wszystkim samodzielnych badań społeczności internetowych, choć naturalnie nadal mogą mieć zasto- sowanie także wtedy, gdy po prostu internet jest jednym z wielu narzędzi badawczych, a także gdy analizuje się zjawiska zachodzące zarówno online, jak i offline. Stosując netnografię możemy np. badać zachowania użytkowników serwisów społecznościowych, forów przez analizę materiałów przez nich udostępnianych. Można wówczas dokonać jakościowej oceny zdjęć i innych materiałów upublicznianych przez internautów, oceniając przy tym stopień autoujawnienia (inaczej samoujawnienia – self- disclosure) i stopnia zachowania prywatności w określonych sytuacjach online. Wydaje się, że w przypadku badań nad prywatnością w internecie dla uzyskania najbardziej miarodajnych efektów korzystne jest zwiększanie rzetelności badań poprzez tzw. triangulację metod (zob. Denzin, 2006), a więc łączenie różnych sposobów pozy- skiwania danych. Takie podejście zastosował np. Ł. Kołodziejczyk (2014), przeprowa- dzając najpierw badanie kwestionariuszowe, a następnie badanie jakościowe w postaci wywiadów pogłębionych, dzięki czemu udało mu się uzyskać lepszy wgląd w badane zagadnienie, zwłaszcza w kwestii definiowani pojęcia prywatności przez badanych, a także określenia jej istotności. Projektując badanie nad prywatnością online, prócz wspomnianej już świeżości za- gadnienia, należy wziąć pod uwagę także dużą dynamikę zachodzących procesów. Warto zatem w badaniach przyjmować podejście indukcyjne (Babbie, 2008, s. 67–70). Należałoby wówczas przeprowadzić badania jakościowe, które pozwolą na sformułowanie wstępnych kategorii badawczych, pól konceptualnych, a także ewentualnych hipotez, które będzie można następnie weryfikować dalszymi badaniami. Stosując podejście dedukcyjne, można bowiem wpaść w pułapkę przyjmowania niewłaściwych założeń i testowania błędnie po- stawionych hipotez. Dla przykładu – często badając sposoby zarządzania informacją pry- watną, zakładamy a priori, że badani przez nas użytkownicy usług online cenią sobie pry- watność i uważają ją za istotną wartość, tymczasem niekoniecznie musi tak być. Częstym problemem jest też odwoływanie się do teorii innych badaczy, które straciły na aktualności. Przestrzeń internetowa dynamicznie się zmienia, zatem twierdzenia sprzed kilku czy nawet kilkunastu lat, mogą współcześnie mało przystawać do rzeczywistości. Podejście indukcyjne pozwala na uniknięcie tego typu pułapek. Malwina Popiołek, Jędrzej Wieczorkowski 237 4. Podsumowanie i kierunki dalszych badań W celu rozpoznania problematyki postaw wobec prywatności w internecie autorzy prowadzą obecnie eksploracyjne badanie kwestionariuszowe, którego celem jest okre- ślenie subiektywnego poziomu zagrożenia prywatności respondentów związanego z przetwarzaniem wskazanych danych prywatnych, w szczególności tych pochodzących z internetu. Badany jest także poziom akceptacji ankietowanych dla naruszenia prywat- ności, związanego z różnorodnymi podanymi celami przetwarzania, w tym biznesowy- mi i ogólnospołecznymi, a także ewentualnie stosowane metody służące zachowaniu prywatności lub anonimowości. Szczegółowe omówienie dotychczasowych wyników wykracza poza cele niniejszego artykułu, lecz warto wskazać wybrane problemy już widoczne przy realizacji badania. Ponieważ w badaniu ankietowym bardzo istotne są dobór, precyzja i jednoznacz- ność pytań, a także ustandaryzowanie odpowiedzi, posłużono się zamkniętymi pytania- mi z odpowiedziami w klasycznej 5-stopniowej skali Likerta. Przykładowo, w pierw- szej części kwestionariusza z góry podano listę metod i rodzaj przetwarzanych danych, np. poziom poczucia zagrożenia związany z gromadzeniem informacji o zachowaniach użytkowników w internecie, w tym informacji o odwiedzanych stronach. Przy konstru- owaniu pytań dokonano uprzednio analizy literatury przedmiotu oraz prasy popularnej (a więc, jak wcześniej zaznaczono, źródeł, które szybko mogą ulec dezaktualizacji). Wątpliwość może zatem budzić to, na ile postawione pytania odpowiadają rzeczywiście tym problemom naruszania prywatności w internecie, które są ważne dla respondentów w momencie przeprowadzania badania. I przede wszystkim, jak ankietowani w ogóle tę prywatność rozumieją, czy jest to dla nich istotna wartość. W praktyce przyjęto więc założenie, że tak – prywatność jest wartością. Analizując odpowiedzi respondentów, niewątpliwie dużo można dowiedzieć się o ich postawach. Możliwe jest tworzenie róż- nych statystyk, np. korelacji pomiędzy poszczególnymi elementami. Przy próbie odpo- wiedniej wielkości można też grupować ankietowanych według różnych cech demogra- ficznych (np. płci). Mimo zalet powyższego ilościowego podejścia eksploracyjnego, należy zdawać sobie sprawę z jego ograniczeń, np. takich, że badanie nie pozwala w pełni zrozumieć przyczynowości zjawiska oraz poznać motywacji respondentów. Biorąc pod uwagę charakterystykę poszczególnych podejść metodologicznych przy badaniu prywatności w sieci, należy rozważyć pracę w interdyscyplinarnych ze- społach badawczych. Możliwe jest łączenie metod stosowanych w różnych dziedzinach, takich jak socjologia i demografia. W szczególności wskazane jest umiejętne łączenie metod ilościowych i jakościowych. Podejście takie pozwala bowiem spojrzeć na badane zagadnienie w sposób szeroki, uwzględniając szerokie spektrum istotnych czynników. W tym kierunku autorzy zamierzają prowadzić swoje dalsze badania. Zarządzanie prywatnością w internecie – wybrane problemy metodologiczne 238 ONLINE PRIVACY MANAGEMENT – SELECTED METHODOLOGICAL ISSUES Keywords: privacy, cyber security, online privacy management Summary. The article describes major methodological problems associated with the study of online privacy management. It deals with the problem of theoretical contexts of the subject, as well as the possible research approaches in different fields of science, e.g.: computer and engi- neering sciences, legal sciences and social sciences. The paper describes possible ways of empiri- cal data collecting (quantitative and qualitative methods), as well as the types of research in this area divided by purpose: exploration, description, explanation. It describes also new ways of research which are possible to study privacy management on the net, e.g. netnography. Translated by Malwina Popiołek and Jędrzej Wieczorkowski Bibliografia Warren, S.D., Brandeis, L.D. (1890). The Right to Privacy. Harvard Law Review, 5 (5). Westin, A. (1967). Privacy and freedom. New York: Athenaeum. Westin, A. (1967). Privacy and freedom. New York: Athenaeum. Westin, A. (1996). Privacy in the Workplace: How Well Does American Law Reflect American Values? Chicago-Kent Law Review, 72 (271), 271–283. Wieczorkowski, J., Pawełoszek, I. (2016). Big data privacy concerns in the light of survey results. IADIS International Journal on WWW/Internet, 14 (1), 70–85. Pobrano z: http://www.iadis portal.org/ijwi/papers/2016141106.pdf. Bibliografia Babbie, E (2008). Podstawy badań społecznych. Warszawa: Wydawnictwo Naukowe PWN. Czopek, J. (2016). Bezpieczeństwo i ochrona prywatności młodzieży w internecie w kontekście edukacji medialnej. Zeszyty Naukowe Wyższej Szkoły Humanitas. Pedagogika, 12, 67–74. Denzin, N. (2006). Sociological Methods: A Sourcebook. Aldine Transaction. Dopierała, R. (2013). Prywatność w perspektywie zmiany społecznej. Kraków: Zakład Wydawni- czy Nomos. Flick, U. (2012). Projektowanie badania jakościowego. Warszawa: Wydawnictwo Naukowe PWN. Giermak, M., Popiołek, M. (2014). Media społecznościowe a prawo do informacji publicznej. W: W. Opioła, M. Popiołek (red.), Media społecznościowe. Nowe formy ekspresji poli- tycznej. Toruń: Wydawnictwo Adam Marszałek. Jemielniak, D. (2013). Netnografia czyli etnografia wirtualna. Prakseologia, 152, 97–116. Klimek, M. (2015). Bezpieczeństwo teleinformatyczne i bezpieczeństwo informacji w przedsię- biorstwach budownictwa inżynieryjnego. Zeszyty Naukowe Uniwersytetu Przyrodniczo- Humanistycznego w Siedlcach, 105, 95–102. Kołodziejczyk, Ł. (2014). Prywatność w internecie. Warszawa: Wydawnictwo SPB. Kozintes, R.V. (2012). Netnografia. Badania etnograficzne online. Warszawa: Wydawnictwo Naukowe PWN. Kuczyński, G. (2009). Ochrona prywatności w internecie. Marketing w Praktyce, 3, 30–32. Lakomy, M. (2013). Zagrożenia dla bezpieczeństwa teleinformatycznego państw – przyczynek do typologii. E-politikon, 6, 100–141. Malko, J., Lis, R. (2016). Cyberbezpieczeństwo systemów zabezpieczeń i sterowania. Przegląd Elektrotechniczny, 1, 182–185. Mazurek, P, Zając, J., Rakocy, K. (2007). Między inwigilacją a uwiedzeniem: Użytkownicy internetu wobec praktyk gromadzenia i przetwarzania danych. Studia Socjologiczne, 3 (186). Nissenbaum, H. (2004). Privacy as Contextual Integrity. Washington Law Review, 79, 101–139 Papińska-Kacperek, J., Polańska, K. (2016). Publiczne zasoby informacyjne w mediach społecz- nościowych. Ekonomiczne Problemy Usług, 122, 93–102. Polańska, K., Wassilew, A. (2015). Analizy big data w serwisach społecznościowych. Nierówności Społeczne a Wzrost Gospodarczy, 4 (2), 117–128. Prensky, M. (2001). Digital Natives, Digital Immigrants: Do They Really Think Diơerently? On the Horizon, 9 (6), 1–6. Pryciak, M. (2010). Prawo do prywatności. Studia ErasmianaWratislaviensia, 4. Raport z badań na zlecenie GIODO: Postrzeganie zagadnień związanych z ochroną danych i prywatności przez dzieci i młodzież. Pobrano z: http://www.giodo.gov.pl/ 1520124/id_art/6914/j/pl. Sanecka, E. (2014). Samoujawnianie online. Przegląd teorii i wyników wcześniejszych badań. W: K. Tucholska, M. Wysocka-Pleczyk (red.), Człowiek zalogowany. Kraków: Biblioteka Jagiellońska. Malwina Popiołek, Jędrzej Wieczorkowski 239 Sienkiewicz, P. (2016). Metodologiczne aspekty badań nad rozwojem społeczeństwa informacyj- nego. Ekonomiczne Problemy Usług, 123, 311–319. Silverman, D. (2012). Prowadzenie badań jakościowych. Warszawa: Wydawnictwo Naukowe PWN. Surma, J. (2013). The Privacy Problem in Big Data Applications: An Empirical Study on Face- book. ASE/IEEE International Conference on Social Computing. Świeboda, H. (2013). Problem prywatności w społeczeństwie informacyjnym. Zeszyty Naukowe Uniwersytetu Szczecińskiego. Ekonomiczne Problemy Usług, 105/2, 93–106. Translated by Malwina Popiołek and Jędrzej Wieczorkowski Cytowanie Popiołek, M., Wieczorkowski, J. (2017). Zarządzanie prywatnością w internecie – wybrane pro- blemy metodologiczne. Ekonomiczne Problemy Usług, 1 (126/2), 231–239. DOI: 10.18276/epu.2017.126/2-23.

Subsets and Splits

No community queries yet

The top public SQL queries from the community will appear here once available.